The principles and practice of screening for disease

WORLD HEALTH OR-GANIZATION ' ' ORGANISATION MONDIALE DE LA SANTt . ~. PRINCIPIES AND PRACTICE. OF SCREENING FOR DISEASE J. M. G. Wilson and G. Jungner Dr J. M. G~ Wilson is a Principal Medical Officer at the Ministr,Y of Health~ London; Dr G. Jungner is .Chief., Clinical.Chemistry .Department., Sahlgren' s Hospital., Gothenburg., Sweden. The views expressed in-this paper are their own and do.not necessarily rep- resent those of either the World Health Organization or., in the ca5e of Dr Wilson., the Ministry of Heaitli •. PA/66.7 The Issue of this· document · d~s ·not constitute formal publication. It .sh?uld not. be r.evl.ewed,_ abstracted or quoted without the agreement of the World Health Organization. Authors alone are responsible for views C\Xpressed In signed article~. Ce document ne constftue pas une publfcatfon. II ne dolt falre !'objet d'aucun compte rendu ou r~sum~ nl d'aucune citation sans l'autorfsatlon de !'Organisation Mondlale de Ia Sant6 Les opinions exprlm~es dans les articles. slgn~s n'engagent que leurs auteurs. 1. 2. 3. 4. - 2 - CONTENTS INTRODUCTION •••• TERMS OF REFERENCE DEFINITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 6 7 3.1 3.2 3.3 3.4 3.5 3.6 3.7 Screening . . Mass screening • Selective screening Multiple screening ••• . . . . .. . Case-finding • • • Population or epidemiological surveys Early disease detection . . . . . 7 PRINCIPLES • • • • • • • • • • 4.1 General considerations • . . . . . . . . . . . . . 7 8 8 8 8 8 8 8 8 4.2 4.3 4.1.1 4.1.2 4.1.3 The aim of early disease detection • Pattern of screening development • The use of different forms of screening Selective screening Mass public health screening • . . Surveillance • • • . . . . . . . . . . • 10 . . . . . 13 . • • . . 13 . . . • • 14 • 14 4.1.3.1 4 .1.3.2 4 .1.3.3 4.1.3.4 4 .1.3.5 Screening hospital patients •••••••••• 15 Screening in industry . . . . . . . . . . . . . 16 ,, --.. ~-~- ................... . Evaluation of results of screening •• . . . . . • • • • • 16 4.2.1 4.2.2 4.2.3 General . . . . . . . . . . Evaluation of screening procedures • 4.2.2.1 Validity . · . . . . . . . 4.2.2.2 Reliabfli ty . . 4.2.2.3 Yield 4.2.2.4 Cost . . 4.2.2.5 Acceptance . 4.2.2.6 Follow-up services . The "borderline" problem • • • • . _. . . . . . 16 . . . . . . . •• 16 • 16 . . . . . . • • • 16 . . . . . . • 16 • • • • • 16 . . • 16 • 16 • • • . . . . 20 Principles of early disease detection . . . . . . . . • 22 4.3.1 4.3.2 Impqrtant problem for the individual and the community •• 22 Accepted treatment • • • • • • • • • • • 22 5. - 3 - 4.3.3 Facilities for diagnosis and treatment should be available 22 4.3.4 4.3.5 4.3.6 4.3.7 4.3.8 Recognizable latent or early symptomatic stage • Suitable test or examination • . . . . . . . . Acceptability to population Need for surveys • • .. . . . . . . . . Groups to be treated in case-finding programmes •• 22 22 • • 22 22 22 4.3.9 Economic balance of the cost of case-finding in relation to total expenditure on medical care • • • 22 4.3.1:0 · Case-ffnding should be a continuing process 4·.3.10.1 · Concept of "surveillance 11 PRACTICE • 5.1' · A'utomatic dat·a handlfng . 5.1.1' . The' particular ne.eds 'o{ screeni'ng · 5.1.2· ·Basic concepts • · 5.1.} Data ·collection · 5·.1.4 . 5".1.5 · 5.1.}.1' · P\mch card 5.1.3.2 Paper tape • 5.1.3.3 · Maghetic tape · Data processing Data storage • • . . . 5.2 . Present' screening practice (including multiple screening) 5.2.1 ·5~2~2 ·5.2.3 Introduction • ·• ~ Comparison of studies ·5~2~3~2 . 5~2:3~3' '5:2:3:4 '5~2.3:5 Hunterdon county San'Francisco longshoremen Chicago'B6ard.of Health' : ·common conditions.for screening 5.3 ·Epidemiological·studies· 5.4 'Periodic nealth·examinations • '5~4.1 5.4.2 5.4.3 Introduction Relationship to general practice • The industrial health examination 5.4.4 Routine examinations through life 22 • 34 • 35 .•. 35 • 36 .•• 37 • 37 ••• 38 • • 38 • 39 •• 39 •• 39 • 39 • 48 • 54 • 54 • 54 ••• 57 • •• 64 •• 64 •. 65 • 66 • 67 - 4 - 5.5 The place of screening in the provision of medical care . . . 68 5.5.1 Justification within a country . . . . . . • 68 5.5.2 Cost of multiple screening . . . . . • . . . . 69 5.5.3 Inte~-country differences 70 5.5.4 Example of gycaecological cytology . . 71 5.5.5 Conclusion • . . . . . . . . . . . . . • 73 5.6 The place of education in the early detection of disease . . . 74 5.6.1 Education of the medical profession . . . 74 5.6.2 Education of the public . . . . . . 75 6. STUDIES OF CERTAIN CONDITIONS . . . . . . 76 6,1 Diabetes mellitus . . . . . . . . . . . 76 6.1.1 General . . . . . . • . . . . . . . 76 6.1.2 Value of early treatment . . . . • . . . . . . . 81 6.1.3 Diagnostic criteria . . . . . • 82 6.2 Heart disease . . . . . . . . 84 6.2.1 Rheumatic and congenital heart disease 84 6.2.2 Ischaemic heart disease . . . . . . . . 86 6.2.2.1 Mortality . . . . . . 86 6.2.2.2 Morbidity . . . . 86 6.2.2.3 Screening 87 6.2.3 Conclusions . . . . . . . . . 92 6.3 High blood pressure . • . 93 6.4 Overweight . . . . . . . . . . 95 6.5 Respiratory diseases . . . . . . . . . . 96 6.5.1 Pulmonary tuberculosis . . . • 96 6.5.2 Non-specific respiratory disease • 100 6.6 Cancer . . . . . . . . . .. . . . • . . 102 6.6.1 Lung cancer . . . . . . . . . . . . • 102 6.6.2 Cancer of the cervix . . . • . . . . . . • 107 6.6.3 Breast cancer . . . . . • . • . . . . uo 6.6.4 Other cancers . . . . . . • . . . . ll4 7. 8. 6.7 6.8 6.9 - 5 - Diseases of the eye . . . . . ' . . 6.7.1 Chronic glaucoma 6.7.2 Otner eye diseases Diseases of the urinary tract 6.8.1 General . . . . . . . - . 6.8.2 Bacteriuria • Rheumatic diseases 6.9.1 Rheumatoid arthritis . . . . . 6.9.2 Gout • • • • 6.10 Mental illness METHODOWGICAL TRENDS IN SCREENING • • • • • • • • 7.1 Clinical or technical screening • 7.2 Participation by doctors • • • • 7.2.1 The doctor's physical examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 7 .2·.2 History-taking by doctor or specially trained medical staff 7.3 ·7.2.3 ;uest:onaries • • • • • • • ••••••• 7.2.4 Clinical methods and laboratory tests used for screening Development of screenin13 facilities .• 7.3.1 Simplification • • • • • • • • 7.3.2 Automation •••••••• . . 7.4 Automated multitest laboratories ad modum Collen •••• CONCWSIONS • • • • . • • • • • • . • • . • . • • • • • • • • • • 8.1 The need for further epidemiological investigations and, allied w:~t:·L ·ci;.is, the need to acjlieve standardization •••••• 8.2 8.3 8.4 8.1.1 8.1.2 8.1.3 8.1.4 8.1.5 8.1.6 Records Carcinoma of the uterus • Glaucoma simplex Mental illness Asymptomatic bacteriuria Cancer of the breast • • • Lung cancer . . . . . . . . Economics • Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 li6 121 122 122 122 124 124 125 126 133 133 133 134 135 135 135 139 139 140 145 146 147 148 148 148 149 149 149 149 150 150 - 6 - 1. INTRODUCTION This paper has developed from the original request of Dr F. Grundy1 Assistant Director General 1 World Health Organization, Geneva1 for a briefing paper on the early detection of-disease~ to be limited mainly'to the chronic diseases of adults in developed' countries~ · Even'within these limitations w~ found we were unable~ with a' subject matter that'not.only covers· a· great deal of medicine but is also itself breaking new ground'rapidly~ to'write a short paper and it has now grown to its present length. The subject of early disease detection is vast and it would clearly be beyond our capacity to b~ comprehensive. This account, then, represents only our personal viewpoint of a rapidly developing aspect of medicine and the examples we have chosen . . . . are ones that have, for one reason or another, appealed to us personally. There may well be other examples, equally good or better, that we have omitted. (We . ' have not, for instance, included the long-established practice of early disease detection in the maternity or child welfare field, largely because_ that practice is already so well established.) To use a simile, we have made a number of preliminary sketches rather than attempt a complete picture. We are also aware that the subject is controversial and that much still needs to be learned. If anywhere we have appeared dogmatic, we hope this may serve to stimulate discussion on which, in the end, real development depends. There- are three main headings. of the paper •. The first part deals. with the basic principles. of early disease detection; the second with the practical con- siderations, including a number o:f e~arnples from different disease conditions.; and the third and last. part deals with the use of present methods and possible development. 2. TERMS OF REFERENCE The. terms- of refe;rence of- this paper are ."to discuss the general. principles of the. administrative and scientific. aspects. of screening· proceduresn •. For the purposes. of this. paper. the. definition of "screening11 proposed by the United States of America. multi-sponsore~Commission on. Chronic Illness {C.C~I.) {see section 3.1 below1.) and accepted by the Regional. Committee. for. Europe, 2 is.adopted~. Periodic phy-sical examination is also. included in the. revie:w and we._ refer to both screening - 7 - and periodic physical examination generically as "early disease detection". Epidemiological surveys to establish the prevalence and incidence of conditions as well as to study longitudinally the natural history of developing di~eas~, are not considered as falling within the terms of reference, which are regarded as confined to case-finding. However, frequent reference is made to surveys which throw light on our attitudes to case-finding. Screening for the chronic non-communicable diseases forms the main subject of the paper; but the problems facing countries at differing stages of develop- ment and having different standards and types of medical care are also discussed, and this to some extent includes communicable disease detection. 3. DEFINITIONS 3.1 Screening The Commission on Chronic Illness Conference on Preventive Aspects of Chronic Disease, held in 19511 defl.ned screening as 11the presumptive identification of unrecognized disease or defect by the application of tests, examinations or other procedures which can be applied rapidly. Screening tests sort out apparently well persons.who probably have a disease from those who probably do not. A screening test is not intended to be diagnostic. Persons with positive or sus- picious findings must be referred to their physicians for diagnosis and necessary treatment". It should be noted that, by definition, unrecognized symptomatic as well as presymptomatic disease is included, and physical examination as well, so long as it is classed as rapid. "Other Procedures" can also embrace the use of questionnaires, which are assuming an increasingly important place in screening. Finally, tests may be diagnostic, though not necessarily so intended; so that, for example~ a gynaecological examination could be covered by this definition provided it is rapidly carried out. ·rn general, we have taken the definition to imply a relatively simple method of case-finding. 3.2 Mass screening This is a term used to :indicate large-scale screening of population groups. In this paper we use this term to refer to population screening where no selection of groups is mad~. - 8 - 3.3 Selective screeniDg We use this term for the screening of selected high-risk groups in the population. It may still be large-scale and can be considered as one form of population screening. 3.4 Multiple screening This has evolved by combining single screening tests, and is the logical corollary of mass screening. Where much time and effort has been spent by a population in attending for a single test (e.g. mass radiography) it is natural and economical to offer other tests at the same time. Multiple (or multiphasic) screening has been defined as "the application of two or more screening tests in 1 combination to large groups of people". 3.5 Case-finding Throughout this paper this term is applied to that form of screening, the main object of which is detecting disease and bringing patients to treatment. 3.6 Population or epidemiological surveys Whilst screening tests may well be used in population surveys (e.g. sphygmo- manometry for blood pressure or tonometry for intraocular tension), the principal aim of surveys is not to bring patients to treatment but to elucidate the preva- lence, incidence and natural history of the variable being examined, though case- finding is a natural by-product of surveys. A good example of an epidemiological survey is the Framingham study of ischaemic heart disease.3 3.7 Early disease detection It is sometimes useful, we think, to use a term that refers to all forms of early disease detection whether by screening, physical examination or other means; and this is meant when we use the term 11early detection11 • 4. PRINCIPIES > 4.1 General considerations 4.1.1 The aim of early disease detection The aim of early disease detection (sometimes called secondary prevention) is simple. Primary prevention seeks to abolish disease by protecting the individual and the population from attack before the challenge has been made. Early detection - 9 - (case-finding) aims. at-discovering and curing conditions which have already produced pathological change but which have not so far reached a stage at which medical aid is sought spontaneously. Fig. 4 .1. These stages are shown diagrammatically in "Well" ' "Ill" FIG. 4.1 I PreJention No prevention I Irmnunity Early I pathological change Early detection I . Treatment I I No early detection I Developing pathological change I Symptoms I Treatment I I No treatment I Gross pathological change Example: Pulmonary Tuberculosis Mantoux test, B.C.G. . M.M.R. Clinical Assessment X-ray sputum .For screening, as defined, there is the second, economic, aim of achieving more for unit expenditure by saving the time of highly-trained professional people. Part of their trained time can be substituted by less highly-trained personnel, able to carry out screening tests, whether by hand or automated (see diagram under "Periodic health examination's II, section 5.4). However, it seems likely that the total cost of screening in a community is higher, not lower, than the cost of conventional medical care since more people will be found in need of treatment (and these largely 'elderly and liable to be under care for a long time). We shall say more about the economic aspects of screening under the proper section (4.3.9). However it is worth noting here' that in some circumstances screening may be altogether uneconomical. A' condition (e.g. helminth infestation) may be - 10 - almost universally prevalent and mass treatment without soreening may in these conditions be the course of choice, thus avoiding the high cost of preliminary confirmation of a virtually certain diagnosis. 4.1.2 Pattern of screening development Sociological factors have been important in the development of screening. ~~-s~_ :r~]_.E?_v~:t __ faci:;()_!"S .c:tr~ clqse_ly relate.d to the degree of sophistication of .. the population at risk; for example, the level of education and awareness of the population, the amount and form of medical care available and the general standard of living_. In highly-developed societies a rise in the level of these factors has been accompanied by a decline in communicable disease and by an apparent in- crease in degenerative and genetically determined disease. Under these conditions, also, there-is a tendency towards diagnosis at_a less advanced stage than in the past. In less developed countries, however, the communicable diseases remain largely paramount and the stage at diagnosis of chronic illness is_often later; whilst living conditions, nutrition, the education of the public and medical care all need to be improved. For these reasons the early detection of illness presents different problems in highly-developed and less well developed countries. The historical development of screening can best be examined by observing the measures taken to control endemic communicable disease; measures which are now to a large extent no longer needed in well-developed areas, but the operation of which is still vitally important in many less-developed countries. Fig. 4.2 sets ·out the··progre·ssive stages (whether iri time or in terms of development) in diagrammatic form. Studying the reasons behind these earlier forms of screening is helpful to the understanding of the later developments in highly developed countries for screening chronic disease. In tropical and subtropical areas of the world parasitic diseases like malaria, schistosomiasis and ankylostomiasis have long been the subject of mass detection. One of the ·main reasons has been the need to control these maj~r causes of disease by attempting to stamp out the human reservoir. Similarly, chest radiography was first introduced primarily as a public health measure to help in controlling the spread of pulmonary tuberculosis; )rolonging the life and health of the individual was at that time a secondary objective. A further example can be found in the attempt to control the spread of ·syphil:Ls by mass serological examination of the population. - 11 - FIG. 4.2 '. J •••• ' .. Screening era Examples of conditions solle;ht Malaria Early Nematodes Leprosy Trachoma Pulmonary tuberculosis Middle Venereal diseases Diabetes Late Ischaemic heart disease Iron-deficiency anaemia. Only when the prevalence pool of endemic communicable disease has'largely been emptied has the objective of early detection been turned chiefly towards the second aim of chronic disease detection. Clearly, economic factors play a large part; controlling the spread of disease is vital to economic prosperity and even survival; whilst prolonging individ:ual life and health is less economically mandatory. These differences are important when considering the p~inciples upon which population screening should be based, for the relative importance of individual considerations varies in the two cases. In a smaller way the same story can be told of industrial health examinations. In the early_ days of industry there were endemic industrial diseases, like mule spinners' cancer, for which primary prevention was developed (in that case by . . removal of the causative agent). In other conditions, for example silicosis and lead pdisoning, where·tne cause could not be wholly eradicated, early detection techniques were dev'eloped (chest x..:.ray, blo?d film and urinary lead content). With advancing sophistication in' industry the idea grew of not only monitoring the health of workers in relation to known environmental health risks, but also of anticipating non-industrial hazards to the health of individuals by periodic .. .· . ~ '- .· ... , : . health examination. These two types of e~amination are comparable to the two stages of growth of general population screening. . . .. ~- ' ·;· . . . - 12 - Whilst the benefits of the first type of mass detection (to control the spread of communicable disease) have been, at least to some extent, demonstrated . ··-· ~- - ~ -.. ----~~- -·- (e.g. tuberculosis) the value of the second needs as yet largely to be ascertained (e.g. diabetes·mellittis, chronic simple glaucoma). Why this should be and how answers to outstanding questions might be found will be examined later in this paper. Much screening practice evolved in the United States of America during the 1950's in the form of multiple screening programmes and has been reviewed else- where4•5•6•7•8•9 and will c~ly be discussed briefly here. A useful bibliography was issued by the Bureau of Chronic Diseases of the California State Department of Public Health.10 The motivation for screening has been dealt with at some 11 . 12 13 length in three papers, respectively by Chapman, Mount1n and Smillie published between 1944 and 1952 at the time of maximum growth of this concept in its application to chronic diseases. Some of the chief points made in their papers were : (a) case-finding by multiple screening is a technique well suited to public health departments, whose role.is changing: (b) The need for provision for diagnosis, follow-up and treatment. This is vitally important; without it case-finding must inevitably fall into disrepute. (c) The need for validation of the tests before applying them to case- finding; harm may result to public health agencies' relationships with the public (not to merition the direct harm to the public), and with the medical profession, from large numbers of fruitless referrals for diagnosis. (d) The danger that multiple screening might lead to the neglect of other aspects of community medical care because of the competing cost and possibly also because a sense of false security might be propagated. (e) The need to evaluate the effect of multiple screening by its results in reduced morbidity and mortality. In 1957 the Commission on Chronic Illness accepted the value of multiple screening as "contributing to good medical practice" and considered that it "constitutes a practical means for early detection of a number of important - 13 - chronic diseases and impairments". The conditions for which the Commission on Chronic Illness. considered at that time.might profitably be screened were: pulmonary tuberculosis visual defects (including chronic glaucoma) hearing defects syphilis diabetes cancers of skin, mouth, breast, cervix and rectum hypertensive disease ischaemia heart disease (possibly). In 1960 the American Public Health Association strongly endorsed multiple screening in a publication "Chronic Disease and Rehabilitation; a Program Guide for State and Local Health Agencies 11 • 7 \rfuile recognizing that screening should, where possible, take second place to periodic health examination as an effective technique for early disease detection, the authors of the "Program Guide" considered that "the sheer weight of economic reality ••• dictate recourse to procedures that conserve the time and energy of highly trained personnel such as physicians and dentists 11 • With the formation, in 1961, of the Chronic Diseases Division of the Public Health Service of the United States of America came the ability to provide State services with project grants for setting up demonstration scree~ng progranmes. A relativeiy large nUmber of these projects have been carried out, but for various reasons it has been difficult for them to satisfy all the above points. 4.1.3 The use .of different forms of screening 4.1.3.1 Selective screening Screening tes'ts can, of course, be used in differ~nt ways, varying from single examinations applied ~o individuals to batteries of t~-sts offered to whole popu- lations. They may also, as already'indicated, be either indiscriminate or selective. From the viewpoint of both the individual and the economy there are obvious advantages in combining a number of tests and applying them all at the one examination, providing each has been shown 'to be medically worth its place. - 14 - However, in-practice, there may be drawbacks to combining certain tests, for· example when each only gives reasonable yields in selected population groups of different age, sex or occupation. Much screening is selective, of long standing and well established. This type of screening is practised, for example, at antenatal, post-natal and infant welfare clinics, where conditions like pre-eclamptic toxaemia, the anaemias of pregnancy and congenital conditions are sought by the application of simple tests. 4.1.3.2 Mass public health screening More recently, public health.agencies have tended to extend their screening activities from these kinds of clinics to the general public. Multiple screening has been offered at ad hoc clinics staffed by ancillary workers, positive results being notified to general practitioners. However, it has been recognized that this approach gives rise to difficulties (some of which are discussed below) and of late it has declined in popularity as a means of early disease detection. 4.1.3.3 Surveillance A third method of screening is the individual approach, as opposed to mass screening. At first sight there is little difference between screening the individual and the ordinary good practice of clinical medicine. The physician examining his patient's urine or blood-pressure or even the weight when he has no special reason to suspect illness related to these findings, is simply com- plying with an accepted standard of good diagnosti .. c practice. He may, in fact, extend these activities to other examinations like haemoglobin or electro- cardiography. If he also arranges for these examinations to be carried out by an auxiliary helper he can then be regarded as·submitting his patients to a form of multiple screening which may or may not be selective, according to. .whether·· only those patients consulting with some complaint are ex_amined, or whether the PhySician has made arrangements for all his patients to undergo these tests. The essential difference between this form of screening and ordinary -good · medical practice is that in the first instance the person examined is presumptively well, whilst in the second he or she comes to the doctor as a patient with a complaint. There is really a good deal of difference between these two concepts; the economic implications for a general practitioner in terms of time, auxiliary .;.. 15 - pe~!) and the:use of r~cords, as well.~;ts premises, is V!ary different if he is aiming to: carry, o1:1t s~lect~ye screening in his pra~tice, in contrast simply to applying a number of tests to patients reporting with,a complaint. Howev.er.; · as a development in medical care, there are clear pote!ltial advantages in this form ·~ ,'.; of positive surveillance. The general practitioner···c:iari-:;-In-·fb.is .. way,-· be enabled to practice personal preventive medicine; and,.· s~condly, the normally clOse contacts between the doctor and his practice should largely avoid communic~tion difficn.llties over the results ·Of tests • Screening hospital patients A particular f.orm of screeningspecial groups of the population is the screening of hospita~ patients. There are at least three aspects to screening patients in hospital. Firstly, hospital patients in general, whether in- or out-patients, constitute a special high-ris1~- ·group~o:f' the population and are likely to give a high yield for conditions like diabetes mellitus, cancer of the cervix and simple glaucoma •. Secondly, patients come to hospital with a complaint of ~ioh the d:i,agnosis may be ~ither in doubt or erroneous. It is then usual to ~equest laboratory and other tests in a sequential fashion, one request often (lepending on the :-result of the. previous one. $ubmit:ting the patient to a number of laborl;l.toryt-ests routinely without exE)rcising individual choice_is a form of·· scr;-eening which may prove its worth _in l~ading to diagnoses ·which would otherwise have been delayed or even missed altoget~er. This performance of a number of laboratory tests on one blood and/or urine specimen at the same time is now becoming quite feasible with the introduction "6:f"laboratory aufomatioi:i {both in carrying-out 'the tests themselves and in th3 equally onerous-matter of processing the data). It may well·prove more economical than the traditional method of seriatim-laboratory requ~sts. Thirdly~· and· following as a consequence of per- forming a number of tests simUltaneously, th~re may be:an·econ6mic gain in hospital stay. One of the most costly items of medical service is the upkeep of a patient in a hospital bed. Rationalizing medical care in hospital so as to minimize the length of stay is one of the chief ways in which the cost of health services can be kept from mounting disproportionately to other expenditure.· ·It is possible (though at the moment evidence is lacking) that screening hospital . patients could shorten length of stay or lower cost in some other way, for example - 16 - by reducing the number of hospital consultations called for. Work on these lines has been carried out in the United States of America and Canada, and trials are starting in.Great Britain and Sweden. 4 .1.3.5 Screening in industry Lastly, industrial populations may offer special advantages for screening, especially in industrialized countries where there is no universal general practitioner service. It is important to remember, of course, that industrial screening examinations are of two sorts; one kind is for special industrial risks, of which examples have been mentioned above; the other kind of examination is aimed at the early detection of diseases which may impair the general efficiency of the worker. This subject is discussed under paragraph 5.4.3. 4.2 Evaluation of results of screening 4.2.1 General The evaluation of screening can be considered from two separate aspects which yet have a certain connexion with each othe_r. These aspects are, firstly, the evaluation of tests or examinations and, secondly, the evaluation of results. The important connecting link is the need to use standard criteria for tests when comparing results between case-finding operations. Associated with these standards is the difficult problem of the "in-between" or "borderline" patient, which will be considered in this section. 4.2.2 Evaluation of screening procedures The Conference on Preventive Aspects of Chronic Disease considered the. evaluation of case-finding tests and programmes in 1951 and the matter has been dealt with at some length in the C.C.I. publication "Prevention, of Chronic - 14 Illness".- The following criteria were discussed. 4.2.2.1 Validity 4.2.2.2 Reliability 4.2.2.3 Yield ~.2.2.4 Cost 4.2.2.5 Acceptance_ 4.2.2.6 Follow-up services -··17-. In this s.eqtion we dea:l only with validity, reliability-and yield; -the other criteria are discussed later, ·under the section on Principles of Early-Disease Deteetion: cost under. paragraphs 4.3.9, acceptability under 4.3.6, and follow-up under 4.3.10~ 4.2.2.1 Validity The C.C.L defines the validity of a screening test as the measure of the fi.equericy with which the reshl t of that t~st is c'onfirm~d by .,im accep·table diagnostic procedur''e, i.e. the ability of tho test to separate those who have the condition soUght froni those who dci not. Applying a scre~ning test to a population will produce four categbries.of result, provided that the whole population is also examined definitively to establish the actu~ prevalence·. of disease. shown di~gra~ticaliy (Fig. 4.3). These.·· four possible results can be Screening result Positive ·Negative FIG. 4.3 THE EFFICIENCY OF A SCREENING TEST modified from Remein, Q •. R. & Wilkerson, H. L~, l5 True disease classification of apparently well population ~------------- ----------------------------------------------~ Unknown cases of disease With disease and with ·pos±tive--test .. · · (true positives) With disease but with - rregatf~~-~j~~~t- -- -- ·--··- ...... · (fals>:c.• ;,.,·o-a~_,.,,..;:,.;.s)~- r ., ·: :c.or.•:"· ·'\':;0-·t~~ .-:%-lfC }'-•·•···· .., ...... ~., 1- Persons without disease ···-·t ....... - ............... . · Without disease.· but \d-th poait:i:v-e-··-t·es-t~- -------- (false positives) Without disease and with negative test (true negatives) * Sensitivity = Diseased persons with .e,ositive test .. - ----··-- · ---- ·-· -·All persons in--population wi·th--d1:sea:se--· · -· ..: .• i..: ·. Specificity * N9~ise~se~ persons with negative test ................................... A~1.E~:r.::>C>~I:} __ !.l1 J?Opul8:(~_9.I?:_ w:itl:.?.~~-~-~s~a.-s.~ * These values are often expressed as percentages. I - 18 -' An ider:.l test would of course cJ..:;t.:'ct only those persr:ms in o. population suffering f'r-om the c6nditi6h lo~ked-fo:r -(as defined by agre~d criteria.) and would not fail to detect :any of-them. The ability of a test to classify as positive those persons with the disease is termed "sensitivity" and the ability to class as. negative those without the disease_, "specificity"; that is, sensitivity is a measure of the false negative rate and specificity of the false positive rate. Sensitivity and spectficity can be varied reciprocally according to the setting . ()f the test. Take_, for example_, the detection of iron deficiency anaemia by determination of the haemoglobin. Let us say that the aim is to diagnose and treat all women in a population with a haemoglobin of 11 g/100 ml or less. ' . . . ' A screening level of 11 g/100 mi will miss a number of cases due to the errors of the te~t and the sensitivity will be low though the specificity will be high. By raiping the screening level to 12 g/100 ml the sensitivity will be raised so that few cases are missed; but the penalty must be paid of lowering the specificity and accePting a rise in the number of false positives. this point: The following example· (Fig. 4.4) illustrates FIG. 4.4 A,-.••'><-• ,_ ·-· ...... ~ ... ·-·r--.-- ''"·• .. ····· Total Anaemic Screening level population . - patients I' . l 11 g % haemoglobin 12 g % haemoglobin ... , . . -~ _ .... ~. 4>· ....... . ·- Positive I Negative -- Positive Negative 100 20 . ~ l ! True False, True lFalse TrueiFalse True False i --.. --~~~--- ... ...... . ·-·. -~. ---~· .. ..... . I I I .. 15 2 I 78 5 19 10 I 70· 1 • • • ~ .. h •• ··~" ••••••••• -. ..... ---·-··•-.. -- .. . . ...... .. ,. Sensitivity 15 100 75 % 19 X 100 95 % X = = 20 20 ......... --~--- . .. ... , ... --· ... * Specificity 7.8 X 100 = 98 %· 70 100 88% 80·· Sox ·=· - .... - •· ""• ... .- - * For working, see Fig. 4.3. - 19 .. 4.2.2.2 Reliability Providing the test ~elected is a good inde~ of th,e -CI.:Lsease sought~ two factors . :,\, - _·,:•,- . , .... are c.once.I"ned in the. reliability or efficiency of the test: the variati()n of the method and the variation of the observ~r. For;exa.znple, in measuring the arterial blood-pressure with an inflatable cuff sphygomanometer there ar~ the variations . . . . . . . . ... ~ connected with the indirect relationship between the method and the true intra- arterial blood.:.pressure and wfth'the 'variability of the blood-pressure itself; and the~e is also the error of the observer (whieh has re~ently been shown to be much more than was previouSly supposeci).'l6,l7 In considering screening techniques there is scope for research into methods. Ideally (as has been said) a test should behighly sensitive and should miss yery ~ . •'_-' . ·' · .. : .. , . .. . · .. ·_. . :·j . : .... ·: : .. · ..:. ':. . ·,>·. : ''_.;h,.· .. ·' :. -~- . ' . : . . . ' •. ' .. few••pet'S.ons with the d_isease, though a rela:tively high_proportion of, false po~i tives - .. ... . .· - ·..... . . ' . . . ' can be accepted; it should be as simple as __ possible _and able t? .. b~,,.Qafi":l.e<;l oou:t ,_, rapidly, often under.improvised field conditions (though in some instances there is ·. -,-. a],so'a case fora 'firmly-based unit to which the population' comes or where speci- mens are sent); thirdly a test must be-acceptable and cause minimal.distui-bance to the subject in its performance. Pain and discomfort, much undressing ~~ the need for a large blood sample may rule out an otherwise_ ~?'?..f!.l!.~t~?.~~~·-····-Ia..§~.!ti';...i t should be a~ _c:qE!_ap . as pos.s~ble • Speed tends to <i.:l.minish efficiency and yice. versa; the local need will probably determine whic:t). factqr::;_ a,re the more important. For example, in detecting diabetes mellitus, urine testing may be chosen for its sim- plicity, cheapness and minimal d:lsturbance to the population. However, testing fo~ glycosuti~ is· inefficient and bl()?.ci: s~ar screening on the spot using capillary · biood might be cho-~'~ri rather than obtaining a ve~ous blC>od 'sample which :has the drawbacks. of co~t in taking, -trans;orting ancf ch~m:f~cilly analysing many spec,imeris ' ' ,, of blood~ ·' , Recezit coininerci~l research h~s ·prodticed a qui~k gluco~e oxidase blc!j~J sug~r ·screening st~ip, 'ihcUgti':rt has not; .as yet, ,,been validated ·by full fieid trials. Here sl:mplici ty. is u~ually. obtained at an uriacceptab1e sacrifice. of accuracy. There has: 'been need for coin~lirison of' a:vailabl'e. Irieth6'as and at least one has riow been publlshed. 18 . It should sdori be poss1bl~ '<to itJake· a rational choic~ related to need. - 20- 4.2.2.3 Yield The yield from screening can be considered as the measure of previously unrecognized disease (whether overt or latent), diagnosed as the result of screening and brought to treatment. Other forms of yield are provided by persons with known disease who have previously lapsed from treatment. The yield is clearly primarily related to the prevalence of disease in the population and to the availability and use of medical care facilities. The highest yields from screening will be obtained from screening for a highly prevalent con- dition in a population where medical care facilities are minimal, e.g. for malari~ carriers in a poorly-developed area. Where medical care is good, though a con- dition may be relatively common, less new disease may be discovered through screening. In highly-developed countries, for example, there appear to be about as many undiagnosed as diagnosed persons with frank diabetes. The other important factor in yield is the efficiency of·the test itself. Thus urine testing for glycosuria will miss large numbers of diabetics in a popula- tion, thereby giving a poor yield. 4.2.3 The "border-line11 problem The .. ne.edocfor epidemiological surveys is perhaps best emphasized by referring briefly to ~ne of the important findings resulting from work of this kind. Measurement in probability samples of a population is tending to show that many physiological varia~les are continuously distribute~round the mean, conforming t~ a normal, or skewed normal, curve. Whether there is a separate diseased popula- tion or not cannot always be determined from the data; but in considering the .. separation between "border-lines" and "diseased" this point is not of prime importance as Fig. 4.5 shows: there is in either case an area of doubt •.. Never~ theless, the distribution of blood-pressure, bl~od cholesterol, blood sugar and intra- oc.ular tension, to give a few examples, all appear to f~vour a continuous disi;;ri::- bution. The. "diseased" part of the population occu~s at the extrerpe ~n<i of the ~stributi?n curve and ~his, as can be seen i~Fig. 4.::;, means there may?~ far more "border-lines" in a population than there are "diseased". Fig. 4-5 DISTRIBUTION OF A VARIABLE IN A POPULATION "Normal" (1) Bi~onal Di~tribution: "Diseased" B E I "Normal" (2) Unimodal Distribution: c D "Diseased" WHO 7003 7 - 21 ·- There is, however, theoretically at least, a difference between the outcome of surveys, depending on the ,dis·C.ribution of the variable or variables measured. If the distribution is bimodal (as might be expected in the case of some genetically transmitted characteristics l:i.ke, for example, phenylketonuria, the "border-line" group will in fuct comprise a mixture of persons vJi th the disease and persons ui thout the disease who:~e level of tho variable falls within the same range (between A and · , . B in the upper diagram of Fig. 4 .. 5). On the other hand if the distribution is un,;lmodal the "bcrder-·line" group v:ill comprise a homogeneous sample of persons, the questioi1 being \vhether the point beh1c-en '1dist:3.se" ar..d "norm<:~.li ty" should be set at C or D (lm'le.':' diag,ram), We m<'ty no·t.e, in passinr.;, that the figure illustrates gr13,phi(!ally ,_t:pe concepts of ser:.si tj_vi ty and sp-3cifici ty. A screening test giving a posi tivc r~adirg at the levE:l of A (-.1pper diagrJ:m) or C (lo-vTer diA.gram) would be higl':\ly sensitive" missing few ca::;e.s _, but yj_elding man~r false positives; by contrast, the "cut-off" points ~t B and D respectiVF;ly indicate a very specific "'Jest. In practice it seem:: likely that· a trial by r~andor.1j_zation of treatment should _enable a reasonable decision on the ·" cut-off11 point _to be made bet,-veen those considc:red _in need of treatment ar.d those who n:ay bo :.~eas9~1r·ed us healthy. Some false negatives would need to be accepted in a bimodal example for the sake of the S!Jecificity of the test (e.g, a ttcut-off" at E), whilst in the unimodal example a similar arbitrary choice would need to be made,· based on the response· to treatment of the "border-line" patients. It is pe~haps worth noting here that the terms 11 sensitivity" and "specificity" whilst having a cJ.ear meaning in the case of the bimodal distribution, have theoreti- cally no mea.ning for a unimodal distribution. Thus, supposing a cut-off level at E in Fig. 4.5 (the bimodal distrib"--ltioE) to be used, both true and false positives would occur at that level. -But withthe unimodal distribution, once a cut-off level has been 2doptcd! all per;;;om.> above that (c .• t;. above level D in Fig. 4.5) would be regarded as diseased, with no fz.lse pcsi t::. ves. In pranticE !3Bnsi tivi ty and specif'ici ty remain importax1t because an indirect index of dic:ea:Je .is usually adopted, subject to varj e.tJ.ons between observers end within pati:mt;:; from one occasion to ax.oth0r, a.s well as to the error of method. This index will cnly have a cer·ta~n p(,wcr L"l diagnos:i.nc~ patients and will miss some and falsely :LncLio.e ethers. - 22 - For a fuller discussion of sensitivity and specificity, as well as the reproducibility and accuracy of a test, the United States Public Health Service Monograph "Principles and Procedures in the Evaluation of Screening for Disease11 , 19 may be consulted. 4.3 ~~inciEles of early disease detection The central idea of early disease detection and treatment is essentially simple. However, the path to its successful achievement, {on the one hand bringing to treat- ment those with previously undetected disease and on the other avoiding harm to those persons not in need of treatment) is far from simple though sometimes it may appear disarmingly easy. For this reason we have devoted this section to a reasonably full discussion of a number of points which could be regarded as guides to planning case- finding. This is especially important when case-finding is carried out as a public health agency programme, where the pitfalls 1nay be more numerous than when screening is performed at a personal doctor level. For ease of description rather than from dogma we have called these points collectively "principles". attempt at elaborating at least some of these principles: 4.3.1 The condition sought should be an important problem. The following is an 4.3.2 There should be an accepted treatment for patients with recognized disease. 4.3.3 Facilities for diagnosis and treatment should be available. 4.3.4 There should be a recognizable latent or early syptomatic stage. 4.3.5 There should be a suitable test or examination. 4.3.6 The test shotud be acceptable to the population. 4.3.7 The natural history of the condition, including development from latent to declared disease, should be adequately understood. 4.3.8 There should be an agreed policy on whom to treat as patients. 4.3.9 The cost of the case-finding programme (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole. Case-finding should be a continuing process and not a "one-time" project. It is now necessary to discuss each of these headings in some detail. - 23 - Add. 4.3.i Impo~~~!:!...P~obl~ for the individual and the community This does ·act nece::sarily mean a high degree of prevalence though that would be a usual.requi:rornent. Thus diabetes mellitus is relatively highly prevalent in western populq.t5.onr. thongh fcaquently of :nild degree. On the other hand phenyl- ketonuria is e:?C.tramely u1J.comm0n but warrants screening on account of the very serious consequen-Jes if it is not discovered and treated very early in life. Clearly the iruport:-mce of the probl.em needs to be considered both from the point of view of the individual and of the community. Thus conditions with serious consequences to. tho indiv;_dual and his or her family in general may warrant relatively unec.onomic s~reoning measures; while certain individually mild conditions, but havlng· serio'vls ~onneq_ucncc::.: for the community if not discovered early and treated, will justify screening on these grounds. An example of the last kind might be the finding and control of over'I'Teight in a population. Add. 4.3.2 Accented treatment -----.. -- ---- Of all the c1iteriawhich a screening test should fulfil the ability to treat the qc:mdi tion adequately, vrhen discovered, is perhaps the most important. In keeping to the pr::.nciplo of above all avoiding doingharm to the pati~nt (the 11 primum non nee er(t of Hippo0rates), treatment must be a first aim. For declared disease tne:re. is_, of 'JO'li'Se, the ethical obligation to provide an accep:ted treatment whethE?r .-~his. is of scientific all;\r proved value or not; but, when new terri tory is being explored by the earlier detection of disease, it is clearly vital to determine ' • • I ,;: ·:· '' • - by experimental r.;urveys wh\)the:r: treating the copdi tions found at an earlier stage than was previously the practice gives a better prognosis. Unless this is so there can be no advantage to the patient and in fact, in alerting hirri or her to a condition w)<ich h<"'-::. not been shown to benefit by treatment at an earlier stage, actua.l harm may be done. This matter resolves itself. into two questions: (a) Does treatment at the pre-symptomatic t_~,roer-line stage of a disease affect its course and prognosis? (b) DoE's t:"eatment at an earlier stage than normal of the developed clinical conditj_on affo0t its course and prognosis? - 24 - Question (a) is referred to more fully in paragraph 4.3.7 below, on the need for adequate survey work. This can be tedious and clifficui t work and it may take years before results become available; but without it there is the danger that, ethically, the investigation stage will be passed and the answer never known with any real precision. Thus we are still in ignorance of the effect of treating the I' lower range of high blood-pressure. If the use of drugs for mild hypertension became general it would no longer be ethical to randomize treatment and we should have to rely· on the unsatisfactory long-term evidence of mortality and the age at which mo~bid changes appear. Border-line diabetes mellitus, ocular hypertension and asymptomatic·bacteriuria are other examples of conditions in this class. Until the needed information has been obtained there is, therefore, no case for alerting . "border-line"persons by case-finding programmes. If persons in the border~line range are to be informed of their findings they should presumably be told they are not diseased. Coming to question (b) we enter the field of accepted clinical practice and the course to be followed is largely pre-determined. Thus it is reasonable to seek in a population and treat persons with signs of clinical diabetes (possibly confined to an elevated blood sugar level) even though they may have no symptomS. It is not known whether early treatment in fact alters the out- look, but this is a plausible assumption (although there is evidence that renal and neurological changes may follow a course independent of treatment). The same argument applies to the early treatment of established chronic glaucoma, where the course of the illness is also long and the treatment unpleasant, and where doubt nrust arise as to the proportion of those prescribed treatment who adhere to it. opinion accepts the value of treatment and there is no ethical alternative. There are also conditions where evidence shows that the prognosis is Medical unaffected, or nearly so, by early recognition and treatment. Lung cancer is one example, mass radiography being the standard method of early detection. In at least one survey it has been demonstrated that the prognosis for life in patients detected at an early stage of bronchial carcinoma was little if any better than that of patients detected later. These patients had made their medical contact on account of symptoms (see paragraph 6.6.1). The same kind of findings have resulted fro~ early detection by cytological sputum examination. Whilst there is a good case for continuing efforts to detect lQ~g cancer early in selected groups, like heavy smokers, knowing that a few patients will benefit, it is doubtful whether any - 25 - useful purpose is served by advocating mass case-finding specifically for lung cancer (tho'ugh, of course, mass radiography carried out for other purposes produces a steady run of ·cases). At present it seems likely that greater efforts should be directed towards educating the public about the risks of cigarette smoking and the need· for investigating persistent cough. It is axiomatic, therefore, that case-finding should only be undertaken when the prospects for treating the condition are at least reasonable. Add. 4.3.3 Facilities for diagnosis and treatment should be available Clearly, .in planning to detect some condition, or group of conditions, ina population i,t. is a prerequisite that persons found in need of treatment should be able __ to obtain it, In general, thelarger the scheme the more this proviso.assumes importance. Thus, in introducing on a national scale cytological screening of all women at risk of uterine cancer, a major part of the scheme must consist.· in ensuring that services are available for the definitive diagno::3is and treatment of those found positive on exfoliative cytological examination. Of even greater magnitude is the problem of providing effective treatment and care in a developing country, where medical services may be extremely thinly distributed, for conditions detected by mass screening. question. In this context pulmonary tuberculosis can pose a difficult Add. 4.3.4 .Recognizable latent orearlY symptomatic stage In order usefully to detect and treat disease at an early stage there must clearly be a'reasonable period in the natural history of the development of the 66ndition.duririg.which symptoms are neither present nor at any rate clamant.· There is in fact a latent stage in many chronic diseases whic~ can be recognized, and also in the carrier state of some acute communicable diseases. However, certain chronic diseases like multiple sclerosis and arteriosclerotic cerebrovascular disease, though there must be a precursor stage, do not have a clinically recognizable latent pericd. Rheumatoid disease, for example, though there·is an early symptomatic stage, has no certainly recognizable presymptomatic state •. - 26 - Add. 4.3.5 Suitable test or examination A number of factors have to be considered. Tests can be divided into diagnos:t;ic ,and screening but this is a matter of degree rather than kind; the screening test, (which of its nature should be easy and quick to perform) is .. allowed to possess a higher margin of error, and may be less valid than a diagnostic test. For some conditions which do have a re~ognizable latent stage there is at present no suitable screening test. For example, barium meal examination for carcinoma of the stomach has been tried but found impracticable, on the ground of radiation exposure, ·discomfort to· the subject, 'a:rtd time ··needed.· (Recently, in Japan, intra-gastric photography has been developed as a screening technique and this may provesatisfactory where there is a high incidence of gastric disease, when used selectively). Similarly, addisonian anaemia could probably be detected at a pre- clinical stage if an easy test for parietal cell antibody was available. On the other hand s~me tests are accepted as suitable for screening because of their simplicity and ready application, though they may not in· fact be very good indices of the condition being looked for; for example, tonometry in the detection of chronic ., ' simple glaucoma. It is important to remember that most of the tests we use are indirect indices of the pathological process we are seeking. In general, it is reasonable to suppose, the less direct the index the greater the liability or error in diagnosis. For example, the examination of the haemoglobin, the actual pigment that is decreased in anaemia, is. a h;ighly_ preci:se .mea§we, of ana.e~ia ( regSU'_gless of type), provided the error of the technique of measurement is small. On the other hand the chest X~ray film in chest disease, the glucose level of the blood in diabetes and the intra-oc1,llar tension in glaucoma are all indirect indices of the conditions sought and, in these instances, there is likely to be an error not only of observation but 81so of interpretation. Thus the shadow, the blood level or the pressure may not always provide a valid guide to the clinical condition • . -,,.-:· It is the task of the investigator alway~ to seek more direct and more valid techniqu~s, ~!,thout sacrificing convenien9e and speed. It would be difficult to replace chest X-rays though ultimately development of a re~iable automated cytological technique applied to high risk groups of the population may prove valuable in cancer diagnosis; while there is a possibility that fatty acid or other biological variable levels in the blood might prove a more reliable index of the clinical - 27- state of diabetes t~ is the bl?od sugar. Regarding chronic glaucoma, there is now evidence that intra-ocular tension may not be a reliable index; in a recent survey as many patients with glaucoma were found within the normal range of tension as there were patients with_raised tension. 20 , 21 Finally, there.is the question ?f the validity of the test, indicated by the proportion of those examined found t.o have. falsely positive tests and the proportion found to have false negative tests, i.e. having the condition looked for but giving a negative response to the test. In case-finding work a fairly high false positive rate is acceptable but the false negative rate should be very low since missed cases can lead to individual disasters and, in the case of a communicable disease like tubercuJ.osis, to the undetected spread of disease. Acceptability topopulation Clearly a test or series of tests must be acceptable to the population to which it is offered. Acceptability is, of course, related to the nature of the risk and to the way in which the ground is prepared previously by health education. For example, the risk attending uterine cancer is by now well known to the more educated sections of western society but we are still pretty much in the dark about the attitude towards prophylactic vaginal examination of women in the lower socio-economic groups. Work is being undertaken {see section 5.6.2) on this particular aspect of ·vaginal cytology, both in Britain and the United States of America, but there is scope for much more investigation of this kind. As an example of suiting the test to existing public attitudes, Davis22 reports a higher degree of acceptance (over eighty per cent.) in a Maryland county for the self-taking of exfoliative cytological preparations using the irrigation-pipette23 than for conventional cervical smear- taking methods. This illustrates the importance of paying special attention to making a test as easy and, as little trouble to perform as possible. There are other forms of examination which, the evidence shows,_ could usefully be carried out in the se13-rch for early disease and its preventioi:l, but which are so unpleasant as to be quite unacceptable. One of these examinations is procto- sigmoidoscopy, tl'w usefulness, of which in detecting pre-cancerous conditions of the recto-sigmoid is accepted. The application of this examination is virtually limited to medical c1:i,nics anci period_ic health examination centres. - 28 - Add. 4.3.7 Need for surveys "·,-· ···r-.r· ... ..... .~ .; .w_e ha:ve ~lready, emphasized.· in both. the "Terms of Reference" and the "Defin,itions11 th:e important differE;lnce i~ concept between c~se-finding and epidemiological surveys. The need for.maintai~ing this distinction is not always clear, since the distinction itself tends to become blurred unless we look at the matter historically. We are apt to ass~me that, b~cause it is possible to carry out useful case-finding by screening for ono condition (sub-clinical pulmonary tuberculosis, for example) without the need for preliminary surveys, the same pritic;.; . . .. iple applies to other diseases, such as diabetes or chronic glaucoma. In making this assumption we forget that much survey work has been carried out on pulmonary .. tuberculosis in the past (as well as at present) and that the natural history of the' early stages of the disease has gradually become established over the course of many years of study. However, when we turn our attention to attempting tocoptrpl thE;_. new epidemi(}S of chronic non-communicable disease by similar case-find!?g techniques, we are. likely to run :;into difficulties unless we are first able to view clear.ly the _natural history, and especially the. precursor stages, of these diseases. The· most. important questions that need answering for conditions like high blood- pressure, ischaemia heart disease, diabetes mellitus and chronic simple glaucoma are: What changes should be regarded as pathological and what may be considered as physiological variations? Are·earlypathological changes progressive? Is there an effective treatment which can be shown either to halt or· reverse the early pathological changes? We-ought t,o note that we dO. not kno~ the answer to the last question even for s.ome E;Stablished clinical conditions of which diabete$ mellitus (in the progress of its complications) and chronic simple gl~ucoma are examples. An important reason for this ignoran~e_is that controlled trials of treatment were not carried out at the time when tlus might have been ethical (this applies specially to glaucoma), the controlled trial technique not having be'Em devel~-ped at the material time. The lesson for pre-symptomatic disease detection is·surely clear: controlled trials of e~rly treatm~ntneed to be. carried out as speedily as possible while they are· still considered ethical. Orice.it is regarded as norlllal'p1:'actice to treat latent disease, whether thls;\1.as been' shown to be beneficial or not, the opportunity for ·since· milch •treatment· - 29'- is both Unpleasant and, in chronic ilness, of lifelong duration following diagnosis, it is clearly important not to treat people unnecessarily. In enthusiastically attacking disease at an early stage the Hippocratic principle, previously mentioned, of "primum non nocere" should not. be neglected. Where it has been decided that a survey of the natural history of a condition and controlled trial of treatment is needed, there is a strong argument against trying to combine the survey with case-finding, as defined. For the sake of the clarity of the results it is usually considered proper that during the course of a survey it should clearly be understood by all taking part, public and investigators alike, that the work is experimental and devoted to finding the answers to certain questions as a preliminary to embarking on the next stage, that of case-finding. ·- ~ '. ·: ~·., Naturally, all persons discovered in the course of a survey to be suffering from clinical disease would be referred for treatment. The border-line group alone would be asked to submit to a randomized trial of treatment. It may be asked why surveys and ~ase-finding should not proceed together at one and the same time. Of course, to the. extent that surveys discover patients with undiagnosed clinical disease .<a.s just mentioned above) they do proceed together. But in general the aims are diff~rent and mixing them can lead -to confusion. Where · c~e-finding, with its implication of treatment, is planned it is necessary to be quite clear who is to be advised to undergo treatment and who not. This means that a decision has to be taken about. the criteria which constitute disease. Making this decision begs two vital questions which the survey side of the programme would be trying to answer, namely the question about patho;Logical significance or physiological variation (4.3.7.1), and the question of whether these early patho- logical changes are progress! ve {4. 3. 7. 2) ~- · ··· Orie ·way out ·of this difficulty, when case-finding, is·to set high criteria for the diagnosis of disease in need of treatment. The· classification by diagnostic techniques of the nborder-line" patients, and the arrangement of a controlled randomized clinical trial (which would constitute the survey side of the mixed programme), entails much clinical and administrative work, as well as long-term scientific follow-up. It might well not be justifiable, on grounds .. of cost and staffing, for a self-selected sample, such as normal case-finding methods would produce. - 30 - 24 The recent findings of a diabetes survey by Butterfield et al. emphasize the need for randomized trials of treatment before advocating case-finding on other than the strictest criteria of clinical diagnosis. A blood sugar survey of all per-sons over 21 years of age found that some 16 per cent. have "diabetes" if the commonly accepted criterion of a blood sugar level of more than 120 mg per 100 ml two hours after 50 mg of glucose by mouth is used. If these figures are extrapolated to the age/sex structure of the population of England and Wales as a whole, there would be found (on the above criteria) a prevalence of 11 diabetes11 of approximately 13 per cent. The survey by Butterfield and his colleagues took the form of case-finding (in the first place, by urine testing for glycosuria) and did not comprise a random probability sample of the general adult population (there was a 67 per cent. response). The figures may, therefore, be to some extent biased. However, it seems unlikely that they are far wrong since they are supported by those of the National Health Survey carried out by the National Centre for Health Statistics of the United States Department of Health, Education and Welfare25 which found 15.5 per cent. of persons aged 18 to 79 with a blood sugar level of 160 mg per 100 ml, or over, one hour after 50 g or oral glucose (see section 6.1.3). Whether some 13 to 15 per cent. of the entire adult population of a country of this type is in need of treatment for diabetes, and, if so, the kind of treatment that is required; urgently needs discovering. Since the treatment of 11 border-line:r diabetics (see 6.1) is not at present known to affect prognosis it seems reasonable in case-finding to advise treatment only for frank, or overt, diabetics until the outcome of surveys, at present in progress, is known. (The question of what should be regarded as declared, or established, diabetes is also discussed in section 6.1.3). Add. 4.3.8 Groups to be treated in case-finding programmes The 11 border~line" problem has already been discussed under Section 4.2.3. I.t is important, in designing a case-finding programme as opposed to a scientific survey, to have a clearly defined policy about "border-line" subjects. If it is agreed that only patients with established disease (wherever the demarcation line of the test used may be drawn} shall be treated, the management of the ~;~)order-line" patient largely depends on the design of the case detecting programmes and on communications. When a personal doctor examines a person, whether in the form of a routine medical examination or by using one or more screening tests, there is no real problem; the results are recorded on a case sheet and the patient need only be informed if there is - 31 - defin,itive disease needing treatment. The doctor in fact makes his own personal decision between what amounts to disease and what merits expectant observation, in the light of his personal knowledge of the person in question. With a·community health scheme, hQwever, great care needs to be taken in passing on information without harm resulting to so~e of the persons examined. In any given scheme there may or may not be personal doctors taking part, according to the local system of medical care. In either case, as long as a clear policy has been agreed, no confusion should result. But if persons with doubtful results are referred to their own personal physicians who have not taken part in agreeing the protocol of the case-finding programme, confusion may result; either those considered to be in need of treatment may not be treated or else those with equiviocal changes may be treated, or both. Arrangements are clearly desirable for the follow-up of the 11bord.er-line 11 cases; this might preferably be done by the personal doctor as part of a routine examination, so as to avoid segregating this group into a qpecial population which,however, is by definition, not in need of treatment. Clinics do exist, of course, for the management of persons not suffering from frank disease but for whom it is .. believed that preventive treatment is indicated. For example, there are clinics for. persons with high-risk factors for ischaemic heart disease 20' 27,28 But tiws.3 • instances there is at l~ast a measure of agreement that expeotant treatment is . effective. Apart from this type of clinic, it seems likely that clinics for following up rroorde!'..:.line" cases should be confined to surveys. Add. ~.3.9 Economic balance of the cost of case-finding in relation to total expenditure on medical care It is often considered that the detection of disease by screening will be economical of a country's resources. In order to examine this question it is perhaps __ .W~!'~~ looking at ::>9111~ qf the reas.ons why screening (specifically, as opposed to other methods of early disease detection) is considered in principle worthwhile. There appear to be two main aims, the one medical, the other economic. The Commission on Chronic Illness in the United States of America, for example, states: "multiple screening, by combining several disease-detection tests, is a stream- lined process assuring speed, efficiency and economy. Hul tiple screening contributes II 20 to good medical practice • -' The medical aim, therefore, is to improve the health of a population by the early detection and treatment of illness; while the - 32 - economic alm is to spare the time of highly-trained people by using technicians, and perhaps automated methods, as a first line in disease detection. Mass radiography, for example, both saves the time of the personal doctor by making good use of .his high index of suspicion and eliminating preliminary history-taking examination; and also may help attain the goal of better health by finding (among other conditions) latent pulmonary tuberculosis .• However, this goal can be expensive, as even highly developed countries find and there comes a point of diminishing returns. In the case of mass radiology, for example, once. the backlog of undetected tuberculosis has been worked off, the economics of advocating mass routine screening is open to question. On the other hand the very economy of a developing eountry can be threatened by uncontrolled communicable or parasitic disease and it may be necessary to carry out a mass programme with priority over medical care needs. It would be helpful to compare th~ economics of medical care provided through screening with the results obtained for similar expenditure on conventional medical care. The probability is that, though the time of individual doctors might be saved during diagnosis many more provisional diagnoses would be made through screening, which would add to the total load of diagnostfc work; and also that the total number of definitive diagnoses would be greater, leading in turn to a greater demand for medical care. As an eXample, in the C.C.I. screening survey in Baltimore,3° the rates of new cases discovered. {u~iM _t,h~ given criteria) w.ere, .per 1000. adul.ts, respectively (Fig. 4.6): Condition Cardiovascular disease Abnormal E.C.G. High blood-pressure Raised blood sugar FIG. 4.6 No. per 1000 adults 20 41 37 27 - 33 - Inall, 32 per cent. of those screened were classified as having one or more "major" abnormalities, not necessarily leading to a new diagnosis, ("major" meaning conditions unconnected with height, weight, hearing or vision). Another example is the screening programme carried out on longshoremen in the San Francisco Bay area in 1951, 31 where screening led to over 19 per cent. new diagnoses. When repeated in 1961 there were still 14 per cent. new diagnoses. The commonest conditions found were again, high blood-pressure (5 per cent.), raised blood sugar (4 per cent.), and abnormal E.C.G. (3 per cent.). Fully to diagnose and treat all these patients must add considerably to the total screening cost. Only a prospective survey to determine whether morbidity and working life were improved, in comparison with a non-screened population, could indicate the ' . ,· ·~ . . saving in cost, or otherwise, to the community. The difficulty of carrying out a trial of the kind is, of course, the need to compare like with l:i,.ke; this means that the sample under study needs to be randomized into screened and control groups. In practice this is extremely difficult.tq carry out in one population since those allocated to the control group will be likely to become"infected" by the screened group (there is good evidence that screening is po~ul~f with the general public) and to demand screening tests too •. One brief comparison is perhaps of interest: the Cortland County (New York) sc~eening project 32 in 1961 screened in nir.e ways: self-administered questionary, haematocrit, height and weight,blood sugat', chest X-ray, urinary albumin and sugar, blood-pressure, oral cytology and 12 lead E.C.G. It cost $ 14.55 (or about £ 5.4.0) per head: .whilst the average general practitioner in the United Kingdom is at the moment paid about$ 5.60 (£. 2.0.0.) per patient seen in the course of one year to cover all medical services (amounting to about $ 1.00 (seven shillings) per consultation). Thus, superficially, screening can be a costly rather than an inexpensive method of providing me.di.cal care. However, with the arrival of automation for both carrying out tests and for data processing, the cost may soon be greatly reduced. Add. 4.3.10 Case-finding should be a continuing process Much screening in the past has taken the form of one-time only "drives" or "weeks". Impetus has been put into arranging for a number of persons to be examined once, but the momentum necessary for making full use of the organization - 34 - called into being~ by .continuing examinations, has not been generated. The. "one- time" examination is clearly only of limited value~ since {a) only a small proportion (often those at. least risk) is likely to be examined and (b) the screening picks up those persons in the population who happen at t4at particular time to have the conditions sought; it cannot touch the future incidence of disease at all. Thus continuing examinations have great advantages. An organization can be built up which .can gradually become more efficient and economical~ and which can take its place as an accepted part of the normal Qedical services. Regular offers of examination are likely {with the help of health education) gradually to cover more and more of the population at risk, including by re-examination new cases of disease. 4.3.10.1 Concept of "[,)urveillanc~~ As we have already noted, many of the difficulties stemming from screening examinations are connected with problems of continuity and of communication between those initiating examinations and the doctors who hav~ personal responsibility for patients. Much of this difficulty may be avoided if arrangements can be made for screening to be carried out under the auspices of the personal doctor himself. this type of care would amoQ~t to is the routine examination of patients in What particular high-risk groups for certain conditions, at regular intervals~ as contrasted with the usual present arrangement cf a patient reporting a departure from normal health to his own·doctor. The routine examination can take the form of a·number of preliminary tests, including completion of a questionary, followed by an interview and examination by the doctor in the event of any abnormality being suggested by the tests. This kind of regular, repeated surveillance at first sight suggests the hopeless involvement of the general practitioner in unproductive routine examinations at the expense of hiB important clinical function. This; however~ should not follow apd~ indeed~ were this the only possibility, surveillance of this sort would be quite impracticable. However, in many countries health centres or group practices of one kind or another either exist or are being developed, and it is in this environment that routine surveillance has the most :favourable prospect of developing. Ancil:J.ary workers, both laboratory and secretarial, can carry out and record screening ' :..., examinations, having first selected from patients' records those persons due for - 3~ -· calling up for examination. To do this, a well-organized system of records is clearly of the first importance and, for large numbers, some form of automatic data processing is mandatory in order to sort persons for examination by age, sex, marital <.. status, parity~. etc. . ' Lookj_ng back to the past, it is interesting that where specialized data handling has not been necessary in order to examine high risk groups of the population, as for example, in maternity and child welfare work and school health services where the high-risk groups are self-selected, screening for departures from health developed at a very much earlier date. This kind of surveillance has the great advantage where a family doctor service exists that continuity of examination and care L'0.n be maintaineC!. ( ti1e pe11 sonal medical record following the patient who moves from doctor to doctor); and that only the personal medical adviser is concerned in the interpretation of tests (assisted if ne.ed be by specialist doctors) so that, in case of doubt, he can record for future information but is not put in the position of having to divulge the findings to the patient. It seems likely that the future development of screening may well be along these lines {so lor-g as a personal doctor service is available) with public health services playing an important part through the provision of premises, ancillary services and data-processing f~wili ties. Laboratory services are, of course, an absolute necessity. These may be provided either at a relatively simple level o.n the practice premises or the services of a hospital laboratory may be employed, special arrangementsbeing made for sending specimens and reports to and from the laboratory. 5. PRACTICE 5.1 Automatic data han~Uing Health investigations result in a considerable amount of data. The main problem is to handle all detailed information in a way that offers possibilities both for surveying the results and for picking out details from each individual screening. The difficulties are much the same as in sick care although the need is comparatively greater in screening. In the past the results were, for practical reasons, oft"en not treated very extensively and, for insta11.ce, vast experience has been 11 buried11 in hospital medical records and has not been easily available for such purposes as research. The use of computers has changed the situation rapidly. It is now gener.~lly agreed there should be no large screening projects without automatic data processing. Important trends may be uncovered through statistical analysis, and new knowledge accrued for medical science and medical practice alike. - 36 - 5.1.1 The particular needs of screening Screening is demanding in some respects. The data-handling equipment has to be chosen for the special needs that arise from the special circ~tances; with work in the field under primitive conditions; when several facts have to be collected on different occasions, to be collected at a later stage; or other circ~tances that are seldom encountered in usual data processing. Of the particular demands, it should be noted that the recording of primary screening results m .. 1.st be extremely simple. The method used should be easy to work with, because the heavy load makes it necessary to reduce the manual work as much as possibleo The manipulation must be easy to learn at different educational levels, both saving the time of highly trained personnel and, concurrently, giving reliable results when used by those given ad hoc training. The data-collection method should aim at recording results of screening in a machine-readable way as early as possible. It should nevertheless allow for visual reading and checking at every step of the procedure< A very large capacity is often necessary, demanding high speed and a large- size computer memory. It is sometimes useful to make arrangements for sorting selected records with results suggesting the need of further investigation. 5.L2 Basic copcepts Adapting computer techniques to screening is mainly a problem of selecting the proper methods for the practical work. pulation, and Etoring of data. The basic components are collecting, mani- For screening purposess the collection of data is the most important part of data management. The choice of methods and means must depend on the particular circumstances of the screening programmeo For instance, the need for temporary storage of primary results until further data are available for completion of the information may determine the way to get a s~table set-up. The processing ot d~ can be done in a conventional way. Although conversion between different media may be determined by the special needs of the screening prqcedures chosen for the particular project, actual manipulation of screening data - the data processing - is done by customary computer technique. - 37 - The storage of' data is also concerned tAli th much the same problems as adminis- trative data handling, and is mostly associated with the difficulties of storing large amounts of data and still having the details retrievable within a short enough access time. Technfcal evolution is proceeding rapidly in this field. The necessary volume for a given amount of data is decreasing, as is the cost. Data collection The problems in data handling for screening are concerned with the collection of data, and of obtaining the information in a machine-understandable shape. In general, the con,ventional means of recording information can be used, e.g. punc,h cards, paper tape or magnetic tape. The choice is, however, dependent . on the partic~ar properties of the media. propertie.s will exemplify this. Some elementary notations of the 5.1.3.1 The punch card is easy to handle and can be punched, checked and sorted .by ;simple machines. For screening, highly effe c.ti ve combinations can be used by combining text en clair with the punched ,holes. A definite advantage is that punch cards allow new information to be added whenever needed, for instance, during different steps in a multiphasic health examination. In order to get enough information on every card (which ordinarily means 80 alpha-numerical signs) a reduction of data is often adopted, for instance, by coding classes of values instead of the individual digital values. Sometimes the vulnerability to damage by handling may be a serious drawback to the use of punch cards for screening; under primitive conditions it may be difficult to avoid such damage. There are many ways of changing punch cards to adapt them for special needs' which are o£ interest for screening purposes. For instance, a combination card, the dual card, offers the possibility of having visual information;-·etc ~; connected to the coded punch card. Punching is somewhat tedious work and needs well-trained personnel, and also introduces errors. Duplicate punching is often necessary, because errors in digital information may be misleading and d~gerous. can be done automatically by machine (a "verifier"). The checking of duplicates - 38 - Another technique of special interest in primary data collection for screening is the use of mark-sensing cards. Instead of punching holes, marks are made on special areas by ordinary lead pencils, or by special ferrite pencils. The marks can then be read by machine. The mark-sensing cards have less capacity, since the marks need more space. They are easy to handle, but it is easy to make faulty notations. For certain purposes, the IBM "Port-A-Punch11 is useful. The cards are partly perforated, and a pencil gives a clear hole in an easy way. The error in mark sensing has always to be considered. However, it has been adapted for laboratory work, etc., with highly satisfactory results. 5.1.3.2 Paper tape as a medium is increasing in popularity. However, a special punch is needed, which is inconvenient. One of the main drawbacks is that, of their nature, paper tape stores sequentially arrange information~ Therefore, addition of data (or sorting, etc.) cannot usually be done without retyping. There are also paper tapes with extra space for adding, for example, text en clair and identification. Paper tape is somewhat unwieldy to handle in quantity. It is nevertheless a cheap and convenient way of storing information, especially for later data processing. Special edge-punched cards provide a simple. form of data recording, and are convenient on a small scale .where, for instance, identification, standardized information or procedures are comparatively often needed. The data can easily be extracted with a minimum of errors. Electric typewriters combined with punch and reader for paper tape (Flexo- writer, Teletypewriter, etc.) constitute self-contained equipment. Such equipment can serve .as a terminal, and is handy and not too costly. Simultaneously, it gives text en clair for visual checking of the information stored on the paper tape. When needed, the information can then be sent to a distant computer centre by using a modulating unit and ordinary telephone connexions. 5.1.3.3 Magnetic tape and other magnetic media, such as discs, will presumably dominate the future scene. The advantages are striking, with high capacity and readily available information. On a l~rge scale, they are economical. For health screening, however, the use of these media is of interest only for the processing and storage of information in a data centre where other factors than - 39 - screening.n.eeds will be decisive. . . ToO.a.y, recording on magn,etiq , tape is still. somewha~ complicated, and requires special equipment. For a .considerable time hence, it will be easier to use other media, such as cards and paper tape, rather than store information on magnetic media. 5.1.4 Data proces~:~ing There-is a discernible trend to use large and fast computers centrally located, instead of smaller ones working at a lower speed. By means of a time-sharing technique, a large computer can perform many different operations concurrently, and is more economical. It is true that small desk-size aomputers are now available, which are comparatively inexpensive and not too slow. However, they are used mostly for data collection on line and as "terminals 11 , or for data reduction. They will presumably not replace large computers to any appreciable extent, especially because the development of data-transmission technique has been very rapid. In a recent number of Scientific American, containing several articles of advanced computer technique, Fano & Corbato33 have described - in a popular way - the possibilities offered by the time-sharing technique. 5-1.5 Data storage Generally speaking, extremely large memory capacity can be achieved today only by magnetic tape. This is a comparatively cheap method and not too space- demanding, but it has the disadvantage of a long access time. Promising evolution is occurring with the production of large disc memories with very high capacity and a very short access time. In the future, it can be expected that computer centres will be able to store enormous amounts of data, and also to distribute them automatically to desired terminals. For a considerable time, however, it may be necessary to work both with easily accessible data which are carefully selected, and cheap means (such as magnetic tape) with a comparatively long access time. 5.2 Present screening practice (including multiple screening) 5.2.1 Introduction As we have seen earlier in this paper, screening developed in the control of endemic communicable diseases, and preventing the spread of conditions like malaria and syphilis still depends in part on screening measures. The development of - 40 - interest in screening for chronic non-communicable diseases since the Second World War has resulted, in the United States of America particularly, in a large number of "programmes" where varying combinations of screening tests have been employed. In Table 5.1 we have listed the principal conditions for which screening has been carried out, sub-divided into communicable (infective and parasitic.} .and non- communicable, and categorized by selective screening by age and type of country. Condition Hearing Vision Phenylketonuria Cong. dislocation hip Rheumatic heart dis. I Congenital heart dis. Hernia Overvveight Diabetes mellitus Anaemia Chronic glaucoma High blood-pressure Renal tract disease Ischaemic heart dis. Rheumatic diseases Cancer of lung Cancer of skin Cancer of bladder Cancer of rectum Cancer of mouth Cancer of cervix Cancer of breast Mental illness Malnutrition Pulmonary tuberculosis V .D., syphilis V .D., gonorrhoea - 41 - CONDmONS SCREENED BY . AGE AND TYPE OF COUNTRY .. Infancy X X X X X X X X X ... Selection by age .. ·-· ..... . Child- hood. Adult ······ ... X X X X X X X X X X X X X X X X X X X ·····--·x-····- r ·· X X X X . I Mlddle l Elderl age y .. . j ... X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Type of development of country High /Middle Least X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - 42 - TABLE 5.1. CONDITIONS SCREENED BY AGE AND TYP~ OF COUNTRY, (continued) .. , ... . "' ' ...... <• .. ·~ . . ' ...... ¥ ·-"''' ' . 0 •••e•' '" ,,,., ... Selection by age Type of development of country ··- .. .. Condition ._ .. Child- I ~d~lt " I . . ..Infancy Middle I Elderl High Middle Least hood 1 age : y : .. ! I ... ... .., . I Urinary tract I infection X X X I X Non-spec. lung I i I I disease X X X X I Infective hepatitis X X I X X )( I Histoplasmosis r X X Coccidiomycosis Trachoma . Yaws X Carrier conditions: Streptococcal X X X Dysentery X X X X X Typhoid X X X Diphtheria X X X Poliomyelitis X X X Malaria X i X Filariasis I X I X I I Schistosomiasis I X X I Trypanosomiasis X I X X Ankylostomiasis X X I Kala-azar ! I I X x I i I Hydatid disease ! X I Trichinosis X X Tinea capitis I X I --43 - 5.2.2 Comparison-:of studies It would greatly lengthen this paper without, in our opinion, being corres- pond·rngry---profitabie·;·to--a-ttempt-a··deta:tleci .. i·ev:i.ew-of'-a:li ··t:his ·-w~~"k·:·--:--Th~·-t;oj;i~~i----- and subti'Q..P:i.cal diseases are a specialized field and might well be treated as the sub~ec--e-or-··a.···separate· ·pa:pe:r-;·----- ca.se:..firia1rig for· the chronic non-communicable diseases-(las been reviewed in the past {see below) and the difficulties have thereby become--a:pparent";···- """These--a:r·e 1ri "the main" twofold;. firstly,. the difficulty of evaluatfng the results of these programmes because of the inability, in general, to follow up the result of screening and determine the effect of early treatment on those diagnosed; and secondly, because the methods employed in different programmes are rarely comparableo. A study of multiple screening was made some years ag_o by the American Medical Association5 which illustrates these points · clearly.. As an example of the type of results obtained in Table 5•2, adapted from a paper by Breslow, 7 is reproduced. It must be noted that the criteria used for diagnosing "new cases discovered" may have varied, the inclusion of more or lessborder-line cases depending, perhaps, finally on the judgement of the physician. (In passing, this number of the J. chron. Dis. also contains other articles on screening for asymptomatic disease, including heart disease, cancer, diabetes, anaemia and glaucoma.) 5.2.3 Examples of multiple screening projects At least three highly organized multiple screening programmes have been carried out in the United_States of America; two under the auspices of the Commission on Chronic Illness, an urban one in Baltimore (already referred to)34 and a rural one in Hunterdon County, New Jersey, 35 and one by the Kaiser Foundation, the Longshoremen's Union, and California State Health Department in San Francisco, also referred to above.31,36,37 5.2.3.1 Baltimore Tables 5.3 and 5.4 demonstrate the main findings of the Baltimore screening from which it is seen that the largest number of previously undetected conditions are found by electrocardiography, sphygmomanometry, blood sugar estimation and chest radiology. Some 63 per cent~ of the sample population had some abuor.ma,:!,Jty dis- ····~ -~~;.oo---·,~oo•••--·--··---·~··· .. •-~---· ~·--·oho ___ _,..._, ................ -yOO••"-"'' ~- ,,,,.,, ....... - ,,,,, _ _.,o~oh>,.o•·--···-.. - ... - .. M·~W,o-oo> -~···•••·<"' ..... 0 ...... ,.._ .. covered by screening, of which half were "major" (i.e. unconnected with height and --44 -·· TABLE 5.2. EXAMPLES OF MULTIPHASIC SCREENJ:NG .RE.~'J;'_$ _ ...... IN CALIFORNIA, 1948-1954 . ··Positive Diagnosis New cases Numbe~ of screening test · confirmed discovered Nature of test persons - ' I tested No. Rate per No. Rate per No. Rate per .. j· 1000 1000 1000 Serologic test for i sy~hilis . 3 974 412 104··· 159 40 23 6 21 733 '1 949 90 425 20 1188 9 Chest X-ray 1 755 001 (a) For tuberculosis (a) 55 210 32 18 939 11 NA NA!! I (b) For heart disease (b) 10 899 ·6 I 3 388 2 697 .. 0.4 (c) For lung canc.er (c) 3500 2 339 0.2 339 0.2 3 990 (a) NA NA 29 7 13 3 (b) i 0.8 21 5 i 9 2 3 I (c) . NA NA >" ·-------~-- ···- • 1 0.3 1 0.3 4 167 (a) 157 38 32 8 3 0.7 I (b) 40 10 I NA .NA NA NA --- ....... -·· ··-· ... -·· -- --···-~-~... ·- .. •• >' . - ~ - ... (c) 15 4 4 1 4 1 Blood·sugar 3 124 Sl+ 17 18 6 11 4 14 863 259 17 127 9 73 5 3 966 1561 I 39 56 14 34 9 . 2 162 311 14 20 10 9 4 3 543 531 15 5 1 5 1 871 : 2 856 30 271 9 16 ' 6 I I ! Urine sugar 3 132 161 5 11 4 -~· 1 1991 - ... _., ... 3 987 50 54 14 29 7 . ' .I Electrocardiogram 2 250 I 395 l 175 110 49 NA .NA I 3 984 666 167 301 l .76 182 46 Blood-pressure 3 989 837 210 3691 92 1207 52 !! NA = not available •. - 45 -. ·-~ TABLE 5 .• 2 •.. EXAMPLEs· OF MULTIPHASIC SCR.E:E!NmG REsutl.rS IN CALIFORNIA, 1948-1954 (continued) '· . - -·· -· ,_ -· ... ·- ........... -- ...... ~·-· ~- ... ~- •o 0 M- o> ~· .. Positive Diagnosis New cases Number of' screening test conf'irmed discovered Nature of test persons , ... ---...---·· . ··- ~~ . ._ ... · ·tested ······No. · ... ; .. ..... Rate per No.· Rate per Ncr~ Rate· per 1000 1000 1000 Haemoglobin 3 986 5 1 1 <1 1 <1 Urine albumin 3 988 92 23 35 9 16 4 Vision . :? .972. -< •• .... 9~t .. ... ~38. . 395 99 205 5~. ___ ..., _____ ,., ... ~- ... ·-· , . ., ...... ... .. ... Height and weight Overweight. 3 992 36o 90 241 6o 74 '19 Underweight 2 86o 90 32 90 32 NA~ NA a - NA = not available. ..... - 46 - TABLE 5.3. ABNORMALITIES FOUND IN THE BALTIMORE SCREENING PROJECT, 1957 Number of abnormalities present Total number of persons screened .. ·-··· . . -· .. - No abnormalities Minor abnormalities only One or more major abnormality Occurrence of major abnormalities: One major abnormality Two major abnormalities Three major abnormalities Four major abnormalities Five or more major abnormalities Number - --. 2 024 -- 743 629 652 4.34 140 . - . - 55 17 6 · ------Percentage·· o:f --- Total with Total major abnormality 100.0 - ·•· - -· - . 36.7 - 31.1 - 32.2 100.0 21.4 66.6 6.9 21.5 .... ··~-···~--~- ..... ----····· .... 2.7 8.4 0.8 2.6 0.3 0.9 - 47 - TABLE 5.4. RESULTS OF SCREENING TESTS ON ADULTS · EVAWATED, AND DIAGNOSES BY PHYSICIANS ' Test results Confirmed by diagnosis Number Test of persons Previously Previously (unweighted) Negative Positive Total known to unknown to ·'. patient patient I Relative rates per 1000 persons screened 70 min' chest X-ray: Tuber'culosis 537 984.4 15.6 3.9 0.5 3.4 Cardiovascular r· d,isease · 537 943.6 56.4 .. 32.9 12.9 20.0 " .. Other ' 537 976.0 24~6 .· 5.1 1.3 3.8 EKG .. 571 874.3 '125. 7 70.3 29.7 40.6 Blood-pr~ssure 6o8 887.9 '112:1 •105.5 68.9· 36.6 Blood suga:r .. ' ·. :-·· .. 601 936.9 6::2.:.l~ 39.5 12.4 27.1 ,_,_ Urine sugar 586 995.1 4.9 4.7 4.3 0.4 Urine albumin 572 991.6 8.4 4.7 3-9 0 •. 8 .. ····¥. ···- . ., . - 48 - weight, hearing or vision, as tested). The most frequent disorder previously unknown to the person screened was heart disease (with electrocardiograph changes and raised blood-pressure as the indicators).· A·critiqu~ ana follow U:p of this screening pro~~~ has been carried out by Wylie. 38•39 He found that only 29 per cent. of the invited sample attended for tests. The participants represented that part of the population less likely to be in need of medical care than the non- participants. The five-year mortality of those screened, age for age, was the same or higher than the five-year mortality experience of the non-participants, varying from six per 1000 at risk for those less than 35 years of age to 149 per 1000 at 50 and over. There was no evidence in the secular trend that persons screened experienced a lower than expected mortality in the first years after screening. Moreover, Wylie found that the age-adjusted mortality of those screened with "major" defects was nearly twice as high as that for persons with "minor" defects and over 18 times as high as persons with negative tests. This is, of course, not unexpected and only demonstrates that people who are ill die sooner than those who are not. It does not show whether or not earlier diagnosis improves the prognosis. 5 • 2. 3 • 2 · .. Hunt-ardon County In the Hunterdon County Survey a sample of the population was examined clinically ("Clinical Evaluation Sample") and a second sample was submitted to multiple screening procedures, as was done in the Baltimore Survey. There was, however, a difference from the Baltimore Survey in that part of the 11Clinical Evaluation Sample" was also examined by multiple screening so that it was possible to compare the results of screening with the known prevalence of disease. The findings are interesting, though difficult to interpret because of the clinical "border-line" problem raised by the use of the technique of screening. Table 5.5 (copied from page 273 of the Hunterdon County report) shows the proportion of abnormal findings from screening, discovered in the "Clinical Evaluation Sample". It is noticeable, as with the Baltimore Survey, that cardiovascular abnormalities are the commonest, even though only 40 per cent. of the sample was aged over 44 years. Table 5.6 (Table 11-1 facing page 304 of the report) gives the most complete analysis available of the Hunterdon screening clinical evaluation comparison. - 49 - TABLE 5.5 Screening test VDRL Mazzini-lipoidal antigen test for syphilis X-ray for chest disease other than tuberculosis (70 mm) Chest X-ray for tuberculosis (70 mm) 1 Blood glucose above 130 mg I I Urinalysis for albUmin - positive or doubtful Urinalysis for sugar- positive or doubtful Diastolic blood pressure - level of 100 mm Audiometer X-ray for cardiovascular disease (70 mm) Weight status Far vision Electrocardiogram (12-lead) Systolic blood-pressure - level of 150 mm Haden-Hausser test for haemoglobin ElectroGardiogram (lead 1) Near vision Less than Abnormal findings (weighted percentage of screened persons) 1 1 1 3 3 4 5 6 11 12 18 18 19 19 19 22 29 TA BL E 5. 6. 1. M U I! rif 'L E SC RE EN IN G RE SU LT S IN T HE H UN 'l'E RD O' l ST UD Y A ll er gi c, I Sc re en in g r e s u lt s a c e . In fe ct iv e a n d N eo pl as m s M al ig na nt e n do cr in e sy st em , D ia be te s B lo od a n d bl oo d- ' Pe rn ic 10 U !I Ir on pa ra si te d is ea se s S yp hi li s pu lm . m e ta bo li c a n d O be si ty fo rm in g o rg an s . de fi ci en cy to f in al d ia gn os is w he re of w he re of n e o pl as m n u tr it io n al d is ea se s m e ll it us w he re of a n a e m ia a n a e m ia ' w he re of T es t a n d N o. o f ' 00 2- 13 8 02 4- 02 :9 14 0- 23 9 16 3 24 0- 28 9 26 o 28 7 29 0- 29 9 29 0 29 1 to ta l n u m be r po s. t e st s ' W ei gh t jud ge me nt ' ; O ve rw ei gh t 12 3 1 1 ' l - 10 7 11 96 3 - - U nd er w ei gh t 6 - - - - - - - - i - , , I - N o. o f te st s 59 7 i M M R -t ub er cu lo si s Su sp ic io us 43 8 2 i 2 l 6 l 5 3 - ! - N o. o f te st s 75 B : ' : ' M M R- ch es t di se as es i l Su sp ic io us f in di ng s 17 - - ; 1 1 3 2 1 1 - - N o. o f te st s 67 2 I I M M R -c ar di ov as cu la r ; Su sp ic io us f in ai ng s 15 4 3 l l - 47 13 33 6 1 1 - N o. o f te st s 74 6 I E le ct ro ca ra io gr am A bn or m al /b or de rl in e 25 5 3 l I 2 - 94 23 70 6 1 1 N o. o f te st s 73 7 i ' D ia st ol ic p re ss ur e ; ~ O ve r 10 0 nu n 11 4 1 - - - 61 lO 50 1 - ' - N o. o f te st s 82 4 - - - S ys to li c pr es su re O ve r 15 0 m m ~5 3 4 2 1 - 10 2 26 75 4 1 1 N o. o f te st s 82 4 H ae m og lo bi n ' P os it iv e 98 5 3 - - 6 3 3 8 - 1 N o. o f te st s 8o 8 - - Sy ph il is -M az zi ni Po s. o r do ub tf ul 6 4 4 - - 2 - 2 1 - - ' N o. o f te st s 75 6 . B lo od g lu co se ' ' • \h ov e 13 0 m g/% 40 - - _ 31 20 11 l - - N : •• o f t. es t: • r. c - - - - ' U ri na ly si s- su ga r ' P os . o r ri ou bt fu l 47 2 1 l l' 30 22 6 5 - - N o. o f te st s 82 3 ' - U ri na ly si s- al bu m in Po s. 0 r do ub tf ul )Q - - - - 8 3 5 1 - - ' N o. o f te st s 82 0 ' " I TO TA L CO ND IT IO NS 10 4 4 l DI AG NO S~ ") 20 2 32 16 7 18 1 3 - - - - - - TA 3L E ' • . · i., . '· !U LT I? LE 3 ':i >• :i' :!I IN :} :iE SU TJ TS I t! T HE h l!N TE :'\D ON S TU DY I .3 cr ~' ~n in .z ; c< ?.- ;-J t. c:: :l c .· :: :: ., ·i. :e as es o : tn .-; · . . e u r: 1a t. ic- · Ct t~ e, ., .. -. ~a r· ~ , , , ce~ los cle :·o ·.J c :. e a r· .r 4r '. e ri os cl er oc ic I A cu te I An gi na I Es se n ti al I I O th er c i r c u la to r' y s ys te rr ; 1i .:; ea se a n d . c o ro n a ry . u e n ig n H yp er te ns io n hy pe rt en si ve 7' i.! 1a l ,- ;i a ;:; !l. l. ~-; w he re of fe v er ji. 3e .:l ::: e c o ro n a ry d is ea se n e a r t U ls ea se o c c lu si o n pe ct or lS hy pe rt en si on d is ea se T es t a n d NC >. o f 40 0- <~ iJ ti IJ O Q- 40 1.3 41 (, 42 0 I 42 0. 0 I 42 0. 1 I 42 0. 2 I 44 4 I 44 6 I 44 7 :o ta l n u m be r po s. te st s W ei gh t ju dg em en t O ve cw e1 gn t 12 5 I 10 3 I l I 4 I 19 I 10 I 4 I 2 I 7 I I 23 U nd er w ei gh t G 3 2 N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sl -J :c io us I 43 I 2- ) I I I I 6 I 7 I I 2 I - I 6 N o. o f te st s 7~ 8 M M R -c he st d is ea se s Su sp ic io us f in di ng s I 17 I 16 ! l I 1 I 3 I 2 I I 2 I 2 I I 3 N o. o f te st s 67 2 M M R -c ar di ov as cu 1a r Su sp ic io us f in di ng s I 15 4 I 16 5 I 2 I ' !1 1 I 32 I 28 I 6 I 2 I 6 I I 17 N o. o f te st s 74 6 t-- E le ct ro ca rd io gr am A bn or m al /b or de rl in e 25 5 ~~ 9 l 16 65 30 10 7 10 - 27 N o. o f te st s 73 7 ' D ia st o li c p re ss u re I \.} 1 ,_ . O ve r 10 0 m m 11 1+ b 9 - c 26 11 4 7 5 1 44 N o. o f te s ts 8~ U - · S y st o li c p re ss u re Ov e~ 1~ )0 m m I c: ~3 I c: 9~ I I 13 I t> l I 35 I ll I 10 I 22 I 2 I 57 ~o . o f te s ts 82 4 - - - i-l .a cr r:o gl ob in P os i t..i v e I 9" I )0 I I l I 9 I 10 I - I 1 I 2 I - I 2 N o. 8 f te st ,.: :; . SQ 3 - 3y p hi l i s -r ~a zz in i Po s. o r do ub tf ul I I 4 I I I I I I I 1 N o. o f te st s 75 6 r - Bl] od . . :: lu co se .~h ove i. so m e;, /% I • iO I 3~ I I I 10 I 7 I I ' ! I 1 I I 4 :J o. o f t~ st s '7 50 Ur in al y~ is -s ug ar P o3 . o r do ub tf ul I I l N o. o f te st s 75 0 1- -- U~ in al Y3 is -3 lb um in P o s. o ;· ; J u b tf u l I ~0 I 1..: .:) I 4 I 10 I " I I 3 I I I 6 N o. o f te s~ ,s ? -;:. o TO TA L SO NI JIT IO NS I I l.t 3'~ 2 I . 39 I 9~· I DI A} NO St. 'D 50 I ll t I 12 I . 37 I 2 I 83 TA BL E 5. 6. 3. - O th er O th er Sc re en in g r e s u lt s a c e . r e s pi ra to ry c hr on ic to fi na l di ag no si s d is ea se s in te rs ti ti al w he re of pn eu m on ia T es t a n d N o. o f 51 0- 52 7 52 5 T ot al n u m be r po s. te st s W ei gh t ju dg em en t O ve rw ei gh t 12 3 6 1 U nd er w ei gh t 6 - - N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sp ic io us 43 16 2 N o. c if te st s 75 8 M M R -c he st di se as es Su sp ic io us f in di ng s 17 4 - N o. o f te st s 67 2 M M R -c ar di ov as cu la r Su sp ic io us f in di ng s 15 4 9 1 N o. o f te st s 74 6 E le ct ro ca rd io gr am A bn or m al /b or de rl in e 25 5 19 2 N o. o f te st s 73 7 D ia st ol ic p re ss ur e O ve r 10 0 m m 11 4 iJ 1 N o. o f te st s 82 4 S ys to li c p re ss u re O ve r 1:: ;0 m m 25 3 20 2 N o. o f te st s 82 4 H ae m og lo bi n P os it iv e 98 7 - N o. o f te st s Bo B Sy ph il is -M az zi ni P os . o r do ub tf ul 6 2 - N o. o f te st s 75 6 B lo od g lu co se A bo ve 1 30 m r;/ % 40 2 - N o. o f te st s 75 0 U ri na ly si s- su ga r Po s. o r do ub tf ul 47 3 - N o. o f te st s 75 0 U ri na ly si s- al bu m in Po s. o r do ub tf ul 50 5 - N o. o f te st s 82 0 TO TA L CO ND IT IO NS 45 5 DI AG NO SE D M UL TI PL E SC RE EN IN G RE SU LT S IN T HE H UN TE RD ON S TU DY D is ea se s o f th e H er ni a o f D is ea se s o f th e B ro nc hi ec ta si s Em ph ys em a a bd om in al ge ni to -u ri na ry di ge st iv e c a v it y s ys te m , w he re of sy st em 52 6 52 7. 1 53 0- 58 7 56 o- 56 1 59 0- 63 7 2 2 - - 6 - - - - - 2 8 - - 2 - 3 - 1 - 1 s 1 2 3 1 2 1 2 6 1 3 - - - 5 12 1 l 6 - ~ - 2 1 - l - - 1 1 1 - 1 - 1 2 - 1 - 1 3 - - 4 7 26 2 3 32 P y el it is C y st it is e tc . 6o o. o 6o S l 2 - - - - - - - - - 1 - - - 2 - 1 - 1 - - - - - 2 1 17 C on ge ni ta l m a lf or m at io ns 75 0- 75 9 2 - 1 - 3 4 1 2 1 - - - - 9 Sy m pt om s, s e n il it y a nc i il l- d ef in ed c o n < i.i tio ns 7B o- 79 5 1 - 1 - 2 3 2 1 - - - - 6 7 1. ]1 1 \) Sc re en in g r e s u lt s a c e . ~o fi n al d ia gn os is ' re s t a n d T ot al n u m be r W ei gh t jud ge me nt O ve rw ei gh t U nd er w ei gh t N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sp iC iO U S· N o. o f te st s 75 8 ~B -~ he st d is ea se s Su sp ic io us f in di ng s No • . o f te st s 67 2 M M R -c ar di ov as cu la r S u sp ic io u s fi n d in gs N o. o f te st s 74 6 E le ct ro ca rd io gr am A bn or m al /b or de rl in e N o. o f te st s 73 7 D ia st ol ic p re ss ur e O ve r 10 0 m m N o. o f te st s 82 4 S ys to li c pr es su rP . O ve r Fi O r m n N o. o f te st s 82 4 H ae m og lo bi :o N o. o f po s. t e st s 12 3 6 43 l7 15 4 25 5 11 4 25 3 P os it iv e I c· c N o. o f te st .:: :; ~( )? Sy ph il is ~M az zt Gl . f- -- -- -- -. -- -- -· - B lc cJ t ~li ..t ·': :JS !': P os . o r do .Jb ',! .'u l t N: :~ o f " Se .-s t.s -;· ~~ A. bo ve j _ "' JO . . ~,.; ..'I _ 10 N o. 'J r t. e_ st s 7 · U ri :-. . :1 !. ~/ s t s - 5 ' r .,_ ,r- 1: · P os . o r a o u u tf u l N o. o f te st s 75 0 U ri n al ys is -a lb u m in P os . o r d ou b t: u l N o. o f ~e ~t s 32 0 TO TA L. C ON DI TI ON S DI AG NO SE D 47 50 TA BL E 5. 6. 4. M UL TI PL E SC RE EN IN G RE SU LT S IN T HE H UN TE RD ON S TU DY Sy m pt om s, s e n il it y a n d il l- de fi ne d c o n di ti on s, w he re of 78 o- 79 5 l l 2 3 2 1. . l 0 7 I I I Al bu m in ur ia u n qu al if ie d A lb um in ur ia o r th o st at ic 78 9. 0 78 9. 1 l I l 1 1 l I - l I 4 l if 1 G ly co su ri a U ra em ia 78 9. 6 79 2 l l I - I 1 I 1 I I I ~ 1 1 l l - 54 - 5.2.3.3 San Francisco longshoremen In 1951 nearly 4000 longshoremen (dockers) were examined by a multiple screening t hni . S F . 31 ec que 1n an ranc1sco. In 196o the mortality and morbidity of the original group (of whom the records of over 3000 were available) were followed up and a repeat examination was given a sample of 818 persons. In the 1951 screening (which included hearing and vision tests, 70 mm chest X-ray, E.C.G., blood-pressure, serology for syphilis, haemoglobin, urinary albu~in test, post-prandial glucose tolerance test, measurement of height and weight and a self-administered questionnaire) 63 per cent. of persons examined had positive tests which had led to the finding of 35 per cent. with clinical disease. Of these, over half (19 per cent.) had previously undiagnosed disease; (that is, about one fifth of those examined were found to have disease previously undiagnosed). 5.2.3.4 Chicago Board of Health More recently, the Chicago Board of Health has initiated a Demonstration Chronic Disease Project on three city housing-sites, under the direction of 40 Dr J. Stamler. This project combines a battery of screening tests with a physical examination and it should provide useful information on the contribution to diagnosis and prognosis made by screening as an aid to physical diagnosis. 5.2.3.5 Common conditions for screening In this part of this paper we do not propo~e to discuss in any detail con- ditions for which it is common to screen; some of the more important of these conditions are treated in some detail in section 5.8. It will suffice here simply to note the chronic conditions for which screening is commonly carried out, either singly or as a multiple operation, and to see to what extent they satisfy the main criteria we have noted under 11 Principles 11 • An analy~is is set out in Table 5.7. In deciding in any given instance about the value of screening,local circumstances of course play a large part. Naturally, as indicated in Table 5.1, selective screening by age-groups will in all cases give higher yields and in certain instances (e.g. phenylketonuria and congenital dislocation of the hip) this is mandatory in order to be of use. In individual instances, also, it is possible to avoid some or all of the objections indicated in the Table and it would. of course be wrong to dogmatize. Allowing, however, for the objections to current case-finding, the conditions which score most heavily in favour are seen in Table 5.7. R el at iv e j P re na ta l w o rt h + pr eg na nc y G re at er L es se r To xa em ia Rh f ac to r A na em ia Sy ph il is A sy m pt om at ic ba ct er iu ri as D ia be te s m e ll it us TA BL E 5. 7. ES TI M AT E OF C ON DI TI ON S FO R CA SE -F IN DI NG B Y VA LU E N eo na ta l Lo co m ot or (C on ge ni ta l di sl oc at io n o f hi p) In bo rn e rr o rs o f m e ta bo lis m (p he ny l- ke to nu ri a) C on ge ni ta l de fe ct s (h ea rt) In fa nc y A na em ia V is io n (a mb ly op ia) H ea ri ng (c on ge ni ta l de af ne ss ) C hi ld ho od M en ta l de ve lo pm en t C on ge ni ta l he ar t di se as e A du lt Sy ph il is - - - - - - - G on or rh oe a - - - - - C an ce rs - u te ru s - bl ad de r - s ki n - m o u th Pu lm on ar y O ld a ge A na em ia s - - - - - - - - - - - - - - - - V is io n (C at ar ac t - s e n il e m a c u la r de ge ne ra - ti on ) H ea ri ng - - - - - - - - - - - - - - - - - (O tit is - c o n du ct io n de af ne ss ) Lo co m ot or (a rt hr iti s) H er ni a - - - - - - - - - - - - tu be rc ul os is - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - O ve rw ei gh t - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C an ce r - br ea st D ia be te s m e ll it us (s tr ic t c r it er ia ) Is ch ae m ic h ea rt di Se as e (s tr ic t c r it er ia V l V l R el at iv e w o rt h L es se r TA BL E 5. 7. ES TI M AT E OF C ON DI TI ON S FO R CA SE -F IN DI NG B Y VA LU E (c on tin ue d) P re na ta l N eo na ta l In fa nc y C hi ld ho od A du lt O ld a ge + pr eg na nc y H ig h bl oo d- pr es su re C hr on ic s im pl e (s tr ic t c r it er ia ) gl au co m a M en ta l il ln es s ( Se le ct iv e - - - - - - - - - - - - - s c re e n in g - - - - - - - - - - - - - - - - o n ly ) ~ - - - - - - - - - - - - - - ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - V1 0 \ - 57 - This is not a very long list and it will at once be noticed that certain conditions, glaucoma for instance, have been given a low priority despite the fact that they may constitute more of a public health problem than some of the conditions listed. This kind of variation from the more usual emphasis placed on certain conditions is explained by selection in the above list on grounds of all-round feasibility. Tne pros and cons for screening for some of the major conditions listed in Table 5.8 are discussed subject by subject in Chapter 6. 5.3 Epidemiological studies Having stated that much of the practice of case-finding poses problems which have not been solved (and this can be discerned from a study of the case-finding project mentioned above) it is, we believe, worth while at this point considering briefly what is being done to remedy this situation. Finding answers to the questions of the value of early treatment and the policy to be adopted over the border-linepatientinvolves, as we have stressed earlier in this paper, epidemio- logical surveys (Evaluation of screening procedures, 4.2.2). We have attempted to set out in the accompanying Tables (Tables 5.9.1 to 5.9.5) in the form of examples only, some of the work at present in progress in different fields. Under "Comments" we have tried to indicate where there are gaps in our knowledge which might be filled by further studies. Much of this work is still in progress and at the moment unpublished. It is more than possible that the choice of examples may appear unbalanced and, if so, this can be attributed to the fact that we have chosen them from our own knowledge only and have not sought to be exhaustive. C on di ti on s fo r w hi ch e a rl y P ub li c he al th d et ec ti o n i s e m pl oy ed pr ob le m D ia be te s m e ll it us + H ea rt d is ea se : Is ch ae m ia + R he um at ic + C on ge ni ta l ? H ig h bl oo d- pr es su re + O ve rw ei gh t + Lu ng d is ea se : tu be rc ul os is + n o n -s pe ci fi c + c a n c e r + R en al d is ea se : n e ph ri ti s ? ba ct er iu ri a + A na em ia (ir on d ef ic ie nc y} + A rt h ri ti s: rh eu m at oi d + go ut + B re as t c a n c e r + U te ri ne c a n c e r + R ec ta l c a n c e r + O ra l c a n c e r + B la dd .e r c a n c e r + Sk in c a n c e r + H ea ri ng ( in he rit ed a n • + a c qu ir ed d ea fn es s} V is io n: gl au co m a + c a ta ra c t a n d s e n il e + m a c u la r de ge ne ra ti on H er ni a + C on g. d is lo ca ti on h ip ? V ar ic os e v e in s ? P he ny lk et on ur ia ? V en er ea l di se as es : s yp hi li s + go no rr ho ea + M en ta l il ln es s + ! N .A . • n o t a pp li ca bl e. TP .B LE 5. 2. C: O l'D I'" I·. JN S SC ?c EE NE D BY A B IL I'r v TO M EI •;T SC :l. EE N IN G C R IT ER IA N ot ur al h is to ry o f pr e- ·· R eq :o gn iz ab le S ui ta bl e te st T es t· a c e e pt ab le c u r s o r ;; ta ge d .e li ne at ed la te n t s ta ge a v a il ab le to p ub li c . . ? '· •' + + + ? + + + ? + + + - + + + ? + + + ? + + + + + + + ? ? + + ? ? + + ? + · + + ? + + + + + + + - + - N .A .! ? + + + . , - + + + ? + + ? ? + + - + + + + + + + + ? + + + ? + + + ? ? + + - - + + - - + + - - + + . . - - + + + + + + - + + + - - + + - ' - N .A . A cc ep te d tr ea tm en t a v a il ab le + ? + + ? + + ? + + + + ? + + + + + + + + + + + + + \· + + + ? A pp ro ve d po li cy o n l at en t . s ta ge - - . . . + + - + + ' + + ? + - + + + + + + + ? - N .A . N .A . N .A . N .A . + + - 'N .A , - - - · . • I \1 1· . 00 I ' TA BL E 5. 9. 1. EX AN PL E.. S OF ' CU R." 'l..E NT S TU DI ES O N Tl li: E PI DE M IO LO GY O F CH RO NI C D IS EA SE S C on di ti on D ia be te s M el li tu s Is ch ae m ic H ea rt D is ea se L - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - - 1 W or k in P ro gr es s 1 B os to n U .S .P ,H .S . St ud y o f M at er na l P re -D ia be te s. 2 B ed fo rd S tu dy o f B or de rl in e D ia be ti cs . 3 I C ol le ge o f G .P s. S tu dy o f B or de rl in e D ia be ti cs . 4 I B .n .A . R an do m iz ed t ri a l o f tr ea tm en t in P ot en ti al a n d s u b- cl in ic al d ia be ti cs . 1 2 3 U .S .P .H .S . Fr am in gh am S ur ve y o f I. H .D . in a de fi ne d po pu la ti on . L .S .H . + T. M . S tu di es o f I. H .D . qu es ti on na ir e. Te cu m se h St ud y o f e pi de m io lo gy o f I. H .D . in a de fi ne d po pu la ti on . 4 I W .H .O . C o- op er ac iv e S tu di es . 5 I N um er ou s s tu di es o f bl oo d li p id l ev el s in po pu la ti on s; c o m pa ri so ns b et w ee n po pu la ti on s. 6 I W or k o n a u to -a n al ys is o f E .C .G . w it h in cr ea se in d ia gn os ti c po w er . 7 I W or k o n e m o ti on al f ac to rs i n I. H .D . 3 I U .S . C o- op er at iv e D ie ta ry T ri al . 9 I M .R .C . T ri al o f P ol yu ns at ur at ed F at s in I .H .D . P at ie nt s. 10 I E di nb ur gh T ri al s o f S te ro id s in I .H .n . Co m m en ts C on ti nu e s e a rc h fo r o th er d is e~ se in di ce s. F ur th er t ri a ls o f ra pi d B .S . s c re e n in g te ch ni qu e n e e de d. W or k is d es ir ed o n r e li ab le s c re e n in g te st s fo r in di vi du al , r a th er t ha n gr ou p, d ia gn os is . M or e c o n tr o ll ed w o rk i s n e e de d o n pe rs on al it y a n d e m o ti on al f ac to rs in I .H .D . M or e w o rk w o u ld b e u s e fu l o n fa ct or s a c u te ly i nf lu en ci ng b lo od l ip id s a n d c lo tt in g t im e. 11 [ C hi ca go B oa rd o f H ea lt h St ud y o f D ie t in p re ve nt io n o f - - - _j__ ~·~~ D-. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + -- - ~ ' :'A BL E 5. 9. 2. - - - ~~,:~ i_';; n = = = t-+= --== : =:~ kin !_" '_" '_' "-- --- --- --- --- --~ --- E ss en ti al H yp er te ns io n l I M .R .C . St ud y o f pr op os it i a n d r e la ti v es i n S • . W al es A r a n do m iz ed t ri a l o f tr ea tm en t fo r Co m m en ts C hr on ic B ro nc hi ti s C an ce r o f Lu ng 1 e a rl y e s s e n ti al h yp er te ns io n ha s n o t I s o fa r be en c a r r ie d o u t. 2 Fr am in gh am S ur ve y 3 Te cu m se h P ro je ct 4 L .S .H . a n d I'. H • . St ud y o f Sp hy gm om an om et ry . l 4 5 l P ro sp ec ti ve s u rv e y o f c li n ic al w o rk er s a n d m en w o rk in g in t he e n gi ne er in g in du st ry ; F le tc he r, L on do n. P ro sp ec ti ve s u rv e y o f in fa nt s a n d yo un g c hi ld re n in Lo nd on - H ol la nd . P ro sp ec ti ve S ur ve y o f s c ho ol ch il dr en i n K en t - H ol la nd . St ud y in S te el w o rk er s, S . W al es - Lo w e. P ro sp ec ti ve S tu dy o f M ed ic al S tu de nt s a t S t. 3 ar th ol om ew 's H os pi ta l fo r e a rl y c o r r e la te s o f c r1 ro ni c br on ch it is - M .n .C . P hi la de lp hi a C he st C li ni c - B ou co t + W ei ss . 2 B al ti m or e St ud y - L il ie nf el d. 3 A lb an y St ud y - Ro m e + v o yl e. S pe ci al s tu dy i n pr og re ss o f s e n s it iv it y o f to ba cc o. T ri al o f a n ti b io ti cs r e a dy fo r pu bl ic at io n. S tu di es a re be in g s ta rt e d o f tr ia ls o f in te ns iv e a n ti -s m ok in g pr op ag an da i n hi gh -r is k gr ou ps i d en ti fi ed i n t he po pu la ti on . W or k is i n pr og re ss a im ed a t a gr ee m en t in te rn at io na ll y o n de fi ni ti on s w it h th e o bj ec t o f m a ki ng v it al s ta ti st ic s a n d re s e a rc h c o m pa ra bl e be tw ee n c o u n tr ie s, e .g . W .H .O . Sy m po si um R ep or t EU RO - 21 2. M or e s tu di es o f m o ti va ti on i n c ig ar et te sm o ki ng a re n e e de d. - - - · - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ J 0 \ 0 TA BL E 5. 9. 3. r - - - - - C o n d it io n _ _ _ _ - T -- -- -- -- ~a rk i n ;;o ~·e ·s- ~-- - - - - - - . - J . - - - - Co m m en ts f - - - - - - - - C an ce r o f U te ru s C an ce r o f B re as t - - - - - - 1 - - - 1 M em ph is, T en ne ss ee S ur ve y. 2 Sa n D ie go S ur ve y. 3 I S t. L ou is S ur ve y. 4 B ri ti sh C ol um bi a Su rv ey . 5 C ar di ff , s. W al es S ur ve y. 6 I A be rd ee n Su rv ey . 7 M an ch es te r - St ud y o f pu bl ic a n d pr of es si on al a tt it u de s. 8 I C op en ha ge n, B al ti m or e, S to ck ho lm , Lo nd on , M an ch es te r, B irm in gh am - s tu dy o f D av is C yt op ip et te . 9 I C ar di ff a n d Lo nd on - St ud y o f en zy m e te s t fo r c yt ol og ic al d ia gn os is . 10 P hi la de lp hi a St ud y o f c o m pu te r s c a n n in g fo r c yt od ia gn os is . 11 I R .C .O .G . St ud y o f pr og no si s in r e la ti on t o b io ps y di ag no si s a n d tr ea tm en t o f c a n c e r o f c e rv ix . 1 2 3 4 St ud y o f X -r ay M am m og rA ph y in p op ul at io n s c re e n in g in N ew Y or k S ta te . P ro sp ec ti ve S tu dy o f e n do cr in e s ta tu s in r e la ti o n t o de ve lo pm en t o f br ea st c a n c e r - Je rs ey - G uy 's H os pi ta l. St ud y a t A lb er t E in st ei n C en tr e in c li n ic al M am m og ra ph y - Eg an . St ud y a t M .D . A nd er so n H os pi ta l, H ou st on , T ex as i n c li n ic al m am m o gr ap hy - v a ri ou s o th er c li n ic al s tu di es . T he re i s ro o m fo r m o re in te rn at io na l s tu di es o f m o rb id it y a n d m o rt al it y in r e la ti o n t o t he u ti li za ti o n o f e x fo li at iv e c yt ol og ic al s e rv ic es . Th e pr ob le m s o f th e u s e o f e x fo li at iv e c yt ol og y by t he p ub li c c o u ld b e s tu di ed m o re in te ns iv el y. D ev el op m en t o f di ag no st ic s c a n n in g by c o m pu te r is b ei ng u n de rt ak en . To d et er m in e it s v a lu e, p op ul at io n s c re e n in g by X -r ay m am m o gr ap hy o u gh t to be d on e w it h ra n do m iz at io n, s o th at o n e gr ou p a re tr e a te d a t e a rl y di ag no si s, a n d o n e tr e a te d a t n o rm a l c li n ic al d ia gn os is , w it h hi st ol og ic al c o m pa ri so n a n d m o rb id it y- m or ta li ty s u rv e y. M at er ia l sh ou ld b e a v a il ab le f or c o -o pe ra ti ve s tu di es o f s u rv iv al f ro m e a rl y de te ct io n c li n ic s, e .g . U ni ve rs it y o f M in ne so ta C an ce r D et ec ti on C en tr e. I 0 \ . . . . . L C on di ti on B ac te ri ur ia s R he um at oi d D is ea se (in cl ud in g o th er a r th ri ti d es ) M en ta l Il ln es s 1 2 3 )t 5 1 2 -,_ / TA ..S LE C ,.C 1. 4. W or k ir ' P ro gr es s B os to n - K as s. Ja m ai ca - M ia ll a n d K as s. R ho nd da , S. W al es - C oc hr an e, M ia ll a n d K as s. Ed gw ar e - B ru m fi tt a n d M on d. C ha ri ng C ro ss H os pi ta l, D e W ar de ne r. B ri ti sh E m pi re r u 1e ur na tis m C ou nc il S ur ve y - D r L aw re nc e. Te cu m se h Su rv ey . U ni ve rs it y o f P io ts bu rg h- Si dn ey C ob b. 4 I U .S . N at io na l H ea lt h Su rv ey . 5 I O th er P re va le nc e S tu di es i n Sc an di na vi a, B ra zi l, N ew Z ea la nd . r_: .; 1 2 S tu di es o f rh eu .m at oL i fa ct or i n v a ri ou s po pu la ti on s - V al ke nb er g, L Ed de n. In st it u ce o f P sy ch ia tr y St ud y o f P re va le nc e o f R ep or te d M en ca l Il ln es s in G en er al P ra ct ic e. Sh ep he rd e t a l. (B .M .J. 1 96 4, 2 , 13 59 ) M .R .C . S oc ia l P sy ch ia tr y U ni t St ud y o f M en ta l Il ln es s in C ar nb er w el l. M .R .C . S oc ia l P sy ch ia tr y U ni t P re va le nc e o f ps yc hi at ri c sy m pt om s in r e la ti o n t o s o c ia l a tt it u d es i n S. W al es r u r a l po pu la ti on ( Ra wn sle y a n d c o ll ea gu es ). Co m m en ts D es pi te a ll w o rk d on e, th er e is s ti ll a gr ea t n e e d o f s tu di es t o d et er m in e th e im po rt an ce o f ba ct er iu ri as ( as ym pt om at ic ) in t he a e ri ol og y o f py el on ep hr it is b y da nd om iz at io n o f tr ea tm en t. P ro sp ec ti ve s u rv e ys t o d et er m in e pr og no st ic e ff ec t o f tr ea tm en t in l on g- te rm a re la ck in g. Th e po ss ib le r e la ti on sh ip o f v e s ic ou re te ra l r e fl ux to i nf ec ti on a n d o n s e t o f py el on ep hr it ic c ha ng es , is n o t s e tt le d . F ur th er s u rv e y w o rk i s to b e e x pe ct ed o n th e li n es s u gg es te d a t C .I .O .M .S . Sy m po siu m i n R om e, 19 61 , a n d by T ec hn ic al C on fe re nc e o n th e P ub li c H ea lt h A sp ec ts o f C hr on ic R he um at oi d A rt h ri ti s a n d R el at ed D is ea se s, h el d in R om e in 1 96 3. l'h er e is r o o m fo r s tu di es o f u n re po rt ed m e n ta l il ln es s in a c o m m u n ity w it h ra n do m iz at io n o f tr ea tm en t. R;' C on di ti on A na em ia (ir on d ef ic ie nc y) 1 2 3 4 C hr on ic G la uc om a 1 2 3 4 - - ~ - - - - - - - ~ ~ - ~ - - - TA BL E 5. 9. 5. W or k in P ro gr es s M .R .C . Pr ev al en ce S tu dy . M .R .C . T ri al o f H ae m og lo bi no m et ry i n G en er al P ra ct ic e. M .R .C . T ri al o f P re ve nt iv e T re at m en t in A do le sc en t G ir ls . M .R .C . T ri al o f P re ve nt io n in A du lt s by a dd it io n o f ir on t o b re ad . M .R .C . Su rv ey o f in tr a- oc ul ar t en si on a n d gl au co m a. B ed fo rd S ur ve y o f in tr a- oc ul ar t en si on a n d gl au co m a. M .R .C . St ud y o f pa ti en ts ' at ti tu de s~ t o M ed ic at io n. R an do m iz ed t ri a l o f tr ea tm en t o f o c u la r hy pe rt en si ve s is i n pr og re ss - M .R .C . Co m m en ts E xt en si on o f ~i el d su rv e ys i s v e ry m u ch d ep en de nt o n th e de ve lo pm en ts o f s u it ab le h ae m at ol og ic al t ec hn iq ue s fo r u se in t he f ie ld . T he re a re c o n tr ad ic ti ng v ie w s o n th e v a lu e o f a dd it io n o f ir on ( an d fo r in st an ce v it am in s) t o f oo d fo r pr ev en ti on o f ir on d ef ic ie nc y a n a e m ia . Fi nd in g a be tt er i nd ex t o e a rl y c hr on ic gl au co m a th an t on om et ry w o u ld fa ci li ta te c a s e -f in di ng a s c o n tr as te d w it h s u rv e ys o f in tr a- oc ul ar t en si on . T he re i s a la ck o f m a te ri al o n ra n do m iz ed t ri a l o f m e di ca l tr ea tm en t o f c hr on ic g la uc om a a n d o f o c u la r hy pe rt en si ve s. 0 \ V I - 64 - Much of the work referred to here has been mentioned in the text7 where references have been cited. For permission to cite the work, not referred to elsewhere in the text and not yet published~ we acknowledge our gratitude to the authors. 5.4 Periodic ""ealth examinations 5.4.1 Introduction It will not have escaped notice that some of the conditions mentioned under the heading of early detection are detectable not by means of specific screening tests but by physical examination. Rectal cancer and, to a lesser extent, uterine cancer fall into this category9 so does examination for hernia and cateract. Other examinations, like blood-pressure and electrocardiography7 may or may not form part of a physical examination. The idea of the periodic health examination goes back a long way. Dobell published a monograph advocating "periodical examinations" in 1861. In 1925 the American Medical Association published a manual for physicians. The Commission on Chronic Illness41 was in favour of "all persons having a careful health examination including selected laboratory tests at appropriate intervals". The Commission advocated screening tests only as a substitute to personal medical examination 7 recognizing that shortage of medical manpower made universal routine medical examinations impossible. However, the difference between these two types of examination is quantitative rather than qualitative. Under periodic health examination the responder attends the doctor who examines him, determines \!hat (if any) laboratory tests are needed, and arranges for a second interview end re-examination if necessary, i.e. there are two physician interviews and one set of tests. With screening the responder undergoes a set of tests 7 then sees a doctor (only if necessary, the preliminary sorting of abnormal from normal having been already carried out by the screening tests)3 the doctor may then order more laboratory tests after examining the responder ~d the doctor sees him once again. Under periodic health examination the doctor sees all responders and himself acts as a selective screening agent ordering what laboratory tests he deems necessary. Finally, he sees the selected responders yet again, with the results of the laboratory tests in front of him. This can be represented in model form (Figure 5.1.). The advantage of the periodic health examination is, of course, the introduction of the trained analysing brain of the doctor into each examination, which is most probably a more efficient selector mechanism than the selection for further examination Fig. 5-l HEALTH EXAMINATION D • Doctor L • Labora- tory HE+ • Positive Examination HE- • Negative Examination NUMERALS • Number ot physician Contacts SCREENING EXAMINATION S + • Positive ~crecning S - • Negative Screenin~ WHO 70038 by a battery of tests, rigidly (however ingeniously) for~.~chosen on general probability grounds. 'l:'he: c;l;i~advantagf? is equally clearly the impracticability of routine physical examina.:tions for all. There is a second disadvantage as well7 the blunting of clinical acumen that. may occur in certain circumstances of repetitive physical examination, particularly where the yield of abnormal results is expected to be low. School physical examination,.without selection, is an example. On the other hand, the yield of abnormalities reported from the examination of middle-aged male business executives is .· . . 4"' . .. high (e.g. ~anco S.C. et al., ~; see also Periodic Health Examinations- Abstracts ... . . . . 43 from the Literature).· The subject of periodic health examination has been dealt within some detail in the "Prevention of Chronic Illness" 44 and it is not intended to deal with the , matter in any way exhaustively in this paper. However, it is worth noting two differing aspects of this kind of examination in relation to medical care. Relationship to general practice The first aspect of interest is in relation to the development of general practice. Where general practice has become relatively highly oli'ganized, whether by concentration on health centres or into group practices, or by voluntary limitation of the number of patients cared for (facilit~ted by a high,per capita firi~ncia~ reward), there has been a tendency towards the elaboration of routine medical exam;ination. Apart from actual physical examination, there may be a full examination of the blood, X-ray of the chest, E.C.G., procto-sigmoidoscopy, as well as testing of the urine and examination of the faeces for occult blood. There is special sco.P~ in tpis kind of environment for the dev~lopment of relatively sophisticated tes~s. The various biochemical estimations possible with the Gray Wedge photometer are a case in point: the recent development of an electronic apparatus for recording the knee jerk (as an index of thyroid function) is another. While information is available on the number and type of abnormalities revealed by this kind of search for early disease we do not have comparisons with the results of finding and treating disease by more conventional methods. Perhaps all that need be said is that it is clearly better to diagnose developed disease at the earliest possible stage and that the only comparison ethically:feasible would be the proportion ()f patients under treatment, by condition, in otherwise comparable general practices. - 66 - The second aspect of interest is the commercial or industrial periodic health examination. In this context we are considering examinations .for the general health of the worker and not statutory examinations for industrial hazards such as silicosis, lead poisoning, industrial X-ray exposure or extensive noise. General medical examinations have, of course, been demanded for a very long time by bodies which set their own terms of entry, for example insurance companies and the armed forces and railways and air lines. In this case there is either a commercial interent in the medical findings, based on actuarial calculations, or an interest in the public or other safety, or both. Thus it is perfectly legitimate, and desirable, that a man seelcing tn become an air pilot should be refUsed on the grounds of equivocal abnormali- ties in his E.C.G., though these changes would not be considered adequate causes on which to base a prognosis in civilian life. Again, if changes develop and are found at routine examination of a trained pilot this may well be a reason for taking him off flying duties. In industry, too, it is necessary that specific medical examinations be made on persons at special risk. For example, the periodic routine cytological screening of the urine in persons who have in the past been exposed to beta-naphtylamine is obviously needed? and other industrial examples can be thought of. It is when, as often happens, industrial firms offer medical examination as a "fringe benefit 11 to their employees, that doubts must arise. The legal term cui bono is applicable in these instances and it is well to be clear that the employee who submits to the examination derives at least as much benefit from it as does the employee. That both sides can banefit from periodic staff medical examination is probably true; it enables persons with early disability sometimes to be found more suitable occupations but there can also be the risk that employees may incur a loss of earnings and responsibility because of a risk that never materializes. It is salutary to remember that the Olympic 100 metre record was once held by a man with an aortic regurgitation that led to his rejection for military service. In some countries, periodic health examinations are cnrried out on all employees every year. In other countries, such examinations are carried out only on selected groups, such as those exposed to occupational hazards, groups with special work demands, such as crane operators, truck drivers, foremen and executives; or groups with higher .morbidity, such as middle aged and old personnel (above 45) and young workers (below 18); or thJse with high, long-term and short-term sickness absences; or those who, according to their foremen, etc., deviate from normal behaviour, absence record or productivity. - 67 - While medical care arrangements v~ry greatly from co1:1:~try.to country there can be special advantages in industrial health examinations. The industrial pbysician is often well placed to follow the same individuals for a long time and thus can detect early deviations from health because he is able to compare the results of the physicial examina- tion and screening with the findings of pre-employment health examinations. Industrial health services offer certain particular advantages for studying early disease detection in that it is possible to ensure the regular attendance of a relatively large population and that continuity of attendance is usually of a high order. The interest and full support of both management and employees has been achieved in several countries through national or local agreements on industrial health services or through legislation. The ILO Recommendation on this subject (No. 1121 1959) must also be mentioned in this connexion. 5-4-4 Routine examinations through life We should not forget~ of course, that there is in most developed countries already a well-established pattern of routine medical exe~inations 1 extending through life. If these are wisely spaced the individual can probably be given the maximum protection with the minimum interference vri th the normal pattern of his life. Starting with ante and post~natal and infant welfare examinations the child is nextexa~ined at or soon after school entry and again before leevJ.ng s2hool. After school there are pre- end youth employroent examinations as vrell as examinations before entering certain specified careers. In industry it would seem reasonable to ask for medical examination at key points in a person's career as part of the selection for prom~tion, considering the high· degree of investment in the individual necessary for senior appointments. However, the results of these examinations need interpreting with latitude~ taking into full account all that has been said aboutborder-lineconditions and giving the individual the benefit of any doubt there may be in a particular case. Finally, as vJilliamson, 45 among others, has shown there is a fruitful harvest of remediable defects to be found in the elderly and aged by routine physical examination. Defects of the special senses and of locomotion constitute some of the greatest handicaps of old age and these tend, in the elderly, to fall short of a level of clinical urgency necessary to compel the patient to seek medical advice. There seems to be considerable scope for clinics for the elderly where special attention would be paid to their medical needs. Certainly some of these clinics appear to be filling a real need. - 68 - 5. 5 Tne pla_c:;c' _ _?_f__:o;creenil'}0 j_n ti1e :')Povision of medical care 5. 5.1 ~ust_:1f_i.c_a.t),.:.C'!t witi-.in a country At the start of this paper we gave some space to a consideration of the aims of the early detection of illness. These aims need considering along i-Jith the policy aims which are central to any system for providing medical care. With greater prosperity more money will become available for personal health services. This will mean that greater efforts can be made to extend the disease-free period of life by all possible means and this will include the detection and correction of early departures from normal health: while poorer countries must needs manage with the minimal medical service necessary to prevent the major epidemic disasters and to maintain the working population in as fit a state as possible. The specific aims of early disease detection are firstly, the control of communicable diseases like pulmonary tuberculosis or bilharziasis; and secondly, the promotion of better health and less human suffering as the result of disease. Both aims are governed by the economic nexus. In the first case, it may well be in the national interest to control a disease or group of diseases and a national campaign may be mounted. In the second case, the advance to better health may be marginal only and may be left to local effort where that is forthcoming. Ideally, and in areas where the economic aspect is all-important, the cost to the community of early detection and treatment of e. condition should be less than the cost of treatment at a later stage. For example, it has been calculated that the cost of diagnosing and treating one patient with pre-symptomatic pulmonary tuberculosis costs about~ 400 a sum which, in relation to the cost of treating a developed case 5 represents a worthwhile saving even without taking into account the probable saving of a human family from anxiety and suffering. On the other hand, serological screening of certain communities for venereal disease, which used to be considered economically worthwhile, may no longer be so considered. Similarly, the returns on screening patients for lung cancer are so poor (vide 6.6.1.) costly as they are, that this procedure is probably neither acceptable on economic nor medical grounds. In working out cost a large number of factors may need to be taken into account. One important factor that comes into the calculation is the possible saving of the time of highly-trained people, representing the cost in time and money of their education and experience. If a screening procedure can, by automation or by employing e. less highly trained person, act as a substitute for some of the time of a highly-skilled person, there will then be a saving. However, it must be remembered that the overall cost to the - 69 - community is ustially greater than before because something is being done 1.,rhich was not being done previously. For instance, the detection of unascertained disability among the elderly and its treatment and rehabilitation undoubtedly costs the community extra even though the ascertainment may be carried out by questionary or ancillary workers rather than doctors. An example in which the overall cost may be lowered can be found where the treatment of a condition is in any case essential. Thus the detection and treatment of carcinoma-in-situ of the cervix uteri should cost considerably less than does the diagnosis and treatment of established invasive ce.ncer. While the diagnosis of carc~noma-in-situ may cost more than the diagnosis of the invasive lesion the cost of treating carcinoma-in-situ is nmch less. But again it should be remembered that there is a period when the overall cost is increased~ from the admission to hospital both of the newly.-detected carcinoma-in-situ patients together with the normal quota of invasive cancer patients during the time when it is still too early to see a fall in the.incidence rate of invasive lesions. Cost of multiple screening The actual cost of multiple screening programmes naturally varies with the examina- tions made. A typical· one, that of the San Francisco longshoreman, is given by Breslow 6 as $5.04 per person screened. In this case, twelve tests were ca-rried out (as cited above). He also quotes two other schemes in which chest X-ray and blood sugar only were done and that the cost lay between $1.50 and $1.75 per person. Other costs are given in the A.N.A. 11 Study of Multiple Screeningn, already referred to. What is difficult, however, is to relate costs like this to what benefits would have derived from spending the same sums in other forms of medical care. It is, perhaps, of interest to consider these services in relation to what :i.s spent generally on medical care. In USA dollar equivalent terms 9 an uverage British general practitioner earns from all sources about $8 000 per year and he looks after, on an average, 2 350 pa tiEmts. This allows between $3.00 and $3.50 only per head for all general medical services. The cost of the whole National Health Service for England and Wales in 1962-63 was £605 million, which allows an expenditure of £13 (or $36) per head of the population over hospital, public health and general medical. California, Union had a total health budget (Federal and State) of over the year for all services, the wealthiest State in the $43 million in 1960-61; 46 this allows expenditure on health and welfare services of about $3 per head in the year, (of which 36 per cent was spent on hospital construction and seven per cent on preventive medical services, about $0.20 per head). Thus in - 70-. relation to total health expenditure multiple screening is costly and its use would need to be judged on the benefits to. health 1.-1hich it could provide. (It should, of course, be remembered that this health budget is for public health only (apart from the contribution hospital construction) and personal medical services are supplied under private contract.) 5.5.3. Inter-country differences Besides justifying itself on "within-country11 economic grounds screening procedures need to fit into existing systems of medical care and the variation of these between countries may influence the pattern of screening. Unlike the usual sequence in clinical medicine when the patient, feeling unwell, consults a doctor and thus submits to diag- nostic examination, early detection entails an appeal to the public to come and be examined. An appeal of this kind is most easily organized through the public health services, and not through the clinical services, (though it may well be to clinicians that the public go for their actual tests as, for example, cervical smears). Where there is no unity of the medical services and where they may be financed in different ways the efficient practice of screening becomes difficult and there is the twinfold danger of a failure in communication and a failure to follow-up, dangers which have been discussed earlier in this paper. On the other hand, where there is a unified medical service it is relatively easy to carry out the whole screening operation, beginning with an appeal to the public, through screening tests, definitive diagnosis, treatment and follow-up, without breakdown in communications at any point. Figure 5.2 represents the two dimensions of economic development and unification of the services for medical care and the straight line passing through the origin indicates an average axis of development of the two variables. Near the point of origin of the curve is the area of poor countries with little in the way of organized medical services. In this area is indicated a medical aid team which might visit the country at its government's request for some special screening operation, (for example, the control of bilharziasis). The other end of the curve is the area of wealthy countries with highly-organized systems of medical care where integrated screening operations carried outunder national arrange- ments may.be expected. At the opposite extremes are countries HXn and "Y". nxn is wealthy but has non-integrated medical services and expensive sporadic screening exercises may take place here, with poor communication and follow-up. Country "Y" on the other hand, is poor economically but has a unified system of medical ce.re. It would comm£.nd ·ideal conditions for the control of the kind of disease which is a problem still in the less prosperous parts of the world 5 for. instance pulmonary tuberculosis. Unfortunately, Fig. 5-2 RELATIONSHIP OF MEDICAL CARE SYSTEM TO ECONOMIC DEVELOPMENT Econ~ic Develop~ent 3 2 1 1 2 3 WHO 70039 Integration of Medical Care Services - '71 - there are·few economically underdeveloped countries which heve achieved integrated m8dical services since this is in itself costly. 5.5.4 Example of gynaecological cytology Soms lessons on the medical care problems and economics of screening can perhaps be learned from experience with initiating service gynaecological cytology in England and Wales, .where. there is a tripartite, decentralized, national health service with semi-independent general practitioner~ public health and hospital services. Here~ the responsibility for the examination is regarded as primarily belonging to the general practitioner~ as he who practices personal preventive medicine. Either he, a public health authority doctor or a doctor from a voluntary family planning organization may, in tt'e case of nwell" women, actually take the cytological material and it is generally regarded as best that this examination should be carried out by a doctor. This poses one problem of screening~ ideally examinations should be carried out by a doctor since women have considerable confidence in the examination and are likely to believe that a negative test indicates that they are not only free from cancer but unlikely to develop it. Whilst a trained assistant~ of nurse or similar standing 9 can easily learn to take cytological material she is not so well trained in observing local pathological changes. Yet 9 if doctors only were to carry out this examination~ not only would a heavy economic burden need to be borne but the service would be likely to· remain limited due to a shortage of medical manpower. If screening (taking material) is to be carried out for uterine cancer by non-medically qualified people, therefore, as to some extent it almost certainly has to be, it is ex-tremely important that the public should understand that this form of medical care is provisional and not definitive. A problem, in a way similar to the above, has to be faced in the examination of cytological specimens. It is of the first importance that those responsible for deciding whether specimens are negative or positive, both pathologists m1d technicians, shall be well trained. Inadequate training can easily bring a whole service into disrepute;, wrong decisions on smears and biopsy material will have far-reaching effects since the definitive treatment for intra-epithelial (carcinoma-in-situ) cervical lesions is normally total hysterectomy. However~ insistence on training in cytology being limited to laboratory technicians, while admittedly ensui'ing a high standard of technical service, is liable to clog the rate of development of a service. The USA and some British Commonwealth countries (among others) have succ·essfully adopted the procedure of ad hoc training in cyto-technology for persons of US university graduate or similar education. This is important because, with all advanced countries, competition for man and women- - 72 - power from all sides has become extremely keen and an over-insistance on formal qualifica- tions prior to training in cyto-technology could be another factor slowing up the development of a service. The number of technicians needed for carrying out not only primary examinations, but also regularly spaced re-examinations, is considerable; as a rough calculation will show~ there e.re about 50 million persons in England and Wales with~ approximately 15 million a.dul t women. Supposing cytologica.l examination were ce.rried out at an average of three-yearly intervals by techiiicians who could examine 10 000 women/specimens per year, a total of 500 technicians would be needed. The problem of the future may however be that of persuading those women at greatest risk to attend for examination (and,. incidentally, perhaps to give advice on health measures likely to diminish the chances of developing cervical cancer); and it·is likely that a rather large proportion of the population may fail to submit to regular examina- tion. Even with smaller numbers the recording and efficient recall, with follow-up if necessary, of all these women presents formidable problems. A possible means of recall that is being explored is the registration centrally of all cytological examinations followed by automatic data processing and the issue of recall appointments. Such a central registration system clearly also has great epidemiological potentialities. By reason of the man-power shortage the search for automated techniques for cytological screening is clearly of great interest and would be acceptable even with a high false positive rate. Provided that the cost per examination could be kept low, automation could be combined with a high degree of centralization of screening services 9 with a consequent saving in overhead costs. However, it would need to be remembered that the lengthened lines of co~~unication would be more vulnerable and it seems likely that on-the-spot re-examination of reported positive specimens would still be necessary. It still remains to be seen whether cytological screening will be accompanied by a significant reduction in deaths from uterine cancer 1 but it appears that already the incidence of invasive cervical cancer can be lowered by this technique. INhere this is economically possible it is no longer therefore a question of whether one form of treat- ment is economically more desirable than another since it can hardly be argued that women might wait until they have developed early invasive cancer, rather than have it diagnosed and treated in the pre-invasive stage. It is, however, of some interest to examine briefly the consequences in hospital bed usage and the following simple calcula- tion may give some idea of the effects of widespread population screening for uterine cancer: - 73 - ; At present there are in England and We.les about 10 500 discharges from hospital annually for carcinoma of the cervix and the average duration of stay is 21 days~ giving . 47 a total of 220 000 hospital days. These discharges refer to admission and not persons. The approximate total time spent in hospital as an in-patient by a woman with carcinoma of the cervix during her lifetime is in the region of 50 days. The average length of admission for cone diopsy for a patient with carcinoma-in-situ is in the neighbourhood of seven days or roughly one seventh of 50 days. On the other hand, perhaps twice the number of women that would ultimately have developed invasive cancer, would be subjected to cone biopsy. Supposing cervical screening to be nearly 100 per cent effective, this would still cut down total hospital bed usage by a factor of three or four. However, admission for biopsy during the early years of a screening prograinrne would occur simultaneously >vi th the same number of admissions for invasive cancer as previously experienced. From the Hospital In-Patient Enquiry reports of England and Wales it cen b8 estimated that in a population of 250 000 there wo11ld be !oughly 35 admissions of cancer of the cervix per year. If one-fifth of the female population over 20 years of age were screened annually, on the l'1eophis, Tennessee rates 50-60 carcinomas-in-situ would be discovered, thus nearly trebling the existing admission rate. At 10 days in hospital for 'each case this would need, on average, e.n extra two beds for a population of 250 ooo, over and above the two to three that would be needed in any case. 5.5.5 Conclusion Clearly some way is needed for measuring the value to a community of the early detection and treatment of both pre-symptomatic and declared disease. The 11value" may be either of the 11cost-benefit'1 kind (e.g. productive of better health or comfort), or of the "cost-effective" variety (i.e., of economic value to the community). Obviously, in terms of health 5 the two sorts of value cannot be fully separated - the man or woman who enjoys better physical or mental health can work better, though it may be impossible to measure the difference in output between loss and more healthy people. The important point, probably in determining whether a particular measure for early disease detection is, or is not, worthwhile, is the degree to which the positive results from screening has been established. For example, it seems doubtful whether it is of value to screen a whole community for high blood·pressure, other than what is already done by the clinical services, in the absence of good criteria and a proved and acceptable form of early treatment. On the other hand, as we have seen 5 it does seem possible that .the early·detection and treatment of cancer of the cervix may lead to its virtual - 74 - eradication5 and this, in terms of benefit to the individual woman, is well worthwhile. However, the economic cost of a campaign to eradicate cervical cancer must, if it is to succeed fully, inevitably be relatively high. The cost must therefore be considered by any community in terms of its own particular economy. If5 as is usual, there is a limit to the health budget, the benefit and effectiveness of the proposed programme needs to be compared with other desirable objectives. For example, an intensive campaign to prevent young people starting to smoke, or a nation-wide attempt to prevent overweight? could improve the life and health of a far lo.rger number of persons in the active span of life than the cervical cancer programme. In the event communities often decide that every feasible form of early disease detection end prevention must be offered to the population with the result that the overall cost cannot satisfactorily be met, and a number of programmes is carried out in a less than effective way, instead of a few that are well-planned and well-executed. 5.6 The place of education in the ea:clv detection of diseqscs 5.6.1 .Lch'_<?2-tion of the medical profession Undergraduate medical education is traditionally disease-orientated. Education in public health has in the past been largely concerned with environmental health and the primary prevention of communicable disease. The teaching of hygiene and public health in Europe has been reviewed in a WHO monograph by Grundy and Meckintosh. 48 Tne dominant role in teaching medical students is still to too large an extent filled by the clinician whose treining, in turn, has not concerned him to any great extent with preventive medicine. At present there remains a tendency for a self-perpetuating system of teaching which allows too little emphasis on new thinking in preventive methods. In recent years there havo been signs of a new epproach, by teaching both the preventive and clinical aspects of medicine at the same time. The preventive side of medical practice in thE:. community needs to be learned both by the undergraduate student and in the pre-registration and post-graduation periods. Thus a ltJHO Expert Committee49 considered that the preventive aspects of medicine should be -taugnt at all stages in the curriculum, including the basic medical sciences and pathology courses. Interest in the earlier detection of disease must largely stem from an understanding of the epidemiological approach. As the WHO Expert Committee report points out, a number of new and special subjects are needed in training for preventive medical services, of which the epidemiological method takes first place, and which should include medical statistics, the social sciences, genetics and the organization of health and welfare services. These subjects are best taught by participation (involving some degree of personal - 75 - respohsibfli ty) in laboratory work, field surveys, out-patient .md peripheral services, domiciliary care and rural health programmes. For fostering the idea of early disease detectidn it is important to introduce the student to e type of general medical practice organized in such a way as to make this work feasible, e.g. well-organized group practice or health centre. Similar views have been set out in a WHO study "The Teaching of the Medical Student for Comprehensive Medical Practice".so This study particulerly E;lmp~asizes the.value of the health centre for this type of teaching. It is, in fact, now realized that it must be for the doctor practising in the community to undertake the first-line prevention of disease, whether at the primary or secondary stage. He is best placed to see the beginnings of illness and to be aware of local genetic and environmental factors. University faculties of general practice are now springing up and their role in relation to early disease detection can fruitfully be exploited in a.ssociation with their fellow departments of epidemiology, social medicine and medical statistics. The place of the general practitioner in the prevention of disease was recently considered by the College of General Practitioners.51 In addition to doctors it is important to remember that there are other workers in the practice of medicine who also need educating towards more positive attitudes to the early detection and treatment of illness, for example, nurses, health visitors, chiropodists and pharmacists. 5.6.2 Education of the public In order to successfully carry out the early detection of disease it is not only essential to have an alert and prevent'ion-orientated medical profession; the public needs also to co-operate in the operation. We know from experience that, for example, in the reporting of symptomatic cancers, the public tends to defer seeing a doct::>r until a condition is relctively far advanced. Knowledge about the value of early diagnosis and treatment needs to be spread to the community (and this, in its turn, needs. to be based on sound scientific studies). Thus it is accepted that education of the public is essential7 what is not always known is the form which this health education should take, both in its contentand in its application.· Early disease detection as a means of prevention in the community needs, in order to succeed, to reach all, or nearly all, those at risk regardless of cultural and economic differences between the various parts of the population. A common experience is that persons with the best general education make use of facilities of which they may perhaps, in reality, be in little need? while others, at a much higher risk of disease but with a less good education, fa.il to utilize a service out of ignorance of their specie.l need (the example of high-risk groups in - 76 - cancer of the cervix is briefly referred to in sections 5.5.4 and 6.6.2 ). Planning health education to reach all kinds of people is therefore essential and~ in order to do this? special studies of public attitudes to disease may be needed in the first place. Studies of ~his kind? have, for example? been carried out for gynaecological cytology in California, 52 nationally in the USA; 53 and one is in progress in Manchester.54 No doubt much still needs.to be learned about the best ways of reaching populations in providing health education. One point of importance is the need for proper organiza- tion. "Delivering the message 11 ? whether in print, by direct speech, radio or television, is a specialized technique and if carried out in an amateurish way 5 can lose most of its potential impact. There is a place in public health administration for specialist health education officers to assist the medical and nursing staff. In addition to local officers the central planning and teaching of health education me.y be valuable? as suggested in Great Britain by the Committee on Health Education. 55 As well as staff employed by public health authorities spreading education on health, other specially-trained persons have an important part to play in educating the public. General practitioners are well placed to influence their patients }n particular high- risk groups to attend for examination) and nurses and health visitors, for example, can influence the public beneficially during the course of their work. 6. ILLUSTRATIVE EXAMPLES OF SCREENING DISEASES The previous sections have dealt almost entirely with the geJ:ler?-1 conE3ide_rations of early disease detection. It now seems useful to consider some specific conditions in rather more detail. In the following section we have therefore chosen certain examples to illustrate specific points made in the general soctions. 6.1 Diabetes mellitus 6.1.1 General Diabetes detection, according to Joslin and others 56 goes back to 1909 when Barringer reported the findings on over 70 000 persons examined for life insurance purposes. However, the concept of population screening for diabetes dates from the survey carried out by Wilkerson and Krall in 194757 on the people of Oxford, Massachusetts. Post-prandial blood sugar estimation together with glycosuria testing was carried out. Since that time an increasing amount of case-finding, both by urine and blood examination, or both together? has been practised in the advanced countries of the world. For typical examples see 1.Jalker and Kerridg8.5B Despite all this work it is still difficult to evaluate the results in terms of benefit to the populations - 77 - screened. Some of the criteria for case-finding discussed above remain unsatisfied. This question will be dealt with later in this section. There is little question that the best documented case-finding has been carried out during the past decade in the USA. Under the Diabetes and Arthritis PrograJn of the Chronic Diseases Division of the US PHS information has been gathered together and published. Thus NcDonald and his co-workers 59 give figures for 1958-1963; during 1963 nearly 530 000 persons were screened by blood sugar estimation, as compared with 190 000 in 1959, an incree.se from one per 1 000 total population to three per 1 000. Of 338 500 screened in 1962, 10 300 tests were positive~ 9 000 (2.7 per cent) of these were referred to their physician, and of these 2 500 (0.7 per cent) wore known to be diagnosed as diabetic. The US National Health Survey finds a rate of nine per 1 000 population of all ages for known cases of diabetes, while thG estimate of unsuspected 60 diabetes is at a rate of e,ight per thousand for all ages. This would yield a total in the USA of 1 500 000 known and 1 400 000 unknown diabetics, so that screening, even in a country where a relatively large amount h&s b0en done, has still a great way to go. In general, blood sugar examination has superseded urine testing, rapid screening being pos:;lible using either the 11Clini tron" apparatus or the simpler slowe.r but much cheaper Glover-Edwards kit, both employing the Wilkerson-Heftmann screening test. 61 Remein and Wilkerson 15 show that, on a. study of 580 persons of varying age and race, the following specificity and sensitivity was obtained for blood sugar and urine screen~ng respectively: :Bi~'Oct- ·(s~~~-gyi-Nel son) 160 mg/100 ml, 1 hour after meal ~- - ... ~~ . . . ~ . Urine (Dreypack, glucose oxidase) 2 + TABIE 6.1.1 test Sensitivity %positive .. 52.9 45.9 Specificity % negative 99.4 90.3 I It is c::lear. that_:ne;ither :ti:Jchnic::ue is .. good .a.t.picking. up a .. high.pr~portion of persons with diabetes though the blood sugar method has the advantage of finding very few false positives. As a practical example of tho application of the above blood sugar sensitivity and specificity, in a. population of 10 000 with a true prevalence of 150 diabetics, there - 78 - would be 70 false negative results (i.e. 70 missed cases) and 60 false positive tests. The practice of case-finding is discussed by the C.C.I. in "Chronic Illness in the United States 11 , Vol. IV, under methodology (Appendix D). The Baltimore screening survey 62 may be of interest~ as an example. The urine was first tested 30 to 50 minutes after 50 g of glucose given as a drink. The vt:nous blood sugar was screened at 160 mg per 100 ml using the Wilkerson-Heftman technique? 45 to 75 minutes after glucose. If positive, a two hour specimen of blood was screened by the same technique at 130 mg per 100 mi blood. Out of 1 916 persons of all ages 33 were found with positive one and two hour values (or not done at two hours)? 14 males and 19 females preponderantly in the 45-64 age-group. In nine persons only was follow-up completed) five were confirmed DS diabetics, of whom four were previously undetected. The rate in which both tests were abnormal (33 in all) was eight per thousand persons examined. This is fairly typicel of the results of case-finding. The cost of this type of operation is estimated by the US PHS as being in the neighbourhood of $0.84 per person screened~ providing this is carried out by an adequately staffed loca.l health department and not including overheads and publicity. 61 At a pick-up rate of eight per thousand population the cost per diabetic discovered would be about $105. A difficulty in comparing case-finding progre~es has been that similar criteria for diagnosing diabetes have not been used from place to place. Tho US PHS "Diabetes Program Guide" lists nine different s~ts of criteria. The Committee to the Public Health Servi·ce. Diagnostic Study recommends the following criteria? with a point value allotted to each reading:- TABlE 6.1.2 Time Minimum Point positive B.S. value of positive Fasting 110 1 1 hour 170 1/2 ... 2 hours 120 1/2 3 hours 110 1 A total of two or more points is considered to indicate a diagnosis of diabetes. - 79 - Despite these attempts at defining diagnostic criteria it is still very difficult to know in practice which are being used in a particular case-finding programme. Where separate forms of medical care co-exist side by side~ follow-up of screened positive persons can be fruitless and it is not very uncommon for this to consist only of a urine exrunina tion following on a positive blood test. Reid's report to WHO 63 makes a number of cogent points: (a) Differences of opinion about the value of generalised~ versus selective~ screening. (b) In the case of selective screening~ the criteria which should be used for selection. (c) The relative merits of urine versus blood screening, although opinion seems to have hardened in favour of the latter. (d) Varying conditions under which screening is carried out. Thus it is sometimes undertaken on the basis of random blood samples) sometimes on tests carried out at special times in relation to meals; sometimes with prior carbohydrate loading; and sometimes with a substantial period of dietary preparation. (e) There is a. wide divergence of opinion about the interpretation of screening levelsj and there are, of course, varying standards according to whether a true glucose or other type of blood sugar estimation is carried out. (f) Follow-up arrangements for th_ose who screen positive e,re variable and many a.re not adequately investigated. (g) Even if full tolerance tests are arranged, there are again variations in such metters as dietary preparation, carbohydrate loading~ and the interpretation of results. (h) There is further confusion (and this applies not only to the USA) because many of the medical publications reporting diabetic screening campaigns do not give enough information to purmit judgements to be formed~ e.g. no details may be given of the preparation of the patient for testing~ of the type of blood used; or of the variety of quantitative estimation employed. (i) Pre.ctically all the surveys which have been carried out point to· the .. conclusion that diabetes mellitus is difficult to define in terms of blood chemistry as any series of tolerance tests will show, at one end of the scale, - 80 - those who are clearly not diabetic, at the other, those who are definitely diabetic 7 but in the middle, a substantic..l group who fall into neither category. In many ways the position resembles that which applies to the diagnosis of hypertension, and greater knowledge is required of the natural history of diabetes bGfore anything of a more definite nature can be said about the intermediate group. (j) This last point is, in turn 1 related to the question of the preventability of diabetic cardiovascular complications, which is again a subject which has not yet been finally answered and which is 5 indeed, unlikely to be decided for another ten or twenty years? by which time various longitudinal studies will have been completed. Before advocating e. general policy for diabetes screening, therefore, more needs to be learned about the disease and the results of treatment. This will be discussed in the next part of this section. However, the detection of overt diabetics can well be encouraged, and high blood levels for screening could be agreed as an interim measure, so as to exclude"border-line 11 cases. Selective screening of the high-risk groups could be more economical since the yield is greater, but a real difficulty, and a potentially expensive one 9 is to make; contact with these groups. What is also needed is the better education of the public and the medical and nursing profession. As Reid 64 has pointed out, enquiry shows that public knowledge of the symptoms of diabetes is extremely vague:, and a proportion of diabetics diagnosed in surveys have already had symptoms for some time, according to Redhead 7 65 Wilkerson and Kral1, 57 College of GGneral Practitioners. 66 In the USA emphasis is now being given to educating the medical profession itself in the diagnosis and control of diabetes. Ancillary workers are trained as "public health representatives" to organize local detection work. They are based on local health departments and visit local doctors or other professional person- nel. 63 L1 some US projects self selectior:. plays a part a.nd c:.s much as an eight per cent yield of dia.bctics may be obtained in this way. Another high-risk group is that of patients attending hospital. The Dundee workers 67 have shown a surprisingly high proportion of undetected diabetics in hospital wards and there is scope for increasing the awareness of doctors and nurses to the sad fact that admission to hospital is in itself no guarantee that the asymptomatic diabetic will be diagnosed. The evidence of the value of screening In screening populations for diabetes it is important to QSk whether the criteria discussed under "Principles" have been fulfilled. Two of these principles merit further discussion: - 81 - (1) the question., whether it can be accepted that ea~ly treatment is valuable, and (ii) the question of the criteriafor diagnosis. 6.1.2 Value of early treatment Early treatment cau be considered at two stages; firstly its effect on early clinical diabetes and secondly, its effect at a later stage of the condition. It is not possible to obtain irrefutable evidence for the value of early (or late) treatment of developed clinical diabetes because of the ethical veto on withholding treatment from a control group. However, clinical evidence has accumulated that care- fully-controlled diabetics experience less complications than badly-controlled patients. Rundles68 has noted that neuropathy is particularly prone to develop in poorly- controlled patients. Garlanc169 states that amyotrophy is totally reversible with full diabetic control. In Ashton's view70 11 good control" of diabetes over many years 71 72 may have some beneficial. effect on retinal microaneurysms. Both Dunlop . and Marble have reported that diabe:t.~cs under good control are much less likely to develop nephropathy: than poor;Ly-controlled patients; while Johnsson73 found significantly less. nephropathy and severe retinopathy in a series of patients treated by strict dieting., with .an att,E;lmpt to keep the urine sugar-free., compared with. a series allowed a morel ··-' : . liberal diet, i:o. whicn only control of polylJ!ia and ketonuria was attempted. 11>/olff Salt 74 found, in acute e:Jeperiments, that hyperlipaemia returns to normal in children when diabetes is controlled, thus suggesting a possible causal relationship between diabetes and early arteriosclerotic changes; and Keen75 has -reported that cataract is more common in poorly-controlled diabetics and >'lhen arteriosclerosis is als<> present. Newburgh and Gonn76 have pointed out, among others, that weight reduction in the obese middle-aged diabetic may result in the glucose tolerance test returning to normal. The relationship between diabetes and obasity has recently been briefly reviewed.77 Finally; as a general finding, patients still living free of complications after a 25 year history of diabetes have been found by Joslin 1 s clinic to be those vrho had controlled· their diabetes meticulously.78 The subjeet has been ably reviewed by Beckett.,79 who urges the need to reduce delay· in the diagnosis of diabetes. There is, therefore., a considerable body of evidence in favour of the benefits of treatments .in minimizing diabetic. complications. However, it must also be allowed .that there is also some evidence to the contrary., particularly on the progress of retinopathy 8o 81 82 and renal changes. " . " When treatment at the earlier, pre-clinical stage is considered the matter becomes in on'1 sense more complicated because of the diffj.culty in deciding where to draw the arbitro.ry line between 11 health11 and 11 disease11 ; though on the other hand, investigation is e:ade simple by the real uncertainty which exists about the value· o:f'"'"fre·atment, thus and t H t 83 . . ethical~y allo;-ring randomization into treated con rol groups. oe nas snown that "pre-d:lt.betic" women put on insulin during pregnancy have significantly more live- birt~.s follo\'rtng t::..~ee.tment than before; and Wilkerson 84 finds insulin in 11 pre-diabetic" wo:::,:m p::>e·vents overweight in the infant, with a probable reduction in perinatal mortality: Reid63, 64 reports on a visit to the US PHS Diabetes Field Research F?,cjlity in Boston, where this work is being carried out, that more members of the control group of 11 pre-diabetica women (not treated with insulin) have developed clinical di.:1botes than have women in the treated group, though the numbers are not statistically 11 Prc·-diabetes" in women is designated by Vvilkerson as a group which includes "a high inc~.dc:;:we of big babies, stillbirths, neonatal deaths, spontaneous abortions, premature del:.:.veries, toxaemia and congenital abnormalities", as well as showing "transient 85 ~0no~rr.alities of carbohydrate metabolism during pregnancy". Fajans and Conn regard the 11 potential" diabetic as a person with a normal oral glucose tolerance curve but shcwl.ng .::. positive response to the cortisone-glucose tolerance test. They advocate· apply5 .. ng th:t.s test to those with an increased risk of diabetes, i.e. the women in the cc::cegories just cited as well as the relatives of diabetics. 6 , "· • ..J...,..,. Tnis question ~.s connected with that of the value of early treatment in that criter:l::> adopted det8rmine the s·Gage of blood sugar abnormality at which the diagnosis of diabetes j.s made and thus the stage at which treatment might be started. A.n irr:vm.~tant factor in considerj_ng the nosology of diabetes mellitus is the part heredity plays. The familial aggregation of diabetics has been attritbuted to the ir.he:ritnnce of a Mendelian recessive gene86, 87 For9. and Glen88 found the prevalence of unrecogn:i.sed diabetes was five times as high among relatives of known diabetics as in . th~ general population. Other surveys have confirmed these findings but it is not: of cslwse~ necessar,y to invoke a single gene to account for this. HarrisS9,90 considers thsrG ls insufficient evidence to determine the type of transmission. Recent blood 21~, 91 sugar surveys have demonstrated a continuous distribution of this index with no visible tendency to bimodality, though this could be present but concealed. On balance, present evidence probably favours a multifactorial inheritance. The question has been - 83 - 92 revievved recently in the British f'iledical Journal. 1'lh8ther single gene or multi- factoriably transmitted, the index blood sugar fails to segregate a discrete group of 11 diabetics11 from "non-diabetics" and therefore the problem of the "border-line" case of Fig. 4.5 (lower drRwing) occurs. This is the problem that lies at the root of discussion of diagnostic crj.tc::-ia and of "pre-diabetes". The best diagnostic criteria are based on the experience of those treating diabetic patients and one such group has advised the US PHS on the set of criteria already cited.61 However, it should be noted that these criteria are not ~;he same as those a¢1.opted by the American Diabetes Association and there ~s no goneral agreement as yet. International agreement on a set of criteria could be a v<:?.J.uable contribution to the comparability of survey work. Recently, a 1tlli0 Expert Commit~ee ras reviewed current knowledge about diabetes mellitus and has - ~ recommended work:illg criteria for single blood sugar levels after glucose loading./ Whilst it. must .be agreed that a diagnosis .. based on a glucose tolerance curve is arbitrary, dividin::.~ "diabetics" from 11 non-diabetics 11 by means of a single blood sugar (or~ even more 8o, urine test) examination is even more arbitrary arrl subject to doubt. Apart from patients actually found with the symptoms of diabetes$ there will be at one end of the scale those with such high blood sugar levels that there can be little if any doubt of the diagnosis; whilst at the other end there will be those persons with lew blood sugar levels, in whom there can be no suspicion of di.abetes. The intermediate group, as studies Jn Birmingnam66 and Bedforo24 and by the US National Health Survey, 91 have Fhown, comprise a formidable proportion. In the US National Health Survey, for example, 15.5 per cent. of persons had a blood sugar level of 160 mg per cent. or more one hour after 50 g of glucose by mouth { se.e. Table. E. 1.3), a level which Remein 15 and Wilkerson showed to be 99.4 per cent. specific for diabetes by the accepted criteria of their advisory panel. This proportion increased greatly with increasing age as the table shows~ thus, nearly half of the women over 65 could be classified as diabetic on this criterion. Similar findings have been reported from Birmingham and Bedford. The US National Health Survey report suggests that current standards for a normal blood glucose level are unrealistically low. As the report says, it is outside the scope of the survey to answer this question but within the scope to raise it. A follow-up of -~-he border-line group found in the Bedford survey is in progress, with randomization of treatment, so that in due course information should become available on the criteria to be adopted for diabetics in need of treatment. Until that time comes it might be legitimate to screen at a higher blood sugar level of, say, 18o-200 mg>; at two hours after 50 g glucose, in order to detect only unequivocal diabetics. There are hints, however, that relati.vely mild hyperglycaemia is associated with cardiovascular - 84 - changes; 94,95 though it remains to be seen whether measures aimed at lowering the blood sugar level are able to prevent or delay these changes. Blood suga.r screening could be greatly facilitated in practice if the newly available glucose oxidase strip paper technique for blood proves its worth. This technique would make much more simple the screening by individual practitioners of the high-risk patients known to them. A large scale field trial of this method has yet to be published~ at least one is in progress in the United Kingdom. There has been hope that other and better diagnostic indices than the blood sugar might be found 5 for exe~ple non-esterified fatty acids (NEFA). But~ so far~ nothing more relie. ble than blood sugar as a clinical index has been demonstrated" TABLE 6.1.3 I Percentage of Adults having Blood Sugar Level of 160 mg% i or more one hour after 50 g of Glucose by Mouth I ~ ! I I t ; I j Age group~ i Total I ! i I ! j years ! 18-79 18-24 I 25-34 35-44 45-54 t 55-64 I 65-74 I i i i I I I I ! I I I I ! Males 1L8 1.3 5.0 11.0 ! 13.2 I 17.7 27.4 I I i ; I 1 Females 18.8 5.1 7.3 11.0 I 19.6 34.5 43.0 i I l Persons 15.5 J.4 6.2 11.0 16.5 26.4 36.0 ; ! i , I t Reference: US National Health Survey. 6.2 Heart disease 6.2.1 Rheumatic and congenital i·1eart disease 75-79 24.7 58.2 4L5 In developed countries rheumatic fever and its consequences appears to be diminishing though convincing figures are not available. Convincing proof can only be provided by incidence figures which in turn depend on the condition being notifiable; and notifi- cation has only been carried out in localized areas" However, deaths from chronic rheumatic heart disease provide some support; and death-rates are diminishing. In England and Wales for example, the crude death-rate fell from 255 to 130 per million persons between 1950 and 1964.96 , 97 Some of this fall is, of course, likely to be due to improved treatment. In developed countries 1 again~ the prevalence rate for chronic rheumatic heart disease is relatively lm.,r. We do not know what it may be in developing -parts of the world. The British general practice survey 98 gives a patient consulting rate per - 85 - annum of 1.4 per thousand persons of all ages (2.J per thousand aged 45-:-64). The US National Health Survey 99 on; examinations carried out on a national sample between 1961-62, found a prevalence rate of 1.1 per cent of persons aged 18-79 with rheumatic heart disease, the highest rate (3.8 per cent) occurring in men aged 75-79. In women 100, 101 .. of 45-54 the rate was 1.8 per cent. \AJHO set enqu~r~es as to world prevalence on foot in 1963 through the cardiological societies of member countries but data are not yet available. Preventive steps can be taken at two stages: firstly by treating prophylactically all persons who have hE:.d at least one attack of rheumatic fever:; and secondly, by administering continuing prophylaxis to persons and who have developed rheumatic heart disease. In areas where the prevalence of sequelae is not high it is probably not economic to attempt mass detection~ (by measuring anti-streptolysing titres, by taping heart sounds and by electrocardiography) and detection and prophylaxis can safely be left to the personal doctor~ An example of detection by taped heart sounds is the study of heart disease in tho Chicago public school child population. 102 In this case the yield proved so small that the cost per case found would surely be prohibitive. The heart sounds of 27 000 elementary school children were recorded on m~~!lE:tic tape a..11d these were interpreted by two physicians. A rate of two per 1 000 children were discovered to have organic heart disease, of whom one per 1 000 were previously unrecognized. Of these previously unrecognized cases 0.2 per 1 000 were due to rheumatic fever and 0.8 per 1 000 to congenital heart disease. An exception to.this principle may be the case where many young people are living together, as in schools and the armed forces. Trials of mass prophylaxis have, for example, been carried out in the US armed forces where·it was found extremely difficult to maintain continuous prophylaxis. It may well be, however, that much higher rates for acute rheumatism and rheumatic heart disease prevail in Africa and Asia, for example, and that prophyla~s in these area.s should be tried, despite the difficulties of dispersion and supervision that would be met. We understand that a trial of case-finding for rhe~atic heart disease is in progress in Russia, based on the National Rheumatological Institute in Moscow. 101 Little more need be said about congenital heart disee.se. Serious cases will normally be discovered at an early stage in life at infant welfare examinations, whether carried out by the family doctor or at health authority clinics. At a later age the school health service takes on this screening role. \~ether so much medical man- and woman-power is best used in carrying out repeated routine health examinations is a - 86 - matter of current discussion. The answer may perhaps li0 partially in streamlining the examination end partially in cutting the number to the minimum, which would be carried out at key times in the life of the schoolchild~ e.g. at the start, at puberty and at school leaving. 6.2.2 Ischaemic heart disease 6.2.2.1 Mortality Of all diseases in developed p~rts of the world arteriosclerotic and d0generative hee.rt disease accounts for the greatest number of deaths, by far the greatest number having specified involvement of the coronary arteries (I.C.D. 420.1.). The death rate has been steadily climbing. For example, 103 in England and \.Jale s, the crude death rate per million living from this assigned cause rose, for men, from 808 in 1940 to 2 731 in 1962 and from women, from 374 to 1 613. There is thus not only about e. three- fold rise for both sexes but an approximately two to one male to female ratio, (with the ratio diminishing the later the year). The evidence that this recorded increase is real is reasonably firm. 103 6.2.2.2 Morbidity For ischaemic heart disease (I.H.D. ), too, reasonably good measures of prevalence and incidence are provided by various prospective surveys. Various studies give mean incidence figures, for men (which of course rise with age) from two to 12 per thousand which seven per thousand as the mode? while prevalence estima.tes vary even more widely from overe.ll rates of 24 per thousand men aged 30-59 (Framingham) 3 to 24 per cent in men aged 55-64 (Annandale). 104 A recent publica- tion of the National Centre for Health Statistics99 finds an overall prevalence of definite I.H.D. in a probability sample of white males aged 18-79 years of 38 per thousand. Epstein and persons over the age of ischaemic heart disease thousand females. 105 his co-workers, in a survey totalling 90 per cent (8 641) of all 16 in the town of Tecumseh, Michigan, measured a prevalence of (including probE,ble disease) of 49 per thousand males and 33 per The prevalence for white men between 65 and 74 years was 12.2 per cent with definite I.H.D. and a further 5.1 per cent with suspected I.H.D. The diagnosis was based on a medical history and cardiovascular exa~ination, including a 12 lead ECG and a full scale chest X-ray. The subject has recently been well reviewed by Epstein. 106 - 87 - In 1963 the WHO, Regional Qffice for Europe held a Technical Heeting on Surveys of - . 1~ the Prevalence of I.H.D. in certain European countries. At this meeting Rose reported on I.H.D. prevalence in working men in England aged 35-59. An overall prevalence of 10 per cent was found, using his questionary 108 and ECG' s read by the Minnesota code. If diagnostically weaker ECG items (ST depression and T inversion) were included as indicating I.H.D. the prevalence increased to 20 per cent. This variability in ECG criteria may account for much of the difference between the prevalence rates cited above. At the same meeting Hostmann and Thomson reported findings for I.H.D. prevalence from Odensa. They found an overall prevalence of six per cent for men aged 40-59, with f, rate of 10 per cent in the 55-59 age group. 6.2.2.3 Screening One of the primary needs for screening, a high prevalence rate, is therefore satisfied. The next important question is, whether a recognizable pre-symptomatic stage exists. This question will be examined in two ways; firstly, is there a latent stage of actual myocardial ischaemia; secondly, can a satisfactory high risk group be defined. (This distinction is referred to in the WHO Technical Report on the Preventive Aspects in Arterial Hypertension and Ischaemic Heart Disease. 109 (a) The latent stage of myocardial ischaemia. The development of the cardio- vascular questionary 6.2-15, 16 has led to the recognition that many people actually have symptoms of I.H.D. without being under medical care. Rose found four per cent of the men aged 35-59 who were examined by questionary had symptoms of angina pectoris and 4.5 per cent h~d symptoms of possible myocardial infarctium. About three-quarters of this angina had been previou~ly undiagnosed. 112 Thus in the symptomatic group alone there is scope for screening in this way. Electrocardiography is the other chief screening technique. Routine examination in this way shows that in a high proportion of people there are changes consistent with I.H.D. in the absence of symptoms, quite apart from the appreciable proportion of those developing actual myocardial infarction without pain - 21 per cent in the Framingham series. l13 I th An d 1 . 1 d f d t 104 . t t f th t t n e nan a e ser1es, a rea y re erre o, 1n en ou o e wen y four men aged 55.64 found to have I.H.D. there was ECG evidence only. In the Ach~son' s 114 gr~up of 53 men aged 65-85 with I. H. D. 27 had ECG changes only. The US National Health Survey,found five per cent of persons aged 18-79 years with either definite or suspected I.H.D. In addition there was a further - 88 - 6.4 per cent who had electrocardiographic abnormalities falling "just short of the rather severe survey criteria for myocardial infarction". 99 In the series, quoted by Rose, 108 of 1 848 English working men aged 35-59, two per cent had definite EGG changes in the absence of symptoms (in answer to the questionary) and a further 10 per cent had the less definite ST depression and/or T inversion. It should be noted here that Hinkle and l] 5 others, · in studies on the variation of the EGG under conditions of daily life, report that changes in the ST segment and T wave occur so frequently in people of all ages in association with ordinary activities that they believe it would be hazardous to assume that these changes necessarily indicate a pathological process. The prognostic significance of some of these EGG changes in the absence of other signs of I.H.D. can be accepted from the results of prospective studies.ll6 We may ask how successful is the ECG in only detecting disease when it is present, i.e. what is its specificity? what proportion of false positives mt=ty we expect? This question cannot properly be answered because we do not know enough about series of persons having abnormal EGG's who later died and were found not to have myocardial disease post mortem. Regarding sensitivity (indicating the proportion of false negatives) of electrocardiography, there is more evidence. The six-year follow-up of the Framingham Survey revealed that 88 men had developed myocardial infarctionj of these, 15 (all of whom died suddenly) had had no previous evidence of I.H.D. 117 An earlier report from Framingham showed that, out of 44 EGG's on patients with I.H.D. 22 were normal. llB Findings of this kind indicate ti1at the val.ue of the ECG by itself as a screening method must be strictly limited. Moreover, as with other tests, the ECG is subject to the drawback of variabi- lity in interpretation between observers. ll9 The Minnesota Code 120 is a help in .improving observer agreement. A further help will be the training of observers (who need not be medically trained) on the set of reference tracings being prepared by WHO Headquarters, Geneva. 107 Inter-observer variation amongst selected observers can be rewardingly low, while inter- observer variation is of the order of twice the variation in reading of the ll('\ 111 same observer. 1 We can conclude, then, that the EGG though highly (b) - 89 - significant when positive 9 is far from being a certain diagnostic technique when used by itself. A third method of screening for I.H.D. (as well as other forms of heart disease) is by mass radiography. Mass radiography is sometimes usE::d in case-finding for heart disease, including arteriosclerotic disease. A good e~mple is the study by Thompson and his colleagues, carried out in Oklahoma. 121 When screening for pulmonary disease the searci:1 for heart dis- ease in addition is undoubtedly vwrtnwhile and it is perhaps best viewed in this wo.y. C8.rried out by itself ci1.e yield for all i1eart disease· is small ~nd coi;l)C.:i.'ed with other methods of finding I.H.D. ~ very small. Nor is radiography likely to find early I.H.D. or hypertensive disease. In Thompson's Oklahoma project 8 126 persons over the ege of 15 were screened by MMR, about one quarter of the whole population of Carter County in this age group. There were 917 (11 per cent) 11 suspect 11 films of which 302 (four per ·cent) were confirmed by 2_ physician. Fifty-two of these cases proved to be previously undiagnosed (60 per cent). Of these 52 cases 45 were either arteriosclerotic, hypertensive~ or both. Lastly, screening for I.H.D. is sometimes carried out by estimation of the serum lipids (usually serum cholesterol level). The predictive value of a high cholesterol, while highly correlated with I.H.D. for groups of the population, is not of a high order in the individual and is therefore not of high value, by itself, as e sc;reening test. There is indeed danger thEJ.t, used unexpectedly, thE: serum cholesterol might be harmfully used by leading to rest:dction of life in persons vri th rele.tively high levels but no other signs of I. H. D. Meanwhile,. (;pidemiological studies of the significance or rise in the se1~ lipid ·1eyels in different populations are in progress. A WHO stande.rdization programme is centred at Atlante, Georgia, JJSA, under :the direction of G.R. Cooper. 122 A review of cardiovascular survey methods is at present being prepared for vJHO by Dr Henry Blackburn of the Laboratory of Physiological Hygiene) University of Minnesota~ and Dr G.A. Rose of the London School of Hygiene and Tropical Medicine. High-risk groups. A number of definite risk factors can now bE: recognized as a result of the various studies, prospective and otherwise, of the epidemiology of I.H.D. These factors will be discussed briefly:. - 90 - (1) Age: The death rate from I.H.D. increases with age and so does the evidence of disease. For example, in the Framingham survey, 117 the six-year incidence for men aged 30-44 was 24.9 per thousand population while it was 90.6 per thousand for men aged 45-62. (2) Sex: Men, as is well recognized, suffer a higher incidence of I.H.D. than women. In the Framingham study the male to female ratio of incidence at ages 30-44 was 25:2 whilst at 45-62 it was 91:45. The tendency for this ratio to approach unity with age is well-recognized end is considered to be related to the hormonal changes at the menopause. 123, 124 (3) Physical activity: Morris and his co-workers 125, 126 have shown thc:tt it is likely that I.H.D. develops more frequently in those Hhose work is physically inactive than'in those Hith physically active occupations. (4) Occupation: There is a gradient amongst men (not amongst married HOmen) in mortality by social class. The Registrar General's Decennial Supplement for England and tvales (1958), 127 gives a Standardised MortPlity Ratio (SMR) gradient from 147 for males on Social Class I to 89 for males Social Class V and the rBnge is even greater for specific occupations (registered medical practitioners, 159); farmers, 62). (5) Serum lipids: The Framingham survey, amongst much other evidence, clearly relates group susceptibility to I.H.D. to the serum cholesterol level. There is no evidence of bimodality in the distribution, except perhaps for the very small group of familial hypercholosterol-aemias. The risk of I.H.D. has been shown to increase with cholesterol level. 117 Even more, perhaps, than with other continuously distributed variables the predictive value for myocardial .infarction of the serum cholesterol in the individual is low. (6) Blood pressure: There is a firm relationship between high blood pressure and I.H.D. The Framingham Survey shows that the. risk of developing I.H.D. increases as the level of the blood pressure (both systolic and diastolic) increases) the observed incidence of I.H.D. being not far short of twice the expected in persons with definite hypertension (systolic B.P. over 159 mm Hg or diastolic B.P. over 95 mm Hg). 117 When left ventricular hypertrophy is present the ratio was more than trebled. (7) Body .1<reigi1t: The independent contribution of overweight as a factor in I. H. D. is not supported by the Framingha~ Survey. 117 The Framingham work and - 91 - also that of Doyle128 2nd others - but not that of Oglesby Pau1116 - does show a positive correlation between I.H.D. and gross overweight; but it appears that this contribution to morbidity is mainly due to the cross- correlation of overweight with serum cholesterol level and blood pressure. Excess weight is also associated with diabetes mellitus which in turn increases the liability to I.H.D. Again, the association may be indirect. However, since overweight is more readily identified than blood pressure or cholesterol level, weight remains an important correlate (even though indirect) for I.H.D. Apart from body weight itself there is certain evidence that body-build may be directly correlated with I.H.D., in that mesormorphic persons are at higher risk than exo- or endormorphic. 129 Spain found a higher prevalence rate for I.H.D •. in endomesomorphic persons which was not altered by excluding h t . . d" "d 1 130 p 116 yper ens1ve 1n 1v1 ua s. aul while not:i.ng n.:-' significant difference .in weight between I.H.D. patients and controls, did find a signifi- cant difference in skinfold thickness, the I.H.D. patients having a higher fat content. (8) Smoking: There is ample evidence that cigarette smoking is in some way correlated with an increased risk of I.H.D. The study of Doll and Hilll3l of British doctors and the prospective surveys of Hammond and Horn# 132 Dawber and his associates133 A.t Framingham, and of Doyle and others.134 as well as the follow-up of the San Francisco 1ongshoremen28 all support this conclusion. The important conclusions are: (a) that the excess risk of I.H.D. increases with the number of cigarettes smoked5 (b) no relationship between risk and the duration of exposure to cigarette smoking has been shown; (c) there appears to be no excess risk from pipe and cigar smoking? (d) there appears to be a rapid reversion of the excess risk of I.H.D. to normal in persons who have given up cigarette smoking. No causal factor in cigarette smoke has been identified but it appears probable from these findings that the excess risk of I.H.D. may pe due to the repeated adrenergic effect of the nicotine content of cigarettes. The Framingham workers estimate that, if cigarette smoking ceased, mortality and morbidity from I.H.D. might be reduced by almost one half. 135 6.2.3 - 92 - Towards the end of 1965 a WHO Working Group met in Prague. 136 The group considered the present possibilities for preventive measures in I.H.D. and outlined the high-risk factus, in order of importance, as: Conclusions high serum lipids high blood-pressure diabetes mellitus cigarette smokjng overweight stress We have seen that more studies have been made into the epidemiology of I.H.D. than into most other non-communicable chronic diseases and that a considerable amount is now known about· the early natural history of the condition (though tantalizing gaps in our knowledge of causation still exist). Much is also known about the available early diagnostic tests~ (though we do not yet know the sensitivity and specificity of the cardiac questionary). We should now ask whether, in the light of this knowledge, enough is known with sufficient accuracy about the natural history of early I.H.D., the tests for its detection in the individual, and its treatment, to justify case-finding programmes. i.Ve have noted above that the practical possibilities for early detection and treat- ment have been recently reviewed by vmo. 136 At present we must perhaps regard the value of early diagnosis in I.H.D. in a tentative manner in view of our ignorance of causation and of methods of prevention. However, there appears to be at least three risk factors which can, through individual effort be reduced~ cigarette smoking, overweight and sedentary habits. It is possible that, with the development of a simple cardiovascular questionary, combined with electro- cardiography, blood-pressure recording, chest X-ray and weighing in older persons, a particularly high-risk group of the population could be identified for which there would be a special incentive to stop smoking, reduce weight and increase physical activity. At least one trial on these lines is in progress. The other line of prevention which has received much attention is the reduction of serum lipid levels by dieting. At present firm information is lacking on the value of diets low in saturated fats and/or with added poly-unsaturated fats in preventing myocardial infarction. However, this is regarded as a reasonable clinical preventive . - 93 - measure in persons at particular risk and it has been recommended on a large scale (e.g. in Norway~ in Sweden by the Swedish Board of Health; and in the USA by the American Heart Association). 1Jhether or not it is justified on present knowledge as a general recommendation 7 alteration (where necessary) of the diet in persons identified in the population through screening measures as being at particular risk may well appear reasonable. In conclusion~ the matter of communication with the public 1 and how this is to be handled 7 is of grea.t importance. A health education policy that presents the public with a reasonable and balanced view of the risks of I.H.D. and what can be done about them needs to be evolved at the same time as the epidemiological work is carried out and not left until actual case-finding is being widely practised. It could be disastrous if persons, knowing themselves to belong to a high-risk group were to take too gloomy a view of what could be done about this situation) it would be equally if not more 5 disastrous if peop~e tried radically to alter their way of living on the basis of inadequate epidemiological evidence. Though a jest9 Myers' "Thumbnail Sketch of the Man least likely to have Coronary Heart Disease" does point a lesson. 137 6.3 High blood-pressure Sphygmomanometry is usually carried out as part of a screening programme. It is of course of value to diagnose severe degrees of hypertension as early as possible. Undoubtedly cases of severe hypertension occur with absence of minimal symptoms though this must be the exception rather than the rule. Perhaps the most valuable reason for taking the blood-pressure is in the case of the young person with a high pressure in whom a. remediable cause may be present 7 like coarctation of the aorta or phaeochromocytoma. Even when the cause lies in the kidney remediable measures may well improve the outlook. There is evidence 5 too? that treating severe essential hypertension improves the prognosis and may reverse retinopathy:3S, 139 There is, :i10wever, the large popu.:..ation of persons havin;; mo~erately "raised" blood-pressure for their age and sex in which the value of treatment has not been established. This is work 1-1hich needs to be done. It could well turn out that drugs 5 while lowering blood-pressure and marginally improving prognosis, were so unpleasant to take for the rest of a patient's life 5 · and induced such a sense of invalidism, that the overall benefit could not be considered worthwhile, despite the incontrovertible evidence that life expectation in the actuarial sense diminishes 1.-ri th rising blood-pressure. 140 The Rostural hypotension and sexual impotence induced by some modern drugs are both severe disabilities in themselves. - 94 - Apart from our present ignorance of the value of treating the early essential hypertensive (and of what criteria to adopt)~ there is the methodological difficulty of the blood-pressure recording itself. Two factors contribute to this difficulty~ (a) interpreting the level in relation to the circumstances under which the reading was taken~ (casual, resting, circumstances of the examination): recent work relating average casual pressures to average continuous recorded pressure in normal and hypertensive persons carrying out normal duties? indicates that there are wide departures from the average relationship in individual instances, and that the average recorded pressure is considerably lower than the average casual 16,141 pressure. (b) the observer variation in recording blood-pressure: To obviate this? sphygmomanometers are now becoming available which remove this 142 source of error. These points are clearly important in practtce since they can seriously affect the validtty of blood-pressure examination. Recently~ the subject of measuring blood-pressure has been well reviewed in the Lancet. 17 We do know that "high blood-pressure" is com:.10n~ increasing with age. There is argument as to whether transmission might be by a single gene or whether it is multi- factorial. Population surveys by Boe et al., 143 Miall and Oldham, 144 Hamilton et 145 146 J47 al,, Kagan et al., and the US National Health Survey- indicate a skewed normal distribution of blood pressure with no sign of the segregation of a "disease" group of hypertensives. The Framingham workers arbitrarily define "high blood-pressurerr as readings by two examiners of 160 mm Hg or over systolic or 95 mm Hg or over diastolic. By these criteria 391 out of 2 024 men aged 29-62 were hypertensive (19 per cent) and 410 out of 2 445 women in the same age-group (17 per cent). Using the same criteria the US National Health Survey found a total of 15 per cent males and 16.7 per cent females between the ages of 18 and 79 years had blood-pressure at or above these levels. The numbers with high blood pressure rose with age in the proportions shown in the Table 6.1.1 (from 147). 105 Epstein and his collaborators in the Tecumseh Survey, found overall a. prevalence of hypertensions heart disease (HHD) of seven per thousand males and nine per thousand females (defining HHD as a systolic blood-pressure of 160 mm Hg or more or diastolic pressure of 96 mm Hg or over, or both together with suggestive evidence of left ventricular hypertrophy). As would be expected, the prevalence rises steeply with age and was higher in females than in males; for example, in the decade 60-69 it was 39 per thousand males and 71 per thousand females. - 95 - We do not yet know whether treatment would be effective and worthwhile in improving prognosis. It is difficult, as Pickering148 has pointed out, to compare the prognosis in hypertensive patients with that expected for the whole population because of the almost inevitable selection in the case of the hypertensive series. Bechgaard,149 however, has compared a group of persons witn raised arterial blood-pressure with the population of Denmark and shows that the mortality for men in the group 40-49 with systolic BP over 180 and less than 200 is nearly five times that of the whole population. (For women it is only 1-1/2 times as high as normal.) The records of insurance companies, too? show the relationship between levels of blood-pressure and mortality. There ·are, therefore, good grounds for treating mild essential hypertension provided it can be demonstrated that this is effective at the lower levels of blood-pressure and does not interfere unacceptably with the normal enjoyment of living. What is evidently needed is a controlled trial of early therapy to answer these questions, including that of the acceptability of long-term drug administration in the absence of symptoms. 6.4 Overweight 6.4.1 In multiple screening programmes it has been usualto record the weight, height e..nd determine, by reference to standard weight tables, whether a person is 11 overweightn. If so, he is told of the risks of overweight and advised about weight reduction. This advice is normally based on the judgement that the excess.weight is due to obesity, since above-average muscle or bone mass may not be associated with a reduced life expectation. Life tables, like those published by the .American Society of Actuaries, show that expectation of life falls with increasing degrees of overweight when weight is standardized for height, build, age and sex and that expectation can in fact be improved by weight reduction. 15° The main risks associated with overw_eigh.t are .cardiovascular disease (only with gross obesity, accordinng to Doyle et al., 129) diabetes, chronic ·- .. -~' respiratory disease and degenerative arthritis. But, as with other continuous distribution conditions, there is difficulty in deciding what should be regarded as overweight~ anything from 10-30 per cent above the mean has been regarded as the lower 151 clinical limit in different studies. Kemsley and his co-workers suggest the upper and lower quartiles as reasonable practical limits, for overweight and underweight. An important factor, not fully allowed for in most standard tables, is the relative adiposity .. Measuring this adds to the complication of the examination, though it may be done relatively simply with an instrument like the Harpenden Calipers. 152 - 96 - Another difficulty about using tables is that normal weight (and~ of course 7 height) for men and women varies from country to country and within the same country at different periods of time. For example, comparison shows that the average height and weight of persons in the USAl53 has been increasing during the present century and US weights (for heights) were greater than those for some other countries (Canada and Norway). In the CCI Baltimore study, 154 in the physician evaluation of a sample by physician interview, the prevalence of obesity as judged by observation was 129 per 1 000 popula.tion~ or nearly one in eight persons. Of the sample submitted to screening 40 per cent of males were 10 per cent or more above the Canadian normals used 9 18 per cent were 20 per cent above and eight per cent were 30 per cent above. As the report says? 11 the high prevalence of obesity •.•.•• points up the need for further research to clarify the relationship between obesity and health ..•..• The size of the problem suggests that the approach may need to be directed to the public at large as well as to the individual obese patients". Strong motivation is indeed needed to induce obese people to reduce their weight by dieting, as those who have had clinical experience of this particular problem know. To some extent this recommendation, made in 1956, is beginning to be translated into action. Norway, for example~ has recently issued a public statement, through the Norwegian Board of Health, advising moderate dieting~ primarily aimed at lowering the intake of polyunsaturated facts but also recommending keeping to a moderate total calorie intake. The problems of definitions and of weight control in obesity were recently reviewed in the British Medical Journal, l55 when the need for further longitudinal studies of the effects of treatment was pointed out. There is no doubt that weighing is one of the simplest and most precise screening tools available and that the information it gives is important for health~ what we still need to learn is how best to act on the information it gives. 6.5 Respiratory diseases 6.5.1 Pulmonary tuberculosis 6.5.1.1 Whilst deaths from pulmonexy tuberculosis began to fall dramatically with the introduction of streptomycin and chemotheraphyl56 notifications have fallen at nothing like the same rate and indeed~ in the older age-groups have even tended to rise. This effect may be attributable to improvement in methods of detection and especially to MMR. Having reached low levels in developed countries there is a tendency for the notification rate to level off. In England and Wales at the end of 1963 there were 340 000 persons under supervision for pulmonary tuberculosis, a ra.te for the population of about 0.75 per cent. The Director-GenerRl of WHO has pointed out that not a single country had reached - 97 - the poi11t qf c_o1;11:,rol wher@, there is less. than one per cent prevale11ce of nfJ,j:;ural. reactors .:,,~ ' ~· '> ' . "- '. - . . to tuberculosis a.moi1g cll.ildrE?n in the 14 year age~group, a cri:ter;i.pn~et by 1;1. previous 1.<JHO expert comm:l_·ctse as indicating that tuberculosis is no longer a serious _public health problem. 6.5.1.2 Pulmonary tuberculosis is perhaps the classical condition for early, often pre-symptomatic, detection (as well as prima.ry prevention with BCG) and it .meets the screening criteria well in that - 1. ~lmonary tuberculosis is an important public health problem. 2. · Facilities for diagnosis and treatment should be ava.ilable, otherwise there is . J:lO benefit only harm, from screening. : 3. The natural history of the precursor stage of the disease he.srbeen' elucidated? early_ infiltration does lead -to. overt disease. 4. There is a recognizable lat~ntstage (positive tuberculin reaction and · infiltration). 5. Suitable tests are available, the tuberculin reaction and mass miniature radib~faphy··· (MMR). 6. The tests are acceptable to the population . .. :·.; .' 7. There is .an accepted (and effective) treatment. . - .. "·'· . '' . . ,~·, .. P~}i',~,0ns with9ut rec_ognised disease (small f~~:r:_osed infiltrations) are :qot treated .as pat~ents. ·:.i..''.' 9"'. ,Th~ cost is t.olerable where MMR is used, by the saving in cost of treating a flqrid. case of tuberculosis. 10. Long-term·follow-up is a built-in part of schemes for the control of tuberculosis.- 6.5.1.3. This brief summary makes no .attempt to review the bulky lite'rature on ·the early detection of tuberculosis. · The subject has been reviewed recently by a World Health Orgai1~Za,1;.ion Expert Committee.~57 Only a few salient points will be discussed: . . 1. Unlike cancer there ic evidence that small tuberculous lesions in the lung equate with early lesions. 158 In a five ·year ·p;ospective survey of civil servants and other workers in the London area the size of lesion'at first X.;..:rayhas proved of the greatest prognostic significance. - 98 - 2. Small infiltrative lesions can only be ignored at peril; persons with these lesions are in fact in need of treatment and between 70 per cent. and 80 per cent. will have further trouble if the lesion is ignored. 3· Proof of the effect of early treatment: this is very difficult to demonstrate since it would be necessary to show that the rate of fall in registrations was influenced by earlier detection and that a significant change in the slope of the curve in a downward direction had occurred. What has in fact occurred as a result of introducing M.M.R. (and this has been seen best in intensive surveys such as those of some Scottish cities) is an early rise in registrations at the time of introduction~ followed by a compensatory fall in registrations afterwards; and with a third relatively rapid swing back to the level of the falling base-line after which the fall in registration continues at its "original" rate (Figure 6.5-1). Thus in England and Wales in 1938-39 the mean annual notification rate per 100 000 persons was 88; in 1947-50, following the impact of M.M.R. it rose to 100 per 100 000 and this was followed by a l5ri fall in 1954-55 to 80 per 100 000 persons • ·· In the comparison carried out in two Welsh mining valleys the fall in registration rate was practically the same after the completion of the M.M.R. Campaign. 4. As with other tests radiological screening sets the problem of sensitivity and specificity~ linked with that of observer variation. Yerushalmyl59 found a false negative rate (a missed lesion of some kind) of nearly one third in addition to a false positive rate of about 2 per cent. By dual reading one third of the lesions missed by a single reader could be picked up~ but this need necessarily adds to the cost of case-finding. 5. As radiological screening progresses the number of persons needing con- tinuing supervision increases cumulatively; few are discharged but additions are always being made. This need for arrangements to keep the intermediate type of patient under surveillance of course adds to the expense. The experience of the Danish Tuberculosis Index160 is, that for every one person needing treatment between five and 10 will need to be kept under observation. Similarly, Styblo 161 has some 20 000 - 30 000 persons under supervision in Czechoslovakia~ in whom the natural Vl c: 0 -~ -Vl C'l ~ ....... 0 ..... Q) ..0 E ::;:) z Fig. 6.5-1 t Time----. Introduction of M.M.R. WHO 70C'IO history of the border-line lesion is being followed (thus critari6n "3" · · above, has not been completely met,; there are early pulnlonary stages · · which require watching rather than treatment). 6. Economics: (a) In a developed country the yield from M.M.R. rapidly declines and it then becomes a matter for decision as to whethe.r to make it . . . . .. .. ~ ; ' more and more selective; or to continue offering it to the whole a,dult population in the expectation that the rare missed case may otherwise go '. . . . longer undetected tharl had M.M.R. nev;er been introduced, (beqause .of the ._.:. very rarity of the disease)#. thereby prolonging the danger of contact cases occurring. One alternative is to make M.M.R. selective for high' risk . ·. -~ , ;.'. --~-c;·~~:· ... · .. :.::c.·:.~·-· ·groups in the population (the adolescent, the elderly, the economically depressed# person's. exposed to special' industrial or professiona.i 'riskS:. persons with cough); or to link M.M.R. w~th ~antoux tes_tinE. ~<? .. that all contacts of Mantoux positive children are X-rayed. ;-;, ·~. ' . In countJ;>ies .. where the co:;;t of a. oountrywide service might be prohibi:. ·•···· ·. . • • . . ·' . • ~ . .l... . • . • . • • . tive and where m~dical services woul,<i not be adequate to carry-out extensive follow-up of border-line cases. In the first place~ it goes without saying that the prerequisite for a detection programme is an., a.deqlla.te mecii~al service for the definitive dle.gnosis and treatment ancl.afte;t'- .. . .. . . : ·' care of the. patient. There is clearly no point in diagnosin_g if facilities for treatment and the prevention of contacts are not available .• Secondly, the stage at which tuberculosis is diagnosedmay be varied according to· .· :' :··resources. M.M.R. is expensive and, as we have noted, produc~s ~"large number of border..:li~~ patient~ f~r costly follow-up. There is a case, now that chemother~py.''k{cl.. antibiotic treatment is so. effective, for pas.sing over the earliest stag~~ of the dise~eand endeavouring only to detect. the f) • .•. ··.. 162 condition when the sputum has become positive. Clearly. there are .. '".·,-:· -· .:.c-~ risks of contact c~es associated with this technique but it is. perhaps the •• r -.. •', ··,• ' t, . . . . .. ' ~· ,.i~ ... . ~ . -- - - most economical secondary preventive step available and therefore capabl:.e of. being widely used. ·,-, ... '·' ;.·•' :".;. i_.: ... t . ~. ' . - 100 -· 6.5.1.4 In sunimary, the words of a memorandum of the Chief, Tuberculosis Unit, WHO, Geneva (1/2/65) :put the world position on screening into perspective: "·••••• symptoms provide a much more selective criterion for achieving high yields than other epidemiological screening criteria, single or combined. ·. 6.5.2 ••••••• where pre-symptomatic case-finding is considered justifiable and feasible, screening procedures dictated by the epidemiological situation can increase the efficiency of a community examination, but the latter's contribution to the total case yield will always remain marginal." And" ••••••• follow-up (after screening) may well be the most important aim from the application of screening procedures, as there are indications that it yields, in the long-run, quite substantial returns." Non-specific r.espiratory disease Chronic non-tuberculous disease of the chest presents rather different problems from those discussed under tuberculosis. By far the largest contributor to this category of illnesses is chronic bronchitis and it is with this that this section will be concerned. There is now overwhelming evidence that the prevalence of chronic br<?nchi tis is associated with the twin factors of atmospheric pollution and cigarette smoking; macro- and micro-air pollution. 163 Wide international differences in mortality and prevalence are found and much of this appears to be related to the factors first mentioned. Undoubtedly, also, the diagnostic customs in the countries concerned play their part. For example, the mortality attributed to bronchitis in England and Wales in 1958 for all :Persons was 65 per 100 000 compared with a rate of only 2.3 per 160 000 persons in the. United States of America (white). However, it is doubtful whether differences in diagnostic habit or coding of the cause of death could account for such large differences. Comparative studies of pathology are now in progress~ in Chicago and London and these should help to resolve these ··,( doubts. As an example of the type of epidemiological work needed in order to understand the real meaning of international differences, the study of Mark of comparative prevalence of respiratory disease in England and Wales and in Norway should be - 101- mentioned. 164 Mork foun<i that, while the prevalence of minor symptoms was nearly the same in the two countries, the prevalence of severe symptoms of chronic respiratory illness was considerably higher in England and Wales. Studies of similar working groups of postal and telegraph workers in London, rural England 165 and United States cities have been carried out by Holland, with comparable results to those of Mork. In England and Wales chronic bronchitis appears to be more prevalent than in 166 any other country. A survey by the College of General Practitioners showed a prevalence of 17 per cent. in men aged 40-64; and in an earlier survey bron- chi tis was found to be the commonest re.ason for consulting a doctor in general practice (260 consultati.ons per 1000 patients per year), second only to the common cold in accounting for numbers of patients consulting their doctor.97 Though much has still to be learned about the aetiology of chronic bronchitis, a great deal of epidemiological work has by now been carried out which has enabled us to identifY some at least of the precipitating factors. Use of this knowledge, largely by abating atmospheric pollution and by persuading people not to smoke cigarettes, should go far towards preventing the disease. However, it will take a long time to achieve these aims and in the meanwhile many young persons are insidiously developing chronic bronchitis. We should try to discover what might be done to arrest the course of this pathological process and, if possible, to reverse it. In the absence of a specific aetiological agent or of specific pathology the assessment of chronic bronchitis has to depend on recording symptoms and on measuring functions like the forced expiratory volume (F. E. V.) or peak expiratory flow rate (P.E.F.) and on sputum volume. Records have proved their worth in cross-sectional epidemiological studies. Work is now in progress on prospective studies aimed at identifYing personal factors of susceptibility to chronic bronchi tis. Preliminary evidence 167 seems to incriminate atmosph~ric pollution at a much earlier stage of life than has been previously recognised. If it proves a practical proposition, by using a short respiratory questionary and a simple measurement of lung function, to identify a high risk group of individuals at a very early stage of respiratory function abnormality, the next stage would be to conduct a trial of treatment. In this particular case the most effective treatment we know is stopping smoking and a - 102- comp~rLson between early chronic bronchi tics still smoking and h'lving stopped smoldng shou.l<i give helpful information abbut the reversibility of the condition. A d'i.fficu}ty wc.u.1..d certainly be thCJ.t c.f cbta:lning test g:!:"'oups compe.re.ble in other respect-s. At the same tj_me, nabn"al~.y, it wcJt.~ld be useful to study the effect of g:t.ving up :::moking on the incid.anee of ischaem:i.c lnart d:Lsease, as referrE:d. to in the sect.ion of this paper which deals "tv·ith that condJtion. Studies of this kind a:re now getting vnder :·u:;;y but it will be a con::::;ide:t'ai)le tim0 before we can pt>rso:::.s c~n l:e identified who would speciaJJ.y benefit from ::nt·ms:i.ve preventive measures, over cmd above those normaJ.J_y <;.dvif>ed. li'inaJ.ly_, industrial risks should also be considered. Th8 P>~·0cise relation:'lhip b::;,t~'!ecn ccPtain industrial hazards~ for exrunple in tlJc stGeJ. i.nc.:ustry_, is not fuJJ_y u·.!.derstood. Her.: again work is needed to discovt>r h'"hcther a high-risk group exists Khich coFld b0 identified and advised en other em:;:>loym'?.nt~ A.t loa;:;t two studies of -:-::1.is kind are in progress in -~:1e United Ki.ngdcm and nv doubt 0+.hers are in progress Th0 need for internatic;J.ally compar£:ble nomenclature and meth~:io].ogy ih chronic non-specific Ju."'lg disease, nnd for further epidemiologic:;.}. r.:;tucl5.cs..., ·.was discusos:i at. .-- 168 a. Hl:IO European Symposium held in Moscow in 19o2, We do not have accur·atc morb:i.di ty data on lung c.:.ncer ~·Jhich indicate the :2:tzE: of the problem. Ravertheles:::. c>.n idea can be. obtained from the general prnctic8 Durvey carried ou·:~ by the College of G~mera::_ Pract:i.t:ioners -:tnd ·c-,~ Gcn8ral Regl;;;.ccr Office ir~ Engl3.!1cl '"'nd Wale:::; in 1955·~56, s:i.nce i·~ i3 likely that all. paticmts Hi th lung c~cE-~r consulted ·::.heir family doctor ;J.t least once d.ur).n::; the observation yca.r. The rate for 11 pn/~ients consulting11 for.- neoplasm of tho lung., bronchus &nd tr·achea is Oo5 pe:::> t..lJ.ousund pe;rsons (1 .. 0 per thotJ..5c-.!ld fO.i" n:::t.lf".l3 &'1<'!. O.l per thousand fm.' fe;;12.les); for men betiJeen the age.:; of 45 and 64 :l t JB 2.1 per.- i.;hou;.:;r.nd. In thi:::; high-risk group luns cancer is found, thc:J:C!fc:::'e, at a rather lo;•J8r prevalence ro.tc ths.n is ca."t'cinoma-il1-si tu of the ut;::rina c0rvix in r,lJ. adul +. v;ome:::J. (3 per thous~.nd)" Hc:·rdver, lung cancer is so lethal. that deA.tbs frcm this nonditicn ['.ro more· than fi•re - 103- time.s as .. frequent as those :from cancer of the cervix. Moreover~ deaths have been increasing annu~ly at an alarming rate in western countries~ associated with the earlier increase in cigaJ."ette smoking; in England and Wales deaths in males have increased by nearly two thirds between 1954 and 1964~ having risen from 14 000 to 24 500 malignant neoplasms of the trachea, bronchus and lung (6.6-.1). Such is the nature of lung cancer~ therefore, that if early detection can be made effective~ mass screening of the adult population in highly-developed countries would be indicated. Unlike the accessible cancers, (cervix~ lip and skin) lung cancer is customarily at an ad.V:i:Uloed ·stage when diagnosed with the help of X-rays and t..,_e prognosis is nearly always bad.:_ The corrected survival rate at five years for males in England and Wales regi5.tered during 1945-1947 was only 14 per cent. for. early cases sub- mitte<;ltoradical treatment {6.6.2). Forall early cases) whether treated or not, the fiye . .,.year. survival· rate was only two per cent. Seven out of eight cases were in men and of these only. 13 per cent. were classified as early at diagnosis while nearly one quarter had metastases. The median duration of· symptoms Nas about six months for the early cases but only four and a half months for those already with metastases. The five-year survival rate of neither early nor late cases treated radically was affected by the duration of the symptomatic history~ being for the early cases 12 per cent~. with symptoms of from 0 to 2 months duration and 10 per cent. for personswith a 12 months and over history. This is a gloomy picture but it is necessary to remember that it is based on persons developing the disease twenty years ago~ Since then mass miniature radio- graphy (M.M.R.) has been practiced widely and it is necessary to study its effect in combination with advances in thoracic surgery. M.M.R. was, of cOurse, developed for case-finding in the epidemiological control of pulmonary tuberculosis. With the decline.of tuberculosis in many countries attention has turned towards the possibility of using M .. M.R. more for diagnosing other lung conditions, of which cancer is the most important~ rather than just for seeking out tuberculosis. We need therefore to see as clearly as possible what is the evidence for benefit so that a sensible policy may be evolved. Four surveys will be considered; that of Posner, McDowell and Cross in Birmingham; 171 Cuthbert 1 s · review of the Glasgow X-ray c~paign; 172 - lo4 - \vaddington 1 s comparison of the Liverpool survey with his own and Gifford'' s routine hospital admissions; 173, 174 and Boucot, Cooper and Weiss' experience with the Philadelphia Pulmonary Neoplasm Research Project. 175 6.6.1.1 (1) Birmingham: Posner and his colleagues analysed all cases of lung cancer diagnosed by M.M.R. units in the Birmingham Hospital Region during one year, 1955-56, 238 in all being investigated. As with the Glasgow series patients diagnosed by 11 conventional11 M.M.R. {"routine M.ri.R. cases") were compared with patients referred to the units by general practitioners. There were rather more older patients in the general practitioner referred group. At one year the survival of the routine M.M.R. group was better than that of the general practitioner referred group - 50 per cent. compared with 36 per cent., though this difference is, not significant. More routine M.M.R. cases proved to be resectable than were the g.p. referred - 44 per cent. compared with 30 per cent._; also there was a higher proportion of lobectomy as opposed t~ pneumonectomy in the routine M.M.R. group, which is usually considered to equate with a b~tter prognosis. As with the Glasgow series 85 per cent. of the routine M.M.R. group had had symptoms at the time of presenting for examination. These workers concluded that the smallest cancers were very easily missed and recommended the selective screening of men over 35 referred by their own doctors, irrespective of symptoms, to static 100 mm camera units. They considered it would be a major mistake if a relatively costly programme of this kind were to be allowed to deflect attention from the importance of primary prevention. 6.6.1.2 (2) Glasgow: 48 patients with proved bronchogenic carcinoma were found through the operations of one of the city chest clinics,and these patients.were compared with 48 consecutive patients referred to the same clinic by general practitioners. The average age of the two series was about the same with a similar range, most of the patients being between 50 and 65. Of the 48 M.M.R., cases 36 were found to have had one or more of the cardinal symptoms of chest ill- ness; but on several counts the disease in the patients referred by general practitioners was more advanced. As might be expected, the patients in both series who were considered suitable for surge~ had tl}e better survival at 18 months; but the M.M~R. group did better than the g.p. referral group, 13 of the M.M.R. group surviving at 18 months compared with only eight of the g.p. group. - 105- 6.6.1.3 (3) Liverpool: Out of more than 450 000 persons over the age of 15 X-rayed in the Liverpool Campaign of February/March, 1959, 235 were admitted to surgical wards for investigation. Of these, 163 were suspected of suffering from bronchial neoplasm, a rate of 0.36 per thousand persons examined (0.5 per thousand males) which is somewhat lower than the general practice survey rate quoted above. Of the 163 suspects there were 118 proved to have a primary bronchogenic carcinoma and Waddington compared these with his earlier series of patients admitted to the Liverpool Thoracic Unit. 80 out of the 118 (68 per cent.-) from the Liverpool Campaign were resected, 90 per cent. of all those surgically·explored~· This is a rather higher rate than was found in the hospital series where o_nly 70 per cent. of those explored c-oUld be resected. They were also able to treat a higher proportion: of the Liverpool Campaign patients by lobectomy (as opposed to pneu- monectoiny) than were so treated in the hosp:ltcil series. HoWever, at the time of the report, while only. 61 per cent. of all the 'operated patients fwho had survived the first two ~onths) had survived one year 67 per cent. of the hospital· series had so survived. There 'seems therefore to nave been no advantage .. at the time of writing, for those -diagnose-d presumptively earlier by M.M.R. 6.6.1.4 (4) Philadelphia: In this well-known project 6137 men of 45 years of age or more were enrolled in an experimental prospective survey with the aim of following them by means of six monthly 70 mm chest X-rays and a short medical history. A previous M.M.R. campaign in :Philadelphia had shown a prev'a.lence rate for lung cancer of 2. 7 per thousand men over 45 which is rather more than the Coll~ge of General Practitioners su.rVey :found in a similar age group. During the course of the study, 26 men developed lung cancer, in whom no radio- logical evidence of neoplasia had been discovered on entry to the project. Only five were without symptoms up to the time of the first positive X-ray and only two were asymptomatic actually at the time of the examination. The other finding of direct interest is the survival: only two out of the 26 had survived to the time . _, ·- of writing the paper, which could have given a maximum survival time of seven and a half years, though there is no statement as to actual survival times. From these surveys it is clear that the prognosis of lung cancer is little influenced by detection by routine radiography and that, in fact, most of those who are detected in this way are symptomatic at the time of their X-ray. It seems - lo6- probable that~ as at present carried out~ routine chest X-ray at any interval greater than six monthly would be of little use. More frequent x~ray examination would not only be uneconomic but would also pose problems of persuasion to attend and of possible harm from the frequency of examination. The best present use of routine radiology is likely to be for selectively screening the middle-aged person~ particularly males, with a persisting cough. The examination of sputum by exfoliative cytology has also been employed as a means of detecting early cancer of the lung. Compared with exfoliative cytology for cervical cancer there are drawbacks. The bronchus is not accessible in the way the cervix is; and examining the sputum is considerably more time-consuming than examining cervical smears if it is to be at all acceptably reliable. At ,Tohns Hopkins Hospital 176 the positive sputum rate in patients with bronchial carcinoma rose from only 20 per cent. when one sputum was examined to 56 per cent. when five specimens (3 smears of each~ i.e. 15 slides) were examined (or from 42 per cent. to 95 per cent. when suspicious reports were included). In another series of 144 177 patients with suspected lung cancer at St. Bartholomew's Hospital,- · London, there were 10 per cent. false· negative results from examining 3 smears from one specimen of sputum. The time needed to achieve this degree of accuracy was unacceptably long. A rapid ·sputum cell concentration technique is needed to shorten examination time and work on these lines is proceeding. Lilienfeld 178 has reported on a comparison between sputum cytology and radiology in the early detection of lung cancer in persons living in u.s. l~terans Administration homes. Up to 1960 over 12 000 persons, aged 45 and over, had been subrrdtted to a 6 monthly X-ray and sputum cytological examination at least once and some 4000 had been screened between three and four times. Of 43 cases of lung cancer diagnosed by follow-up, cytological screening contributed to the diagnosis in 15 and this diagnosis would have been missed if X-ray screening alone had been used. On the other hand, if cytology alone had been used, 21 out of the 43 would have been missed. Unfortunately survival at six months was no better than previous experience where repeated screening was not carried out. It appears, therefore, that sputum cytology~ while it can add to the diagnostic accuracy of X-ray examination alone, is not atcpresent able to offer an improvement to the prognosis of lung cancer and would not be economically justifiable. - 1-07 - .Clearly -the prevent:ton ·of lting cancer would be far better than early detection and .it is interesting that" the American Cancer Society and Veterans Administration Study just refepred to showed that one per cent. of-present smokers had positive or suspect sputum cytology as compared with 0.35 per cent. of those who had never smoked and 0.47 per cent. of past smokers. 6.6.2 Cancer of- the cervix The evidence for the value of detecting cervical cancer early is relatively strong. The earlier the stage at diagnosis the better is the survival rate. It is reasonable to suppose that diagnosis at the pre-invasive, carcinoma-in-situ, stage would very greatly improve on existing sUrV-ival though it is as yet too soon for . . .. ·. statistical data on this point to have accumulated, nor though mortali~y from cervicar ~~6e~- is in general decreasing at a slow but steady rate in advanced countries, is there evidence of a reduction in the death-rate directly attributable to diagrl.osis and treatment of the pre-invasive lesion. It is, indeed, probably too soon to expect to see this effect, since intensive commUnity screening for can.cer of the cervix has only been practised for a few years; and that only in a relatively few centres. A significant fall in the incidence of invasive cervical cancer has however been observed in British Columbia since the introduction of widespread screening by exfoliative cytology. Unlike other pre-symptomatic conditions- the natural histo:cy. of pre-invasive - .. cerv.ical canc~r has been reasonably closely studied. Firstly, there is relatively good evidence that the in-situ lesion pre~-~des and turns into invasive cancer of the-cervix. The direct evidence for this is both retrospective179 and pros~dtive~lSO, 181; 182 >. 183 There is also indirect evidence based ori the age- -.. . . 184, 185 specificdistl:-ibution of in-situ and invasive cancers. Secondly, estimates are available of the proportion of in-situ lesions which. b . . 180, 181, 182, 185, 186 t t . l f ecome ~nvas~ve though these es ima es vary w~de y, rom one quarter to two-thirds. Another important feature of the pre-symptomatic lesion about which there is admittedly incomplete information is its duration. Dunn187 has done useful work on this point, based on his study of the U.S.P.H.S., Memphis, Tennesee, survey. : ' . . 188 He and his colleagues now estimate __ an average duration of aboy.t 10 years - 108- (calculated from the age-specific prevalence and incidence rates). This accords reasonably well with the estimate of Boyes, Fidler & Lock, 185 of 12 to l3 years, based on mean age of onset of both in-situ and clinical invasive carcinoma. Some of the picture presented by pre-invasive cancer of the cervix has therefore been filled in, but there are still important gaps in our knowledge. We need to know about the effect on mortality as quickly as possible since it is just possible that incidence could fall without a reduction in the death rate. To obtain this information mortality data related to numbers of examinations in populations screened heed to be collated. It is also important, for practical screening, to discover how frequently cytological examinations need to be carried out. To learn this it is necessary to note the time elapsing between the last negative examination and the first positive in as large a number of reported screenings as possible, the screening interval being varied in different groups of women. In this way it should be possible to construct a frequency distribution of the rate of progression from negative, through in-situ, to invasive cancer in the small population of women in whom this progress is relatively fast. Depending on the shape of this distri- bution the optimum screening interval could be selected. Registration of cyto- logical examination and linkage with invasive cancer registration is one way in which this might be done. There are other practical problems associated with population screening, the solution of which depends on epidemiological knowledge. One of the problems is that of ensuring that examination is offered to women at greatest risk. The evidence shows that incidence of cervical cancer increases with age; that it affects parous rather than multiparous women; that early coitus is an aetiological factor rather than parity; that there is a sharp s0cial class gradient with the higher incidence in the lower social classes; and that there are marked cultural and geographical differences. 189,l90 It is important therefore, in organizing population screening, to take steps to ensure as far as possible that parous women in the lower socio-economic groups are not only offered, but accept, examination. This means a 11 free" service and organized health education along lines planned from prior attitude studies. Without this approach it may be doubted whether merely providing facilities will in practice get to the root of the matter. In the United States of America, - 109- California State Health Department has adopted this approach in Alameda County52 and, more recently, a National United States Sample Study has been carried out. 53 In the United Kingdom an attitude study is in progress in Manchester.Slf A recent development, at present still in the evaluation stage, shows signs of overcoming the difficulty of persuading women at high risk to submit to cytological examination. This is the irrigation-pipette, 25 a plastic pipette which can be inserted by the woman herself into the posterior fornix region of the vagina. The pipette contains fixative which is eXPelled into the vaginal pool and sucked back together with exfoliated material. The pipette is then placed in a container and mailed to the laboratory where the material is centrifuged before making smear preparations. Davis has reported as much as a 90 per cent. acceptance of this 2") technique among 11 semi-indigent" women in King County, Maryland; ·· and a similar acceptance rate is being found in general practices in /1_ ·)srdeen, to give only 101 two examples.- J-· With the general acceptance of the value of cytological screening for uterine cancer attention has naturally turned to the possibility of developing automated techniques. Two may be mentioned. 6.6.2.1 A number of workers are examining the value of estimating the 6-phosphogluconate dehydrogenase {6-PGD) level in vaginal aspirate as an index of the presence of malignant cellso So far this technique has been proving too unreliable for case-finding. Though apparently reliable as regards sensitivity as a test for invasive cancer of.the cervix, it has a false negative rate in the region of 50 per 192 cent. for carcinoma-in-situ. · The false positive rate is also high, at between 20 and 40 per cent. 192,193 The value of varying the technique is now being examined. · 6.6.2.2 The development of an automatic electr9nic scanner of preparations of vaginal cells has been in progress since the 1950's. The "Cytoanalyser" has not so far been developed into a practicable instrument but work on these lines is advancing. There are numerous possible uses for an instrument of this type and more than medical interests are involved. The cost of developing a prototype instrument is considerable and there might well be a case for co-operative effort in this field. More recently, the possibility of detecting cervical cancer cells in vaginal - 11'0 - aspirate, using the Coulter Counter, has been reported. 194 being o~itically investigated. The technique is still 6.6.3 Breast Cancer 6.6.3.1 In England and Hales carcinoma of the breast in 1964 accounted for 20 per cent. of all female cancer deaths and for nearly four per cent. of all female deaths. 195 It is easily the commonest cause of death from cancer in the female and mortality has remained steady from year to year regardless of treatment. Thus the death rate for all females in England and vJales per million living (standardized) was 158 in the decade 1901-10 and in 1950-54 was 182. This trend is carried through all age-groups, as rates for 1936-1954 demonstrate. 196 Nulliparous women are more prone to the disease than women who have lactated and there is evidence that some tumours are hormone-dependent; in one group of tumours the peak registration age is reached at the ·time of the menopause and tl1is is followed by a temporary fall, though another group show a steady increase in the registration rate throughout life.197 6.6.3.2 There is no question, therefore, about there being a group at higll. risk involving large numbers of middle-aged women, at least in certain countries (some countries, Japan for example, have an extremely low prevalence of breast cancer).198 Unfortunately the results of treatment are discouraging; the corrected five-year survival rate for all stages in Encland and l"Jales, 1945-47, was only 37 per cent., though it was 67 per cent. for the earliest stage when radically treated.199 This form of cancer appears to a large extent to run its own course and to be not' greatly 200 influenced by treatment, whether surgical or radiotherapeutic. Park & Lees have estimated that, at most, treatment accounts for cure in no more than five to 10 per cent. of women. Lewison, 201 from the breast clinic at John Hopkins Hospital found little difference in survival when differing forms of surgical treatment, carried out t d .ff t t· . d d Be " R bb. 202 . 20 f 11 a ~ eren ~me per~o s, were compare • rg ~ o ~ns ~n a -year o ow-up of breast cancer at the Memorial Hospital, New York, considered the cure rate by surgery for anaplastic duct carcinomas was 12 per cent. 6.6.3.3 For a highly prevalent and lethal condition that tends to run its course whatever the treatment the question of early diagnosis, with a view to more complete - lll- eradication of the tumour and consequent improvement in the prognosis, assumes high importance. We should enquire wnether, if breast cancers can be brought to treat- ment at an earlier stage than is the present practice, by health education, frequent self-examination of the breast or by scre~ning by soft X-rays (or possibly by infra- red or ultra-sonic scanning), the prognosis is likely to be improved. Delay may occur for two main reasons: firstly, fear; ru1d secondly, failure to be aware of a small lump in the breast. The first of these reasons for delay is gradually diminishing pari passu with a rising level in general education. Better general education enables people to reason more clearly and. to plan mtional steps to meet a situation. More specifi- cally, ignorance and fear go cheek by jowl and health education, in seeking to over- come ignorance, at the same time aims to dispel the fear which can prevent a woman from consulting her doctor as soon as she notices something wrong. The second reason for delay, the fact that the lesion is smail enoUgh to escape attention at all unless looked for in some special way, needs highly organized mass scr~ening to overcome it. The methods are- costly, use up valuable resdurces~-both human and material, and involve the female public in a co-operative effort which cou*q._ .have bad, as well as good, effects on morale. Therefore, before advocating case-finding of this kind, it is necessary to examine the case for detecting breast cancer when the lesion is as small as possible. The question of the value of detecting the lesion when small may be examined in two ways. Firstly,· it is worth looking at some of the evidence for the benefits of di~osing bre~st cancer at an early stage in its course (which is linked with the grade of malignancy). Secondly, the evidence for the relationship between size of lesion and stage and degree of malignancy needs to be examined. 1. Prognosis by Stage: (a) There is a considerable correlation between stage at registration of breast cancer in the female and duration as reported by the patient (6.6-36). Nevertheless in this series of registrations in England and Wales in nearly 20 per cent. of women with a history of more .than two years the growth had neither spread nor apparently invaded the lymphatics; while in over 25 per cent. of those with a declared duration of less than one month the growth had invaded extra-mammary tissue - 112 - or produced distant metastases. Therefore length of history alone will not account for the facts; there must also be a wide variation in degree and nature of malig- 207 nancy between tumours. ~ (b) Survival: the average prognosis for breast cancer is poor: the Registrar General's Supplement finds, as we have seen, a corrected five-year survival rate of 37 per cent. and concludes that the chances of survival depend almost entirely on the clinical stage when treatment is begun and that, independently of this, neither the duration of the t~our before diagnosis nor the age of the patient ... .. 04 seriously affect survival. 2 Among others Bloom, in a recent report 205 has demonstrated, by careful grading of breas.t carcinoma according to histological type, that survival correlates well not only with stage but also with grade. The stage two cancers, for example, bear a f;ive..;year survival rate of 71 per cent. for grade I tumours down to a 26 per cent. survival rate for grade III tumours. 2. Relationship between size of lesion and stage and degree of malignancy: th . h b . 1 d 206 ~s question as een rev~ewed by Suther an • with small tumours experienc better survival rates. There is evidence that women Hat-fkins, 207 for example, in a series of over 3000 found a five-year survival of 86 per cent. when the lesion was less than 1.5 em in diameter; and Taylor end Wallace 208 have reported an inverse relationship between the size of the primary grotNth at operation and five year survival, varying from 89 per cent. when the tumour was less than 2 em in diameter to 18 per cent. when it was over 4 em. Small size, however, is not uniformly related to a good prognosis. I\reyberg and Chri~tiansen 209 considered, in a review of nearly 1000 patients, that small cancers did not have a specially favourable prognosis. They concluded that more than half of the patients who present with . ,_ tumours no bigger than 1 x 2 em are likely to die from their cancer. The evidencE! therefore supports the case for a degree of correlation, well short of absolute, betvfeen size of lesion and prognosis. Wh~t then can be said about the bverall value of early treatment? We have seen that, though littie affected by duration of symptoms, survival depends on ~e;' stage and tumour grade. The effect of age can be accounted for mainly by the later clinical stage at registration in these patients; and this in turh is associated· with the longer duration of history found in older women. There is also evidence that, to some ext·ent, the smaller the lesion the better the prognosis (though not to - 11.3 ... an extent that could be considered satisfactory). ' ' ' The conclusion is# therefore# that a proportion of patients would in fact be treated at an earlier stage if delay in duration of symptoms could be reduced; but this proportion may not be very high beci:i.tise' 56 many patients are found to have a later stage of cancer or a highly invasive type even when the history is very short. 6.6.3.lJ X-ray mammography For the above reasons it has been suggested that earlier diagnosis of breast carcinoma at a pre-clinical and perhaps impalpable stage should improve the prognosis. This is a reasonable hypothesis but one difficult to test. In addition to the development of diagnostic X-ray mammography a certain amount of screening of non- defined" pupui'ations.has been carried out. For example, Gershon - Cohen 210 and his cO-worke~s discovered 28 cancers in 1100 women over the age of 35 examined every six' months f'or· eight years. In a hosp~tal series of 2500 women with unr~lated breast pathology Egan detected .. 2ll 58 malignant growths. In order to determine the value of X-ray mammography in population screening a controlled trial on a defined population is needed, with a comparison between the results of standardized treatment in a group of women diagnosed by X-ray mammography and a second group in whom the diagnosis of breast cancer has been made in the usual way. A very large population is needed in order to provide adequately sized treatment groups. The annual incidence of carcinoma of the breast in England and Wales ~s of the order of 64 per 100 000 women aged between 35 and 74, so' that a popu:J.at_ion of this size, representing a total population of some quarter million persons would only yield individual treatment groups of 30 patients each, regardless "'' ?, .- of age, breakdown and histological type. Further difficulties in a therapeutic trial of this kind are its very long-term nature (follow-up fa~ a great many years is es::r~htial, with the consequent loss :to the trial from migration) and the large ainouht;of skill~dradiographic time and appari3.tus that needs to be available. IIdt-fever, a stil-vey o:f this kind is being Undertaken in the State of New York, in as~ocia£ion 'with the Hospital Itisu:r~ce Plart j~f Greater New York. 212 . - ·.· .. Until it can be discovered whether there is value in pre-symptomatic diagnosis in this way it wo~ld seem that X-ray mammography should be limited to its use as an adjunct to diagnosis. - ll4 .. - .Other techniques are being explored besides.· that of radiology. Infra-red photography is one possible method and it has been shown to be possible to produce outlines of tumours having an increased blood flow compared with the surrounding tissue (or vice_versa) by means of an infra-red scanning device. Another possibi- lity is the use of ultra-sonics and.this field~ too~ is being explored though it has not as yet reached a practicable stage. 6.6.4 Other cancers .6.6.4.1 · Screening for other cancers has been shown to be of value in certain groups in the population at special risk~ Of these~ perhaps the foremost are workers in certain industries, of which the rubber and electric cable industries are the most important. In the past benzidine and beta-naphthylamine or allied substance.s' ,were used in the manufacture of rubber articles or electric cable insula- ting material. These substances are now known to be highly carcinogenic, especially for the bladder. Pre-cancerous polyps and early cancers of the bladder can be accurately detected by means of exfoliative cytology. For those at risk routine cytological examination of the urine at six monthly intervals is recommended. 6.6.4.2 Exf<_:>liative cy-tology has also been shown to be valuable as a diagnostic aid in the early detection of cancers of the oro-pharynx. Dental inspection often shows small lesions of the tongue or cheek which would not normally be suspected of malignancy. However, by routine scraping of these lesions and cytological examina- tion of the material a proportion can be shown to be carcinomatous and radically exterminated. This is probably the best present use for oral exfoliative cytology. It is sometimes suggested that mouth washings or scrapings from the cheek should be routinally examined for all persons undergoing dental inspection~ but it is likely that the yield of unsuspected malignancy would be very low and the use of resources in this way uneconomic, 6.6.4.3 Cancer of the stomach is another form of malignancy where the methods of early de:tection can be employed. Unfortunately the prognosis for stomach cancer is appallingly bad when diagnosed by normal clinical means, at whatever stage. Indeed it is a paradox that. the shorter the history of symptoms the worse the prognosis in terms of survival. The best hope of improving the results of surgery is by diag- nosis at a pre-capcerous stage and this may be done either by gastric cytology or gastric photography, or both. However, the stomach is inaccessible and any kind of - ll5 .- mass screening is hardly practicable. The exfoliative cytological technique for examining the stomach requires particular skill and needs to be carried out under hospital conditions; its use is therefore virtually limited to aiding clinical di~1osis in patients with suspected lesions. It is used for screening high-risk populations, particularly elderly men, for example in Japan, where there is a high incidence of carcinoma of the stomach. Perhaps the particular use of these tech- niques will be found in a special high-risk group of persons who have already given positive results to a preliminary screening test. The haemoglobin level could be used as a preliminary test,. since persons with gastric atrophy tend to develop macrocytic anaemia. Another possibility is a tubeless gastric analysis for the presence of free hydrochloric acid, using an electrolyte-absorbing resin. For the future,. it may prove possible to develop a simple test for gastric parietal cell antibodies. 6.6.4:4 In ·England and Wales, in 1963, there were 5393 deaths from cancer of the rectum,. which accounted for 5.3 per cent. of all cancer deaths. In addition, there were over 9000 deaths from intestinal cancers, many of which will have ?..risen in the sigmoid colon. This comprises,. therefore, a very considerable cancerproblem. Although during the past 20years operative mortality for colon and ~ectal cance.r has fallen considerably, it is an unfortunate fact that survival is still about the 213 same as it was 20 years ago. The best hope, therefore, in improving the outlook may well lie in earlier deteetion (as indeed,. for other forms of cancer). Routine procto-sigmoidoscopy .. as part of any general medical examination, offers the possibility of detecting not only early invasive rectal and sigmoid cancers but also 1 . h 1 v .d 11 El d h" k. 214 f precancerous es~ons, sue as po yps. !~en a scm an ~s co-wor ers, or example,. at the University of Pennsylvania Diagnostic Clinic found, on routine procto- sigmoidoscopy of 10o6 persons,. 105 (10 per cent.) with rectal polyps. Carcinomatous change in the polyp was established by biopsy in three cases. Similarly, Hertz_ and 215 .. others reported from the Strang Cancer Prevention Clinic a finding of cancer in 2.2 per thousand out of 26 000 persons examined over the age of 45 by procto- sigmoidoscopy (i.B per 1000 in women and 3.1 per 1000 in men). 52 per cent .• of the patients were asymptomatic and 48 per cent. had only minimal symptoms (persons with ' . . '• definite preceding histories having been excluded) • More needs to be learned of the risk of ca~cer developing in polyps. The risk in familial generalized intestinal polyposis is k11own to be very considerable; the - ll6- conversion rate to invasive cancer of the various forms of sporadic polyp is at present not so well known. However, there are difficulties in offering procto-sigmoidoscopy to persons undergoing routine health examinations. In the first place there is difficulty in ensuring that faeces do not obscure the view at the time of examination; and, in the second place, there is reason to believe that the nature of the procedure may deter people from attending. Whatever the difficulties may be at routine health examinations there is little doubt there is good reason to advocate routine procto- sigmoidoscopy in persons at risk who are in any case undergoing medical examination at clinics or in hospital. 6.6.4.5 Lastly, the importance of searching for early pre-cancerous and cancerous skin lesions needs to be remembered. This is particularly true in countries where the sunlight is strong and the population largely of Caucasian stock and employed in agriculture. 6. 7 Diseases of the e're Chronic elaucoma It is only in the last two decades that chronic open angle glaucoma, or glaucoma simplex, has come to be recognised as a separate nosological entity from angle closure glaucoma. Chronic simple glaucoma, which for practical purposes only occurs after the age of forty, has been said to at tack 11 like a thief in the night" because of the long period during which physical signs gradually progress and the perhaps equally long period when subjective visual changes are minimal and only detected by 216 careful examination. For all glaucoma patients registered as blind,Sorsby found only 0.4 per cent. below the age of 40 at registration and 65 per cent. were between the ages of 60 and 79. Glaucoma (both acute and chronic) accounted in England and Wales, for the period 1951-54, for 13.6 per cent. of registered blindness, or 4200 persons. There is reason to believe that registration of blindness is incomplete and it is likely that there are far more persons in England and Wales either blind or with impaired eyesight due to glaucoma. There is a considerable literature on the prevalence of chronic glaucoma and its distribution is apparently world-wide. The accepted treatment of chronic glaucoma is by miotic drops administered daily - 117- for an indefinite period~ starting if possible in the early stage of the disease. If interference with vision becomes ser;tous an operation aimed at increasing the aqueous outflow is carried out. A commonly rec~mmended treatment consil?tS of pilocarpine hydrochloride (0.5 - 2 per cent. solution) administered two to four times daily. 217 One difficulty is that the value of this medical treatment has not been effectively evaluated; it is also an uncomfortable treatment which can be dangerous in that it severely limits night vision~ so that there must be real doubt about the degree to which patients follow treatment instructions. Established chronic glaucoma is diagnosed by finding characteristic visual field changes~ cupping of the optic disc and retardation of the normal outflow capacity of the aqueous fluid~ as measured by tonography. A rise in intra-ocular tension is regarded as the usual accompaniment of chronic wide angle glaucoma and this is the accepted sign by which early glaucoma is detected. Intra-ocular tension is measured by tonometry~ the Schiotz Tonometer being the most popular instrument used in screening. The principle followed in screening is that a rise in intra-ocular tension is the first sign of early glaucoma~ preceding the onset of other signs by some years (perhaps from 10 to 20 years). A population is examined by tonometry and perhaps six or seven per cent. are found to have pressures in the higher range (over 25 mm, of mercury). Of these~ about two pe.r cent. are considered to have probable glaucoma~ with an intra-ocular tension of over 30 mm of mercury (e.g. Glaucoma Program Guide~ pp 10-11). The U.S. P.H.S. Monograph No. 67218 on screening for disease illustrates the bimodal distribution of a disease attribute and takes the intra-ocular tension in glaucoma as a possible example~ showing separate normal distributions for diseased and non-diseased population. There is no good evidence that this is so and population samples of intra-ocular tension that have been recorded 20 .. 21, 219 do not support the bimodal model. Evidently, as with height and weight (and probably blood pressure and blood sugar) there is a continuous distribution of intra-ocular tension, with the probability of glaucoma increasing in the higher range of pressures. As with other conditions like diabetes and arterial hypertension, therefore, the "borderline" problem occurs~ posing the question: "does ocular hypertension indicate an early, pre-symptomatic, stage of chronic glaucoma?"· - ns- There is another question, too, which does not arise with diabetes or high blood-pressure: "does chronic glaucoma occur in the absence of a rise in intra- ocular tension? 11 • There is an authoritative body of ophthalmological opinion which 220 answers this question in ·the negative (e.g. Goldman). Howe rer, there are those who believe this does happen221 and a recent survey in the Welsh Rhondda20•21 supports this view (though admittedly the numbers are not great enough to be statistically significant). In the Ferndale, Rhondda survey, 13 cases of chronic simple glaucoma were found out of the population examined (between the ages of 40 and 74 years) of 4246 (92 per cent. of the total Ferndale population); this is a total glaucoma prevalence of 0.28 per· cent. Six of the 13 cases vrere already known glaucoma patients and only seven were n~wly discovered. All persons underwent not only applanation tonometry but also examination of the optic fundus and a one in three random sample visual field perimetry as well. Glaucoma-was diagnosed by finding a characteristic visual field defect with optic disc cupping. In addition, seven further suspected glaucoma patients had intra-ocular tensions of less than 21 mm of mercury and were .·,. diagnoseq on the basis of optic disc and visual field defects alone. Of these seven cases, two have since been observed to have raised pressure. Adding these seven low tension glaucoma cases to the 13 found by tonometry raises the prevalence rate for chronic simpl~ glaucoma in this population to 0.43 per cent. which is low compared with other recorded rates but similar to the findings of Stromberg. 219 If these findings are confirmed on a larger sample the inference is that the early detection of glaucoma by tonometry alone would need to accept a false negative rate of the order of more than 50 per cent. If this vvork is substantiated it will be necessary to search for other, more satisfactory, methods for early glaucoma detection. Visual field screening, which only detects patients when undoubted early clinical glaucoma is present (though its course, before subjective interference with vision begins to be noticed by the patient, !nay well be as long as 10 years) is an obviously attractive screening technique. Unfortunately there are two important drawbacks: firstly, the plotting of the visual fields, even with a new electronic flash apparatus (the "Globuck") takes up to five minutes for each examination; and secondly the number of false positive examinations found by this technique and needing full ophthalmological investigation is unacceptably high, - 119- though mapy non-glaucomatous defects were fonnd' ·by this technique (13 per cent. in the Ferndale· Survey though this was fonnd with the Freedman not the Globtick Screener) •.. Apart · frOm :'the possible need to look for other methods of diagnosing chronic glaucoma early there is the practical problem for a.riy commnnity of the ability of the lodal eye s~rvices to meet an increased load of examinations. 'Ihe Ferndale stl:f.ve;y- (the only. survey~ so far as the writers are aware~ carried out on either a complete defined population or a randomly selected popuiation sample~ and therefore representative of that population) has demonstrated a prevalence of 0.43 per cent. confirmed chr9llic sample glaucoma~ (prev:l,o:u~ly diagnosed.plus newly discovered)., a ... - :. ' .. co~~~e~ab~ )ower rate .. than that reported from other~ non-::representative ' :~: " . .. - ' . . i ·:· ~ : . . .. . ··' . • ·. . . • . : . • . ... . . . - . • population surveys. .Qne reason may well, be tl:la.t the rate is spuriously raised in .-. . ... . ~- .. most detection programmes by weighting with persons having an increased likelihood of glaucoma {e.g. relatives of glaucoma patients). Another~ artificial~ factor of defini tioh which may acc~nnt for the low prevale~ce fonnd at Ferndale is the use of the terni "ocular hypertensicin" by the authors of this survey. This in itself would tend to decrease the prevalence compared with other surveys. In contrast to this low true p~valence rate for glaucoma fonnd by tonometry., the false positive rate was high.,· 8.6 per cent. A rate of this order of persons needing further investi- gation by an .ophthalmologist would throw an impossible strain on any conn try's eye serJi6es. Perkins,. 222 for example, has calculated that in England and Wales this cbUid ,.inean a load ·of 4500 persons ·to be investigated by each consultant ophthalmologist Assuming· that by some means all the patients with early glaucoma in a population have been diagnosed, the que~tion of treatment next arises. We have. seen that the~e is doubt about the efficacy of medical treatment and that -'ther~ is ·a 'ha;rilf'l.ll ~lem~nt. Work l·s urgently needed to help decide on the two poirits: (~) ;ci6 pati~nt~ :i.n fact do as they are advised? and (b) if they do this~ (i) does it prevent them developing more advanced glaucoma? and (ii) has it deleterious effects, e.g. do those on miotic drug treatment suffer more accidents than the rest of ·the ·popUlation?. It might well be worth assembling all the evidence on the value of the medic.aL ··treatment of. early chronic glaucoma and deciding on the evidence whether a randomized trial of treatllient is justified. - 120- Lantly, there is the problem of' the "borderlines", those persons with intra- ocular tensions at the high end of' the distribution without symptoms or sign ·of' glaucoma. 222 In some surveys persons with a diminished aqueous outflow, as measured by tonography, are regarded as belonging to the "glaucoma" group and are excluded f'rom_the borderline group. In the Ferndale Survey only those with optic cupping and a field defect were regarded as glaucomatous. The "borderlines" com- prise a gro~p of' individuals to whom a randomized trial of' treatment can ethically be offered; there is real ignorance as to: (a) the risk to the individual of developing chronic glaucoma and (b) whether prophylactic medical treatment will lower what risk there is and, if so, to what degree. A trial of' this kind is in progress as part of' the Ferndale Survey but of' course it will be several years at least before significant results can be expectcdo Loolclng to .the future, it would be good to see work in hand to discover more about predisposing factors in chronic glaucoma Besides the relationship with a raised intra-ocular tension we know there is a familial trend; there is a possibility th~t there may be some correlation with iris colour. This is little enough to go on. Chronic glaucoma is one of the conditions in which useful correlations with other physiological variables or with other morbid processes might be revealed by a wide- f'ronted prospective study on the lines of' the TectUnseh experiment, where a complete J.. • popu1ation _is followed over time and its <Usease experience correlated with recordings of as many variables as possible. There is obvious scope for biochemical and digital automation in ma,k.ing leasible this form of large study. On ~he. tech,p.ical side, it was found in the Ferndale Survey that, while the coefficient of variation of the Goldman Applanation Tonometer was greater than that of thr::l Sch5.o-t.z instrtUne.nt, neither was considerable in terms of mean pressure levels. Applanation tonometry was more acceptable tq the population that the Schiotz technique~ The Mackay Marg tonometer is commercially available and it is claimed that this c~~ be used without anaesthetizing the cornea. If reasonably accurate readings can be obtained in this way it would clearly be a great advantage~ since using a local anaesthetic on the cornea inevitably adds some element of risk to the procedure. - 121 - Until the answers to some of the problems discussed have been provided by experimental work, what can usefully be done for the detection of chronic glaucoma? It seems that tonometry is the only practicable screening method we have at the present. In order to avoid an uneconomic and unbalanced use of the eye services available in most advanced countries· it would seem wise for the present to concent- rate on those at highest risk~ i.e. persons over 40 related to patients with known glaucoma,· since a familial tendency to glaucoma has been demonstrated. 223 What is done· in ·practice will depend a· great deal on the type of medical care prevailing in the area in qtiest{on.- · In the United States of America much glaucoma screening is at present befug carried out and efforts are being ·made to avoid the "drive" and to try to run a continuing ·:programme. In the San Jose, California., programme for 1961, for example 3286 persons were examined and, of these, 27, .were found to .be suffering from glaucoma at a cost of $ 175 per case found. This is compared with the estimated cost of $ 1200 per year needed to support one blind person under the United States "categorical'.'aid ·program :for the bifud". This report states that "it is apparent t~t eariy case di~ScQvery is of significant econOrilic benefit to the taxpayer". But . this is only true if . earf:Y detectiori arid treatment do in fact prevent or delay blindness and this still needs to be ascertained. In countries where health services are unified it is even more mandatory than in countries where me<tle.al. oar~ ,is _largely financed by the individual, that the facts .be ascertained befo:re ;·fo:rm\llating a country-wide policy and undertaking universal .glaucoma detection ... While glaucoma accounts for 13.6 per cent. of the registered blindness in England and Wales, cataract is responsible for 26.2 per cent. and senile macular lesions for 21.6 per cent.. "Congenital abnQrmalities" accotmts for a further 4.2 per cent. of blindness, 38 per cent. of it in children under four years. Much of this blindness, as well as impairment of vision short of blindness, is remediable. In infancy the condition of amblyopia, present from birth, can lead, unless corrected, to blindness in one eye. Detection of the condition needs to be carried out early in life, usually'before the child can read, so that orthoptic measures may be instituted. - J.22 -~ .. In California, for example, the public health nurse service is used to provide information and instruction to parents on screening young children at home :for amblyopia. At the other end of life impairment of vision from cataract, senile macular degener~ti.on or other cause can be detected by a visual acuity test and by examining the lens with a spot-light.. Cataract is of course. remediable by surgery and senile macular degeneration can be helped by pr~viding the patient with a visual aid. As · with glaucoma3 the medical care aspect needs examining carefully. The proportion of hospital beds available for eye surgery.is usually small and waiting lists for cataract operations long. Examination in the elderly.. as with the young, might well be carried out by the home nursing service after special.-:training .. 6.8 Diseases of the c.U'inarv tract 6.8.1 General Urine specimens are easy to obtain in screening projects and urinalysis is simple and can be indicative of a number of conditions.. of which diabetes and kidney .· ·· disease are the most important; though as an index of sub-clinical icterus in epidemic jaundice uribilinogen examination can be extrem3ly useful. paragraph is concerned with the urine in renal disease only. However, this rn t:he :Baltimore screening study a prevalence of albuminuria of 8. 4 per thousand parsons examined was found. In 4. 7 per thousand this was confirmed by diagnostic . tests~ and of these only 0.8 per thousand had been previously unknown to the patient. In his five year follow-up Wylie 38 found this test of high prognostic significa:n,ce !, . ·:·.,· .. , 6.8.2 ·-.Bacterillri:a , ·_, . _;._- '· ·· · In screening for urinary system disease the aetiology of the condition sought is of great importance from the preventive point of view.· While patients with chrtn1ic'o.g:tomerulo.-nephritis can indeed. be helped (particularly, perhaps_, by guarding against recurring streptococcal infectiQiiJ., ,, interest has in the past decade. turned to chronic pyelo.:..nephri tis and specially to its possible source in recurrent . attacks of urinar,r:Hifection, whether accompanied by symptoms or asymptomatic. Here agaJ.n is the classical screening situation: a potentially lethal. condition preceded by a. - 123 - possibly latent and.reversible stage. What is not yet certain is-the role played by asymptomatic_ bacteriuria infections, particularly in the female, both in child- hood and adult life. The natural history of these infections (which are certainly common) still need3 to be worked out in relation to the development of kidney damage, hypertensive disease and abnormalities of t.he urinary tract like vesico-ureteric reflux. Clearly the matter is potentially one of great public health .importance. Accurate prevalence figures for pyelo-nephritis are not available. The College of General Practiticners survey 5l showed a patient consulting rate of 13 per thousand persons for urinary tract infectj,ons. The possible precursor of much pyelo-nephritis, asymptomatic bacteriuria in pregnancy, has been found· in a number of surveys to be prevalent in about 5 per cent. of pregnant women; the inc1dence. of acute pyeio-nephritis in those women with . persistent bacteriuria· has be€m discovered to be ten times greater than in women without init±B:l. bacteriuria. 224 · Miail ·and Kass and theii- colleagues 22 ') have found a bacteriuria prevalence of 4.4 per cent. in the general female population aged over 15 years in Jamaica, but re;..examinations show that, while the prevalence remains the same, d~fferent women are affected at each survey.· ·-So far, the evidenc~ for an aetiological connection between the asymptomatic bacteriuria and later chronic pyelO- nephritis i:l,ppears to be incomplete. What is needed is a prospectj_ve- tri.al treat~ent; thoughthe·evidence for the immediate value of treatment in pregnancy, in preventing attacks of symptomatic urinary infection, prematurity and foetal loss, is now con- siderable~ suggests that untreated asymptomatic bacteriuria in pregnahcy i·s- accompanied by increased foetal loss and low birth weight; whilst . . -. . . ShouJ.d these findings be Sl:l.bstantj_ated thera are grobnds ':for -f1id.espread cai5e-finding, both in the female child and in pregnancy. Ttie'ineans fordoing this a::f.e already being worked out; the·triphenyl tetrazolium chloride (T.T~c"f screening test, described by Siriunons and Williams 227 shows promise· of p:i--oviding a useful screening test which can be used in the field, prcrviding facilities for refrigeration of urine specimens are available, Catheterization is· not required and indeed is contra-indicated ; 228 a "catch" mid-stre.ain :specimen of urine:. pasSed into a sterile container and immediately cooled by refrigeration; :is all that is required:, Doubtless .. n wider evaluation of the T.Jr. 0 •. test and the - 124 - possible development of improved tests is needed, since other workers have been unable to obtain comparable results in comparison with culture methods. 6.9 Rheumatic diseases 6.9.1 Rheumatoid arthritis "Arthritis and Rheumatism" accounts for a large part of the illness which presents to the general practitioner. In the British general practice enquiry51 it accounted for a patient consulting rate of 65 per thousand, (not including acute rheumatism) second only to acute nasopharyngitis in order of prevalence and of course causing far more disability. Population surveys for rheumatoid arthritis and for the presence of rheumatoid factor have been relatively extensive. 229 The prevalence of definite rheumatoid arthritis in different sample populations in Northern Europe varies from 2-3.5 per cent. for persons aged 55-64, with an additional 1-6 per cent. of persons probably affected. In North America the Tecumseh, Michigan, project finds a prevalence rate in all persons over the age of six years of 0.39 per cent. definite and 0. 85 per cent. probable rheumatoid g,rthritis. 230 "231 Tne world-wide distri~mtion of this dise2cse i12"s not j•et been defined by clinic2.l surveys and the ve.lue of international morcalit2r data is limited. However, serological testing for rheumatoid factor shows its distribution to be wide- spread in both Africa and Asia; though the exact relationship of the presence of positive tests (Latex fixation and Waaler-Rose) to the development of rheumatoid arthritis has still to be determined. There is therefore good evidence of a high degree of prevalence of rheumatoid arthritis (its diagnosis being based on the American Rheumatism Association criteria). This provides a strong prima facie case for early detection. However, the other criteria for justifying screening are not fulfilled, notably: (1) there is no clear pre-symptomatic stage (despite the somewhat obscure relationship with positive tests for rheumatoid factor): (2) there is.no specific treatment. The best accepted treatment for the early stages of rheumatoid artl1ritis is to protect the affected joints from excessive use. Case-finding before the patient presents with disability does not therefore appear to be at present justifiable. More needs yet to be learned about the aeti~ logy of the condition and then, perhaps, about its early treatment. Apart from -.125- basic work on a. possible aeti.ological factor {such as a virus;; as currently reported to be suspected by workers in Birmingham) continuing survey work can add to our . knowledge. of the rela,tionship of the various faqtors so far uncovered to the disease process itself. A possibly useful contribution to survey work might be a concomitant continuing population sur.ve:r of hiochcm:;.cal va:L'iables in blood. This would be feasible \'lith automated laboratory i'ncilitj.es. In this way early deviations from the "normal" biochemical pattern might be noted in relation to early rheumatoid cha.l1ges. Gout Although gout is not a highly pr·evalent condition {only 0.8 ·patients consulting per 1000 persons in the British general practice survey (though up to 4.5 per thousand in the prevalence studies quoted by Kellgren)229 .it is a condition {a) with a recog11i~}l,ble pre-clinical st,age (hyperuricaemia), (b) where there ia a;;high-risk group in the population (r.::latives of gouty probands) ancl. {c) for which an effective treatment exists (lo~i purine diet,. avoidance qf. exQess dietary fat and alcohol and the l1Se of uric~_suric dr.ugs). There is therefore a good case for screening for this condit~on, I:JOSsibly in a selective manner. Gout: e,ppen.rs to be one of the few clinical conditions which is with little doubt transmitted by one or, at most_, two genes and where a bi- or tri,..mooal distribution of se!'Uffi uric· .acid levels is found .. in n population. Kellgren 232 has suggested that an intermedia~e peak in the distribution curve_, at 6 mg per 100 ml_, may represent ··the heterozygotes in t.he population.. the gouty homozygotes having a peak distribution in the gouty. range-" at about 8.5 mg per 100 ml. A ·sub-·committee of the C.I.O.M.S. Symposium held in Rome i...'"l 1961 agreed that a serurn uric acid levels in males of over 7 mg per 100 ml and jJl females of above 6 mg per 100-ml should be regarded as one of four criteria fo:;." the G.iagnosis of gout • . It seems likely that .those at genetically high risk may be brought to the clinical level by variations in the habits of the population. A high purine diet, excessive intcl.l;:e of dietary fat and alcohol (as experienced by .well-to-do city dwelle~s as comp8~red with poorer country dv-rellers) and the use of drugs such as salicylates in lo>'l dosage_, _hypotensive &.gents <md diuretics may promote a higher prevalence of gout; whilst restriction in the diet and in the use of drugs (as - 126 - occurred during the 1939-45 war) tends to cause the prevalence to wane. Biochemical screening of the uric acid level will indicate those persons at high risk and the risk can then be lowered by alteration of the diet and advice on the use of drugs. In Great Britain a survey by Lawrence and his colleagues233 showed a prevalence of five and a half per cent. of males over the age of 15 in the rural area of Wensleydale with serum uric acid values of 6 mg per 100 ml or more. Among male inhabitants aged 55-64 of the town of Leigh, in Lancashire, the prevalence rate of men with serum uric acid levels above 6 mg per 100 ml was as high as 18 per cent. Thus selective screening would give a high yield of persons to whom useful advice could be given. 6.10 Mental illness On grounds of magnitude the early detection and treatment of mental illness in a community should have a high priority. In the British survey of general practice51 the r"'. te fo~' p::"tients consul tin;; their family doctor for psychoneurotic disorders ranked next only to acute nasopharyngitis, rheumatic disorders and bronchitis in frequency; the annual patient consulting rate was 46 per 1000 patients and the total consultation rate 166 per 1000, or more than 4 per cent. of the total consultation rate for all causes. In a survey directed specifically to mental 234 illness in general practice, Shepherd and his colleagues found, in a random sample of London practices, a total patient consulting rate for psychiatric morbidity of 140 per 1000 persons (176 per 1000 females and 98 per 1000 males) over a twelve month period. Of these, neuroses accounted for by far the greatest number (89 per lbOO) whilst'psychoses were diagnosed in six per 1000 persons. For women, psychiatric disorders ranked as the third commonest cause for consultation after respiratory and orthopaedic or traumatic conditions, in that order. A review of 50 surveys carried out in different parts of the world showed rates for all mental disorders from one to 370 per 1000 population.. The differences are mainly due to variations in the techniques employed in the surveys. It is clear therefore that mental ill health constitutes a great burden on society and it is reasonable to aim at early detection and treatment as a buttress to conventional methods. However, when examined more closely, difficulties appear. - 127 - ·. Ili;'contrast to most other branches of medicine it has not so far been possible to reach an aetiological classification of mental disease and for this reason there are very few specific treatments (neuro--vascular syphilis and phenylketonuria are examples of exceptions). Thus there is no clear agreement about the diagnosis and treatment of much overt mental disease. "Treatment" may consist of drug therapy, as in depressive illness, but it also includes the management of the total social situation of the patient and his immediate family group. In certain conditions it is not clear what should be the "best" treatment of the whole situation. It may be in the immediate interest of the patient himself to carry on as a member of the community, possibly remaining undiagnosed. On the other hand this situation may lead to breakdown of the family situat'ion, when diagnosis and removal for a time from th.e·sbcial.environment is the treatment of choice. It is therefore not necessarily beneficial for the psychoneurotic, say, to be the subject of early diagnosis; his best treatment may be that of continuing the struggle to integrate himself with society. After a fashion, this is the same problem of borderline diseas~ met with else- where; the problem of what to designate as "disease" in need of treatment, allied with the question of whether early treatment will be beneficial. When all this has bee8 said ;it still appears t_J;a.t there is a group of persons with relatively mild depre~sl:ve i:LlP:ess who could benefit from,,.~,a.rly diagnosis and treatment with, among other method~r modern ataractic drugs. T~ere is a shortage of evidence of the value of various treatments in depressive illness, especially in its earlier stages. The Medical Research Council has in progress a trial of treatments (including drugs and electroconvulsive therapy) of more advancied..depression; 235 but evidently a great deal of mild depressive illness is not refe:t'red for specialist advice and is treated by general practitioners with ataractic drugs. A trial of treatment of this milder illness might be valuable and could be combined with a survey of early, unreported, mental illness in the community. Work has recently been published on ascertained mental disease in general practice (as reported above, Shepherd and others,)234 but there is perhaps need for further work to identify persons who might be helped b;j early treatment (including social adjustment within their family group) and for a trial of such treatment. This would entail a trial, inter alia, of the benefits of using social workers in the pre-breakdown stage of psychiatric disease. - 128- Population surveys of psychiatric illness in the community are being undertaken (e.g. by the Medical Research Council in Camberwell and South Wales) but we do not kno\'r of trials of early treatment of the kind described though they may of course exist. 6.10.1 Mental retardation Examples of specific conditions where screening is carried out A detailed critical review has been made in 1964 by Gruenberg of a number of community wide surveys on mental retardation. 236 Remarkable variations are found between and within age-groups. At the age of peak-prevalence, about 14 years, Gruenberg found a 10-fold variation from about 1.0 per cent. to about 10 per cent. Such variations may be largely accounted for by differences in the limits and scope of studies, diversity of defL~itions and availability of measuring instruments. Some type of screening of school-age children has been widely carried out in count- ries with a well-developed school system, based largely on the criterion of ability to keep up with the classwork, but frequently supplemented by the use of IQ tests. Where suitable services exist, the children suspected of retardation are often referred to school psychological services for further testing and investigation of the possible cause of the retardation. Stress is increasingly being laid on the importance of using batteries of tests combined with socio-psychiatric investigation since failure to keep up with the school system has frequently been found to have causes other than low level of intelligence: moreover most specialists now doubt the prognostic value of IQ tests used alone. An example of the screening of a school-age population is that carried out by Jae~i and Jaeggi in Geneva237 as a preliminary to the re-organization of the Cantonal socio-medical educational service for the mentally retarded. The primary screening was carried out through consultation of institutional, educational and medical registers and reports. It was considered that practically all suspected cases could be found by this means, these reports being very thorough and kept up to date. (An indication of the reliability of the data is given by the fact that in 1964 10 per cent. of the total school population of the Canton was seen by the socio- medical service). A detailed questionary was then completed for each suspected case providing information on social, psychological and medical conditions as well as - 129 - educational achievemen.t. It .Has found that 0.8 per cent. of the children of school age l'fere· :in need of special social, medical or educational service. 6.lo.2 Phenylketonuria (PKU) PKU is one of the few disorders leading to mental retardation of which the cause is known - a genetically transmitted me~abolic abnormality. It. the defect is detected very early in life mental retardation can be prevented or favourably modified through a special diet low in phenylala~ine. A bacterial "inhibition assay" screening test for blood phenylalanine (Guthrie test) was used routinely by. State Health. Departments in the United States of America in 1962 and ·1963 on more than 400 000 newborn · infants born in about 500 hospitals. The frequency· of PKu was found to be· one per 10 347. Such routine testing has now become mandatory in some of the States (e.g. New York, Massachusetts). It is also used widely in the United Kingdom and.some.other countries. cannot be used in areas without a wide network of health services. Obviously it 6.10 .. 3 Galactosaemia This disorder is also caused by an inborn error of metabolism, but if untreated it leads to early death, unlike PKU. A diet loN in galactose prevents the develop- ment of the clinical condition which includes severe mental retardation. Promising attempts are being made.to attain simple screening methods, e .. g. paper strip tests. By quantitative biochemical screening procedures it seems possible to detect, with some measure of accuracy, the heterozygous state of galactosaemia thus making it possible to detect paren·ts who might produce galactosaemic children:38 . 6.11 Anaemia 6.11.1 Introduction There are, of course, many causes of anaemia but iron deficiency is by far the commonest ana'only this form of anaemia is· considered here. This deficiency is essentially ·~·imbalance between intake and absorption, and excretion or loss in other ways; it can be prevented by correcting either deficient intake or abnormal loss. Thus one of the priricipal criteria :for screening is met, in that the condition can be prevented by treatment (whether primary or secondary; early or late). - 130- A second feature of iron deficiency anaemia in connection with screening is the test for its detection. Measurement of the haemoglobin level is one of the few examinations which directly estimate the variable in question. There is therefore no doubt about the significance of what is being measured (as there is, for example, in measuring intra-ocular pressure as a test for pre-symptomatic glaucoma); though the other difficulties of definition and the variability of the method and the observer are the same as for other tests. 6.11.2 Prevalence Iron deficiency anaemia was the subject of a HEO Study Group in 1958.239 After considering the haematological data on apparently normal persons throughout the world the Study Group adopted the following criteria for haemoglobin values below which anaemia could be considered to exist: TABlE. 6.11.1. Age in Years Sex Hb g/loo ml 0.6 - 4 10.8 ) ) 5 - 9 11.5 ) 11.5 ) 10 - 14 12.5 ) Adults Male 14.0 Female 12.0 Pregnancy 10.0 i I The Group's Report reviewed the available studies of prevalence and considered that more of these studies were needed. It pointed out that in some parts of the 'I'Torld anaemia constituted a major public health problem. An investigation in Mauritius had shown that 50 per cent. or more of certain groups of the population were probably affected. In Western countries the prevalence is clearly much lower. 240 Using the above criteria, the survey by Berry and others of London housewives in 1951 showed nine per cent. of the women to have haemoglobin levels of less than 12 g - 131- per_lOO ml. Ki~patriok Cilld Hardisty 1 2·41 in a study in South Wales and North Englandc of men and women in age groups respectively of 35-64 and 55-64, found 14 per c~nt. of the women with haemoglobin levels of under 12 g per 100 ml and three per cent. of the men with levels of less than 12.5 g per 100 ml. Kilpatrick and Hardisty's female group were over the age of menopause. Iron deficiency anaemia is, of course, most prevalent in women who menstruate and it is likely that women in that age period would be found to be anaemic even more frequently. For this reason iron deficiency anaemia is a condition specially suited to selective screening of women between the age of, say, 20 and 44. More need·s to be learned, however, about the criteria constituting anaemia and the relationship of haemoglobin levels to symptoms. Though we may have a reasonable idea of what is clearly anaem:i.a .. we do not know much about optimum levels of haemo- _g:).obin. To this purpose there is need pf' further survey work: (a) to ascertain with the greatest possible accuracy the distribution of haemoglobin levels in a probability sample of a large population; (b) tq relat~ carefully, and without observer bias,· symptoms to haemoglobin leve],.s in iron de1"ici~ncy anaemia, as well as symptoms to the effects of trea"t;ment; (c) to discover the cheapest and most effective way of treating and preventing the recurrence of iron deficiency anaemia. Studies in this field are at present in progress under the auspices of the M.R.C. Epidemiological Research Unit (South Wales), and there may of course be other studies under way. 6.11.3 Methodology Where samples of venous blood are being withdrawn for other purposes, there is advantage in centralising the estimation of haemoglobin, so that it can be carried out on an accurate calibrated photo-electric instrument. On the other hand, collection of venous samples in tubes specially for this purpose is expensive and in many field circumstances it may be both cheaper and more convenient to estimate the haemoglobin on a sample of capillary blood. The simplest and cheapest way of screening for anaemia is the Phillips-Van Slyke specific gravity method (which is ---132 - used by-the British National Blood Transfusion Service for screening potential blood donors). However, its accuracy for use in general or selective screening is 6Peri 242 to question. A recent trial carried out by the M.R.C. Epidemiological Research Unit (South Wales) /18/ has shown the American Optical Company's Spencer haemo- globinometer to compare well with the M.R.C. Gray Wedge, the EEL photo-electric photometer and the Sahli method. The A.O. haemoglobinometer has the advantage that it works without· dilution of the blood sample; ·a drop of bloOd is simply placed on a special slide, haemolysed with a stick impregnated with saponin, and covered with an optical coverslip. The slide and coverslip are -optically prepared so as to present a standar4 depth of blood between them, th~ colour of which is compared with a st~dard. -:Another, and reasonably accurate, ·screening technique for capillary blood is the micro-haematocrit method. This method, of course, requires a centrifuge and is therefore best labor.~tory-based. 6.11.4 Screening Screening for anaemia is carried out, with highest yield, as we have seen, on women in the menstrual age ·groups. W.Omen in the· child-bearing age very largely attend maternity and child welfare clinics and this offers an excellent venue for screening. 'I'here are, however, the other causes of anaemia which may be picked up by routine haemoglo~inometry, such as other blood diseases, malignant.dis~ase, peptic ulceration and other gastro-intestinal disease, rheumatic and renal disease. Routine haemoglobin estimation is therefore probably one of the most profitable ·S:6fe·Eining tests f~r all ages in pointing to unsuspected disease; Jungner and Jungner243 found 1.4 per cent. of 30 000 persons in the ~rmland project were con- firmed as being ana.emic, the commonest previously undiagnosed condition. Other- than- part of a case-finding population screening programme, haemoglobino- metry is clearly a most useful addition to the normal examination ofpatients in a general practiitioner's work.· It can easily be ~arried out by an ancillary helper. Fry~_lj-~ has reported, for example, the results of routine haemoglobin estimation in an Outer London general practice. He took a 10 per cent. sample of all his adult patients and found 18 per cent. had haemoglobin values of less than 12 gjloo ml which wa.S ten timE3s the rate of "clinical1' anaemia in his practice. · - 133 - 6.11.5 Conclusions In conclusion, therefore, anaemia is probably one of the more acceptable con- ditions for screening that at present exist; it is highly prevalent~ can be sufficiently accurately detected· and, when due to primary iron deficiency, responds excellently to treatment. The haemoglobin level is also a sensitiv~ index to a number of other conditions, of which anaemia may be one of the earlier signs. Naturally, once anaemia is discovered, it is of the fil?st importance that a complete ~at.ological ~vestigation is then carried out, leading to a definiti~e diagnosis. 7. METHODOLOGICAL TRENDS IN SCREENING n1 the foliowing1 we will consider some common procedures and tests which have been applied in health screening progr~es, or in similar screening surveys in a population. There seem to be certain intE!rest.:i!}g t~nds in the tecpniques which may dominate in the futu~,.in contrast. with the wide variations in acting and think;Ln.g .. today .. 7.1 Clinical or technical screen~ Firstly, there is a strong trend to use automated methods, particularly for chemical tests. The. ~xtensive use of laboratory methods has, however~ ·been- criticized., and the value of impersonal and highly standardized methods may be questioned. Only if the resources are limited, can the medical value justify the cost aJ;ld,. effort P.f. automated or mechanized. tecl:miques with high capacity. ' -·· ' . - ... - '·" . .· . It can be noted that ~:volution may occur stepwise, ·when laboratory tests._ and certain . ,. . ··- -·' . . . . . ... ·· ' clinical examinatic;>Il$ may be applied in. "drives", but are later used as special facilities at the praotitioner's disposal for periodical health examin{ltion. Another obvious trend is the increasing use of data-processing machines. '!his matter will not be considered here. Health Services will adapt these principles used for sick eare, but in some respects - registration of data and storage - the work in- health examinations may lead the · evolution. 7.2 Participation by doctors The most essential and decisive factor for the extent and organization of health screening has been whether or not medical examination by a physician has been included in the survey. It is debatable whether a physical examination is to be regarded as a screening procedure. Depending on this., we can divide health investigations into two groups. One is a complete health examination., which includes a doctor's physical examination; the other group consists mainly of a ba~tery of clinical and laboratory tests, when the contribution of physicians is limited to evaluating certain tests., or is com- pletely lacking. Inclusion of the doctor's findings., ·as well as his interview and evaluation~ greatiy widen the scope of screening, and make it comparable to a general medical ex&nination. As an important part of the final medical conclusion, the examination by a physician is extremely valuable .. The signal factor is the time that is devoted to the check-up. Simple inspections and fast examinations, as well as c~prehensive physical examinations., have been practised. 7.2.1 The doctor's physical examination This examination is 'expected to be a complete physical examination, with inspection., palpation and auscultation, and should include measuring· the blood pressure and rectal palpation .. The_ physical examination is sometimes carried out as one of the first steps in a health screening programme, together with checking the questionary and ordering special tests, which are not carried out in all patients. made bytbe doctor after all tests J:lave b~en completed~ The final evaluation is Obviously, ·there is not always a sharp distinction between the technique used for physical examination in health and in sick care. Especiallywhen seeking· such conditions as malnutrition, the technique may be similar to methods-used for children'' s care in general, as seen for instance from the report by a WHO EXpert Committee on malnutrition. 245 It can be anticipated that some. of the work, which today. is done by the doctor:, may be taken over by technical personnel with special equipment. -By means of proper organization, many. clinical tests can be done more easily., and on a fairly large sc~e. This will make the goal easier to reach: always to have a complete medical examination. Cance.r det~ction is sometimes one of the main objects of health screening. This is despite the_ fact that no guarantee of freedom from malignancy can be given, . -', . .. . ., :· . .. ·~ ' . and that the poss_ib::ti.It~es-o:f detectinf!; ·cancer--- in an asymptomatic stage - are extremely limited. - 135 - Several suggestions·.· have been made about how to perform the cancer-detecting examination most effectively. Reference is made to,· for iilstance, the well known description byDay. 246 In a WHO Expert Committee's R;port, 247 the following !"" examination is mentioned: inspection of the entire skin area and all accessible body cavities, urine tests, chest X-rays and proctoscopy;. in males prostatic examination, in females cervical smears and palpation of the breast. Sputum cytology, gastrointestinal radiography, blood counts, colposcopy, and possibly mammography may be added. History-taking by doctor or specially trained medical staff • The history is very important, and can be obtained by proper questionaries. It has Qeen reported from many investigations that the medical history and the physi- cian's physical examination give the greatest contribution to the diagnosis. However;, most of the diagnoses are then !mown before the health screening. How much medical value is affoPded by the notation of earlier known disease remains to be seen. Obviously, the fuformation is most useful· the first time that an examination is undertaken. The ttalue of the history is tremendous and the advantage of questionaries is great. Questionaries · Many usef'l.ll questionari_es have been suggested. Best known is the Comell :r.1edical Index, a simple checklist medical history form (" self-screener"), which has guided many other attempts. The size and contents of the questionary depend on the 'purpose and the facilities available. Although the number of questions can be very small, a general survey of the state of health may require 200 - 500. questions, arranged in groups. Some workers use still more comprehensive systems, where positive findings, if any, are more carefully investigated. As a rule, the questionary is checked by a doctor or by specially trained hospital staff. The gain in time is al9o appreciated in connexion with the physical examination. 7.2.4 Clinical methods ~d laboratory tests used for screening .. Health examinations originate from·and resemble in many ways the traditional visits to a doctor's office. This was natural when health examinations started. There has, however, been a gradual change in some respects, depending on the lack of definite signs of a disease. - 136 - Looking at the procedures used, it is obvious that for a long time conventional diagnostic procedures or functional tests will be used. In Table 7.1,. an attempt is made to give a rapid idea of the examination programmes used. .A comparison is made between more conventional and traditional screening programmes {column 1) and some new attempts {columns 2 and 3). In the first column the frequency of the test is apparent from the numbers given. The material is taken from the study of 33 screening surveys 4 . in the United States during 1946-1954. The table shows that none of the tests was used in all 33 surveys but a few - chest X-ray or MMR, simple serology, and blood sugar determination - were most popular • . , . .. .. ~ ...... --.. ·-··-·----··· The other two colums in Table 7.1 give specific examples of more recent projects, according to somewhat different principles. The first (column 2) is a Swedish study in the county of ~land, where 100 000 people were offered health screening in connexion with a traditional survey for tuberculosis by MMR. 242 Apart from MMR, the basic examination consisted of blood pressure measurement, urine analysis and a multiple blo()d examination with a number of ·tests: · haemoglobin, haematocrit, serum analysis of serum iron, creatinine, the transaminases GC1I' and GPI' (e.g. the aminopherases glutamic - oxaloacetic and glutamic - pyruvic transaminase), cholesterol and beta-lipoprotein, zinc sulphate test for gamma-globulin, thymol rubidity test and, finally, the determination of total protein-bound hexoses and of sialic acid, both non-specific tests indicating many kinds of inflammatory reactions. The screening was carried out by a field group, who also took the blood samples. Analysis was done in a centrally located automated laboratory, and the results handled by computer. This kind of investigation was performed to find objective grounds for recommending examination by a doctor. In such an investigation, the methods chosen are obviously less diagnostic, but sensitive to various diseases. The blood sample taking - as well as supplementary clinical tests - is simple and can be done under primitive conditions, using disposable material (sterile needles, sample containers etc.). The last column (column 3) gives an ·example of- a periodic health examination in California by the Kaiser Foundation (see section 7.4). ( It gives a survey _of tfte selected methods with advanced techniques, which are used on a large scale. - 137 -· TABLE.7.1.1 .. SURVEY OF SCREENING PROCEDURES The examples chosen are taken from: 1. A summary of 33 studies in. the U.S.A. 4 24 2. A .?wedish pilot stud,y witt!. a b~tte:ry of chemical blood ~nsts; 3 3. A multiphasic study in Cal:l.fomia by Collen et al.248, 9 .. Physical examination History taking or questionary Oral and/or dental examination Intraocular tension, tdnometry Visual acuity Retinography Hearing tests Procto-sigmoidoscopy Biometric measurements,_ as height, weight, skinfold measurement Blood pressure ECG Lung function tests Cytology: vaginal smear Chest X-ray or MMR .. . .... X-ray mammography Serology: VDRL, etc. blood groups, Rh Haematology: haemoglobin haematocrit cell count, differentials count Blood in faeces Erythrocyte sedimentation rate 1. N'\lmber of stu:.;; dies using the test {from 33 surveys in USA, 1946-1954) 4 15 5 1 18 12 .. 2. 3. 'Ihe Vlmn1and. · Multiphasic study, in screening Sweden by·Collen et al. 1965 1962-1964 ·+ + + + + : + + + + + + + '· + ~ 7 , ... + .. + + + --- ... - ~···· + + + + + + - 13?- SURVEY OF SCREENING. PROCEDURES Chemistry Urine: sugar Urine: protein ... ~. · · ··· Urinary deposit· · ·· · · ··· Bacteriuria Blood sugar Blood sugar after test load Protein-bound hexoses, sialic acid Cholesterol Beta~ lipoproteins Serum albumin and/or total protein Gamma-globulin (Kunkel) Thymolturbidity or similar test Transaminases Creatinine and/or BUN Uric acid Calcium Serum iron and/or IBC 1. Number·of stu- dies using the test (from 33 ····surveys~ iri .. USA.,· 1946-1954) 16 16 3- .. . ..~ .... 30 2. The Varmland study, in Sweden 1962-1964 + + + + + + + + + + Multiphasic screening by Collen et al. 1965 + + ~ --. ' ·+ + + + + + + -- 139 - From the examples given can alsp-be :derived the possibilities of investigating only certain diseases. ·Although the variation of different projects is considerable, the main featur~~ in common are found in .. the:, 'table. One of the striking characteris- tics of the development is that the more .the ,eost .. can be. reduced per test, the more investigations are made • Whatever the real cause may be, there seems to be a wish . -'·' to increase such health studies... ;The willingness to follow up the investigations is not increasing to the same degree, although today some serious attempts can, fortunately, be noted. 7.3 Development of screening facilities Tec~ical evolution will have a drastic influence on all these f~ctors for screening. Generally speaking, the considerable efforts to improve _he~th screening ~Y technical means can be said to follow two _different lines of evolution. 7.3-.1 Using simpler techniques for laboratory tests. 7.3.2 Laboratory automation. Simplification The first_ trend represents the use of tests that are extremely simplified, for instance, paper-strip tests. . ~.' ··. ··. . . and.~ill, ~crease in number. Such techniques will certainl:y: be greatly improved, Simplification of complicated method,s, such as those for determining sugar and urea in blood also seem promising. There are, however, definite limitations to such procedures, as well as certain difficulties in getting reliable registration, sample identification and reporting of results, etc. The analytical cost decreases when complicated procedures are done by extrem~ly simple methods, although by no means to the degree that might be expected. The manual handTing ana·· sorting is too time-consuming to be economical on a large scale. " So far, the cost of the material is not negligible. A good example of simplification in another r.espect is the laboratory routine worked out by s. Suchet in Paris.250 Using a well-planned routine, it is possible for one technician in a w·orking day" to perform 100 of each of the following analyses: erythJ:>ocyte sedimentation, rate, haema~~rit, lipoproteins, antistreptolysin titre, urin~ sugar, protein and blood, urea in serum and 300 serological tests !or syphilis. - _140 - The techniques can be nonsidered as semi-quantitative. In some instances, it may be permissible to sacrifice some accuracy in order to obtain a high capacity and low cost.- Although it remains to be decided how precise methods should be for screening purposes, it is certainly not necessary to have the same accuracy as in hospital-laboratory work or research. There are too many difficulties in sampling and handling specimens in large-scale investigations, so that a semi-quantitative level might well be justified. 7.3.2 Automation The second important trend is laboratory automation,; on which intense work has been done with promising results. The first clinically important analytical robot was the AutoAnalyzer. In health screening, the most common way of achieving sufficient capacity has been to have several AutoAnalyzers working in parallel. In such a way, large systems have been as.sembled, capable of carrying out a large battery of well-known tests. Many improvements that are of interest for health screening have been made on the AutoAnalyzers. Multi-channel equipment is now available, where several of the inconveniences have been overcome that 'were troublesome in a large-scale investi- gation. ·Examples are: patient identifidation with numberirig of sainples, peak detection and reading, calibration, analogue-digital conversion and auto~tic print- out. The analytical programme has gradually become directed towards health check-up needs, and for AutoAnalyzers there is now a large variety of procedures to choose from. In Sweden, experience of an automatic system for chemical mass analysis - partly based on AutoAnalyzers - led to a new equipment, called the AutoChemist, working with discrete samples and including a ~mall computer. 251 Being a machine for mass analysis, the A.ut6bhemist has a very ruih capacity, theoretically some 1/2 - 1 million samples per year,- with 20 or more analyses on each ·sample. In practice, other factor~ than th~- analytical work load are presumably decisive for normal operation, but with a fixed analytical programme - as for health screening - the conditions are especiall;i favourable. · The ma.xilnU.m efficiency, up to 150 specimens per hour - 141- and up to 40 different analyses on each sample1 will be used only periodioally1 but is advantageous for managing a ·temporB.ry high load... . The AutoChe!Jlist. has . 24 fixed anal~ical channels for different tests1 chemical as well as bacteriological and serol:oglcal. · · Handling of such an apparatus is extremely simple and personnel-saving. Loading is done on one side of the apparatus 1 and on the other side1 the samples are returned ··on a similar ·transport belt after analysis. The inclusion of a small, desk-size computer is very useful for screening purposes: e.g. editing the format 1 sorting and checking. Recording takes place "on line" by automatic print-out on a Teletypewriter with a paper-tape punch. It seems likely that technical evolution for automatic analytical apparatus will aim at an increasingly high speed of analysis. The essential feature will, however, presumably be an increase in the number of types of analysis. There seems to be a distinct tendency to efface the borderlines between chemical1 bacteriological1 serological1 haematological and other procedures. This implies that·. the equipment1 as an electro-mechanical device for automation of analyses 1 will be relatively similar for very different tests. The decisive factor for the apparatus system will be the choice of tests for health screening. Theoretically, almost any method could be automated. The analytical technique may be based on well-known procedures but has to be modified and adjusted to the automatic machine. In practice 1 however1 the difficulties and costs often become so great that the evolution seeks other ways1 tries completely different methQis to get the desired information. In order to give an idea of what might be of interest for screening purposes~ various methods are listed in Table 7.2. ~mportant advances are being made. The list is by no means complete, and In some cases, screening needs may promote automation. For instance, micro- biological determination of vitamin B 12 may be used in a large scale as a screening procedure. - 142 - TABLE 7.2 .1. SURVEY OF SCREENING PROCEDURES . . .l ~··. , .. ~ :. '. 1. ! 2. j .t 3· Manual method I Fully automatic i Comments ! on available for ; procedures automatic " i large-scale ! available with methods use : high speed ' BACTERIOLOGY ' I Growth (selected media) + + ; Turbimetric Bact. count i : (+) + Indirect counting CHEMISTRY ; : Urine l I ' Sugar .. I Automation + ' + Ketone bodies i ' economically I + I + I Protein I + + reasonable ( ! only i Urinary sediment. ! (+) ' - in a large i I j scale ' Bacteriuria l + + ' PKU (see .. also phenylalanine ' in blood) ' ; (-) + Fluorimetric I Blood: carbohydrates ' i I I SUgar I + + Galactose I (-) + Enzymatic Total protein-b. hexose I - l + I I Sialic acid ! (+) l + BloOd: lipids i : I ' Cholesterol + + ! Beta-lipoprotein + + Turbidimetric Total lipids·· I + + ! Phospnolip:i,ds ! - (+) ! ' Non-esterified fatty acids NEFA i - i (+) I Triglycerides ' (+) i (+) fluori-j '. i Evt. I i metric .. 1 (~) Denotes that by expensive means and special organization a fairly high effectiveness may be achieved. (+) Denotes that under certain circumstances large capacity can be reached. - 143- TABLE 7. 2. 2. SURVEY OF SCREENING PROCEDURES -·- ... '•• .. Serum protein r- Total protein Al'Puniin zinc sUlphate (gamma-globulin) Electrophoresis Haptoglobin Transferrin Thymol turbidity Ceruloplasmin Enzyme activities Transaininases Phosphatas~s Lactic acid dehydrogenase Non-protetn nitrogenous compounds Creatinine Urea, BUN NPN Uric acid Phenylalanine Electrolytes, etc. so~i~/Pdf~ssium Calcium/phosphorus Serum iron Iron-binding capacity l. Manual method Fully automatic available for procedures large-scale available with use high speed + (+) + ( +) ( +) (+) + - + ( +) (+) + ( +) (-) (-) ( +) (+) (-) ( -t-) (+) + + + - + + + + + + + + + + + + ( +) + + Comments on automatic methods Turbidimetric Non-specific test, still on research level Fluorimetric, better than urine test Of little value for screening As supplementary ~. haematological !i test r! (-) denote's that by exp€msive means and special organization· a fairly high effectiveness may be achieved (+) denotes that under certain circumstances large capacity can be reached - 144 - TABLE 7 .2.3. SD'RVEY OF SCREENING. PROCEDURES CYTOLOGY Vaginal smear Urine Cancer cells in blood Sputum P..AEMATOLOGY Haemoglobih · · Cell counts Differential count Erythrocyte sedimentation rate Haematocrit SEROLOGY WR VDRL .. etc. RA test ASrfA MISCElLANEOUS Bl-ood (faeces_, urine) PBI PoJ.ar!ography 6-phosphoglucose dehydrogenase (vaginal washing) 1. Manual method available for large-scale use (t) (-) (-) (-) + + (-) (+) ( +) ( +) + + + + - - (-) ·- 2. . Fully automatic procedures available with high speed + ( +) (-) { +) ( +) + ( +) - ( +) + + -. 3. Comments on automatic methods Results from electroscanners still experi- mental Can be replaced by protein-bounc carbohydrate tests Semi-automatic - electronic - methods available Large scale automatic methods may be available at high cuts Semi• automatic methods have been tried Non-specific test Still on a research level (-) denotes that by expensive means and special organization a fairly high effectiveness may be achieved (+) denotes that under certain circumstances large capacity can be reached - 145 - 7.4 Automated mUltitest laboratories ad modum Collen The most important methodological progress in this field has been made by Dr Morris ·F. Collen and his coilaborators. Automated multitest laboratories in.the KaiseJ?-Permanente medical centres have been arranged when advanced examination techniques have been ad~pted ·for periodic .he~th e~aminations on a large scale. 248 .. 249 In many ways., the organization and equipment indicate new trends., but are safely grounded on long experience. .The se~uP. is expensive, but inany details are of interest even when resources are limited. ·The examination progr~e includes the following: · · · 1. · ECG with 6 leads, combined with phonocardiogram. Results are recorded by mark.;.;sense cards, but the intention is to evaluate the ECG by computer analysis~ 2. · Glucose load test: 75 g of glucose in 240 ml of water; blood sugar level determined after 1 hour (evt. also after 2 hours). 3· Chest X-ray with 70 mm film., postero-anterior projection., read by a radiologist. 4. x-ray mammography on women over 45 years of age. Cephalocaudad and lateral views of· each breast are taken, mammographs read by a radiologist. Supine pulse rate and blood · pressllre. Recorded manually on mark-serise CardS• 6. ·Visual acuity by·reading a wall chart, as well as pupillary escape test. Mark~sertse card recording. Tonometry. The intraocular pressure is noted on mark~sense cards. "' ~ :. ~ (At the same time the left pupil is dilated for later retinal photography). ·-8. Vital capacity and one-second forced expiratory rate, recorded manually on mark-sense card. '. - 146 -~ 9. Hearing test by automated audiometer, with the graphed readings transferred to a mark-sense card. l,O. Questionar:i,es: (1) one medical, self-ac1ministe~ed in the form of 207 pre-punched cards, each with. a single question. The patients drop the cards in boxes with "yes" or "no", and: a card-reading machine records the results. (2) One psychological questionarY, also self-administered. 11. Blood tests: haemoglobin, white cell count, venereal disease research laboratories' test for syphilis (VDRLL rheumatoid factor (latex fixation slide test), bloo~ groups and eight chemical tests (serum glucose, creati- nine, albumin, .total protein~ cholesterol, uric acid, calcium and transa- minase) made by AutoAnalyzer and the results directly p~ched on cards. 12. Urine_ tests: for bacteriuria (chemically) and for pH, blood, glucose, and protein by paper-strip tests. 13. Retinal photograp~ read b~r a,:.--1 ophthalmologist. 14. Weight and skinfold thickness are measured manually • Height and trans- . verse body measurements are automatically recorded on punch cards, Routinely :I all patients above 40 years of age are recommended sigmoidoscopy.. and for women __ also a gynaecological examination w.ith cervical smear. The automated multitest laboratory has its own data centre. The ·computer plays an important role as an integrating part of the laboratory. Automatically, t.Q.e computer prints out the summary report for the phy:sician. The capacity is exceedingly high, and routine examinations for 4000 people a month are rep9rted. For detaiis, the reader is referred to the reports in the literature. Generally speaking, this is an indication of a possible evolution, at any rate in countries with good resources. One difficulty may be satisfactory co-operation with the physicians. In California, this has been the case, but in other parts of the world the conditions may differ. It is interesting to note that such a laboratory centre can adopt equipment and methods that otherwise would be impractical or too costly. This can help to develop and test methods that are specially suited to screening .. and for detecting diseases in a very early stage. Such possibilities for research are certainly most welcome. - 147- 8. Conclusions Reviewing the subject of early disease detection~ it is clear to us that, although for a number of years there has been an increasing interest both among the medical profession and the public, we are still at a very early and comparatively primitive stage in the systematic detection and treatment of early disease. For some conditions we have powerful methods of detection, but we do not yet know the effect of early treatment (e.g. diabetes mellitus); for others we are still experimenting to find satisfactory tests (e.g. chronic simple glaucoma); for only a relatively few condi- tions are there already established well-tested and successful means of pre-sympto- matic detection and treatment (e.g. cancer of the cervix, and even here there'are certairi qualifications). It is important therefore to ask ourselves in what particular respects is further investigation, promotion, or education needed; and, since these are big problems affecting whole populations and carrying the. implications of radical changes in emphasis in the practice of medicine in the direction of prevention, what part WHO might play as catalyst in this reaction. It may be helpful to consider some of the possible ways in which these aims may be realized, point by point. 8.1 The need for further epidemiological investigations and, allied .witb thi.$, · the need tp>,achieve standardizat;i,on ;(' Tooof:ten~ in,the past, work has been undertaken in different countri~s pn the ' \ . ' same ostensible condition, only for it to be found at a late stage th,at the workers •. . .. ' were using differing definitions,. This has happened, for instance~ over the condi- tions of carcinoma-in-situ of the cervi~ uteri and of glaucoma simplex. At the same time there is so much work to be done where there is need to examine large populations at a low risk for the condition under ~tudy, that there would be great advantage 'if' results could be pooled or at least intelligently compared. WHO has already played a large part in fostering co-operative international studies; we suggest there is yet room for extending this role. As we have seen~ WHO e~rt committees or seminars have considered many of the chronic diseases, notably tuberculosis, non-specific respiratory disease, anaemia, malnutrition, high blood-pressure~ ischaemic heart disease, diabetes mellitus, mental illness and the cancers. In some instances (e.g. anaemia, cardio-vascular disease and diabetes) special attention has been paid to reaching internationally acceptable definitions. TWo conditions, we believe, still needing internationally accepted definitions are carcinoma-in-situ of the cervix and glaucoma simplex. - 148- Ofcourse, rElaqhing acceptable definitions, diagnosing and then discovering the effect of treatment are lengthy processes entailing much painstaking work by teams of investigators. The needs of establishing the sensitivity and specificity of screening tests and the effect of early treatment are the reasons for the epidemiological surveys to which we have paid such attention in this paper. Surveys are, however, undertaken for a number of reasons, but the object of investigating the feasibility of pre-symptomatic diagnosis and treatment may not always be one of the reasons. Wr10 could, in our view, play a valuable part in both acting as a clearing house for information about epidemiological studies directed towards the development of screening techniques; and in acting as a watchdog to remind those considering studies that future screening possibilities might be borne in mind during the planning stages. While there is, in our opinion, everything to be said for the value of screening being developed by ~ hoc studies of individual conditions, so broad is the field of early disease detection that we consider there may be a place for WHO itself to keep the whole field under continuous review, stimulating and fertilizing here and there as the need appears. Conditions for which, in our view, more organized epidemiological work is· urgently needed are:- 8.1.1 Carcinoma of the uterus: more statistically controlled-trials of the effect of cervical cytology on mortality are badly needed. At present we are largely relying on British Columbia~ some United States cities like Memphis, San Diego and Louisville, Norwegian and British studies, all of iihich have drawbacks either of size of population or epidemiological acceptability. · 8.1.2 Glaucoma simplex: two or three small studies have suggested that ,previously accepted screening criteria are inadequate. are needed. More similar, and comparable, studies 8.1.3 Mental illness: is quantitatively a vast world problem yet little can at present be done about its early detection and treatment. Surveys aimed at defini- tions of early disease, acceptable diagnostic techniques, and incorporating controlled trials of treatment are badly needed. - 149 - 8.1.4 Asymptomatic bacteriuria: there is at least a suggestion that undetected ... urinary infections early in life may lie at the root of much crippling arterial hypertension in later life. Early diagnosis and prompt treatment could potentially have a very important impact. Surveys are ~eing carried out but, in view of the importance of the ultimate condition, the matter should perhaps be attacked on a , .... larger scale. 8.1..5 Cancer of the breast: we. are still ignorant of the effect of ma,king the effort to diagnose b.reast cancer before symptoms are reported, although advances in clinical treatment have had disappointing results and mortality remains little affected •. At least one large-scale mammography su_rvey is in progress. In view of th~ :l,()n& time that must elapse before an effect. on deaths can be seen, it. may perhaps be wise to ensure that valid results will be obtained by at least duplica;tipg the work elsewhere. The preventive value of breast self-examination has never b_een, adequately assessed, as far as we are aware, and~ since this is much cheaper, safer and generally more practicable than X-ray mammography, it seems most important that its value (or otherwise) should be known for ~el:'tain. If the technique were shown to be t'l:'tily beneficial it' should become much easier to propagate' its wider spread. ;::·l'-Lt':..:~ c:.". ; 8.1.6 Lung cancer~ the appalling prognosis for lung cancer even when detected at ·. <,;:, the earliest possible stage radiologically calls for more work directed towards • ' - i .. :~ ' .•. .. : . ·. better screening methods and, _if possible, prevention. ~ ,. ) : .. ~ :·· . ' '' ' ·' It seems likely that cyto- logical examination will show abnormal cells in the sputum at a pre-cancerous stage . ..... . .. , !.'· . \'<hich could lea~ to locali~ation of cancer in a bro~chu~ .. be~ore it appears radio- logically. There is need for more work on these lines, wi.th agreed standards of . •"'' " .;· • :•< ... . . cytological nomenclature, diagnostic procedures and follow-up arrangements. We consider WHO might usefully act in stimulating and co-ordinating work along these, and perhaps other, lines. 8.2 Recof'ds· For the ftelective screeniug of high-risk groups of the population a sine qua non is the ability to know who ··c6~stitutes these groups. It may, of course, be enough to issue a genera:! iri\riia:tion to a particular group (e.g. adult women for cervical cytology); but there :i::{ alWays a risk that those persons at highest risk may opt out. lt is 'there~iore often more successful (as well as better public relations) to - 150- ask people as individuals to attend. In the type of surveillance from a health centre or group general practice~ to which we have refer!'ed, individual invitations are~ essential. In order to draw particular-groups in large numbers from the population~ records capable of being dealt with by automatic data handling methods are needed. In practice this can be difficult, both from the point of view of the design of a suitable record and what should be put on it; and from the point of view of confidentiality. In general practice, for example, the patient's medical record is likely_only to be handled by doctors or those in his close c.onfidence. But for the future, it seems likely we must expect general practitioners' records to be handled at data processing centres. How is this to be done without . a breach of co!lfidentiality? It appears to us, therefore, that records the desigp and handling is an important and urgent problem and that WHO might usefully consider this question. 8.3 Economics Considering its importance surprisingly little is known about the economics of early disease detection. Most probably this is because screening has so far been largelyexperimental and the question of its economic cost has not arisen. Certainly, for screening for tuberculosis, a good deal is known about the cost since this was introduced as a service long ago. It would be valuable to study the cost of screening for· different conditions, either alone or in combination, tinder differing s~stems of medical care. Undoubtedly the arrangements for. medical' care in some. countries are at present more suited to early disease detection than those in other countries. Initiating comparative, and comparable, economic studies of this sort is something WHO might consider useful to undertake. 8.4 Education Practicable techniques for screening for early disease are now available but still, as a profession, we continue to see clinical conditions for the first time at a-late·· stage of development when treatment is less likely to be successful. Also clirticiaris,· in.ge~eral, are too oriented towards conventional diagposis and treatrri~nt to think readily in preventive terms. · Thus there is much room for improving both the attitude of the public and the attitude of the profession to the early detection of disease, as we have pointed out in Section 5.6. Hitherto - 151 - diagnosis has been the province of the clinician who has been concerned mainly with the individual patient. It has been possible to have a high standard of clinigal practice co-existing with poor overall medical care due to unequal distribution of ' -· . resource-s: ... rmf)iiciF_iii ... eariy~disease ~et~-<::~~-0.1?-:·.JcervicS;l cancer screening# for example) is the idea of extension of preventive clinical medical services to a whole community and this in turn postulates a new way of thinking in the-medical and ancill.a.:ry .. prQfess-:Lons ... The- medical pro-fession itself# and the public# may therefore be on the verge of a revolution in its attitude to clinical (as opposed to traditional preventive) medicine and there should be good opportunities for WHO to influence the course of medical and public education. -so as to further# to the best advantage# this changi.tlg approach. New departments of general practice# postgraduate institutes# • , I ~- " ' . ·.· . . . departments 'of soclar med.ic:fri.e and the like. are growing up. These should facilitate the training of doctors and ~issemination of new ideas through attachments and -~~J~ . . . .. feliowsnips~--iri. all of which 'WHO" cail-play a part both by facilitating interchanges and_ through the influence of its committees. - 152 - REFERENCES Chapters 1-5 1. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass.; Harvard University Press, Chapter 5, p. 45 2. World Health Organization, Regional Committee for Europe (1964) The Pre-symptomatic Diagnosis of Diseases by Organized Screening Procedures, EUR/ Rc 14/ techn. Disc., (unpublished working document) 3. Dawber, T. R., Moore, F. E. & Mann, G. V. (1957) Coronary Heart Diseases in the Framingham Study, Am. J. Publ. Health, 47, Suppl., p. 4 4. Council on Medical Service, American Medical Association (1955) A Study of Multiple Screening: Descriptive Data on Thirty-three Screening Surveys, (Revised) Chicago 5. Commission on Chronic Illness (1957-1959) Chronic Illness in the United States, Vol. I-IV, Cambridge, Mass., Harvard University Press 6. Breslow, L. (1955) Multiphasic Scree·ning in California, J. Chron. Dis., g, 375 7. American Public Health Association (196o) Chronic Disease and Rehabilitation: A Programme Guide for State and Local Health Authorities, The American Public Health Association, Inc., New York 8. Kurlander, A. B. & Carroll, B. E. (1953) Case-finding through Multiple Screening, Publ. Health Reports, 68, 1035 9. Wilson, J. M. G. (1962) Report on Multiphasic Screening, Wld Hlth Org. Rep. Travelling Fellowship to the USA, mimeographed paper, 62 R/UK - 13 10. Chronic Illness & Aging Unit, Bureau of Chronic Diseases, State Department of Public Health (1963) Bibliography on Disease Detection: Health Maintenance, Periodic Health Examination and Multiphasic Screening, Berkeley, Calif., mimeographed document 11. Chapman, A. L. (1949) The Concept of Multiphasic Screening, Publ. Health Reports, 64, 1311 12. Mountin, J. W. (1950) Multiple Screeening and Specialised Programmes, Publ. Health Reports, 65, 1359 13. Smillie, W. G. (1952) Multiple Screening, Am. J. Publ. Health, 42, 255 14. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 48 - 153 - 15. Remein~ Q. R. & Wilkerson,_ H. L .• (1961) The Efficiency of Screening Tests for . piabet~s, J-•. Chron •. Dis._,. :]..3, 6 16. Holland, W. W_. & Humerfelt, S. (1964) Comparison of Blood Pressure: Comparison of Intra-Arterial arid-·cuffValues, Brit. med. J., ii, 1241 17. Lanc-et· (T966) Measuring Blood-Pre-~sU:r-e_! Leadi~ J1.~~~-?.£.e~ Lancet; !· p. 414 18. Elwood, P. c. & Jacobs, A. (1966) Haemoglobin Estimation: A Comparison of Different Tecbrlique9, :Srit. med. J.,·i, 20 . -· ~--···-~--· ·-··~ 20. 21. 22. 24. 25. 26. 27. Thorner, R. M• & R~mein, Q. R. (1961) US Department of Health, Education and Welfare, Principles and Procedure§in the_Evaluationof Screening for Disease, Public Health Monograph No. 67, Public Health Service Publication No. 846, Washington, D.C. - Hollows, F. C. & Graham, __ P. A. (1966) The Ferndale Glaucoma Survey, in: criiiuc-oma, 'Ei:iideirifoiogy~- EarlY. Diagnosis and Some Aspect's of Treatment; Proceedings of a Symposium held at the Royal College of Surgeons of England,_ p. _ 24, Edinburgh, E. & ·S. Livingstone Graham, P. A. & Hollows, F. C. (1966) ibid, p. 103 Davis, H. J. (I965),The ·:rr:dgation Smear: Results in Population Screening for- Cervfca:r·c-a.ncer, --find International Congress of Exfoliative Cytology, Paris Davis~ _H. -J: (1962} -The ":trr"igation Smear - A Cytologic Method for Mass PopUiatlori-Screening by Mail, Am·~-;r. o'bstet~ GynecoL, 84, 1017 Butterfield, w. J. H., Keen, H. & Sharp, c. L. E. H._ (l964).Diabetes Survey in Bedford, l96~,.Proc. Roy. Soc. Med., 57, 193 United States Department of Health, Education and Welfare'(l964). Glucose Tolerance of Adu1ts, United States, 196o-1962, Diabetes Prevalence and Results-of ·a-Glucose Tolerance Test, by Age and Sex, Nat:'..onal Center for Health Statistics, Public Health Service Publication., No. 1000, Series 11, No. 2, Washington, D.C. St8.!lller., J., Berkson, ., Young, Q. D., Hall, Y. &-Miller, w. (1963) Approaches to the Primary Prevention of Clinical Hear:t T!.i~ease ~n :H:i,@-risk, Middle- aged Men, Ann. New York Acad. Med., 97, 932 Jolliffe, N., Rinzler, S. H. & Archer, M. (1959) The Anti-Coron,a.cy--_Cl_Ub: Including a Discussion of the Effects of a Prudent Diet on the Serum Ch~~~st:::-ol LeV:e~- _'?f_ Middle:..aged Men, Am. J. Clin. Nutr., l• 451 - 154 - 28. Christakis, G., Rinzler, s. H., Archer, M .• , Winslow, G., Jampel, S., Stephenson, J., Friedman, G., Fein, H., Kra._us~_ A;, _Gt._James, G. (1966) The Anti-Coronary Club: A Dietary Approach to the Prevention of Coronary Heart Disease - A Seven-Year Report, A.m. J. Publ. Health, 56, 299 . 29. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 65 )0. Ibid, Vol. IV, Chronic Illness in a Large City, p. 262, 448 31. Weinerman, E. R., Breslow, L., Belloc,.N. B., Waybur, A. & Milmore, B. K. (1952) Multiple Screening of Longshoremen with Organised Medical Follow-up, 1\·m:· i:- Pu"bi.' lleaith,· 42, 1552 - -- · --- · ·- · -- - · 32 New York State Department of Health, Bureau of Chronic Disease & Geriatrics, Annual Report (196o) Report on Cortland.County Screening Demonstration · -(~publisEe<l ·r-ep()~_~Y_ _ __ -····-33. Fa.no; R. M. & Corbato,- F • • Y. (1966) Time-sharing on Computers, Scientific American, 215:3, 129 34. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. IV, Chronic Illness in a Large City,The Baltimore Study, Cambridge, -Mass"., ·Harvard University Press 35. Commission on Chronic Illness (1959) Chronic Illness in the United States, Vol. III, Chronic Illness in a Rural Area, The Hunterdon Study, Cambridge, Mass., _Harvard University Press 36. Beuchley, R. W., Robert, M. A., Dr~e, M. &.Breslow, .L. (l-958) Height, Weight and Mortality in a Population of Longsh9remen, J. Chron. Dis., 1, 363 37. Borhani, 0. N., Hechter, H. H. & Breslow, L. (1963) Report of a Ten-Year ·---Follow-up Study of ~h~--~~1:1 ~a.nci~c?_ IJ<:)~~~~o:s~~~~~--J. Chron Dis_., 16, 1251 38. Wylie, c. M. (1961) PartfciP"ation-I:n-·a: iviuitipie screerl.ing~-ciinic-witil Five- Year Follow-up, Publ. Health Reports, 76, 596 39. Wylie, c. M. (1961) Use of_Death Rat~s in Evaluating Multiple Screening, Publ. Health Reports, 76, llll 40. Stamler, J. (1962) Chronic Disease Control -bivision~-Chicago Board of Health, Ann. Report, p. 12, mimeographeq document. 41. Commission oq._Chron:i,c Illnes:;> (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 29 - 155 - 42. Franco, S. c., Gerl, A. J. & Murphy, ·a. T. (196l) ... P.eriodic Healtl:LExam,ip.~t:i,o_n~: a Long Term Study, 1949-1959, J. Occ. Med., L• 13 43. United States Department of Health, Education andWe.lfare (19q3) :P~:r:-iodic Health Examinations - Abstracts from the Literature, Public Health Service, Division of Occupational Health,Publication, No. 1010, Washington, D.C. 44. ·Commi--ssion on Chronic Illness (1957) Chronic I-llness in the United States, Vol. I, Secondary Prevention Tlwough Perio.dic Healt.h..E~~nations, Cambridge, Mass., Harvard University Press, p. 28 45. Williamson, J ., Stokoe, I. H., Gray, s., Fisher, M., Smith, A., McGhee, A. & Stephenson, E. (1964) Old People at Home, Lancet, _!, 1117 46. California's Health (1961) Biennial Report Edition, 19, No. 11 47~ · -1N1;1son;, J • M. G. (1965) Some Aspects· of the Epidemiology of Cervical Cancer, 'Month. Bull. Min. Health & Publ. Health Lab. Serv., 24, 72 _ . ·48' •. '6rundy;-- F ;· & Mackintosh, J. M. {1957) The Teaching of Hygiene and Public . . Health in Europe, Wld Hlth Org.,.:.Monogr. Ser., 34, .. 109' 49. World Health Organization (1964) Promotion of Medical Practitioners Interest in Preventive Medicine:-. rrwe·lfth Report of the WHO Expert ·Committee on . Professional and Technical Education of·Medical and Awdli~ryPersonnel, Wld Hl th Org. techn. Rep. Ser., 269 50. Brockington~ ·F.-,. Silver, G. & Vuletic, A. (1964) Teaching of the Medical ... · _Student for Comprehensive Medical Practice with emphasis on extramedical institutions) WHO Study, ~ffiO Educ 126, mimeographed document 51. College of General Practitioners (1965) Present State and Future Needs of General Practice, London, Council of the College of General Practitioners 52~ Br'eslow, L. & Hochstilll, J. R. (1964) Sociocultural Aspects·of·Cervical Cytology in Alameda County, Calif., Publ. Health Reports, 79, 107 53. Kegeles, s. S., Kirscht; "J.P., Haefner, D.P. & nosenstock, I .. M.:::.(1965) Survey of Beliefs About Cancer Detection and Taking Papaniqo::l:aou Tests., Publ. Health Reports, §£, 815 54. Wakefield, J. & Baric, L. (1965}= Public. and P:r:'ofessional Attitudes -to a Screening Programme for the Prevention of Cancer of the Uterine Cervix., Brl t.' J.. Prev·. ·Soc. Med •, 1:2, 151 55. Cebtl:-al Health Services Council (1964) Health Education: Report of a Joint Committe~ of the Central and Scottish Health Serv.iQes.Gouncils, p. 10, London,. H.M.s.o. - 156- Section 6~1 Diabetes Mellitus 56. Joslin, E. P., Root, H. F., Wb.fte,-- P~ & Marble, A. (1959) The Treatment of Diabetes Mellitus, (lOth Edition) London., Henry Kimpton, p. 19 57. Wilkerson, H. L. c. & Krall, ·1~ -P.- (1947) Diabetes in a.· New E~gland Town, J. Am. Med. Assoc., .ill,.. 209 58. Wa,.l~e.I', J ~ B. & Kerri~e, D.. (1961) Diabetes in an English Community - A Study of its Incidence an(~~t~ai fl1.~~ocy, Leicester UnivBrsity Press 59. McDonald, G. vJ., Fisher, G. F. & Pentz, P. C. (1965) Diabet"es Screening Activities, July 1958 to June 1963, Publ. Health Reports, 80, 163 60. United States Department of Health, Education & Welfare (1961) Diabetes Fact Book, Public Health Serv.i~~ Publication No. 890, Washington 61. . United States Department of Health, Education & Welfare (196o) Diabetes Program' Guide, Public .Health Serviqe Publication No. 506,· Wa~~gton, p. 51 --·· -~--. -· ····---·-----~-·-··--··--· •····· ··-·.. . . .. . 62. Conimission on Chronic Illness (1957) Chronic Illness in the United States., Vol. IV, Appendix D, Chronic.Illness in a Large Citl, Cambridge, Mass., Harvard University Press, p. 466 63. Reid, J. J. R. (1962) Diabetes Mellitus - Certain Aspects of Preventive, Detection and Treatment, Wld Hlth Org. Rep. Trav~lling Fellowship to the USA, mimeographed paper, 62 R/UK - 13 64, R~J_Q,,t __ ~. J .•.. ~~ (1960) __ Public Knowledge of Diabete~, Med. Off., 103,325 , .. 65. R~dll~-~d. .. ··-i. H~ · (1960}Incidence of ·aiycosuria and. Diabetes Mellitus in a General Practice, Brit. med. J., i, 695 66. Coll~g~ -~f- Gen~~al Pr~ctitioners (1962) A Diabetic Survey, Brit •. med. J ... i, 1947 67. Kirkland, J. A. & Morgan, H. G. (1961) An Assessment of Routine Hospital Urine Testing for. _!'rot_e_il?-__ ~ _ _q_!ucose, Scot. med. J., 6, 513 68. Rundles, R. w. (1945) Diabetic Neuropathy, General Review ~th Report of 125 cases, :Medicine, Baltimore, 24, 111 69. Garland, H. (196o) The Neurological Complications of Diabetes Mellitus, 'Clifiical Aspects., Proc. Roy. Soc. Med ... ..5.2_,. 137 ~ ... ' , . 70. Ashton, N. (1959) Diabetic Retinopathy, A New Approach, Lancet ... .li.,, 625 71. ..~.:.~l_o.p_, .P.• .. M. ( 1954) ·Are Diabetic ·Degenerative Complications Pr~:vep.table? Brit. ~d. J ~ 1 .:I.i~~~-~83---_- ·-- · _ ........ -_ .::.: .......... :. 72. Marble, A. (1955) Coronary Artery Disease in the Diabetic, Diabetes, 1 .. 290 74. 75. 76. 7.7· 78. 80. 81. 82. 8). 84. 85. 86. 88. - 157 - Johnssen, S. (196o) Retinopathy and Neuropathy in Diabetes Mellitus, Comparison of the Effects of Two Forms of Treatment, I)la~.!-~s ... ,.2 .. 1 Wo],ff, 0. lh ,& Sal:t;,_ H. B. (1958) Serum-Lipids and Blood Sugar .Levels in Childhood Diabetes, Lancet, i, 707 Keen, H. (1959) Paper read to Medical & Scientific Section, British Diabetic As:sociation Annua.":!, Clinical Meeting at Guy's Hoqpital, London 0 ~ H •''' • Oo •-• ---·-· ~'' ' '···--·-·• '0 '' 0 '' ••• '- Newburgh, ~. H. :& Conn;-·;r~. w: ( 1939) A New Interpretation of Hyperglycaemia in Obese, Middle-aged Persons, J. Am. Med. Assoc., ]lg, 7 ~cet (1965) Obesity & Diabetes, Leading Article, Lancet, _!, ~60 Root, H. F., ~lirsky, s. & Ditzel, J. (1959) Proliferative Retinopathy in Diabetes Mell;itu:s, :Rev;iew of Eigtlt Hundred & Forty-Seven Cases, J. Am. Med. Assoc., 169, 903 ...... __ _ ;Beckett, A. G. (1962) Ha.I;'111ful Effects of pelay inDiagn,osi.s .of Pi.abe.tes Mell,itus, The Prac.tit,ioner, 189, 57 . . .·r. Brown, I. K. & Jones, A •. T. (19q4) .Retinopathy .and Diabe~i~ .Con.tro:l, Brit. J. QPht~l., .. 48, 148 .· ' . . . Collyer, R. T. & Hazlett, B. E. (1961) Retinopathy and :NeuropathY· in One Hundred Growth-Onset Diabetic Patients, Canad. Med. Ass., I, ~, 1328 Daysog, A., jr, Dobson, H. L. & Brennan, J. c. (1961) Renal Glomerular & Vascular Lesions in Pre-Diabetes and in Diabetes Mellitus, A Study Based on Renal Biopsies, Ann. Int. Med., 2&, 672 Hoet, J. P. (1954) Carbohydrate Metabolism During Pregnancy, Diabetes, 2_, 1 Wilkerson, H. L. C. (1959) Pregnancy and the Prediabetic State, Ann. New York, Acad. Sci., 82, 219 Fajans, s. S. & Conn, J. W. (1954) An Approach to the Prediction of Diabetes Mellitus by Modification of the Glucose Tolerance Test with Cortisone, Diabetes, 2 .. 296 White, P., Joslin, E. P., Root, H. F. & Marble, A. (1959) Treatment of Diabetes Mellitus, (lOth Edition) London, Henry Kimpton, p. 48 Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p~ 166 Ford, N. J. & Glenn, B. (1951) Undetected Diabetes Among the Relatives of Diabetics, Sth. Med. J., 44, 239 - 158 - 89. Harris, H. (1947-1949) The Incidence of Parental Consangiiinity in Diabetes Mellitus~_ .Ann. Eugen., 14, 293 90. Harris, H •. (1950) The Familial Distribution of .Diabetes Mellitus, Ann. Eugen ... ~ .. 95 91. United States of America Department of Health, Education & Welfare (1964) Gluco::;e Tolerance of Adults,.· United States, 1960-62, Diabetes Prevalence and Results of a Gluccse Tolerance Test by Age & Sex, National Center for Health Statistics, Public Health Service, .Publication No. 1000, Series·ll, No. 2, Washington D~C. 92. British Medical Journal (1965) Inheritance of Diabetes Mellitus, Brit. med. ·J., _!, 940 93. World .. Health Organization (1965) Diabetes Mellitus.; Report of a WHO Expert Committee, Wld Hlth Org. tecl:m. Rep. Ser .... 310 ... · .......... - .. , .. 94. Keen, H.·_, Rose.. G., Pyke, D. A. .. Boyns, D. .. Chlouverakis, C. & Mistry.. S. ( 1965) Blood Sugar and Arterial Disease_, Lancet_, ii, 505 .. 95. Ostrander, L. D • ., Francis, T ... Hayner_, N. s ... .Kjelsberg, M. 0. & Epstein, F. H. · - (1965) The Relationship of Cardiovascular Disease to Hyperglycaemia, Ann. Int. Med., 62., 1188 - 159 - Section 6.2 Heart Disease .. 96. Registrar General (1966) The Registrar General's Statistical Review for the Year 1964, Part I, Tables, IVIedical, London, H.M.S.O. 97. Logan, W. P_. D. &_ Cus}:li()n, A~ .A •. {J95§) Studies on Medical and Population Sub,lects, No. 14, Vol. I, Morbidity Statistics from General Practice, London, H.M.S.O. - .. 98. Registrar General (1962) The Registrar General's Statistical Review for the Year 1960, Part I, Tables, Medical, London, H.M.S.O. 99. United States Department of Health, Education and Welfare (1964) Heart Disease in Adults, United States, 1960-1962, Public Health Service Publication, No. 6, Series 11, Washington, D.C. 100. World Health Organization (1965) The Work of the WHO in Cardicvascular Diseases, 1959:1964, (unpublished report) 101. World Health Organizat~on, Car~iovascular Diseases Unit, WHO, Geneva, to be published 102. Miller, R. A., Smith, J., Stamler, J., Hahneman, B., Paul, M. H., Abrams, J., Hai~, G., Edelman, J., Willard, J. & Stevens, w. (1962) The Detecting of Heart Disease in Children - Results of a Mass Field Trial with Use of Tape-Recorded Heart Sounds, Circulation, ~~ 85 103. Wilson, J. M. G. & Heasman, M. A. (1959) Coronary Artery Disease- An Epidemiological Review, Month. Bull. Hin. Hlth. & Publ. Hlth. Lab. Serv., 18, 94 104. · Kennedy, A. C. (1957) Observations on the Incidence of Coronary Heart Disease in a Rural Are.a in S. W. Scotland, Scot. med. J., .£, 420 105. Epstein, F. H., Ostrander, L. D., Johnson, B. C., Payne, M. vJ., Hayner, N. S., Keller, J. B. & Francis, T. (1965) Epidemiological Studies of Cardio- vascular Disease ip a Total Community - Tecumseh - Michigai.l., Ann. Int. Med., 62, 1170 106. Epstein, F. H. (1965) ':I'he Epidemiology of Coronary Heart Disease, J. Chron. Dis., 1§, 735 107. World Health OrganJ.iation (1963) Survey of the Prevalence of Ischaemic Heart Diseases in Certain European Countries - Report of a Technical Meeting convened. by the Regional Office -for Europe of the WHO, Copenhagen, 1963, EURO, 179.3, mimeographed document 108. Rose, G. A. (1962) The Diagnosis of Ischaemic Heart Pain and Intermittent Claudicat.ion~. in Field _Surveys, Bull. Wld Hlth Org., n_, 645 - 160 - 109. World Health Organization (1962) Arterial Hypertension and Ischaemic Heart. Disease, (Preventive & Therapeutic Measures) Wld Hlth Org. tecbn. Rep. Ser., 231.1! 19 . 110. Burgess, A. N., Fejfar, Z. & Kagan, A. (1963) Arterial Hypertension and Ischaemic Heart Disease - Comparison in Epidemiological Studies, WHO, Geneva 111. Kagan, A. (1965) Comparison of Electrocardiographic Data with Particular Reference to the WHO Minnesota Code Trial, Millbank Nemorial Quarterly, USA, .!2 112. Rose, G. A. Personal communication 113, Stokes, J. & Dawber, T. R. (1959) The Silent Coronary: The Frequency and Clinical Characteristics of Unrecognised Myocardial Infarction in the .. ____ "...:::. __ --~ra.nDJlgham Study, Ann. Int. Med., 2Q, 1359 114. Acheson, R. M. & Acheson, E. D. (1958) Coronary and Other Heart Disease in a Group of Males aged 65-85, Brit. J. Prev. Soc. Med., 12, 147 115. Hinkle, L. E., Carver, S., Benjamin, B., Christenson, W. & Strone, B. (1964) Studies in Ecology of Coronary Heart Disease; i, Variations in the Human Electrocardiogram U."1der Conditions of Daily Life, Archiv. Environ. Health, 9, 14 116. Paul, 0., Lepper, M. H., Phelan, W. H., Dupertnis, G. W., MacMillan, A., McKean, H. & Park, H. (1963) A Longitudinal Study of Coronary Heart Dis~~e,. Circulation,. 28, .20 117. Kannel, W. B., Dawber, T. R., Kagan, A., Revotskie, W. & Stokes, J. (1961) -__ , Factors of Risk in the Development of Coronary Heart Disease - Six Year -· Follow-up Experience ofthe Framingham .Study, Ann. Int. Med., _22, 33 118. Dawber, T. R., Kannel, W. B., Love, D. E. & Streeper, R. B. (-1952) The Electrocardiogram in Heart Disease Detection - A Comparison of the Multiple and Single Lead Procedures, Circulation, 2 .. 559 119. Thomas, A. J., Cochrane, A. L. & Higgins, I. T. T. (1958) The Measurement of the Prevalence of Ischaemic Heart Disease, Lancet, ii, 540 120. Blackburn, H., Keys, A., Sir.Jon.sor", E., Rantakarju, P. & Punsar, S. (l96o) . -~~_}!:lectrocardiogram in Population Studies, Circulation, 21, 1160 121. Thompson, W. B., Hudmet, H. B.,jr,Russo, P. E., Brown, F. 'R •. 'i:Mostl..ey, K. T. (1961) A Review and Study of Cardiovascular Disease Screen~ng with the Miniature Chest X-ray, J. Chron. Dis.,~ .. 148 122. Cooper, G. R. (1965) International Comparability of Blood Lipids, Millbank Memorial Fund Quarterly, ii, .!"2, 49 -161 - 123. Oliver, M. F. & Stuart-Harris, (1965) Experiments on Animals, Brit. med. J., ii, 1203 124. Oliver, M. F. & Boyd, G. s. (1959) Effect of Bilateral Ovariectomy on Coronary- Artery Disease and Serum-Lipid Levels, Lancet, ii, 690 125. Morris, J. N. & Heady_, J. A. (1953) Coronary Heart Disease and Physical Activity of Work, Brit. med. J., ii_, 10?5, 1111 126. Morris, J. N. & Crawford, M. D. (1958) Coronary Heart Disease and Physical Activity of Work, Brit. med. J., ii, 1485 127. Registrar General (1958) Registrar General's Statistical Review, Decennial Supplement on Occupational Nortality 1949-53, Part II, London, H.H.s.o. 128. Doyle, J. T., Heslin, A. s., Hillebal, H. E. & Formel, P. F. (1959) Early Diagnosis of Ischaemia Heart Disease, New Eng. J. Med., 261, 1096 129. Gertler, M. M. & White, P. D. (1954) Coronary Heart Disease in Young Adults, Cambridge, Mass., Harvard University Press 130. Spain, D. M., Nathan, D. J. & Gellis, M. ( 1963) Weight, Body Type and the Prevalence of Coronary Atherosclerotic Heart Disease in -Males, Am. J. Med. Sci. , 245, 63 131. Doll, W. R • .& Hill, A. B. (1956) Iung Cancer and Other Causes. of Death in Relation to Smoking, Brit. med. J., 11, 1071 -. 132. Hamnond, E. C. & Horn, D. (1958) Smoking and Death Rates - Report on Forty- four Months of Follow-up of 187 -783 Men, J. Am. Med. Assoc., 166, 1159, 1294 133. Dawber, T. R., Kannel, W. B., Revotskie_, N•, Stokes, J., Kag~_,A. ·&Gordon, T. (1959) Some Factors Associated with the Development of Coronary Heart Disease (Six Years Follow-Up Experience in the Framingham Study) Am. J. Publ. Health_, ~~ 1349 134. Doyle, J. T., Dawber, T. R._, Kannel, W. B., Kinch, S. H. & Kahn, H. A. (1964) The Relationship of Cigarette Smoking to Coronary Heart Disease, J. Am. Med. Assoc._, 190, 886 135. Kannel, W. B. (1964) Editorial: Cigarette Smoking and Coronary Heart Disease_, Ann. Int. Med., 60, 1103 136. World Health Organization (1965) Working Group an Studies of Preventive Measures in Ischaemia Heart Disease - Summary of Discussion_, Mimeographed document_, EURO- 179.3 137. Groom, D. (1961) Population Studies of Atherosclerosis_, Ann. Int. Med._, 221 51 - 162 - 138. Smirk, F. H. (1957) Hi.sh. Arterial Pressure, Oxford, Blackwill Scientific Publications, p. 687 139. Hodge, J. V. &Dollery, C. T. (1964) Retinal Soft E.xudates, .21!.~d-~., 2_2, (N,S.) 117 140. Pickering, G. W. (1965) Hyperpiesis: High Blood I1:•essure Without Evident Cause: Essential Hypertension, Brit. ~~d.~., Ji, 959, 1q21 141. Kain, H. K., Hinman, A. T. & Sokolow, M. (1964) Arterial Blood Pressure IJieasurement with a Portable Recorder in Hypertensive Patients; Iv variability and Correlation with "Casual 11 Pressures, Circulati..P~ )_Q, 882 142. Rose, G. A., Eollal1d, W. W. & Crowley,; E. A. (1964) A Sphygmomanometer for Epidemiologists, J.,ance,:t, l-_, 296 143. Boe, J. ~ -Hi..unerfelt, s. & Weden:~varig-:F. ·.(1957) The Blood Pressure in a Population, Acta Med. B_q§.l1d. Suppl., 321, Bergen . . 144. . Iii:i,all, W. E. & Oldham, P. D. (1958) The Inheritance of Arterial Blood Pressure, Clin. Sci., 11., 404 145. Hamilton, M., Piclcering, G. W., Roberts, J. A. F. &: Sowry, G. S. C. (1954) The Aetiology of Essential Hypertension I, Tne Arterial ~ressure in the General Population, p_lin. Sci.~- .J..l, 11 146. Kagan, A •. , (Jord_9n, T., K~el, W. B. & Dawber, T. R. (1959) Blood .tre~sure and its Relation -to_ .9.9}-"'.onary Hear~ Dis_Et~_ip._B~ Fr~ l'_9.Q11_l.a_lj_on,: HYperten~.s>_r:~,. Vol. III, American Heart Association 147. .United States Department of Health, Education & Uelfare (1964) ID_o_o.§_ Pr~S.§.ill'S, of Adults by __ jt~,.?1ld §_eb United_Si:;.?-.:t~e.s..,._J_$)6Q~-:- l_9..§g, National Centre for Health Statistics, Public Health Service ·Publication, No. 1000, Ser. 11, No. 4, ltJash:i,ngton, D.C. 148. Pickering, G. W. (1955) High Blood Pressure_, London, J .• &; A. Ohurc.hill, p. 299 149. Bechg('l.ard, P. (1946) Arterial Hypertension: a Follow-Up Study ,of One Thousand Hypertonies, Acta T1ed_~l?..9..§ll<!·, .Suppl., 172 .:.;:::·· .... - - 163 - 150. Ungerleider, H. E. & Gubner, R. S. (1958) 1J.J~_l}.§~~~p. MedicJ.ne ... Springfield, Illinois, Charles C. Thomas l5L Kemsley, W. F. F., Dillewicz, W. Z. & Thomson, A. M. (1962) A New· Height-for- Height Standard Based on British Anthropometric Data, Brit. J. Prev_. Soc. ~ ... 16, 189 152 ~ Edwards, D. A. w., Hammond, w. H... Healy I M. J. R., Tanner, J. 111. & vJhi tehouse, R. H. (1955) Desi0n and Accuracy of Oalipern for Measuring Subcutaneous Tissue Thic:~ess, f)r~_. _ __.;r_. Nutr., 2,, 133 153. United States Department of Agriculture, AGriculttu~al Research Service. Human Nutrition Research Division, (l96C) Heizj:l_ts and it/eights of_j\..Q...u..lJ..S. in the Unit_~l St_a_t_e.s ... Home Economics Research Report, No. 10, vlashington 154. Commission on Chronic Illness (1957) Chronic Illp~ss in the UniteSL_State~ Vol. I, Rr_eyep_t_i_op_of_ .. Qb.!'.9.J!.:i~:!Jle_9.§ .... CarJbridge, ll1ass ... Harvard University Press, p. 267 · 155. British Medic.al Journal (1964) Leading article, f3i"it. med. J., .;hi, 1543 - 164 - 156. General Register Office (1957) Tuberculos:h_s S1._ati_s.ti_(fs for ~~-'3ll.Clli~l.e§.. 12)8 .:.. ):95.5 .... studies on Medical and Population, Subjects No. 10, London, H.M.S.O. 157. \vorld .Health Organization (1964) W.H.O. Expert Committee on Tuberculosis, Wld Hl th Or_g_._j;_e..f!lli.1.-:..Rep. Ser ., 290 158. Springett, V. H. ( 1956) _t1inj.rnal Pulmon~'!'..u~""'.£.ll!.o.sJ..e_..£2und by_ rlla.§..lLR_a:_c:l.i..OU.§l..P~ London, Lewis 159. Yerushalmi, J. (1953) The Reliability of Chest Roentgenography and its Clinical Implications, P.J...§._. of the Ohe.§..,i, 24, 133 160. Horwitz, 0.'(1964) Epidemiological Basis of Tuberculosis E.:r'adiqation, 2, Dynamics of Tuberculosis, Morbidity f!: ~1ortali ty, Bull. \Vld Hl tl1....2l'Jl•, 2.Q., 609 - ... 161. Styblo, K. (1964) Identification of High Tl1berculosis Risk Groups in Connection with the Epidemiological and Clinical Study of Tuberculosis in Czec~oslovalda, in Co..~.laboration with HHO, J3_lJ.ll. Int. Un. _C2[lj:;ra. Tu.P.•, _22, 36)'. 162. Erin, L. (1960) Detection of Pulmonary Tu~erculosis by Sputum Survey, ~p~~~ London, 41, 363 163. Scottish Health Sei,.,.vices ColiDcil (1963) J?ronch.itis,., Jie.J2ort by a ~:.Pommittee of the St.anding Hedical Adv~ai tj;ee, Edinburgh 164. Horl;:, T. (1962) A Comparative Study of Respiratory Disease in Engla.'l"ld, Hales and Norway, 1\ct..?:._ £~_9._.__§_q.§llld., Sup pl., J8l~ 165. Holland, W. \l. (1965) Respiratory Disease in :Cn::;land and the United States, Ar.,ch. Envir'_?n_._]!_lth,., lQ., 338 166. College of General Practitioners (1961) ~nl~onic Bronchitis in Great Britain, Jki t. me~. , ).l, 973 167. Holland, W. W. Personal communication 168. World Health Organization, Regional Offic~~ f_ol: :::1..u."..?..P.e (1962j EJJ£..9..P_el!:ll Symposium on_lL~:?ll~~p-Specific LungDj~e~2~~ Moscow, EURO- 212, Mimeographed docuoent - 165 - Section 6.6.1 Lung Cancer 169. Registrar General (1966) Registrar General's Statistical Review of England f': Wales for the Year 1964, Part i, ·Tables, Medical, London, H.M.S .0. 170. Registrar General (1957) Registrar General·'s Statistical Review_of England f': Wales for the Year 1952, Supplement on Cancer, p. 114, London, H .. M .• S.O. 171. Posner, E., McDowell, L. A. f': Cross, K. W. (1959) Mass Radiography and and Cancer of the Lung, Drit. med. J., !, 1213 172. Cuthbert, J. (1959) Bronchogenic Carcinoma: A Mass Radiography Group Compared with a Practitioners Group, Brit. med. Dis. Chest., 53, 217 173. Gifford, .J. H. & Waddington, J. K. B. (1957) Review of 464 Cases of Carcinoma of. the .Lung. 'J:r.eated by Resection, Brit. med. J., --_!;,---7?~·-··-- . 174. Waddington, J. K. B. (1960) Surgical Aspects of the Mass X-ray Campaign, . Liverpoo~, 1959, Med.. . Off. , .10.4, 293 . . .. 175. Boucot, K. R., Cooper, D. A. 8: Weiss, W. (1961) The Philadelphia Pulmonary Neoplasm Research Project, An Interim Report, Ann. Int. Med., 54, 363 176. Frost, J. K. Cytology Laboratory, Department of Pathology., John Hopkins Hospital, Pal tim ore, rllary land, U.S • A. , Personal connnuni cation 177. Canti, G. (1964) Analysis of 100 Cases of Bronchial Carcinoma, Annual Meeting, British Society for Clinical Cytology, London 178. Lilienfeld, A. (1963) ACS-VA Co-operative Study for Evaluation of Radiologic and Cytologic Screening in the Early Detection of Lung Cancer, Progress Report, Unio~ Internat. Contra cancer. Acta, 19, 1330 - 166 - 6.6.2 Cervical Cancer 179. Jones, H. (1952) In Discussion of Paper by Hertig, A. T., Younge, P. A. & McKelvey,· J; L~~ ·entitled 'rA Debate, What is Cancer in situ of the Cervix? Is it the Pre-invasive Form of True Carcinoma?", Am. J. Obstet. Gynecol., 64, 807, 816 180. Petersen, 0. (i955) Precancerous Changes of the Cervical Epithelium in Relation to Manifest Cervical Carcinoma, D-~nish Sci. Press Ltd., Copenhagen 181. Lange, P. (1960) Clinical ~ Histological Studies on Cervical Carcinoma, Precancerosis, Early Metastases, and Tubular Structures in the Lymph-nodes, Acta Path. et Microb. Scand., Suppl. 143 182. Clernrnesen, J. (1962) On the Prognosis of Precancerous Conditions of the Uterine Cervix Proceedings of International Conference, The Morphological Precursors ·of Cancer, University of Perugia, Italy, p. 463 183. Koss, L. G. et al~(l961) A Long-Term Cyto-histologic Study of Untreated carcinoma-in-Situ and Related Abnormalities of the Uterine Cervix, Paper Presented at 1st. International Congress of Exfoliative Cytology, Vienna - - 184. Younge, P. A., Hertig, A. T. & Armstrong_, D. (1949) A Study of 135 Cases of carcinoma in Situ of the Cervix at the Free Hospital for Women, Am. J. Obstet. Synecol., 58, 867 185. Boyes, D. A., Ffd.l:er, H. K. i: Lock, D. R. (1962) Significance of In Situ carcinoma of the Uterine Cervix, Brit. med. J., !, 203 186. Dunn, J. E. (1962) The Use of Incidence and Prevalence in the Study of Disease Development in_a Population, Am. J. Publ. Health, 52, 1107 .. ,._ - 187. Dunn, J. E. (1958) Preliminary Findings of the Memphis-Shelby County Uterine Cancer Study and Their Interpretation, Am. J. Publ. Health, 48, 861 188. Kashgarian, M., Dunn, J. E., Erickson, C. C. r0 : Sprunt, D. H. paper Unpublished 189. Terris, M. (1960) Carcinoma of the Cervix: An Epidemiologic Study, J. Am. Assoc., 174, 1847 190. Boyd, J. T. ;": Doll, R. (1964) A Study of the Aetiology of Carcinoma of the Cervix Uteri, Brit. J. Cancer, 17, 419 - 167 - 191. r1acGregor, J. E., Fraser, H. E. & Mann, E. H. F. (1966) The Cytopipette in the Diagnosis of Early Cervical Carcinoma, Lancet, !, 252 192. Cameron, C. B. ~ Hussain, 0. A. N. (1965) 6-Phosphogluconate Dehydrogenase Activity in Vaginal Fluid: Limdtations as a Screening Test for G:~n:i,ta:L Cance.r "'· _Brit. med. J. , i.o_: __:!529 193. Labrum, A. H • .7: Gibbs, D. F. (1964) Clinical Significance of Levels of _6;:,Pho~phog1ucona:te Dehydrogenase in Vaginal Fluid, "Technicon" Conference: 11Automation in Analytical Chemistry", London .. - -- 194. Ladi,nsky,. J. L., Sarto, ''"• E. e: Peckham, B. M. {1964 ). Cell Size Distribution Patterns as a Means of Uterine Cancer Detection, J. Lab. C1in. Med., 64, 970 - 168 - 6.6.3 Breast Cancer 195. Registrar General·(l966) Statistical Review of England and Wales for the Year 1964; Part I, Tables, Medical, London, H.M.S.O. 196. Registrar General (1957) Studies on f!led~ical and Population Subjects, No. 13, Cancer Statistics for England and ~vales 1901-55, London, H.M.S .0. 197. Registrar General ( 1957) Statistical Review o.f England and Wales for 1952, Suppl. on Cancer; p. 11, Londott; ll.M.S. :J. 198. Lilienfeld, A. M. (1963) The EpidemiolovJ of Dreast Cancer, Cancer Res., ' 23, 1503 199. Registrar C'eneral (1957) Statistical Review of England and Wales for 1952, Suppl. on Cancer, p. 74, London, H.M.S.O. 200. Park, W. W. & Lees, J. c. (1951) The Absolute Curability of Cancer of the Breast, Surg. Gynaec. Obstetr., 22• 129 201. Lewison, E. F. (1963) An Appraisal of Long-term Results in the Treatment of Breast Cancer, Unio Internat. contra Cancra Acta, 19, 1547 202. Berg, J. W. & Robbins, G. F. (1963) TWenty Year Follow-up of Breast Cancer, Unio Internat. contra Cancra Acta, 12• 1575 203. Registrar General (1957) Statistical Review of England and Wales for 1952, Suppl. on Cancer, p. 12, London, H.M.S.O. 204. Registrar 'eneral (1957) Statistical l.eview of England and Wales for 1952, Suppl. on Cancer, Diagram P4 "· : '5, p. 17, London 205. I"-loom, H •. J .. G. (1965) The Influence of Delay on the Natural History and Prognosis of ''reast Cancer, '"ri t. J. Cancer, 19, 228 206. Sutherland, R. (1960) Cancer, The Significance of Delay, London, ,ezutterworth 207. Hawkins, J. H. (1944) Evaluation of ::"reast Cancer as a Cuide to Control ProQ;rammes, J. Nat. Cancer Inst., ~. 1~45 208. Taylor, r'. H. " Wallace, R. H. (1947) Carcinoma of the ~·reast: End Result, Massachusetts C'eneral Hospital, 1933-1935, New Eng. J. Med., 237, 475 209. 210. 211. 212. 213. ... 214~ 215. - 169 - Kreyberg~ L. -": Christiansen, T. (1953) The Prognostic Sig_n~_ficance of Small Size in Breast Cancer, Drit. J. Cancer, z, 37 -. Gershon.::c·ohe·n.; J. -;c !:':arden; .. (. G. };;·:--·(i964) Detection of Unsuspected Breast Cancer by Mammography~ Ann. New York Acad. Sci., 114, 782 -- . ·-·. --··· -··· Egan, R. L. (1962) Mammography, an Aid to Diagnosis of Breast Cancer, J. Am. Med. Assoc., 182, 839 Stapiso, S., Strax, P. 1'\:Venet, L. (1966) Evaluation of Periodic Breast cancer Screening with Mammography Methodology and Early Observation, J. Ain. Med. Assoc., 195, ni· · ·- · --· · · · ... ·· Eger, S. A. (1965) Early Diagnosis in Colon and Rectal Cancer, Ca., 15:6, 275 Clark, T. W., Schor, S. S., Elsom, K. 0., Hobbard, 0. B. ,.,; Elsom, K. A. (1961) Value of Periodic Examinations in Detecting Cancer of the Rectum and Colon, Postgrad. Med., 27, 290 Hertz, R. E., Deddish, M. R. & Day, E. (1960) The Periodic Examination. Evaluation of Routine Tests and Procedures, Ann. Int. Med., 54, 1209 -· -----· -----·------ . . .'• ·-- . --- --··-. --··-··· - 170 - 6.7 Eye Diseases .. ---~-. ----·-···· 216. Sorsby, A. (1956) Ministry of Health, Blindness in England 1951-1954, 217. London, H.M.S .0 •. . . ~--- New York State, Glaucoma Program Guide, P.O. Box 7283, Albany 1, New York, p. 10 Health Education Service, U.S.A. (unpublished pamphlet) 218. United States Department of Health, Education and Welfare. (1961) Principles and Procedures in the Evaluation of Screeninz for Disease, Public Health Service, Monograph, No. 67 _ ..... 219. StrBmberg, u. (1962) Ocular Hypertension: Frequency, Course and Relation to other Disorders occurring in Glaucoma, as seen from J1et~s Survey of all Inhabitants over Forty Years of Age in a Swedish Town, Acta Ophthalm, Suppl., 69. 220. Goldmann, H. (1959) Some Basic Problems of Simple Glaucoma, Am. J. Ophthalm, ~. 213 221. Duke-Elcier, S. (1957) The Bowman Lecture; the Aetiology of Simple Glaucoma, Trans. Ophthalm. Soc. of the_U.K., 77, 205 222. Perkins, E. S. (1965) Glaucoma Screening from a Public Health Clinic, Brit. med. J., i• 417 223. Paterson, G. D. (1966) The Value of Family Studies in the Detection of Glaucoma Simplex; in Glaucoma: Epidemiology, Early Diagnosis and Some Aspects of Treatment, Proceedings of a Symposium held at the Royal College of Surgeons of England, London, Edinburgh, E. & S. Livingstone, p. 51 - 171 - 6.8 Urinary Tract 224. Kaitz .. A~·-··L;;·& Hodder_, E. W. (1961) Bacteriuria and Pyelonephritis of Pregnancy (Prospective Study of 616-Pregnant Women) .. New Eng. J. Med., 265, 667 225. Mi.all_, W. E., Kass_, E. H., Ling, J. & Stuart, K. L. (1962) Factors Influencing Arterial-Pressures in the General Pqpulati.on in Jamaica;_ Brit.· m~d. J., ii, 497 226. Kass, E. H. (1962) Pyelonephritis and Bacteriuria, Ann. Int. Med., 56, 46 227. Simmons., N. A. & Williams., J;n. (1962) A Simple Test tor Significant Bacteriuria, Lancet_, l_, 1377 228. Brumfitt, w ... Davies_, B. I. & Rofiser_, E. ap i (1961) Uretheral Catheter as a Cause of Urinary Tract Infection in Pregnancy and Puerperium, Lancet., ii_, 1059 - 172 - ~ Rheumatic Disease 229. Kellgren, J. H. (Editor) (1963) The Epidemiology of Chronic Rheumatism a Symposium Arra.D.ged by the CIOMS, Vol. I, Blackwell, OXford 230. Mikkelsen, W. M., Dodge, H. J., Duff, I. F., Epstein, F. H. & Naiper, J. A. (1963) Clinical and Serological Estimates of the Prevalence of Rheumatoid Arthritis in the Population of Tecumseh, Michigan 1959-60, In the Epidemiology of Chronic Rheumatism, a Symposium arranged by the CIOMS, Vol. I, Blackwell, Oxford, p. · 239 231. Epstein, F. H., Francis, T., Hayner, N., Johnson, B. C., Kjelsberg, M. 0., Naiper, J. A., Ostrander, L. D., Payne~ M. w. & Dodge, H. J. (1965) Prevalence of Ch:i:-onic Diseases and Distribution of Selected Physiologic Variables in a Total Community, Tecumseh, Michigan, J. Chron. Dis., 81, 307 232. Kellgren, J. H. (1964) Heberden Oration, 1963, The Epidemiology of Rheumatic Diseases, Ann. Rheum. Dis., 23, 109 233. Lawrence, J. S., Hewitt, J. V. & Popert, A. J. (1963) Gout and Hyperuricaemia in the United Kingdom, Symposium arranged by the CIOMS, Vol. I, Blackwell, Oxford, p. 176 - 173 - 6.10 Mental Illness 235- 236. 237. Shepherd, M. (1964) Minor Mental. _Illness in London, Some Aspects of a General SurV'ey ,_ .Bri_t ~ · med. J., ii, 1359 ~., Medical Research Council: (1:965). Clinic~ Trial of Treatment of Depressive illness, Brit~ med •. J., 1, 881 Gruenberg, E. M. (1964) Mental retardat;ion: Epidemiology_, Chicago, Stevens, H. A. & Heber, R., p. 259 Jaeggi, A. & Jaeggi, F. (1965)-.Rimseignement des enfants et .adolescent reputes arrieres dans le canton de Geneve, Psych. 1 'enfant, Vol. VIII: 2, 453 238. Wein'berg,. A. N. (1961) .Detection of Congenital. Galactosemia . and the Carrier State Using Galactose-Cl4 andB1oQd cells·, Metaooiisffi;· ro; ·:p~-728 · -- · ·· --------·-- .. ·- · - 174 - 6 .11 Anaemia 239. World Health Organization (1959) Iron Deficiency Anaemia, Wld Hl th Org • techn. Rep. Ser • , 182 240. Berry, W. T. c., Cowin, P. J. & Mages, I{. E. (1952) Haemoglobin Levels in Adults and Children, Brit .. med. ;J., _!, 410 241. Kilpatrick, G. B. &: Hardisty, _R. M. (1991) The Prevalence of Anaemia in the Community, A Survey of a Random Sample of the Population, Brit. med. J., !' 773 242. Spooner, R. D. (1960) The incidence of Anaemia in General Practice in New South Wales, Med. J •. Australia, _g, 727 .. 243-. Jungner, G. & Jungner, L · (1966) The Health Screening· in Vctrmland, in SurveH.lahce and Early Diagnosis in General Practice, Office of Health Economies, London· 244. Fry, J. (1962) Minor Maladies, Practitioner, 189, 633 - 175 - Chapter 7 275. World Health Organization (1963) Expert Committee on Medical Assessment of Nutritional Status, Wld Hlth Org. techn. Rep • Ser ., 258 246. Day, E. (1960) What is an Adequate "Cancer Checkup"? Postgraduate Med ... 27, 274 247. World Health Organization (1964) Expert Committee on Prevention of Cancer, Wld Hlth Org. techn. Rep. Ser~ 276 248. Collen, M. F., Rubin, L., Neyman, J., Dantzig, G. B., Baer, R. M. & Siegelaub, A. B. (1964) Automated Multiphasic Screening and Diagnosis, Am. J. Publ. Health, 54, 741 249. Collen, M. F. (1966) Periodic Health Examinations Using an Automated Multitest Laboratory, J. Am. Med. Assoc., 195, 830 250. Suchet, A. s. (1963) Methode Active de Surveillance de la Sante des Grandes Collectivites Agricoles et Industrielles, Acta Medica et Sociologica, II, 237 251. Jungner, G. (1966) Data Processing in the Clinical Laboratory, in: Automated Data Processing in Hospitals, International Conference in Elsinor 1966 (in print)

WORLD HEALTH OR-GANIZATION ' ' ORGANISATION MONDIALE DE LA SANTt . ~. PRINCIPIES AND PRACTICE. OF SCREENING FOR DISEASE J. M. G. Wilson and G. Jungner Dr J. M. G~ Wilson is a Principal Medical Officer at the Ministr,Y of Health~ London; Dr G. Jungner is .Chief., Clinical.Chemistry .Department., Sahlgren' s Hospital., Gothenburg., Sweden. The views expressed in-this paper are their own and do.not necessarily rep- resent those of either the World Health Organization or., in the ca5e of Dr Wilson., the Ministry of Heaitli •. PA/66.7 The Issue of this· document · d~s ·not constitute formal publication. It .sh?uld not. be r.evl.ewed,_ abstracted or quoted without the agreement of the World Health Organization. Authors alone are responsible for views C\Xpressed In signed article~. Ce document ne constftue pas une publfcatfon. II ne dolt falre !'objet d'aucun compte rendu ou r~sum~ nl d'aucune citation sans l'autorfsatlon de !'Organisation Mondlale de Ia Sant6 Les opinions exprlm~es dans les articles. slgn~s n'engagent que leurs auteurs. 1. 2. 3. 4. - 2 - CONTENTS INTRODUCTION •••• TERMS OF REFERENCE DEFINITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 6 7 3.1 3.2 3.3 3.4 3.5 3.6 3.7 Screening . . Mass screening • Selective screening Multiple screening ••• . . . . .. . Case-finding • • • Population or epidemiological surveys Early disease detection . . . . . 7 PRINCIPLES • • • • • • • • • • 4.1 General considerations • . . . . . . . . . . . . . 7 8 8 8 8 8 8 8 8 4.2 4.3 4.1.1 4.1.2 4.1.3 The aim of early disease detection • Pattern of screening development • The use of different forms of screening Selective screening Mass public health screening • . . Surveillance • • • . . . . . . . . . . • 10 . . . . . 13 . • • . . 13 . . . • • 14 • 14 4.1.3.1 4 .1.3.2 4 .1.3.3 4.1.3.4 4 .1.3.5 Screening hospital patients •••••••••• 15 Screening in industry . . . . . . . . . . . . . 16 ,, --.. ~-~- ................... . Evaluation of results of screening •• . . . . . • • • • • 16 4.2.1 4.2.2 4.2.3 General . . . . . . . . . . Evaluation of screening procedures • 4.2.2.1 Validity . · . . . . . . . 4.2.2.2 Reliabfli ty . . 4.2.2.3 Yield 4.2.2.4 Cost . . 4.2.2.5 Acceptance . 4.2.2.6 Follow-up services . The "borderline" problem • • • • . _. . . . . . 16 . . . . . . . •• 16 • 16 . . . . . . • • • 16 . . . . . . • 16 • • • • • 16 . . • 16 • 16 • • • . . . . 20 Principles of early disease detection . . . . . . . . • 22 4.3.1 4.3.2 Impqrtant problem for the individual and the community •• 22 Accepted treatment • • • • • • • • • • • 22 5. - 3 - 4.3.3 Facilities for diagnosis and treatment should be available 22 4.3.4 4.3.5 4.3.6 4.3.7 4.3.8 Recognizable latent or early symptomatic stage • Suitable test or examination • . . . . . . . . Acceptability to population Need for surveys • • .. . . . . . . . . Groups to be treated in case-finding programmes •• 22 22 • • 22 22 22 4.3.9 Economic balance of the cost of case-finding in relation to total expenditure on medical care • • • 22 4.3.1:0 · Case-ffnding should be a continuing process 4·.3.10.1 · Concept of "surveillance 11 PRACTICE • 5.1' · A'utomatic dat·a handlfng . 5.1.1' . The' particular ne.eds 'o{ screeni'ng · 5.1.2· ·Basic concepts • · 5.1.} Data ·collection · 5·.1.4 . 5".1.5 · 5.1.}.1' · P\mch card 5.1.3.2 Paper tape • 5.1.3.3 · Maghetic tape · Data processing Data storage • • . . . 5.2 . Present' screening practice (including multiple screening) 5.2.1 ·5~2~2 ·5.2.3 Introduction • ·• ~ Comparison of studies ·5~2~3~2 . 5~2:3~3' '5:2:3:4 '5~2.3:5 Hunterdon county San'Francisco longshoremen Chicago'B6ard.of Health' : ·common conditions.for screening 5.3 ·Epidemiological·studies· 5.4 'Periodic nealth·examinations • '5~4.1 5.4.2 5.4.3 Introduction Relationship to general practice • The industrial health examination 5.4.4 Routine examinations through life 22 • 34 • 35 .•. 35 • 36 .•• 37 • 37 ••• 38 • • 38 • 39 •• 39 •• 39 • 39 • 48 • 54 • 54 • 54 ••• 57 • •• 64 •• 64 •. 65 • 66 • 67 - 4 - 5.5 The place of screening in the provision of medical care . . . 68 5.5.1 Justification within a country . . . . . . • 68 5.5.2 Cost of multiple screening . . . . . • . . . . 69 5.5.3 Inte~-country differences 70 5.5.4 Example of gycaecological cytology . . 71 5.5.5 Conclusion • . . . . . . . . . . . . . • 73 5.6 The place of education in the early detection of disease . . . 74 5.6.1 Education of the medical profession . . . 74 5.6.2 Education of the public . . . . . . 75 6. STUDIES OF CERTAIN CONDITIONS . . . . . . 76 6,1 Diabetes mellitus . . . . . . . . . . . 76 6.1.1 General . . . . . . • . . . . . . . 76 6.1.2 Value of early treatment . . . . • . . . . . . . 81 6.1.3 Diagnostic criteria . . . . . • 82 6.2 Heart disease . . . . . . . . 84 6.2.1 Rheumatic and congenital heart disease 84 6.2.2 Ischaemic heart disease . . . . . . . . 86 6.2.2.1 Mortality . . . . . . 86 6.2.2.2 Morbidity . . . . 86 6.2.2.3 Screening 87 6.2.3 Conclusions . . . . . . . . . 92 6.3 High blood pressure . • . 93 6.4 Overweight . . . . . . . . . . 95 6.5 Respiratory diseases . . . . . . . . . . 96 6.5.1 Pulmonary tuberculosis . . . • 96 6.5.2 Non-specific respiratory disease • 100 6.6 Cancer . . . . . . . . . .. . . . • . . 102 6.6.1 Lung cancer . . . . . . . . . . . . • 102 6.6.2 Cancer of the cervix . . . • . . . . . . • 107 6.6.3 Breast cancer . . . . . • . • . . . . uo 6.6.4 Other cancers . . . . . . • . . . . ll4 7. 8. 6.7 6.8 6.9 - 5 - Diseases of the eye . . . . . ' . . 6.7.1 Chronic glaucoma 6.7.2 Otner eye diseases Diseases of the urinary tract 6.8.1 General . . . . . . . - . 6.8.2 Bacteriuria • Rheumatic diseases 6.9.1 Rheumatoid arthritis . . . . . 6.9.2 Gout • • • • 6.10 Mental illness METHODOWGICAL TRENDS IN SCREENING • • • • • • • • 7.1 Clinical or technical screening • 7.2 Participation by doctors • • • • 7.2.1 The doctor's physical examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 7 .2·.2 History-taking by doctor or specially trained medical staff 7.3 ·7.2.3 ;uest:onaries • • • • • • • ••••••• 7.2.4 Clinical methods and laboratory tests used for screening Development of screenin13 facilities .• 7.3.1 Simplification • • • • • • • • 7.3.2 Automation •••••••• . . 7.4 Automated multitest laboratories ad modum Collen •••• CONCWSIONS • • • • . • • • • • • . • • . • . • • • • • • • • • • 8.1 The need for further epidemiological investigations and, allied w:~t:·L ·ci;.is, the need to acjlieve standardization •••••• 8.2 8.3 8.4 8.1.1 8.1.2 8.1.3 8.1.4 8.1.5 8.1.6 Records Carcinoma of the uterus • Glaucoma simplex Mental illness Asymptomatic bacteriuria Cancer of the breast • • • Lung cancer . . . . . . . . Economics • Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 li6 121 122 122 122 124 124 125 126 133 133 133 134 135 135 135 139 139 140 145 146 147 148 148 148 149 149 149 149 150 150 - 6 - 1. INTRODUCTION This paper has developed from the original request of Dr F. Grundy1 Assistant Director General 1 World Health Organization, Geneva1 for a briefing paper on the early detection of-disease~ to be limited mainly'to the chronic diseases of adults in developed' countries~ · Even'within these limitations w~ found we were unable~ with a' subject matter that'not.only covers· a· great deal of medicine but is also itself breaking new ground'rapidly~ to'write a short paper and it has now grown to its present length. The subject of early disease detection is vast and it would clearly be beyond our capacity to b~ comprehensive. This account, then, represents only our personal viewpoint of a rapidly developing aspect of medicine and the examples we have chosen . . . . are ones that have, for one reason or another, appealed to us personally. There may well be other examples, equally good or better, that we have omitted. (We . ' have not, for instance, included the long-established practice of early disease detection in the maternity or child welfare field, largely because_ that practice is already so well established.) To use a simile, we have made a number of preliminary sketches rather than attempt a complete picture. We are also aware that the subject is controversial and that much still needs to be learned. If anywhere we have appeared dogmatic, we hope this may serve to stimulate discussion on which, in the end, real development depends. There- are three main headings. of the paper •. The first part deals. with the basic principles. of early disease detection; the second with the practical con- siderations, including a number o:f e~arnples from different disease conditions.; and the third and last. part deals with the use of present methods and possible development. 2. TERMS OF REFERENCE The. terms- of refe;rence of- this paper are ."to discuss the general. principles of the. administrative and scientific. aspects. of screening· proceduresn •. For the purposes. of this. paper. the. definition of "screening11 proposed by the United States of America. multi-sponsore~Commission on. Chronic Illness {C.C~I.) {see section 3.1 below1.) and accepted by the Regional. Committee. for. Europe, 2 is.adopted~. Periodic phy-sical examination is also. included in the. revie:w and we._ refer to both screening - 7 - and periodic physical examination generically as "early disease detection". Epidemiological surveys to establish the prevalence and incidence of conditions as well as to study longitudinally the natural history of developing di~eas~, are not considered as falling within the terms of reference, which are regarded as confined to case-finding. However, frequent reference is made to surveys which throw light on our attitudes to case-finding. Screening for the chronic non-communicable diseases forms the main subject of the paper; but the problems facing countries at differing stages of develop- ment and having different standards and types of medical care are also discussed, and this to some extent includes communicable disease detection. 3. DEFINITIONS 3.1 Screening The Commission on Chronic Illness Conference on Preventive Aspects of Chronic Disease, held in 19511 defl.ned screening as 11the presumptive identification of unrecognized disease or defect by the application of tests, examinations or other procedures which can be applied rapidly. Screening tests sort out apparently well persons.who probably have a disease from those who probably do not. A screening test is not intended to be diagnostic. Persons with positive or sus- picious findings must be referred to their physicians for diagnosis and necessary treatment". It should be noted that, by definition, unrecognized symptomatic as well as presymptomatic disease is included, and physical examination as well, so long as it is classed as rapid. "Other Procedures" can also embrace the use of questionnaires, which are assuming an increasingly important place in screening. Finally, tests may be diagnostic, though not necessarily so intended; so that, for example~ a gynaecological examination could be covered by this definition provided it is rapidly carried out. ·rn general, we have taken the definition to imply a relatively simple method of case-finding. 3.2 Mass screening This is a term used to :indicate large-scale screening of population groups. In this paper we use this term to refer to population screening where no selection of groups is mad~. - 8 - 3.3 Selective screeniDg We use this term for the screening of selected high-risk groups in the population. It may still be large-scale and can be considered as one form of population screening. 3.4 Multiple screening This has evolved by combining single screening tests, and is the logical corollary of mass screening. Where much time and effort has been spent by a population in attending for a single test (e.g. mass radiography) it is natural and economical to offer other tests at the same time. Multiple (or multiphasic) screening has been defined as "the application of two or more screening tests in 1 combination to large groups of people". 3.5 Case-finding Throughout this paper this term is applied to that form of screening, the main object of which is detecting disease and bringing patients to treatment. 3.6 Population or epidemiological surveys Whilst screening tests may well be used in population surveys (e.g. sphygmo- manometry for blood pressure or tonometry for intraocular tension), the principal aim of surveys is not to bring patients to treatment but to elucidate the preva- lence, incidence and natural history of the variable being examined, though case- finding is a natural by-product of surveys. A good example of an epidemiological survey is the Framingham study of ischaemic heart disease.3 3.7 Early disease detection It is sometimes useful, we think, to use a term that refers to all forms of early disease detection whether by screening, physical examination or other means; and this is meant when we use the term 11early detection11 • 4. PRINCIPIES > 4.1 General considerations 4.1.1 The aim of early disease detection The aim of early disease detection (sometimes called secondary prevention) is simple. Primary prevention seeks to abolish disease by protecting the individual and the population from attack before the challenge has been made. Early detection - 9 - (case-finding) aims. at-discovering and curing conditions which have already produced pathological change but which have not so far reached a stage at which medical aid is sought spontaneously. Fig. 4 .1. These stages are shown diagrammatically in "Well" ' "Ill" FIG. 4.1 I PreJention No prevention I Irmnunity Early I pathological change Early detection I . Treatment I I No early detection I Developing pathological change I Symptoms I Treatment I I No treatment I Gross pathological change Example: Pulmonary Tuberculosis Mantoux test, B.C.G. . M.M.R. Clinical Assessment X-ray sputum .For screening, as defined, there is the second, economic, aim of achieving more for unit expenditure by saving the time of highly-trained professional people. Part of their trained time can be substituted by less highly-trained personnel, able to carry out screening tests, whether by hand or automated (see diagram under "Periodic health examination's II, section 5.4). However, it seems likely that the total cost of screening in a community is higher, not lower, than the cost of conventional medical care since more people will be found in need of treatment (and these largely 'elderly and liable to be under care for a long time). We shall say more about the economic aspects of screening under the proper section (4.3.9). However it is worth noting here' that in some circumstances screening may be altogether uneconomical. A' condition (e.g. helminth infestation) may be - 10 - almost universally prevalent and mass treatment without soreening may in these conditions be the course of choice, thus avoiding the high cost of preliminary confirmation of a virtually certain diagnosis. 4.1.2 Pattern of screening development Sociological factors have been important in the development of screening. ~~-s~_ :r~]_.E?_v~:t __ faci:;()_!"S .c:tr~ clqse_ly relate.d to the degree of sophistication of .. the population at risk; for example, the level of education and awareness of the population, the amount and form of medical care available and the general standard of living_. In highly-developed societies a rise in the level of these factors has been accompanied by a decline in communicable disease and by an apparent in- crease in degenerative and genetically determined disease. Under these conditions, also, there-is a tendency towards diagnosis at_a less advanced stage than in the past. In less developed countries, however, the communicable diseases remain largely paramount and the stage at diagnosis of chronic illness is_often later; whilst living conditions, nutrition, the education of the public and medical care all need to be improved. For these reasons the early detection of illness presents different problems in highly-developed and less well developed countries. The historical development of screening can best be examined by observing the measures taken to control endemic communicable disease; measures which are now to a large extent no longer needed in well-developed areas, but the operation of which is still vitally important in many less-developed countries. Fig. 4.2 sets ·out the··progre·ssive stages (whether iri time or in terms of development) in diagrammatic form. Studying the reasons behind these earlier forms of screening is helpful to the understanding of the later developments in highly developed countries for screening chronic disease. In tropical and subtropical areas of the world parasitic diseases like malaria, schistosomiasis and ankylostomiasis have long been the subject of mass detection. One of the ·main reasons has been the need to control these maj~r causes of disease by attempting to stamp out the human reservoir. Similarly, chest radiography was first introduced primarily as a public health measure to help in controlling the spread of pulmonary tuberculosis; )rolonging the life and health of the individual was at that time a secondary objective. A further example can be found in the attempt to control the spread of ·syphil:Ls by mass serological examination of the population. - 11 - FIG. 4.2 '. J •••• ' .. Screening era Examples of conditions solle;ht Malaria Early Nematodes Leprosy Trachoma Pulmonary tuberculosis Middle Venereal diseases Diabetes Late Ischaemic heart disease Iron-deficiency anaemia. Only when the prevalence pool of endemic communicable disease has'largely been emptied has the objective of early detection been turned chiefly towards the second aim of chronic disease detection. Clearly, economic factors play a large part; controlling the spread of disease is vital to economic prosperity and even survival; whilst prolonging individ:ual life and health is less economically mandatory. These differences are important when considering the p~inciples upon which population screening should be based, for the relative importance of individual considerations varies in the two cases. In a smaller way the same story can be told of industrial health examinations. In the early_ days of industry there were endemic industrial diseases, like mule spinners' cancer, for which primary prevention was developed (in that case by . . removal of the causative agent). In other conditions, for example silicosis and lead pdisoning, where·tne cause could not be wholly eradicated, early detection techniques were dev'eloped (chest x..:.ray, blo?d film and urinary lead content). With advancing sophistication in' industry the idea grew of not only monitoring the health of workers in relation to known environmental health risks, but also of anticipating non-industrial hazards to the health of individuals by periodic .. .· . ~ '- .· ... , : . health examination. These two types of e~amination are comparable to the two stages of growth of general population screening. . . .. ~- ' ·;· . . . - 12 - Whilst the benefits of the first type of mass detection (to control the spread of communicable disease) have been, at least to some extent, demonstrated . ··-· ~- - ~ -.. ----~~- -·- (e.g. tuberculosis) the value of the second needs as yet largely to be ascertained (e.g. diabetes·mellittis, chronic simple glaucoma). Why this should be and how answers to outstanding questions might be found will be examined later in this paper. Much screening practice evolved in the United States of America during the 1950's in the form of multiple screening programmes and has been reviewed else- where4•5•6•7•8•9 and will c~ly be discussed briefly here. A useful bibliography was issued by the Bureau of Chronic Diseases of the California State Department of Public Health.10 The motivation for screening has been dealt with at some 11 . 12 13 length in three papers, respectively by Chapman, Mount1n and Smillie published between 1944 and 1952 at the time of maximum growth of this concept in its application to chronic diseases. Some of the chief points made in their papers were : (a) case-finding by multiple screening is a technique well suited to public health departments, whose role.is changing: (b) The need for provision for diagnosis, follow-up and treatment. This is vitally important; without it case-finding must inevitably fall into disrepute. (c) The need for validation of the tests before applying them to case- finding; harm may result to public health agencies' relationships with the public (not to merition the direct harm to the public), and with the medical profession, from large numbers of fruitless referrals for diagnosis. (d) The danger that multiple screening might lead to the neglect of other aspects of community medical care because of the competing cost and possibly also because a sense of false security might be propagated. (e) The need to evaluate the effect of multiple screening by its results in reduced morbidity and mortality. In 1957 the Commission on Chronic Illness accepted the value of multiple screening as "contributing to good medical practice" and considered that it "constitutes a practical means for early detection of a number of important - 13 - chronic diseases and impairments". The conditions for which the Commission on Chronic Illness. considered at that time.might profitably be screened were: pulmonary tuberculosis visual defects (including chronic glaucoma) hearing defects syphilis diabetes cancers of skin, mouth, breast, cervix and rectum hypertensive disease ischaemia heart disease (possibly). In 1960 the American Public Health Association strongly endorsed multiple screening in a publication "Chronic Disease and Rehabilitation; a Program Guide for State and Local Health Agencies 11 • 7 \rfuile recognizing that screening should, where possible, take second place to periodic health examination as an effective technique for early disease detection, the authors of the "Program Guide" considered that "the sheer weight of economic reality ••• dictate recourse to procedures that conserve the time and energy of highly trained personnel such as physicians and dentists 11 • With the formation, in 1961, of the Chronic Diseases Division of the Public Health Service of the United States of America came the ability to provide State services with project grants for setting up demonstration scree~ng progranmes. A relativeiy large nUmber of these projects have been carried out, but for various reasons it has been difficult for them to satisfy all the above points. 4.1.3 The use .of different forms of screening 4.1.3.1 Selective screening Screening tes'ts can, of course, be used in differ~nt ways, varying from single examinations applied ~o individuals to batteries of t~-sts offered to whole popu- lations. They may also, as already'indicated, be either indiscriminate or selective. From the viewpoint of both the individual and the economy there are obvious advantages in combining a number of tests and applying them all at the one examination, providing each has been shown 'to be medically worth its place. - 14 - However, in-practice, there may be drawbacks to combining certain tests, for· example when each only gives reasonable yields in selected population groups of different age, sex or occupation. Much screening is selective, of long standing and well established. This type of screening is practised, for example, at antenatal, post-natal and infant welfare clinics, where conditions like pre-eclamptic toxaemia, the anaemias of pregnancy and congenital conditions are sought by the application of simple tests. 4.1.3.2 Mass public health screening More recently, public health.agencies have tended to extend their screening activities from these kinds of clinics to the general public. Multiple screening has been offered at ad hoc clinics staffed by ancillary workers, positive results being notified to general practitioners. However, it has been recognized that this approach gives rise to difficulties (some of which are discussed below) and of late it has declined in popularity as a means of early disease detection. 4.1.3.3 Surveillance A third method of screening is the individual approach, as opposed to mass screening. At first sight there is little difference between screening the individual and the ordinary good practice of clinical medicine. The physician examining his patient's urine or blood-pressure or even the weight when he has no special reason to suspect illness related to these findings, is simply com- plying with an accepted standard of good diagnosti .. c practice. He may, in fact, extend these activities to other examinations like haemoglobin or electro- cardiography. If he also arranges for these examinations to be carried out by an auxiliary helper he can then be regarded as·submitting his patients to a form of multiple screening which may or may not be selective, according to. .whether·· only those patients consulting with some complaint are ex_amined, or whether the PhySician has made arrangements for all his patients to undergo these tests. The essential difference between this form of screening and ordinary -good · medical practice is that in the first instance the person examined is presumptively well, whilst in the second he or she comes to the doctor as a patient with a complaint. There is really a good deal of difference between these two concepts; the economic implications for a general practitioner in terms of time, auxiliary .;.. 15 - pe~!) and the:use of r~cords, as well.~;ts premises, is V!ary different if he is aiming to: carry, o1:1t s~lect~ye screening in his pra~tice, in contrast simply to applying a number of tests to patients reporting with,a complaint. Howev.er.; · as a development in medical care, there are clear pote!ltial advantages in this form ·~ ,'.; of positive surveillance. The general practitioner···c:iari-:;-In-·fb.is .. way,-· be enabled to practice personal preventive medicine; and,.· s~condly, the normally clOse contacts between the doctor and his practice should largely avoid communic~tion difficn.llties over the results ·Of tests • Screening hospital patients A particular f.orm of screeningspecial groups of the population is the screening of hospita~ patients. There are at least three aspects to screening patients in hospital. Firstly, hospital patients in general, whether in- or out-patients, constitute a special high-ris1~- ·group~o:f' the population and are likely to give a high yield for conditions like diabetes mellitus, cancer of the cervix and simple glaucoma •. Secondly, patients come to hospital with a complaint of ~ioh the d:i,agnosis may be ~ither in doubt or erroneous. It is then usual to ~equest laboratory and other tests in a sequential fashion, one request often (lepending on the :-result of the. previous one. $ubmit:ting the patient to a number of laborl;l.toryt-ests routinely without exE)rcising individual choice_is a form of·· scr;-eening which may prove its worth _in l~ading to diagnoses ·which would otherwise have been delayed or even missed altoget~er. This performance of a number of laboratory tests on one blood and/or urine specimen at the same time is now becoming quite feasible with the introduction "6:f"laboratory aufomatioi:i {both in carrying-out 'the tests themselves and in th3 equally onerous-matter of processing the data). It may well·prove more economical than the traditional method of seriatim-laboratory requ~sts. Thirdly~· and· following as a consequence of per- forming a number of tests simUltaneously, th~re may be:an·econ6mic gain in hospital stay. One of the most costly items of medical service is the upkeep of a patient in a hospital bed. Rationalizing medical care in hospital so as to minimize the length of stay is one of the chief ways in which the cost of health services can be kept from mounting disproportionately to other expenditure.· ·It is possible (though at the moment evidence is lacking) that screening hospital . patients could shorten length of stay or lower cost in some other way, for example - 16 - by reducing the number of hospital consultations called for. Work on these lines has been carried out in the United States of America and Canada, and trials are starting in.Great Britain and Sweden. 4 .1.3.5 Screening in industry Lastly, industrial populations may offer special advantages for screening, especially in industrialized countries where there is no universal general practitioner service. It is important to remember, of course, that industrial screening examinations are of two sorts; one kind is for special industrial risks, of which examples have been mentioned above; the other kind of examination is aimed at the early detection of diseases which may impair the general efficiency of the worker. This subject is discussed under paragraph 5.4.3. 4.2 Evaluation of results of screening 4.2.1 General The evaluation of screening can be considered from two separate aspects which yet have a certain connexion with each othe_r. These aspects are, firstly, the evaluation of tests or examinations and, secondly, the evaluation of results. The important connecting link is the need to use standard criteria for tests when comparing results between case-finding operations. Associated with these standards is the difficult problem of the "in-between" or "borderline" patient, which will be considered in this section. 4.2.2 Evaluation of screening procedures The Conference on Preventive Aspects of Chronic Disease considered the. evaluation of case-finding tests and programmes in 1951 and the matter has been dealt with at some length in the C.C.I. publication "Prevention, of Chronic - 14 Illness".- The following criteria were discussed. 4.2.2.1 Validity 4.2.2.2 Reliability 4.2.2.3 Yield ~.2.2.4 Cost 4.2.2.5 Acceptance_ 4.2.2.6 Follow-up services -··17-. In this s.eqtion we dea:l only with validity, reliability-and yield; -the other criteria are discussed later, ·under the section on Principles of Early-Disease Deteetion: cost under. paragraphs 4.3.9, acceptability under 4.3.6, and follow-up under 4.3.10~ 4.2.2.1 Validity The C.C.L defines the validity of a screening test as the measure of the fi.equericy with which the reshl t of that t~st is c'onfirm~d by .,im accep·table diagnostic procedur''e, i.e. the ability of tho test to separate those who have the condition soUght froni those who dci not. Applying a scre~ning test to a population will produce four categbries.of result, provided that the whole population is also examined definitively to establish the actu~ prevalence·. of disease. shown di~gra~ticaliy (Fig. 4.3). These.·· four possible results can be Screening result Positive ·Negative FIG. 4.3 THE EFFICIENCY OF A SCREENING TEST modified from Remein, Q •. R. & Wilkerson, H. L~, l5 True disease classification of apparently well population ~------------- ----------------------------------------------~ Unknown cases of disease With disease and with ·pos±tive--test .. · · (true positives) With disease but with - rregatf~~-~j~~~t- -- -- ·--··- ...... · (fals>:c.• ;,.,·o-a~_,.,,..;:,.;.s)~- r ., ·: :c.or.•:"· ·'\':;0-·t~~ .-:%-lfC }'-•·•···· .., ...... ~., 1- Persons without disease ···-·t ....... - ............... . · Without disease.· but \d-th poait:i:v-e-··-t·es-t~- -------- (false positives) Without disease and with negative test (true negatives) * Sensitivity = Diseased persons with .e,ositive test .. - ----··-- · ---- ·-· -·All persons in--population wi·th--d1:sea:se--· · -· ..: .• i..: ·. Specificity * N9~ise~se~ persons with negative test ................................... A~1.E~:r.::>C>~I:} __ !.l1 J?Opul8:(~_9.I?:_ w:itl:.?.~~-~-~s~a.-s.~ * These values are often expressed as percentages. I - 18 -' An ider:.l test would of course cJ..:;t.:'ct only those persr:ms in o. population suffering f'r-om the c6nditi6h lo~ked-fo:r -(as defined by agre~d criteria.) and would not fail to detect :any of-them. The ability of a test to classify as positive those persons with the disease is termed "sensitivity" and the ability to class as. negative those without the disease_, "specificity"; that is, sensitivity is a measure of the false negative rate and specificity of the false positive rate. Sensitivity and spectficity can be varied reciprocally according to the setting . ()f the test. Take_, for example_, the detection of iron deficiency anaemia by determination of the haemoglobin. Let us say that the aim is to diagnose and treat all women in a population with a haemoglobin of 11 g/100 ml or less. ' . . . ' A screening level of 11 g/100 mi will miss a number of cases due to the errors of the te~t and the sensitivity will be low though the specificity will be high. By raiping the screening level to 12 g/100 ml the sensitivity will be raised so that few cases are missed; but the penalty must be paid of lowering the specificity and accePting a rise in the number of false positives. this point: The following example· (Fig. 4.4) illustrates FIG. 4.4 A,-.••'><-• ,_ ·-· ...... ~ ... ·-·r--.-- ''"·• .. ····· Total Anaemic Screening level population . - patients I' . l 11 g % haemoglobin 12 g % haemoglobin ... , . . -~ _ .... ~. 4>· ....... . ·- Positive I Negative -- Positive Negative 100 20 . ~ l ! True False, True lFalse TrueiFalse True False i --.. --~~~--- ... ...... . ·-·. -~. ---~· .. ..... . I I I .. 15 2 I 78 5 19 10 I 70· 1 • • • ~ .. h •• ··~" ••••••••• -. ..... ---·-··•-.. -- .. . . ...... .. ,. Sensitivity 15 100 75 % 19 X 100 95 % X = = 20 20 ......... --~--- . .. ... , ... --· ... * Specificity 7.8 X 100 = 98 %· 70 100 88% 80·· Sox ·=· - .... - •· ""• ... .- - * For working, see Fig. 4.3. - 19 .. 4.2.2.2 Reliability Providing the test ~elected is a good inde~ of th,e -CI.:Lsease sought~ two factors . :,\, - _·,:•,- . , .... are c.once.I"ned in the. reliability or efficiency of the test: the variati()n of the method and the variation of the observ~r. For;exa.znple, in measuring the arterial blood-pressure with an inflatable cuff sphygomanometer there ar~ the variations . . . . . . . . ... ~ connected with the indirect relationship between the method and the true intra- arterial blood.:.pressure and wfth'the 'variability of the blood-pressure itself; and the~e is also the error of the observer (whieh has re~ently been shown to be much more than was previouSly supposeci).'l6,l7 In considering screening techniques there is scope for research into methods. Ideally (as has been said) a test should behighly sensitive and should miss yery ~ . •'_-' . ·' · .. : .. , . .. . · .. ·_. . :·j . : .... ·: : .. · ..:. ':. . ·,>·. : ''_.;h,.· .. ·' :. -~- . ' . : . . . ' •. ' .. few••pet'S.ons with the d_isease, though a rela:tively high_proportion of, false po~i tives - .. ... . .· - ·..... . . ' . . . ' can be accepted; it should be as simple as __ possible _and able t? .. b~,,.Qafi":l.e<;l oou:t ,_, rapidly, often under.improvised field conditions (though in some instances there is ·. -,-. a],so'a case fora 'firmly-based unit to which the population' comes or where speci- mens are sent); thirdly a test must be-acceptable and cause minimal.distui-bance to the subject in its performance. Pain and discomfort, much undressing ~~ the need for a large blood sample may rule out an otherwise_ ~?'?..f!.l!.~t~?.~~~·-····-Ia..§~.!ti';...i t should be a~ _c:qE!_ap . as pos.s~ble • Speed tends to <i.:l.minish efficiency and yice. versa; the local need will probably determine whic:t). factqr::;_ a,re the more important. For example, in detecting diabetes mellitus, urine testing may be chosen for its sim- plicity, cheapness and minimal d:lsturbance to the population. However, testing fo~ glycosuti~ is· inefficient and bl()?.ci: s~ar screening on the spot using capillary · biood might be cho-~'~ri rather than obtaining a ve~ous blC>od 'sample which :has the drawbacks. of co~t in taking, -trans;orting ancf ch~m:f~cilly analysing many spec,imeris ' ' ,, of blood~ ·' , Recezit coininerci~l research h~s ·prodticed a qui~k gluco~e oxidase blc!j~J sug~r ·screening st~ip, 'ihcUgti':rt has not; .as yet, ,,been validated ·by full fieid trials. Here sl:mplici ty. is u~ually. obtained at an uriacceptab1e sacrifice. of accuracy. There has: 'been need for coin~lirison of' a:vailabl'e. Irieth6'as and at least one has riow been publlshed. 18 . It should sdori be poss1bl~ '<to itJake· a rational choic~ related to need. - 20- 4.2.2.3 Yield The yield from screening can be considered as the measure of previously unrecognized disease (whether overt or latent), diagnosed as the result of screening and brought to treatment. Other forms of yield are provided by persons with known disease who have previously lapsed from treatment. The yield is clearly primarily related to the prevalence of disease in the population and to the availability and use of medical care facilities. The highest yields from screening will be obtained from screening for a highly prevalent con- dition in a population where medical care facilities are minimal, e.g. for malari~ carriers in a poorly-developed area. Where medical care is good, though a con- dition may be relatively common, less new disease may be discovered through screening. In highly-developed countries, for example, there appear to be about as many undiagnosed as diagnosed persons with frank diabetes. The other important factor in yield is the efficiency of·the test itself. Thus urine testing for glycosuria will miss large numbers of diabetics in a popula- tion, thereby giving a poor yield. 4.2.3 The "border-line11 problem The .. ne.edocfor epidemiological surveys is perhaps best emphasized by referring briefly to ~ne of the important findings resulting from work of this kind. Measurement in probability samples of a population is tending to show that many physiological varia~les are continuously distribute~round the mean, conforming t~ a normal, or skewed normal, curve. Whether there is a separate diseased popula- tion or not cannot always be determined from the data; but in considering the .. separation between "border-lines" and "diseased" this point is not of prime importance as Fig. 4.5 shows: there is in either case an area of doubt •.. Never~ theless, the distribution of blood-pressure, bl~od cholesterol, blood sugar and intra- oc.ular tension, to give a few examples, all appear to f~vour a continuous disi;;ri::- bution. The. "diseased" part of the population occu~s at the extrerpe ~n<i of the ~stributi?n curve and ~his, as can be seen i~Fig. 4.::;, means there may?~ far more "border-lines" in a population than there are "diseased". Fig. 4-5 DISTRIBUTION OF A VARIABLE IN A POPULATION "Normal" (1) Bi~onal Di~tribution: "Diseased" B E I "Normal" (2) Unimodal Distribution: c D "Diseased" WHO 7003 7 - 21 ·- There is, however, theoretically at least, a difference between the outcome of surveys, depending on the ,dis·C.ribution of the variable or variables measured. If the distribution is bimodal (as might be expected in the case of some genetically transmitted characteristics l:i.ke, for example, phenylketonuria, the "border-line" group will in fuct comprise a mixture of persons vJi th the disease and persons ui thout the disease who:~e level of tho variable falls within the same range (between A and · , . B in the upper diagram of Fig. 4 .. 5). On the other hand if the distribution is un,;lmodal the "bcrder-·line" group v:ill comprise a homogeneous sample of persons, the questioi1 being \vhether the point beh1c-en '1dist:3.se" ar..d "norm<:~.li ty" should be set at C or D (lm'le.':' diag,ram), We m<'ty no·t.e, in passinr.;, that the figure illustrates gr13,phi(!ally ,_t:pe concepts of ser:.si tj_vi ty and sp-3cifici ty. A screening test giving a posi tivc r~adirg at the levE:l of A (-.1pper diagrJ:m) or C (lo-vTer diA.gram) would be higl':\ly sensitive" missing few ca::;e.s _, but yj_elding man~r false positives; by contrast, the "cut-off" points ~t B and D respectiVF;ly indicate a very specific "'Jest. In practice it seem:: likely that· a trial by r~andor.1j_zation of treatment should _enable a reasonable decision on the ·" cut-off11 point _to be made bet,-veen those considc:red _in need of treatment ar.d those who n:ay bo :.~eas9~1r·ed us healthy. Some false negatives would need to be accepted in a bimodal example for the sake of the S!Jecificity of the test (e.g, a ttcut-off" at E), whilst in the unimodal example a similar arbitrary choice would need to be made,· based on the response· to treatment of the "border-line" patients. It is pe~haps worth noting here that the terms 11 sensitivity" and "specificity" whilst having a cJ.ear meaning in the case of the bimodal distribution, have theoreti- cally no mea.ning for a unimodal distribution. Thus, supposing a cut-off level at E in Fig. 4.5 (the bimodal distrib"--ltioE) to be used, both true and false positives would occur at that level. -But withthe unimodal distribution, once a cut-off level has been 2doptcd! all per;;;om.> above that (c .• t;. above level D in Fig. 4.5) would be regarded as diseased, with no fz.lse pcsi t::. ves. In pranticE !3Bnsi tivi ty and specif'ici ty remain importax1t because an indirect index of dic:ea:Je .is usually adopted, subject to varj e.tJ.ons between observers end within pati:mt;:; from one occasion to ax.oth0r, a.s well as to the error of method. This index will cnly have a cer·ta~n p(,wcr L"l diagnos:i.nc~ patients and will miss some and falsely :LncLio.e ethers. - 22 - For a fuller discussion of sensitivity and specificity, as well as the reproducibility and accuracy of a test, the United States Public Health Service Monograph "Principles and Procedures in the Evaluation of Screening for Disease11 , 19 may be consulted. 4.3 ~~inciEles of early disease detection The central idea of early disease detection and treatment is essentially simple. However, the path to its successful achievement, {on the one hand bringing to treat- ment those with previously undetected disease and on the other avoiding harm to those persons not in need of treatment) is far from simple though sometimes it may appear disarmingly easy. For this reason we have devoted this section to a reasonably full discussion of a number of points which could be regarded as guides to planning case- finding. This is especially important when case-finding is carried out as a public health agency programme, where the pitfalls 1nay be more numerous than when screening is performed at a personal doctor level. For ease of description rather than from dogma we have called these points collectively "principles". attempt at elaborating at least some of these principles: 4.3.1 The condition sought should be an important problem. The following is an 4.3.2 There should be an accepted treatment for patients with recognized disease. 4.3.3 Facilities for diagnosis and treatment should be available. 4.3.4 There should be a recognizable latent or early syptomatic stage. 4.3.5 There should be a suitable test or examination. 4.3.6 The test shotud be acceptable to the population. 4.3.7 The natural history of the condition, including development from latent to declared disease, should be adequately understood. 4.3.8 There should be an agreed policy on whom to treat as patients. 4.3.9 The cost of the case-finding programme (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole. Case-finding should be a continuing process and not a "one-time" project. It is now necessary to discuss each of these headings in some detail. - 23 - Add. 4.3.i Impo~~~!:!...P~obl~ for the individual and the community This does ·act nece::sarily mean a high degree of prevalence though that would be a usual.requi:rornent. Thus diabetes mellitus is relatively highly prevalent in western populq.t5.onr. thongh fcaquently of :nild degree. On the other hand phenyl- ketonuria is e:?C.tramely u1J.comm0n but warrants screening on account of the very serious consequen-Jes if it is not discovered and treated very early in life. Clearly the iruport:-mce of the probl.em needs to be considered both from the point of view of the individual and of the community. Thus conditions with serious consequences to. tho indiv;_dual and his or her family in general may warrant relatively unec.onomic s~reoning measures; while certain individually mild conditions, but havlng· serio'vls ~onneq_ucncc::.: for the community if not discovered early and treated, will justify screening on these grounds. An example of the last kind might be the finding and control of over'I'Teight in a population. Add. 4.3.2 Accented treatment -----.. -- ---- Of all the c1iteriawhich a screening test should fulfil the ability to treat the qc:mdi tion adequately, vrhen discovered, is perhaps the most important. In keeping to the pr::.nciplo of above all avoiding doingharm to the pati~nt (the 11 primum non nee er(t of Hippo0rates), treatment must be a first aim. For declared disease tne:re. is_, of 'JO'li'Se, the ethical obligation to provide an accep:ted treatment whethE?r .-~his. is of scientific all;\r proved value or not; but, when new terri tory is being explored by the earlier detection of disease, it is clearly vital to determine ' • • I ,;: ·:· '' • - by experimental r.;urveys wh\)the:r: treating the copdi tions found at an earlier stage than was previously the practice gives a better prognosis. Unless this is so there can be no advantage to the patient and in fact, in alerting hirri or her to a condition w)<ich h<"'-::. not been shown to benefit by treatment at an earlier stage, actua.l harm may be done. This matter resolves itself. into two questions: (a) Does treatment at the pre-symptomatic t_~,roer-line stage of a disease affect its course and prognosis? (b) DoE's t:"eatment at an earlier stage than normal of the developed clinical conditj_on affo0t its course and prognosis? - 24 - Question (a) is referred to more fully in paragraph 4.3.7 below, on the need for adequate survey work. This can be tedious and clifficui t work and it may take years before results become available; but without it there is the danger that, ethically, the investigation stage will be passed and the answer never known with any real precision. Thus we are still in ignorance of the effect of treating the I' lower range of high blood-pressure. If the use of drugs for mild hypertension became general it would no longer be ethical to randomize treatment and we should have to rely· on the unsatisfactory long-term evidence of mortality and the age at which mo~bid changes appear. Border-line diabetes mellitus, ocular hypertension and asymptomatic·bacteriuria are other examples of conditions in this class. Until the needed information has been obtained there is, therefore, no case for alerting . "border-line"persons by case-finding programmes. If persons in the border~line range are to be informed of their findings they should presumably be told they are not diseased. Coming to question (b) we enter the field of accepted clinical practice and the course to be followed is largely pre-determined. Thus it is reasonable to seek in a population and treat persons with signs of clinical diabetes (possibly confined to an elevated blood sugar level) even though they may have no symptomS. It is not known whether early treatment in fact alters the out- look, but this is a plausible assumption (although there is evidence that renal and neurological changes may follow a course independent of treatment). The same argument applies to the early treatment of established chronic glaucoma, where the course of the illness is also long and the treatment unpleasant, and where doubt nrust arise as to the proportion of those prescribed treatment who adhere to it. opinion accepts the value of treatment and there is no ethical alternative. There are also conditions where evidence shows that the prognosis is Medical unaffected, or nearly so, by early recognition and treatment. Lung cancer is one example, mass radiography being the standard method of early detection. In at least one survey it has been demonstrated that the prognosis for life in patients detected at an early stage of bronchial carcinoma was little if any better than that of patients detected later. These patients had made their medical contact on account of symptoms (see paragraph 6.6.1). The same kind of findings have resulted fro~ early detection by cytological sputum examination. Whilst there is a good case for continuing efforts to detect lQ~g cancer early in selected groups, like heavy smokers, knowing that a few patients will benefit, it is doubtful whether any - 25 - useful purpose is served by advocating mass case-finding specifically for lung cancer (tho'ugh, of course, mass radiography carried out for other purposes produces a steady run of ·cases). At present it seems likely that greater efforts should be directed towards educating the public about the risks of cigarette smoking and the need· for investigating persistent cough. It is axiomatic, therefore, that case-finding should only be undertaken when the prospects for treating the condition are at least reasonable. Add. 4.3.3 Facilities for diagnosis and treatment should be available Clearly, .in planning to detect some condition, or group of conditions, ina population i,t. is a prerequisite that persons found in need of treatment should be able __ to obtain it, In general, thelarger the scheme the more this proviso.assumes importance. Thus, in introducing on a national scale cytological screening of all women at risk of uterine cancer, a major part of the scheme must consist.· in ensuring that services are available for the definitive diagno::3is and treatment of those found positive on exfoliative cytological examination. Of even greater magnitude is the problem of providing effective treatment and care in a developing country, where medical services may be extremely thinly distributed, for conditions detected by mass screening. question. In this context pulmonary tuberculosis can pose a difficult Add. 4.3.4 .Recognizable latent orearlY symptomatic stage In order usefully to detect and treat disease at an early stage there must clearly be a'reasonable period in the natural history of the development of the 66ndition.duririg.which symptoms are neither present nor at any rate clamant.· There is in fact a latent stage in many chronic diseases whic~ can be recognized, and also in the carrier state of some acute communicable diseases. However, certain chronic diseases like multiple sclerosis and arteriosclerotic cerebrovascular disease, though there must be a precursor stage, do not have a clinically recognizable latent pericd. Rheumatoid disease, for example, though there·is an early symptomatic stage, has no certainly recognizable presymptomatic state •. - 26 - Add. 4.3.5 Suitable test or examination A number of factors have to be considered. Tests can be divided into diagnos:t;ic ,and screening but this is a matter of degree rather than kind; the screening test, (which of its nature should be easy and quick to perform) is .. allowed to possess a higher margin of error, and may be less valid than a diagnostic test. For some conditions which do have a re~ognizable latent stage there is at present no suitable screening test. For example, barium meal examination for carcinoma of the stomach has been tried but found impracticable, on the ground of radiation exposure, ·discomfort to· the subject, 'a:rtd time ··needed.· (Recently, in Japan, intra-gastric photography has been developed as a screening technique and this may provesatisfactory where there is a high incidence of gastric disease, when used selectively). Similarly, addisonian anaemia could probably be detected at a pre- clinical stage if an easy test for parietal cell antibody was available. On the other hand s~me tests are accepted as suitable for screening because of their simplicity and ready application, though they may not in· fact be very good indices of the condition being looked for; for example, tonometry in the detection of chronic ., ' simple glaucoma. It is important to remember that most of the tests we use are indirect indices of the pathological process we are seeking. In general, it is reasonable to suppose, the less direct the index the greater the liability or error in diagnosis. For example, the examination of the haemoglobin, the actual pigment that is decreased in anaemia, is. a h;ighly_ preci:se .mea§we, of ana.e~ia ( regSU'_gless of type), provided the error of the technique of measurement is small. On the other hand the chest X~ray film in chest disease, the glucose level of the blood in diabetes and the intra-oc1,llar tension in glaucoma are all indirect indices of the conditions sought and, in these instances, there is likely to be an error not only of observation but 81so of interpretation. Thus the shadow, the blood level or the pressure may not always provide a valid guide to the clinical condition • . -,,.-:· It is the task of the investigator alway~ to seek more direct and more valid techniqu~s, ~!,thout sacrificing convenien9e and speed. It would be difficult to replace chest X-rays though ultimately development of a re~iable automated cytological technique applied to high risk groups of the population may prove valuable in cancer diagnosis; while there is a possibility that fatty acid or other biological variable levels in the blood might prove a more reliable index of the clinical - 27- state of diabetes t~ is the bl?od sugar. Regarding chronic glaucoma, there is now evidence that intra-ocular tension may not be a reliable index; in a recent survey as many patients with glaucoma were found within the normal range of tension as there were patients with_raised tension. 20 , 21 Finally, there.is the question ?f the validity of the test, indicated by the proportion of those examined found t.o have. falsely positive tests and the proportion found to have false negative tests, i.e. having the condition looked for but giving a negative response to the test. In case-finding work a fairly high false positive rate is acceptable but the false negative rate should be very low since missed cases can lead to individual disasters and, in the case of a communicable disease like tubercuJ.osis, to the undetected spread of disease. Acceptability topopulation Clearly a test or series of tests must be acceptable to the population to which it is offered. Acceptability is, of course, related to the nature of the risk and to the way in which the ground is prepared previously by health education. For example, the risk attending uterine cancer is by now well known to the more educated sections of western society but we are still pretty much in the dark about the attitude towards prophylactic vaginal examination of women in the lower socio-economic groups. Work is being undertaken {see section 5.6.2) on this particular aspect of ·vaginal cytology, both in Britain and the United States of America, but there is scope for much more investigation of this kind. As an example of suiting the test to existing public attitudes, Davis22 reports a higher degree of acceptance (over eighty per cent.) in a Maryland county for the self-taking of exfoliative cytological preparations using the irrigation-pipette23 than for conventional cervical smear- taking methods. This illustrates the importance of paying special attention to making a test as easy and, as little trouble to perform as possible. There are other forms of examination which, the evidence shows,_ could usefully be carried out in the se13-rch for early disease and its preventioi:l, but which are so unpleasant as to be quite unacceptable. One of these examinations is procto- sigmoidoscopy, tl'w usefulness, of which in detecting pre-cancerous conditions of the recto-sigmoid is accepted. The application of this examination is virtually limited to medical c1:i,nics anci period_ic health examination centres. - 28 - Add. 4.3.7 Need for surveys "·,-· ···r-.r· ... ..... .~ .; .w_e ha:ve ~lready, emphasized.· in both. the "Terms of Reference" and the "Defin,itions11 th:e important differE;lnce i~ concept between c~se-finding and epidemiological surveys. The need for.maintai~ing this distinction is not always clear, since the distinction itself tends to become blurred unless we look at the matter historically. We are apt to ass~me that, b~cause it is possible to carry out useful case-finding by screening for ono condition (sub-clinical pulmonary tuberculosis, for example) without the need for preliminary surveys, the same pritic;.; . . .. iple applies to other diseases, such as diabetes or chronic glaucoma. In making this assumption we forget that much survey work has been carried out on pulmonary .. tuberculosis in the past (as well as at present) and that the natural history of the' early stages of the disease has gradually become established over the course of many years of study. However, when we turn our attention to attempting tocoptrpl thE;_. new epidemi(}S of chronic non-communicable disease by similar case-find!?g techniques, we are. likely to run :;into difficulties unless we are first able to view clear.ly the _natural history, and especially the. precursor stages, of these diseases. The· most. important questions that need answering for conditions like high blood- pressure, ischaemia heart disease, diabetes mellitus and chronic simple glaucoma are: What changes should be regarded as pathological and what may be considered as physiological variations? Are·earlypathological changes progressive? Is there an effective treatment which can be shown either to halt or· reverse the early pathological changes? We-ought t,o note that we dO. not kno~ the answer to the last question even for s.ome E;Stablished clinical conditions of which diabete$ mellitus (in the progress of its complications) and chronic simple gl~ucoma are examples. An important reason for this ignoran~e_is that controlled trials of treatment were not carried out at the time when tlus might have been ethical (this applies specially to glaucoma), the controlled trial technique not having be'Em devel~-ped at the material time. The lesson for pre-symptomatic disease detection is·surely clear: controlled trials of e~rly treatm~ntneed to be. carried out as speedily as possible while they are· still considered ethical. Orice.it is regarded as norlllal'p1:'actice to treat latent disease, whether thls;\1.as been' shown to be beneficial or not, the opportunity for ·since· milch •treatment· - 29'- is both Unpleasant and, in chronic ilness, of lifelong duration following diagnosis, it is clearly important not to treat people unnecessarily. In enthusiastically attacking disease at an early stage the Hippocratic principle, previously mentioned, of "primum non nocere" should not. be neglected. Where it has been decided that a survey of the natural history of a condition and controlled trial of treatment is needed, there is a strong argument against trying to combine the survey with case-finding, as defined. For the sake of the clarity of the results it is usually considered proper that during the course of a survey it should clearly be understood by all taking part, public and investigators alike, that the work is experimental and devoted to finding the answers to certain questions as a preliminary to embarking on the next stage, that of case-finding. ·- ~ '. ·: ~·., Naturally, all persons discovered in the course of a survey to be suffering from clinical disease would be referred for treatment. The border-line group alone would be asked to submit to a randomized trial of treatment. It may be asked why surveys and ~ase-finding should not proceed together at one and the same time. Of course, to the. extent that surveys discover patients with undiagnosed clinical disease .<a.s just mentioned above) they do proceed together. But in general the aims are diff~rent and mixing them can lead -to confusion. Where · c~e-finding, with its implication of treatment, is planned it is necessary to be quite clear who is to be advised to undergo treatment and who not. This means that a decision has to be taken about. the criteria which constitute disease. Making this decision begs two vital questions which the survey side of the programme would be trying to answer, namely the question about patho;Logical significance or physiological variation (4.3.7.1), and the question of whether these early patho- logical changes are progress! ve {4. 3. 7. 2) ~- · ··· Orie ·way out ·of this difficulty, when case-finding, is·to set high criteria for the diagnosis of disease in need of treatment. The· classification by diagnostic techniques of the nborder-line" patients, and the arrangement of a controlled randomized clinical trial (which would constitute the survey side of the mixed programme), entails much clinical and administrative work, as well as long-term scientific follow-up. It might well not be justifiable, on grounds .. of cost and staffing, for a self-selected sample, such as normal case-finding methods would produce. - 30 - 24 The recent findings of a diabetes survey by Butterfield et al. emphasize the need for randomized trials of treatment before advocating case-finding on other than the strictest criteria of clinical diagnosis. A blood sugar survey of all per-sons over 21 years of age found that some 16 per cent. have "diabetes" if the commonly accepted criterion of a blood sugar level of more than 120 mg per 100 ml two hours after 50 mg of glucose by mouth is used. If these figures are extrapolated to the age/sex structure of the population of England and Wales as a whole, there would be found (on the above criteria) a prevalence of 11 diabetes11 of approximately 13 per cent. The survey by Butterfield and his colleagues took the form of case-finding (in the first place, by urine testing for glycosuria) and did not comprise a random probability sample of the general adult population (there was a 67 per cent. response). The figures may, therefore, be to some extent biased. However, it seems unlikely that they are far wrong since they are supported by those of the National Health Survey carried out by the National Centre for Health Statistics of the United States Department of Health, Education and Welfare25 which found 15.5 per cent. of persons aged 18 to 79 with a blood sugar level of 160 mg per 100 ml, or over, one hour after 50 g or oral glucose (see section 6.1.3). Whether some 13 to 15 per cent. of the entire adult population of a country of this type is in need of treatment for diabetes, and, if so, the kind of treatment that is required; urgently needs discovering. Since the treatment of 11 border-line:r diabetics (see 6.1) is not at present known to affect prognosis it seems reasonable in case-finding to advise treatment only for frank, or overt, diabetics until the outcome of surveys, at present in progress, is known. (The question of what should be regarded as declared, or established, diabetes is also discussed in section 6.1.3). Add. 4.3.8 Groups to be treated in case-finding programmes The 11 border~line" problem has already been discussed under Section 4.2.3. I.t is important, in designing a case-finding programme as opposed to a scientific survey, to have a clearly defined policy about "border-line" subjects. If it is agreed that only patients with established disease (wherever the demarcation line of the test used may be drawn} shall be treated, the management of the ~;~)order-line" patient largely depends on the design of the case detecting programmes and on communications. When a personal doctor examines a person, whether in the form of a routine medical examination or by using one or more screening tests, there is no real problem; the results are recorded on a case sheet and the patient need only be informed if there is - 31 - defin,itive disease needing treatment. The doctor in fact makes his own personal decision between what amounts to disease and what merits expectant observation, in the light of his personal knowledge of the person in question. With a·community health scheme, hQwever, great care needs to be taken in passing on information without harm resulting to so~e of the persons examined. In any given scheme there may or may not be personal doctors taking part, according to the local system of medical care. In either case, as long as a clear policy has been agreed, no confusion should result. But if persons with doubtful results are referred to their own personal physicians who have not taken part in agreeing the protocol of the case-finding programme, confusion may result; either those considered to be in need of treatment may not be treated or else those with equiviocal changes may be treated, or both. Arrangements are clearly desirable for the follow-up of the 11bord.er-line 11 cases; this might preferably be done by the personal doctor as part of a routine examination, so as to avoid segregating this group into a qpecial population which,however, is by definition, not in need of treatment. Clinics do exist, of course, for the management of persons not suffering from frank disease but for whom it is .. believed that preventive treatment is indicated. For example, there are clinics for. persons with high-risk factors for ischaemic heart disease 20' 27,28 But tiws.3 • instances there is at l~ast a measure of agreement that expeotant treatment is . effective. Apart from this type of clinic, it seems likely that clinics for following up rroorde!'..:.line" cases should be confined to surveys. Add. ~.3.9 Economic balance of the cost of case-finding in relation to total expenditure on medical care It is often considered that the detection of disease by screening will be economical of a country's resources. In order to examine this question it is perhaps __ .W~!'~~ looking at ::>9111~ qf the reas.ons why screening (specifically, as opposed to other methods of early disease detection) is considered in principle worthwhile. There appear to be two main aims, the one medical, the other economic. The Commission on Chronic Illness in the United States of America, for example, states: "multiple screening, by combining several disease-detection tests, is a stream- lined process assuring speed, efficiency and economy. Hul tiple screening contributes II 20 to good medical practice • -' The medical aim, therefore, is to improve the health of a population by the early detection and treatment of illness; while the - 32 - economic alm is to spare the time of highly-trained people by using technicians, and perhaps automated methods, as a first line in disease detection. Mass radiography, for example, both saves the time of the personal doctor by making good use of .his high index of suspicion and eliminating preliminary history-taking examination; and also may help attain the goal of better health by finding (among other conditions) latent pulmonary tuberculosis .• However, this goal can be expensive, as even highly developed countries find and there comes a point of diminishing returns. In the case of mass radiology, for example, once. the backlog of undetected tuberculosis has been worked off, the economics of advocating mass routine screening is open to question. On the other hand the very economy of a developing eountry can be threatened by uncontrolled communicable or parasitic disease and it may be necessary to carry out a mass programme with priority over medical care needs. It would be helpful to compare th~ economics of medical care provided through screening with the results obtained for similar expenditure on conventional medical care. The probability is that, though the time of individual doctors might be saved during diagnosis many more provisional diagnoses would be made through screening, which would add to the total load of diagnostfc work; and also that the total number of definitive diagnoses would be greater, leading in turn to a greater demand for medical care. As an eXample, in the C.C.I. screening survey in Baltimore,3° the rates of new cases discovered. {u~iM _t,h~ given criteria) w.ere, .per 1000. adul.ts, respectively (Fig. 4.6): Condition Cardiovascular disease Abnormal E.C.G. High blood-pressure Raised blood sugar FIG. 4.6 No. per 1000 adults 20 41 37 27 - 33 - Inall, 32 per cent. of those screened were classified as having one or more "major" abnormalities, not necessarily leading to a new diagnosis, ("major" meaning conditions unconnected with height, weight, hearing or vision). Another example is the screening programme carried out on longshoremen in the San Francisco Bay area in 1951, 31 where screening led to over 19 per cent. new diagnoses. When repeated in 1961 there were still 14 per cent. new diagnoses. The commonest conditions found were again, high blood-pressure (5 per cent.), raised blood sugar (4 per cent.), and abnormal E.C.G. (3 per cent.). Fully to diagnose and treat all these patients must add considerably to the total screening cost. Only a prospective survey to determine whether morbidity and working life were improved, in comparison with a non-screened population, could indicate the ' . ,· ·~ . . saving in cost, or otherwise, to the community. The difficulty of carrying out a trial of the kind is, of course, the need to compare like with l:i,.ke; this means that the sample under study needs to be randomized into screened and control groups. In practice this is extremely difficult.tq carry out in one population since those allocated to the control group will be likely to become"infected" by the screened group (there is good evidence that screening is po~ul~f with the general public) and to demand screening tests too •. One brief comparison is perhaps of interest: the Cortland County (New York) sc~eening project 32 in 1961 screened in nir.e ways: self-administered questionary, haematocrit, height and weight,blood sugat', chest X-ray, urinary albumin and sugar, blood-pressure, oral cytology and 12 lead E.C.G. It cost $ 14.55 (or about £ 5.4.0) per head: .whilst the average general practitioner in the United Kingdom is at the moment paid about$ 5.60 (£. 2.0.0.) per patient seen in the course of one year to cover all medical services (amounting to about $ 1.00 (seven shillings) per consultation). Thus, superficially, screening can be a costly rather than an inexpensive method of providing me.di.cal care. However, with the arrival of automation for both carrying out tests and for data processing, the cost may soon be greatly reduced. Add. 4.3.10 Case-finding should be a continuing process Much screening in the past has taken the form of one-time only "drives" or "weeks". Impetus has been put into arranging for a number of persons to be examined once, but the momentum necessary for making full use of the organization - 34 - called into being~ by .continuing examinations, has not been generated. The. "one- time" examination is clearly only of limited value~ since {a) only a small proportion (often those at. least risk) is likely to be examined and (b) the screening picks up those persons in the population who happen at t4at particular time to have the conditions sought; it cannot touch the future incidence of disease at all. Thus continuing examinations have great advantages. An organization can be built up which .can gradually become more efficient and economical~ and which can take its place as an accepted part of the normal Qedical services. Regular offers of examination are likely {with the help of health education) gradually to cover more and more of the population at risk, including by re-examination new cases of disease. 4.3.10.1 Concept of "[,)urveillanc~~ As we have already noted, many of the difficulties stemming from screening examinations are connected with problems of continuity and of communication between those initiating examinations and the doctors who hav~ personal responsibility for patients. Much of this difficulty may be avoided if arrangements can be made for screening to be carried out under the auspices of the personal doctor himself. this type of care would amoQ~t to is the routine examination of patients in What particular high-risk groups for certain conditions, at regular intervals~ as contrasted with the usual present arrangement cf a patient reporting a departure from normal health to his own·doctor. The routine examination can take the form of a·number of preliminary tests, including completion of a questionary, followed by an interview and examination by the doctor in the event of any abnormality being suggested by the tests. This kind of regular, repeated surveillance at first sight suggests the hopeless involvement of the general practitioner in unproductive routine examinations at the expense of hiB important clinical function. This; however~ should not follow apd~ indeed~ were this the only possibility, surveillance of this sort would be quite impracticable. However, in many countries health centres or group practices of one kind or another either exist or are being developed, and it is in this environment that routine surveillance has the most :favourable prospect of developing. Ancil:J.ary workers, both laboratory and secretarial, can carry out and record screening ' :..., examinations, having first selected from patients' records those persons due for - 3~ -· calling up for examination. To do this, a well-organized system of records is clearly of the first importance and, for large numbers, some form of automatic data processing is mandatory in order to sort persons for examination by age, sex, marital <.. status, parity~. etc. . ' Lookj_ng back to the past, it is interesting that where specialized data handling has not been necessary in order to examine high risk groups of the population, as for example, in maternity and child welfare work and school health services where the high-risk groups are self-selected, screening for departures from health developed at a very much earlier date. This kind of surveillance has the great advantage where a family doctor service exists that continuity of examination and care L'0.n be maintaineC!. ( ti1e pe11 sonal medical record following the patient who moves from doctor to doctor); and that only the personal medical adviser is concerned in the interpretation of tests (assisted if ne.ed be by specialist doctors) so that, in case of doubt, he can record for future information but is not put in the position of having to divulge the findings to the patient. It seems likely that the future development of screening may well be along these lines {so lor-g as a personal doctor service is available) with public health services playing an important part through the provision of premises, ancillary services and data-processing f~wili ties. Laboratory services are, of course, an absolute necessity. These may be provided either at a relatively simple level o.n the practice premises or the services of a hospital laboratory may be employed, special arrangementsbeing made for sending specimens and reports to and from the laboratory. 5. PRACTICE 5.1 Automatic data han~Uing Health investigations result in a considerable amount of data. The main problem is to handle all detailed information in a way that offers possibilities both for surveying the results and for picking out details from each individual screening. The difficulties are much the same as in sick care although the need is comparatively greater in screening. In the past the results were, for practical reasons, oft"en not treated very extensively and, for insta11.ce, vast experience has been 11 buried11 in hospital medical records and has not been easily available for such purposes as research. The use of computers has changed the situation rapidly. It is now gener.~lly agreed there should be no large screening projects without automatic data processing. Important trends may be uncovered through statistical analysis, and new knowledge accrued for medical science and medical practice alike. - 36 - 5.1.1 The particular needs of screening Screening is demanding in some respects. The data-handling equipment has to be chosen for the special needs that arise from the special circ~tances; with work in the field under primitive conditions; when several facts have to be collected on different occasions, to be collected at a later stage; or other circ~tances that are seldom encountered in usual data processing. Of the particular demands, it should be noted that the recording of primary screening results m .. 1.st be extremely simple. The method used should be easy to work with, because the heavy load makes it necessary to reduce the manual work as much as possibleo The manipulation must be easy to learn at different educational levels, both saving the time of highly trained personnel and, concurrently, giving reliable results when used by those given ad hoc training. The data-collection method should aim at recording results of screening in a machine-readable way as early as possible. It should nevertheless allow for visual reading and checking at every step of the procedure< A very large capacity is often necessary, demanding high speed and a large- size computer memory. It is sometimes useful to make arrangements for sorting selected records with results suggesting the need of further investigation. 5.L2 Basic copcepts Adapting computer techniques to screening is mainly a problem of selecting the proper methods for the practical work. pulation, and Etoring of data. The basic components are collecting, mani- For screening purposess the collection of data is the most important part of data management. The choice of methods and means must depend on the particular circumstances of the screening programmeo For instance, the need for temporary storage of primary results until further data are available for completion of the information may determine the way to get a s~table set-up. The processing ot d~ can be done in a conventional way. Although conversion between different media may be determined by the special needs of the screening prqcedures chosen for the particular project, actual manipulation of screening data - the data processing - is done by customary computer technique. - 37 - The storage of' data is also concerned tAli th much the same problems as adminis- trative data handling, and is mostly associated with the difficulties of storing large amounts of data and still having the details retrievable within a short enough access time. Technfcal evolution is proceeding rapidly in this field. The necessary volume for a given amount of data is decreasing, as is the cost. Data collection The problems in data handling for screening are concerned with the collection of data, and of obtaining the information in a machine-understandable shape. In general, the con,ventional means of recording information can be used, e.g. punc,h cards, paper tape or magnetic tape. The choice is, however, dependent . on the partic~ar properties of the media. propertie.s will exemplify this. Some elementary notations of the 5.1.3.1 The punch card is easy to handle and can be punched, checked and sorted .by ;simple machines. For screening, highly effe c.ti ve combinations can be used by combining text en clair with the punched ,holes. A definite advantage is that punch cards allow new information to be added whenever needed, for instance, during different steps in a multiphasic health examination. In order to get enough information on every card (which ordinarily means 80 alpha-numerical signs) a reduction of data is often adopted, for instance, by coding classes of values instead of the individual digital values. Sometimes the vulnerability to damage by handling may be a serious drawback to the use of punch cards for screening; under primitive conditions it may be difficult to avoid such damage. There are many ways of changing punch cards to adapt them for special needs' which are o£ interest for screening purposes. For instance, a combination card, the dual card, offers the possibility of having visual information;-·etc ~; connected to the coded punch card. Punching is somewhat tedious work and needs well-trained personnel, and also introduces errors. Duplicate punching is often necessary, because errors in digital information may be misleading and d~gerous. can be done automatically by machine (a "verifier"). The checking of duplicates - 38 - Another technique of special interest in primary data collection for screening is the use of mark-sensing cards. Instead of punching holes, marks are made on special areas by ordinary lead pencils, or by special ferrite pencils. The marks can then be read by machine. The mark-sensing cards have less capacity, since the marks need more space. They are easy to handle, but it is easy to make faulty notations. For certain purposes, the IBM "Port-A-Punch11 is useful. The cards are partly perforated, and a pencil gives a clear hole in an easy way. The error in mark sensing has always to be considered. However, it has been adapted for laboratory work, etc., with highly satisfactory results. 5.1.3.2 Paper tape as a medium is increasing in popularity. However, a special punch is needed, which is inconvenient. One of the main drawbacks is that, of their nature, paper tape stores sequentially arrange information~ Therefore, addition of data (or sorting, etc.) cannot usually be done without retyping. There are also paper tapes with extra space for adding, for example, text en clair and identification. Paper tape is somewhat unwieldy to handle in quantity. It is nevertheless a cheap and convenient way of storing information, especially for later data processing. Special edge-punched cards provide a simple. form of data recording, and are convenient on a small scale .where, for instance, identification, standardized information or procedures are comparatively often needed. The data can easily be extracted with a minimum of errors. Electric typewriters combined with punch and reader for paper tape (Flexo- writer, Teletypewriter, etc.) constitute self-contained equipment. Such equipment can serve .as a terminal, and is handy and not too costly. Simultaneously, it gives text en clair for visual checking of the information stored on the paper tape. When needed, the information can then be sent to a distant computer centre by using a modulating unit and ordinary telephone connexions. 5.1.3.3 Magnetic tape and other magnetic media, such as discs, will presumably dominate the future scene. The advantages are striking, with high capacity and readily available information. On a l~rge scale, they are economical. For health screening, however, the use of these media is of interest only for the processing and storage of information in a data centre where other factors than - 39 - screening.n.eeds will be decisive. . . ToO.a.y, recording on magn,etiq , tape is still. somewha~ complicated, and requires special equipment. For a .considerable time hence, it will be easier to use other media, such as cards and paper tape, rather than store information on magnetic media. 5.1.4 Data proces~:~ing There-is a discernible trend to use large and fast computers centrally located, instead of smaller ones working at a lower speed. By means of a time-sharing technique, a large computer can perform many different operations concurrently, and is more economical. It is true that small desk-size aomputers are now available, which are comparatively inexpensive and not too slow. However, they are used mostly for data collection on line and as "terminals 11 , or for data reduction. They will presumably not replace large computers to any appreciable extent, especially because the development of data-transmission technique has been very rapid. In a recent number of Scientific American, containing several articles of advanced computer technique, Fano & Corbato33 have described - in a popular way - the possibilities offered by the time-sharing technique. 5-1.5 Data storage Generally speaking, extremely large memory capacity can be achieved today only by magnetic tape. This is a comparatively cheap method and not too space- demanding, but it has the disadvantage of a long access time. Promising evolution is occurring with the production of large disc memories with very high capacity and a very short access time. In the future, it can be expected that computer centres will be able to store enormous amounts of data, and also to distribute them automatically to desired terminals. For a considerable time, however, it may be necessary to work both with easily accessible data which are carefully selected, and cheap means (such as magnetic tape) with a comparatively long access time. 5.2 Present screening practice (including multiple screening) 5.2.1 Introduction As we have seen earlier in this paper, screening developed in the control of endemic communicable diseases, and preventing the spread of conditions like malaria and syphilis still depends in part on screening measures. The development of - 40 - interest in screening for chronic non-communicable diseases since the Second World War has resulted, in the United States of America particularly, in a large number of "programmes" where varying combinations of screening tests have been employed. In Table 5.1 we have listed the principal conditions for which screening has been carried out, sub-divided into communicable (infective and parasitic.} .and non- communicable, and categorized by selective screening by age and type of country. Condition Hearing Vision Phenylketonuria Cong. dislocation hip Rheumatic heart dis. I Congenital heart dis. Hernia Overvveight Diabetes mellitus Anaemia Chronic glaucoma High blood-pressure Renal tract disease Ischaemic heart dis. Rheumatic diseases Cancer of lung Cancer of skin Cancer of bladder Cancer of rectum Cancer of mouth Cancer of cervix Cancer of breast Mental illness Malnutrition Pulmonary tuberculosis V .D., syphilis V .D., gonorrhoea - 41 - CONDmONS SCREENED BY . AGE AND TYPE OF COUNTRY .. Infancy X X X X X X X X X ... Selection by age .. ·-· ..... . Child- hood. Adult ······ ... X X X X X X X X X X X X X X X X X X X ·····--·x-····- r ·· X X X X . I Mlddle l Elderl age y .. . j ... X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Type of development of country High /Middle Least X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - 42 - TABLE 5.1. CONDITIONS SCREENED BY AGE AND TYP~ OF COUNTRY, (continued) .. , ... . "' ' ...... <• .. ·~ . . ' ...... ¥ ·-"''' ' . 0 •••e•' '" ,,,., ... Selection by age Type of development of country ··- .. .. Condition ._ .. Child- I ~d~lt " I . . ..Infancy Middle I Elderl High Middle Least hood 1 age : y : .. ! I ... ... .., . I Urinary tract I infection X X X I X Non-spec. lung I i I I disease X X X X I Infective hepatitis X X I X X )( I Histoplasmosis r X X Coccidiomycosis Trachoma . Yaws X Carrier conditions: Streptococcal X X X Dysentery X X X X X Typhoid X X X Diphtheria X X X Poliomyelitis X X X Malaria X i X Filariasis I X I X I I Schistosomiasis I X X I Trypanosomiasis X I X X Ankylostomiasis X X I Kala-azar ! I I X x I i I Hydatid disease ! X I Trichinosis X X Tinea capitis I X I --43 - 5.2.2 Comparison-:of studies It would greatly lengthen this paper without, in our opinion, being corres- pond·rngry---profitabie·;·to--a-ttempt-a··deta:tleci .. i·ev:i.ew-of'-a:li ··t:his ·-w~~"k·:·--:--Th~·-t;oj;i~~i----- and subti'Q..P:i.cal diseases are a specialized field and might well be treated as the sub~ec--e-or-··a.···separate· ·pa:pe:r-;·----- ca.se:..firia1rig for· the chronic non-communicable diseases-(las been reviewed in the past {see below) and the difficulties have thereby become--a:pparent";···- """These--a:r·e 1ri "the main" twofold;. firstly,. the difficulty of evaluatfng the results of these programmes because of the inability, in general, to follow up the result of screening and determine the effect of early treatment on those diagnosed; and secondly, because the methods employed in different programmes are rarely comparableo. A study of multiple screening was made some years ag_o by the American Medical Association5 which illustrates these points · clearly.. As an example of the type of results obtained in Table 5•2, adapted from a paper by Breslow, 7 is reproduced. It must be noted that the criteria used for diagnosing "new cases discovered" may have varied, the inclusion of more or lessborder-line cases depending, perhaps, finally on the judgement of the physician. (In passing, this number of the J. chron. Dis. also contains other articles on screening for asymptomatic disease, including heart disease, cancer, diabetes, anaemia and glaucoma.) 5.2.3 Examples of multiple screening projects At least three highly organized multiple screening programmes have been carried out in the United_States of America; two under the auspices of the Commission on Chronic Illness, an urban one in Baltimore (already referred to)34 and a rural one in Hunterdon County, New Jersey, 35 and one by the Kaiser Foundation, the Longshoremen's Union, and California State Health Department in San Francisco, also referred to above.31,36,37 5.2.3.1 Baltimore Tables 5.3 and 5.4 demonstrate the main findings of the Baltimore screening from which it is seen that the largest number of previously undetected conditions are found by electrocardiography, sphygmomanometry, blood sugar estimation and chest radiology. Some 63 per cent~ of the sample population had some abuor.ma,:!,Jty dis- ····~ -~~;.oo---·,~oo•••--·--··---·~··· .. •-~---· ~·--·oho ___ _,..._, ................ -yOO••"-"'' ~- ,,,,.,, ....... - ,,,,, _ _.,o~oh>,.o•·--···-.. - ... - .. M·~W,o-oo> -~···•••·<"' ..... 0 ...... ,.._ .. covered by screening, of which half were "major" (i.e. unconnected with height and --44 -·· TABLE 5.2. EXAMPLES OF MULTIPHASIC SCREENJ:NG .RE.~'J;'_$ _ ...... IN CALIFORNIA, 1948-1954 . ··Positive Diagnosis New cases Numbe~ of screening test · confirmed discovered Nature of test persons - ' I tested No. Rate per No. Rate per No. Rate per .. j· 1000 1000 1000 Serologic test for i sy~hilis . 3 974 412 104··· 159 40 23 6 21 733 '1 949 90 425 20 1188 9 Chest X-ray 1 755 001 (a) For tuberculosis (a) 55 210 32 18 939 11 NA NA!! I (b) For heart disease (b) 10 899 ·6 I 3 388 2 697 .. 0.4 (c) For lung canc.er (c) 3500 2 339 0.2 339 0.2 3 990 (a) NA NA 29 7 13 3 (b) i 0.8 21 5 i 9 2 3 I (c) . NA NA >" ·-------~-- ···- • 1 0.3 1 0.3 4 167 (a) 157 38 32 8 3 0.7 I (b) 40 10 I NA .NA NA NA --- ....... -·· ··-· ... -·· -- --···-~-~... ·- .. •• >' . - ~ - ... (c) 15 4 4 1 4 1 Blood·sugar 3 124 Sl+ 17 18 6 11 4 14 863 259 17 127 9 73 5 3 966 1561 I 39 56 14 34 9 . 2 162 311 14 20 10 9 4 3 543 531 15 5 1 5 1 871 : 2 856 30 271 9 16 ' 6 I I ! Urine sugar 3 132 161 5 11 4 -~· 1 1991 - ... _., ... 3 987 50 54 14 29 7 . ' .I Electrocardiogram 2 250 I 395 l 175 110 49 NA .NA I 3 984 666 167 301 l .76 182 46 Blood-pressure 3 989 837 210 3691 92 1207 52 !! NA = not available •. - 45 -. ·-~ TABLE 5 .• 2 •.. EXAMPLEs· OF MULTIPHASIC SCR.E:E!NmG REsutl.rS IN CALIFORNIA, 1948-1954 (continued) '· . - -·· -· ,_ -· ... ·- ........... -- ...... ~·-· ~- ... ~- •o 0 M- o> ~· .. Positive Diagnosis New cases Number of' screening test conf'irmed discovered Nature of test persons , ... ---...---·· . ··- ~~ . ._ ... · ·tested ······No. · ... ; .. ..... Rate per No.· Rate per Ncr~ Rate· per 1000 1000 1000 Haemoglobin 3 986 5 1 1 <1 1 <1 Urine albumin 3 988 92 23 35 9 16 4 Vision . :? .972. -< •• .... 9~t .. ... ~38. . 395 99 205 5~. ___ ..., _____ ,., ... ~- ... ·-· , . ., ...... ... .. ... Height and weight Overweight. 3 992 36o 90 241 6o 74 '19 Underweight 2 86o 90 32 90 32 NA~ NA a - NA = not available. ..... - 46 - TABLE 5.3. ABNORMALITIES FOUND IN THE BALTIMORE SCREENING PROJECT, 1957 Number of abnormalities present Total number of persons screened .. ·-··· . . -· .. - No abnormalities Minor abnormalities only One or more major abnormality Occurrence of major abnormalities: One major abnormality Two major abnormalities Three major abnormalities Four major abnormalities Five or more major abnormalities Number - --. 2 024 -- 743 629 652 4.34 140 . - . - 55 17 6 · ------Percentage·· o:f --- Total with Total major abnormality 100.0 - ·•· - -· - . 36.7 - 31.1 - 32.2 100.0 21.4 66.6 6.9 21.5 .... ··~-···~--~- ..... ----····· .... 2.7 8.4 0.8 2.6 0.3 0.9 - 47 - TABLE 5.4. RESULTS OF SCREENING TESTS ON ADULTS · EVAWATED, AND DIAGNOSES BY PHYSICIANS ' Test results Confirmed by diagnosis Number Test of persons Previously Previously (unweighted) Negative Positive Total known to unknown to ·'. patient patient I Relative rates per 1000 persons screened 70 min' chest X-ray: Tuber'culosis 537 984.4 15.6 3.9 0.5 3.4 Cardiovascular r· d,isease · 537 943.6 56.4 .. 32.9 12.9 20.0 " .. Other ' 537 976.0 24~6 .· 5.1 1.3 3.8 EKG .. 571 874.3 '125. 7 70.3 29.7 40.6 Blood-pr~ssure 6o8 887.9 '112:1 •105.5 68.9· 36.6 Blood suga:r .. ' ·. :-·· .. 601 936.9 6::2.:.l~ 39.5 12.4 27.1 ,_,_ Urine sugar 586 995.1 4.9 4.7 4.3 0.4 Urine albumin 572 991.6 8.4 4.7 3-9 0 •. 8 .. ····¥. ···- . ., . - 48 - weight, hearing or vision, as tested). The most frequent disorder previously unknown to the person screened was heart disease (with electrocardiograph changes and raised blood-pressure as the indicators).· A·critiqu~ ana follow U:p of this screening pro~~~ has been carried out by Wylie. 38•39 He found that only 29 per cent. of the invited sample attended for tests. The participants represented that part of the population less likely to be in need of medical care than the non- participants. The five-year mortality of those screened, age for age, was the same or higher than the five-year mortality experience of the non-participants, varying from six per 1000 at risk for those less than 35 years of age to 149 per 1000 at 50 and over. There was no evidence in the secular trend that persons screened experienced a lower than expected mortality in the first years after screening. Moreover, Wylie found that the age-adjusted mortality of those screened with "major" defects was nearly twice as high as that for persons with "minor" defects and over 18 times as high as persons with negative tests. This is, of course, not unexpected and only demonstrates that people who are ill die sooner than those who are not. It does not show whether or not earlier diagnosis improves the prognosis. 5 • 2. 3 • 2 · .. Hunt-ardon County In the Hunterdon County Survey a sample of the population was examined clinically ("Clinical Evaluation Sample") and a second sample was submitted to multiple screening procedures, as was done in the Baltimore Survey. There was, however, a difference from the Baltimore Survey in that part of the 11Clinical Evaluation Sample" was also examined by multiple screening so that it was possible to compare the results of screening with the known prevalence of disease. The findings are interesting, though difficult to interpret because of the clinical "border-line" problem raised by the use of the technique of screening. Table 5.5 (copied from page 273 of the Hunterdon County report) shows the proportion of abnormal findings from screening, discovered in the "Clinical Evaluation Sample". It is noticeable, as with the Baltimore Survey, that cardiovascular abnormalities are the commonest, even though only 40 per cent. of the sample was aged over 44 years. Table 5.6 (Table 11-1 facing page 304 of the report) gives the most complete analysis available of the Hunterdon screening clinical evaluation comparison. - 49 - TABLE 5.5 Screening test VDRL Mazzini-lipoidal antigen test for syphilis X-ray for chest disease other than tuberculosis (70 mm) Chest X-ray for tuberculosis (70 mm) 1 Blood glucose above 130 mg I I Urinalysis for albUmin - positive or doubtful Urinalysis for sugar- positive or doubtful Diastolic blood pressure - level of 100 mm Audiometer X-ray for cardiovascular disease (70 mm) Weight status Far vision Electrocardiogram (12-lead) Systolic blood-pressure - level of 150 mm Haden-Hausser test for haemoglobin ElectroGardiogram (lead 1) Near vision Less than Abnormal findings (weighted percentage of screened persons) 1 1 1 3 3 4 5 6 11 12 18 18 19 19 19 22 29 TA BL E 5. 6. 1. M U I! rif 'L E SC RE EN IN G RE SU LT S IN T HE H UN 'l'E RD O' l ST UD Y A ll er gi c, I Sc re en in g r e s u lt s a c e . In fe ct iv e a n d N eo pl as m s M al ig na nt e n do cr in e sy st em , D ia be te s B lo od a n d bl oo d- ' Pe rn ic 10 U !I Ir on pa ra si te d is ea se s S yp hi li s pu lm . m e ta bo li c a n d O be si ty fo rm in g o rg an s . de fi ci en cy to f in al d ia gn os is w he re of w he re of n e o pl as m n u tr it io n al d is ea se s m e ll it us w he re of a n a e m ia a n a e m ia ' w he re of T es t a n d N o. o f ' 00 2- 13 8 02 4- 02 :9 14 0- 23 9 16 3 24 0- 28 9 26 o 28 7 29 0- 29 9 29 0 29 1 to ta l n u m be r po s. t e st s ' W ei gh t jud ge me nt ' ; O ve rw ei gh t 12 3 1 1 ' l - 10 7 11 96 3 - - U nd er w ei gh t 6 - - - - - - - - i - , , I - N o. o f te st s 59 7 i M M R -t ub er cu lo si s Su sp ic io us 43 8 2 i 2 l 6 l 5 3 - ! - N o. o f te st s 75 B : ' : ' M M R- ch es t di se as es i l Su sp ic io us f in di ng s 17 - - ; 1 1 3 2 1 1 - - N o. o f te st s 67 2 I I M M R -c ar di ov as cu la r ; Su sp ic io us f in ai ng s 15 4 3 l l - 47 13 33 6 1 1 - N o. o f te st s 74 6 I E le ct ro ca ra io gr am A bn or m al /b or de rl in e 25 5 3 l I 2 - 94 23 70 6 1 1 N o. o f te st s 73 7 i ' D ia st ol ic p re ss ur e ; ~ O ve r 10 0 nu n 11 4 1 - - - 61 lO 50 1 - ' - N o. o f te st s 82 4 - - - S ys to li c pr es su re O ve r 15 0 m m ~5 3 4 2 1 - 10 2 26 75 4 1 1 N o. o f te st s 82 4 H ae m og lo bi n ' P os it iv e 98 5 3 - - 6 3 3 8 - 1 N o. o f te st s 8o 8 - - Sy ph il is -M az zi ni Po s. o r do ub tf ul 6 4 4 - - 2 - 2 1 - - ' N o. o f te st s 75 6 . B lo od g lu co se ' ' • \h ov e 13 0 m g/% 40 - - _ 31 20 11 l - - N : •• o f t. es t: • r. c - - - - ' U ri na ly si s- su ga r ' P os . o r ri ou bt fu l 47 2 1 l l' 30 22 6 5 - - N o. o f te st s 82 3 ' - U ri na ly si s- al bu m in Po s. 0 r do ub tf ul )Q - - - - 8 3 5 1 - - ' N o. o f te st s 82 0 ' " I TO TA L CO ND IT IO NS 10 4 4 l DI AG NO S~ ") 20 2 32 16 7 18 1 3 - - - - - - TA 3L E ' • . · i., . '· !U LT I? LE 3 ':i >• :i' :!I IN :} :iE SU TJ TS I t! T HE h l!N TE :'\D ON S TU DY I .3 cr ~' ~n in .z ; c< ?.- ;-J t. c:: :l c .· :: :: ., ·i. :e as es o : tn .-; · . . e u r: 1a t. ic- · Ct t~ e, ., .. -. ~a r· ~ , , , ce~ los cle :·o ·.J c :. e a r· .r 4r '. e ri os cl er oc ic I A cu te I An gi na I Es se n ti al I I O th er c i r c u la to r' y s ys te rr ; 1i .:; ea se a n d . c o ro n a ry . u e n ig n H yp er te ns io n hy pe rt en si ve 7' i.! 1a l ,- ;i a ;:; !l. l. ~-; w he re of fe v er ji. 3e .:l ::: e c o ro n a ry d is ea se n e a r t U ls ea se o c c lu si o n pe ct or lS hy pe rt en si on d is ea se T es t a n d NC >. o f 40 0- <~ iJ ti IJ O Q- 40 1.3 41 (, 42 0 I 42 0. 0 I 42 0. 1 I 42 0. 2 I 44 4 I 44 6 I 44 7 :o ta l n u m be r po s. te st s W ei gh t ju dg em en t O ve cw e1 gn t 12 5 I 10 3 I l I 4 I 19 I 10 I 4 I 2 I 7 I I 23 U nd er w ei gh t G 3 2 N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sl -J :c io us I 43 I 2- ) I I I I 6 I 7 I I 2 I - I 6 N o. o f te st s 7~ 8 M M R -c he st d is ea se s Su sp ic io us f in di ng s I 17 I 16 ! l I 1 I 3 I 2 I I 2 I 2 I I 3 N o. o f te st s 67 2 M M R -c ar di ov as cu 1a r Su sp ic io us f in di ng s I 15 4 I 16 5 I 2 I ' !1 1 I 32 I 28 I 6 I 2 I 6 I I 17 N o. o f te st s 74 6 t-- E le ct ro ca rd io gr am A bn or m al /b or de rl in e 25 5 ~~ 9 l 16 65 30 10 7 10 - 27 N o. o f te st s 73 7 ' D ia st o li c p re ss u re I \.} 1 ,_ . O ve r 10 0 m m 11 1+ b 9 - c 26 11 4 7 5 1 44 N o. o f te s ts 8~ U - · S y st o li c p re ss u re Ov e~ 1~ )0 m m I c: ~3 I c: 9~ I I 13 I t> l I 35 I ll I 10 I 22 I 2 I 57 ~o . o f te s ts 82 4 - - - i-l .a cr r:o gl ob in P os i t..i v e I 9" I )0 I I l I 9 I 10 I - I 1 I 2 I - I 2 N o. 8 f te st ,.: :; . SQ 3 - 3y p hi l i s -r ~a zz in i Po s. o r do ub tf ul I I 4 I I I I I I I 1 N o. o f te st s 75 6 r - Bl] od . . :: lu co se .~h ove i. so m e;, /% I • iO I 3~ I I I 10 I 7 I I ' ! I 1 I I 4 :J o. o f t~ st s '7 50 Ur in al y~ is -s ug ar P o3 . o r do ub tf ul I I l N o. o f te st s 75 0 1- -- U~ in al Y3 is -3 lb um in P o s. o ;· ; J u b tf u l I ~0 I 1..: .:) I 4 I 10 I " I I 3 I I I 6 N o. o f te s~ ,s ? -;:. o TO TA L SO NI JIT IO NS I I l.t 3'~ 2 I . 39 I 9~· I DI A} NO St. 'D 50 I ll t I 12 I . 37 I 2 I 83 TA BL E 5. 6. 3. - O th er O th er Sc re en in g r e s u lt s a c e . r e s pi ra to ry c hr on ic to fi na l di ag no si s d is ea se s in te rs ti ti al w he re of pn eu m on ia T es t a n d N o. o f 51 0- 52 7 52 5 T ot al n u m be r po s. te st s W ei gh t ju dg em en t O ve rw ei gh t 12 3 6 1 U nd er w ei gh t 6 - - N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sp ic io us 43 16 2 N o. c if te st s 75 8 M M R -c he st di se as es Su sp ic io us f in di ng s 17 4 - N o. o f te st s 67 2 M M R -c ar di ov as cu la r Su sp ic io us f in di ng s 15 4 9 1 N o. o f te st s 74 6 E le ct ro ca rd io gr am A bn or m al /b or de rl in e 25 5 19 2 N o. o f te st s 73 7 D ia st ol ic p re ss ur e O ve r 10 0 m m 11 4 iJ 1 N o. o f te st s 82 4 S ys to li c p re ss u re O ve r 1:: ;0 m m 25 3 20 2 N o. o f te st s 82 4 H ae m og lo bi n P os it iv e 98 7 - N o. o f te st s Bo B Sy ph il is -M az zi ni P os . o r do ub tf ul 6 2 - N o. o f te st s 75 6 B lo od g lu co se A bo ve 1 30 m r;/ % 40 2 - N o. o f te st s 75 0 U ri na ly si s- su ga r Po s. o r do ub tf ul 47 3 - N o. o f te st s 75 0 U ri na ly si s- al bu m in Po s. o r do ub tf ul 50 5 - N o. o f te st s 82 0 TO TA L CO ND IT IO NS 45 5 DI AG NO SE D M UL TI PL E SC RE EN IN G RE SU LT S IN T HE H UN TE RD ON S TU DY D is ea se s o f th e H er ni a o f D is ea se s o f th e B ro nc hi ec ta si s Em ph ys em a a bd om in al ge ni to -u ri na ry di ge st iv e c a v it y s ys te m , w he re of sy st em 52 6 52 7. 1 53 0- 58 7 56 o- 56 1 59 0- 63 7 2 2 - - 6 - - - - - 2 8 - - 2 - 3 - 1 - 1 s 1 2 3 1 2 1 2 6 1 3 - - - 5 12 1 l 6 - ~ - 2 1 - l - - 1 1 1 - 1 - 1 2 - 1 - 1 3 - - 4 7 26 2 3 32 P y el it is C y st it is e tc . 6o o. o 6o S l 2 - - - - - - - - - 1 - - - 2 - 1 - 1 - - - - - 2 1 17 C on ge ni ta l m a lf or m at io ns 75 0- 75 9 2 - 1 - 3 4 1 2 1 - - - - 9 Sy m pt om s, s e n il it y a nc i il l- d ef in ed c o n < i.i tio ns 7B o- 79 5 1 - 1 - 2 3 2 1 - - - - 6 7 1. ]1 1 \) Sc re en in g r e s u lt s a c e . ~o fi n al d ia gn os is ' re s t a n d T ot al n u m be r W ei gh t jud ge me nt O ve rw ei gh t U nd er w ei gh t N o. o f te st s 59 7 M M R -t ub er cu lo si s Su sp iC iO U S· N o. o f te st s 75 8 ~B -~ he st d is ea se s Su sp ic io us f in di ng s No • . o f te st s 67 2 M M R -c ar di ov as cu la r S u sp ic io u s fi n d in gs N o. o f te st s 74 6 E le ct ro ca rd io gr am A bn or m al /b or de rl in e N o. o f te st s 73 7 D ia st ol ic p re ss ur e O ve r 10 0 m m N o. o f te st s 82 4 S ys to li c pr es su rP . O ve r Fi O r m n N o. o f te st s 82 4 H ae m og lo bi :o N o. o f po s. t e st s 12 3 6 43 l7 15 4 25 5 11 4 25 3 P os it iv e I c· c N o. o f te st .:: :; ~( )? Sy ph il is ~M az zt Gl . f- -- -- -- -. -- -- -· - B lc cJ t ~li ..t ·': :JS !': P os . o r do .Jb ',! .'u l t N: :~ o f " Se .-s t.s -;· ~~ A. bo ve j _ "' JO . . ~,.; ..'I _ 10 N o. 'J r t. e_ st s 7 · U ri :-. . :1 !. ~/ s t s - 5 ' r .,_ ,r- 1: · P os . o r a o u u tf u l N o. o f te st s 75 0 U ri n al ys is -a lb u m in P os . o r d ou b t: u l N o. o f ~e ~t s 32 0 TO TA L. C ON DI TI ON S DI AG NO SE D 47 50 TA BL E 5. 6. 4. M UL TI PL E SC RE EN IN G RE SU LT S IN T HE H UN TE RD ON S TU DY Sy m pt om s, s e n il it y a n d il l- de fi ne d c o n di ti on s, w he re of 78 o- 79 5 l l 2 3 2 1. . l 0 7 I I I Al bu m in ur ia u n qu al if ie d A lb um in ur ia o r th o st at ic 78 9. 0 78 9. 1 l I l 1 1 l I - l I 4 l if 1 G ly co su ri a U ra em ia 78 9. 6 79 2 l l I - I 1 I 1 I I I ~ 1 1 l l - 54 - 5.2.3.3 San Francisco longshoremen In 1951 nearly 4000 longshoremen (dockers) were examined by a multiple screening t hni . S F . 31 ec que 1n an ranc1sco. In 196o the mortality and morbidity of the original group (of whom the records of over 3000 were available) were followed up and a repeat examination was given a sample of 818 persons. In the 1951 screening (which included hearing and vision tests, 70 mm chest X-ray, E.C.G., blood-pressure, serology for syphilis, haemoglobin, urinary albu~in test, post-prandial glucose tolerance test, measurement of height and weight and a self-administered questionnaire) 63 per cent. of persons examined had positive tests which had led to the finding of 35 per cent. with clinical disease. Of these, over half (19 per cent.) had previously undiagnosed disease; (that is, about one fifth of those examined were found to have disease previously undiagnosed). 5.2.3.4 Chicago Board of Health More recently, the Chicago Board of Health has initiated a Demonstration Chronic Disease Project on three city housing-sites, under the direction of 40 Dr J. Stamler. This project combines a battery of screening tests with a physical examination and it should provide useful information on the contribution to diagnosis and prognosis made by screening as an aid to physical diagnosis. 5.2.3.5 Common conditions for screening In this part of this paper we do not propo~e to discuss in any detail con- ditions for which it is common to screen; some of the more important of these conditions are treated in some detail in section 5.8. It will suffice here simply to note the chronic conditions for which screening is commonly carried out, either singly or as a multiple operation, and to see to what extent they satisfy the main criteria we have noted under 11 Principles 11 • An analy~is is set out in Table 5.7. In deciding in any given instance about the value of screening,local circumstances of course play a large part. Naturally, as indicated in Table 5.1, selective screening by age-groups will in all cases give higher yields and in certain instances (e.g. phenylketonuria and congenital dislocation of the hip) this is mandatory in order to be of use. In individual instances, also, it is possible to avoid some or all of the objections indicated in the Table and it would. of course be wrong to dogmatize. Allowing, however, for the objections to current case-finding, the conditions which score most heavily in favour are seen in Table 5.7. R el at iv e j P re na ta l w o rt h + pr eg na nc y G re at er L es se r To xa em ia Rh f ac to r A na em ia Sy ph il is A sy m pt om at ic ba ct er iu ri as D ia be te s m e ll it us TA BL E 5. 7. ES TI M AT E OF C ON DI TI ON S FO R CA SE -F IN DI NG B Y VA LU E N eo na ta l Lo co m ot or (C on ge ni ta l di sl oc at io n o f hi p) In bo rn e rr o rs o f m e ta bo lis m (p he ny l- ke to nu ri a) C on ge ni ta l de fe ct s (h ea rt) In fa nc y A na em ia V is io n (a mb ly op ia) H ea ri ng (c on ge ni ta l de af ne ss ) C hi ld ho od M en ta l de ve lo pm en t C on ge ni ta l he ar t di se as e A du lt Sy ph il is - - - - - - - G on or rh oe a - - - - - C an ce rs - u te ru s - bl ad de r - s ki n - m o u th Pu lm on ar y O ld a ge A na em ia s - - - - - - - - - - - - - - - - V is io n (C at ar ac t - s e n il e m a c u la r de ge ne ra - ti on ) H ea ri ng - - - - - - - - - - - - - - - - - (O tit is - c o n du ct io n de af ne ss ) Lo co m ot or (a rt hr iti s) H er ni a - - - - - - - - - - - - tu be rc ul os is - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - O ve rw ei gh t - - - - - - - - - - - - - - - - - - - - - - - - - - - - - C an ce r - br ea st D ia be te s m e ll it us (s tr ic t c r it er ia ) Is ch ae m ic h ea rt di Se as e (s tr ic t c r it er ia V l V l R el at iv e w o rt h L es se r TA BL E 5. 7. ES TI M AT E OF C ON DI TI ON S FO R CA SE -F IN DI NG B Y VA LU E (c on tin ue d) P re na ta l N eo na ta l In fa nc y C hi ld ho od A du lt O ld a ge + pr eg na nc y H ig h bl oo d- pr es su re C hr on ic s im pl e (s tr ic t c r it er ia ) gl au co m a M en ta l il ln es s ( Se le ct iv e - - - - - - - - - - - - - s c re e n in g - - - - - - - - - - - - - - - - o n ly ) ~ - - - - - - - - - - - - - - ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - V1 0 \ - 57 - This is not a very long list and it will at once be noticed that certain conditions, glaucoma for instance, have been given a low priority despite the fact that they may constitute more of a public health problem than some of the conditions listed. This kind of variation from the more usual emphasis placed on certain conditions is explained by selection in the above list on grounds of all-round feasibility. Tne pros and cons for screening for some of the major conditions listed in Table 5.8 are discussed subject by subject in Chapter 6. 5.3 Epidemiological studies Having stated that much of the practice of case-finding poses problems which have not been solved (and this can be discerned from a study of the case-finding project mentioned above) it is, we believe, worth while at this point considering briefly what is being done to remedy this situation. Finding answers to the questions of the value of early treatment and the policy to be adopted over the border-linepatientinvolves, as we have stressed earlier in this paper, epidemio- logical surveys (Evaluation of screening procedures, 4.2.2). We have attempted to set out in the accompanying Tables (Tables 5.9.1 to 5.9.5) in the form of examples only, some of the work at present in progress in different fields. Under "Comments" we have tried to indicate where there are gaps in our knowledge which might be filled by further studies. Much of this work is still in progress and at the moment unpublished. It is more than possible that the choice of examples may appear unbalanced and, if so, this can be attributed to the fact that we have chosen them from our own knowledge only and have not sought to be exhaustive. C on di ti on s fo r w hi ch e a rl y P ub li c he al th d et ec ti o n i s e m pl oy ed pr ob le m D ia be te s m e ll it us + H ea rt d is ea se : Is ch ae m ia + R he um at ic + C on ge ni ta l ? H ig h bl oo d- pr es su re + O ve rw ei gh t + Lu ng d is ea se : tu be rc ul os is + n o n -s pe ci fi c + c a n c e r + R en al d is ea se : n e ph ri ti s ? ba ct er iu ri a + A na em ia (ir on d ef ic ie nc y} + A rt h ri ti s: rh eu m at oi d + go ut + B re as t c a n c e r + U te ri ne c a n c e r + R ec ta l c a n c e r + O ra l c a n c e r + B la dd .e r c a n c e r + Sk in c a n c e r + H ea ri ng ( in he rit ed a n • + a c qu ir ed d ea fn es s} V is io n: gl au co m a + c a ta ra c t a n d s e n il e + m a c u la r de ge ne ra ti on H er ni a + C on g. d is lo ca ti on h ip ? V ar ic os e v e in s ? P he ny lk et on ur ia ? V en er ea l di se as es : s yp hi li s + go no rr ho ea + M en ta l il ln es s + ! N .A . • n o t a pp li ca bl e. TP .B LE 5. 2. C: O l'D I'" I·. JN S SC ?c EE NE D BY A B IL I'r v TO M EI •;T SC :l. EE N IN G C R IT ER IA N ot ur al h is to ry o f pr e- ·· R eq :o gn iz ab le S ui ta bl e te st T es t· a c e e pt ab le c u r s o r ;; ta ge d .e li ne at ed la te n t s ta ge a v a il ab le to p ub li c . . ? '· •' + + + ? + + + ? + + + - + + + ? + + + ? + + + + + + + ? ? + + ? ? + + ? + · + + ? + + + + + + + - + - N .A .! ? + + + . , - + + + ? + + ? ? + + - + + + + + + + + ? + + + ? + + + ? ? + + - - + + - - + + - - + + . . - - + + + + + + - + + + - - + + - ' - N .A . A cc ep te d tr ea tm en t a v a il ab le + ? + + ? + + ? + + + + ? + + + + + + + + + + + + + \· + + + ? A pp ro ve d po li cy o n l at en t . s ta ge - - . . . + + - + + ' + + ? + - + + + + + + + ? - N .A . N .A . N .A . N .A . + + - 'N .A , - - - · . • I \1 1· . 00 I ' TA BL E 5. 9. 1. EX AN PL E.. S OF ' CU R." 'l..E NT S TU DI ES O N Tl li: E PI DE M IO LO GY O F CH RO NI C D IS EA SE S C on di ti on D ia be te s M el li tu s Is ch ae m ic H ea rt D is ea se L - - - - - - - - - - - - - - - - - - - - T - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - - 1 W or k in P ro gr es s 1 B os to n U .S .P ,H .S . St ud y o f M at er na l P re -D ia be te s. 2 B ed fo rd S tu dy o f B or de rl in e D ia be ti cs . 3 I C ol le ge o f G .P s. S tu dy o f B or de rl in e D ia be ti cs . 4 I B .n .A . R an do m iz ed t ri a l o f tr ea tm en t in P ot en ti al a n d s u b- cl in ic al d ia be ti cs . 1 2 3 U .S .P .H .S . Fr am in gh am S ur ve y o f I. H .D . in a de fi ne d po pu la ti on . L .S .H . + T. M . S tu di es o f I. H .D . qu es ti on na ir e. Te cu m se h St ud y o f e pi de m io lo gy o f I. H .D . in a de fi ne d po pu la ti on . 4 I W .H .O . C o- op er ac iv e S tu di es . 5 I N um er ou s s tu di es o f bl oo d li p id l ev el s in po pu la ti on s; c o m pa ri so ns b et w ee n po pu la ti on s. 6 I W or k o n a u to -a n al ys is o f E .C .G . w it h in cr ea se in d ia gn os ti c po w er . 7 I W or k o n e m o ti on al f ac to rs i n I. H .D . 3 I U .S . C o- op er at iv e D ie ta ry T ri al . 9 I M .R .C . T ri al o f P ol yu ns at ur at ed F at s in I .H .D . P at ie nt s. 10 I E di nb ur gh T ri al s o f S te ro id s in I .H .n . Co m m en ts C on ti nu e s e a rc h fo r o th er d is e~ se in di ce s. F ur th er t ri a ls o f ra pi d B .S . s c re e n in g te ch ni qu e n e e de d. W or k is d es ir ed o n r e li ab le s c re e n in g te st s fo r in di vi du al , r a th er t ha n gr ou p, d ia gn os is . M or e c o n tr o ll ed w o rk i s n e e de d o n pe rs on al it y a n d e m o ti on al f ac to rs in I .H .D . M or e w o rk w o u ld b e u s e fu l o n fa ct or s a c u te ly i nf lu en ci ng b lo od l ip id s a n d c lo tt in g t im e. 11 [ C hi ca go B oa rd o f H ea lt h St ud y o f D ie t in p re ve nt io n o f - - - _j__ ~·~~ D-. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + -- - ~ ' :'A BL E 5. 9. 2. - - - ~~,:~ i_';; n = = = t-+= --== : =:~ kin !_" '_" '_' "-- --- --- --- --- --~ --- E ss en ti al H yp er te ns io n l I M .R .C . St ud y o f pr op os it i a n d r e la ti v es i n S • . W al es A r a n do m iz ed t ri a l o f tr ea tm en t fo r Co m m en ts C hr on ic B ro nc hi ti s C an ce r o f Lu ng 1 e a rl y e s s e n ti al h yp er te ns io n ha s n o t I s o fa r be en c a r r ie d o u t. 2 Fr am in gh am S ur ve y 3 Te cu m se h P ro je ct 4 L .S .H . a n d I'. H • . St ud y o f Sp hy gm om an om et ry . l 4 5 l P ro sp ec ti ve s u rv e y o f c li n ic al w o rk er s a n d m en w o rk in g in t he e n gi ne er in g in du st ry ; F le tc he r, L on do n. P ro sp ec ti ve s u rv e y o f in fa nt s a n d yo un g c hi ld re n in Lo nd on - H ol la nd . P ro sp ec ti ve S ur ve y o f s c ho ol ch il dr en i n K en t - H ol la nd . St ud y in S te el w o rk er s, S . W al es - Lo w e. P ro sp ec ti ve S tu dy o f M ed ic al S tu de nt s a t S t. 3 ar th ol om ew 's H os pi ta l fo r e a rl y c o r r e la te s o f c r1 ro ni c br on ch it is - M .n .C . P hi la de lp hi a C he st C li ni c - B ou co t + W ei ss . 2 B al ti m or e St ud y - L il ie nf el d. 3 A lb an y St ud y - Ro m e + v o yl e. S pe ci al s tu dy i n pr og re ss o f s e n s it iv it y o f to ba cc o. T ri al o f a n ti b io ti cs r e a dy fo r pu bl ic at io n. S tu di es a re be in g s ta rt e d o f tr ia ls o f in te ns iv e a n ti -s m ok in g pr op ag an da i n hi gh -r is k gr ou ps i d en ti fi ed i n t he po pu la ti on . W or k is i n pr og re ss a im ed a t a gr ee m en t in te rn at io na ll y o n de fi ni ti on s w it h th e o bj ec t o f m a ki ng v it al s ta ti st ic s a n d re s e a rc h c o m pa ra bl e be tw ee n c o u n tr ie s, e .g . W .H .O . Sy m po si um R ep or t EU RO - 21 2. M or e s tu di es o f m o ti va ti on i n c ig ar et te sm o ki ng a re n e e de d. - - - · - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ J 0 \ 0 TA BL E 5. 9. 3. r - - - - - C o n d it io n _ _ _ _ - T -- -- -- -- ~a rk i n ;;o ~·e ·s- ~-- - - - - - - . - J . - - - - Co m m en ts f - - - - - - - - C an ce r o f U te ru s C an ce r o f B re as t - - - - - - 1 - - - 1 M em ph is, T en ne ss ee S ur ve y. 2 Sa n D ie go S ur ve y. 3 I S t. L ou is S ur ve y. 4 B ri ti sh C ol um bi a Su rv ey . 5 C ar di ff , s. W al es S ur ve y. 6 I A be rd ee n Su rv ey . 7 M an ch es te r - St ud y o f pu bl ic a n d pr of es si on al a tt it u de s. 8 I C op en ha ge n, B al ti m or e, S to ck ho lm , Lo nd on , M an ch es te r, B irm in gh am - s tu dy o f D av is C yt op ip et te . 9 I C ar di ff a n d Lo nd on - St ud y o f en zy m e te s t fo r c yt ol og ic al d ia gn os is . 10 P hi la de lp hi a St ud y o f c o m pu te r s c a n n in g fo r c yt od ia gn os is . 11 I R .C .O .G . St ud y o f pr og no si s in r e la ti on t o b io ps y di ag no si s a n d tr ea tm en t o f c a n c e r o f c e rv ix . 1 2 3 4 St ud y o f X -r ay M am m og rA ph y in p op ul at io n s c re e n in g in N ew Y or k S ta te . P ro sp ec ti ve S tu dy o f e n do cr in e s ta tu s in r e la ti o n t o de ve lo pm en t o f br ea st c a n c e r - Je rs ey - G uy 's H os pi ta l. St ud y a t A lb er t E in st ei n C en tr e in c li n ic al M am m og ra ph y - Eg an . St ud y a t M .D . A nd er so n H os pi ta l, H ou st on , T ex as i n c li n ic al m am m o gr ap hy - v a ri ou s o th er c li n ic al s tu di es . T he re i s ro o m fo r m o re in te rn at io na l s tu di es o f m o rb id it y a n d m o rt al it y in r e la ti o n t o t he u ti li za ti o n o f e x fo li at iv e c yt ol og ic al s e rv ic es . Th e pr ob le m s o f th e u s e o f e x fo li at iv e c yt ol og y by t he p ub li c c o u ld b e s tu di ed m o re in te ns iv el y. D ev el op m en t o f di ag no st ic s c a n n in g by c o m pu te r is b ei ng u n de rt ak en . To d et er m in e it s v a lu e, p op ul at io n s c re e n in g by X -r ay m am m o gr ap hy o u gh t to be d on e w it h ra n do m iz at io n, s o th at o n e gr ou p a re tr e a te d a t e a rl y di ag no si s, a n d o n e tr e a te d a t n o rm a l c li n ic al d ia gn os is , w it h hi st ol og ic al c o m pa ri so n a n d m o rb id it y- m or ta li ty s u rv e y. M at er ia l sh ou ld b e a v a il ab le f or c o -o pe ra ti ve s tu di es o f s u rv iv al f ro m e a rl y de te ct io n c li n ic s, e .g . U ni ve rs it y o f M in ne so ta C an ce r D et ec ti on C en tr e. I 0 \ . . . . . L C on di ti on B ac te ri ur ia s R he um at oi d D is ea se (in cl ud in g o th er a r th ri ti d es ) M en ta l Il ln es s 1 2 3 )t 5 1 2 -,_ / TA ..S LE C ,.C 1. 4. W or k ir ' P ro gr es s B os to n - K as s. Ja m ai ca - M ia ll a n d K as s. R ho nd da , S. W al es - C oc hr an e, M ia ll a n d K as s. Ed gw ar e - B ru m fi tt a n d M on d. C ha ri ng C ro ss H os pi ta l, D e W ar de ne r. B ri ti sh E m pi re r u 1e ur na tis m C ou nc il S ur ve y - D r L aw re nc e. Te cu m se h Su rv ey . U ni ve rs it y o f P io ts bu rg h- Si dn ey C ob b. 4 I U .S . N at io na l H ea lt h Su rv ey . 5 I O th er P re va le nc e S tu di es i n Sc an di na vi a, B ra zi l, N ew Z ea la nd . r_: .; 1 2 S tu di es o f rh eu .m at oL i fa ct or i n v a ri ou s po pu la ti on s - V al ke nb er g, L Ed de n. In st it u ce o f P sy ch ia tr y St ud y o f P re va le nc e o f R ep or te d M en ca l Il ln es s in G en er al P ra ct ic e. Sh ep he rd e t a l. (B .M .J. 1 96 4, 2 , 13 59 ) M .R .C . S oc ia l P sy ch ia tr y U ni t St ud y o f M en ta l Il ln es s in C ar nb er w el l. M .R .C . S oc ia l P sy ch ia tr y U ni t P re va le nc e o f ps yc hi at ri c sy m pt om s in r e la ti o n t o s o c ia l a tt it u d es i n S. W al es r u r a l po pu la ti on ( Ra wn sle y a n d c o ll ea gu es ). Co m m en ts D es pi te a ll w o rk d on e, th er e is s ti ll a gr ea t n e e d o f s tu di es t o d et er m in e th e im po rt an ce o f ba ct er iu ri as ( as ym pt om at ic ) in t he a e ri ol og y o f py el on ep hr it is b y da nd om iz at io n o f tr ea tm en t. P ro sp ec ti ve s u rv e ys t o d et er m in e pr og no st ic e ff ec t o f tr ea tm en t in l on g- te rm a re la ck in g. Th e po ss ib le r e la ti on sh ip o f v e s ic ou re te ra l r e fl ux to i nf ec ti on a n d o n s e t o f py el on ep hr it ic c ha ng es , is n o t s e tt le d . F ur th er s u rv e y w o rk i s to b e e x pe ct ed o n th e li n es s u gg es te d a t C .I .O .M .S . Sy m po siu m i n R om e, 19 61 , a n d by T ec hn ic al C on fe re nc e o n th e P ub li c H ea lt h A sp ec ts o f C hr on ic R he um at oi d A rt h ri ti s a n d R el at ed D is ea se s, h el d in R om e in 1 96 3. l'h er e is r o o m fo r s tu di es o f u n re po rt ed m e n ta l il ln es s in a c o m m u n ity w it h ra n do m iz at io n o f tr ea tm en t. R;' C on di ti on A na em ia (ir on d ef ic ie nc y) 1 2 3 4 C hr on ic G la uc om a 1 2 3 4 - - ~ - - - - - - - ~ ~ - ~ - - - TA BL E 5. 9. 5. W or k in P ro gr es s M .R .C . Pr ev al en ce S tu dy . M .R .C . T ri al o f H ae m og lo bi no m et ry i n G en er al P ra ct ic e. M .R .C . T ri al o f P re ve nt iv e T re at m en t in A do le sc en t G ir ls . M .R .C . T ri al o f P re ve nt io n in A du lt s by a dd it io n o f ir on t o b re ad . M .R .C . Su rv ey o f in tr a- oc ul ar t en si on a n d gl au co m a. B ed fo rd S ur ve y o f in tr a- oc ul ar t en si on a n d gl au co m a. M .R .C . St ud y o f pa ti en ts ' at ti tu de s~ t o M ed ic at io n. R an do m iz ed t ri a l o f tr ea tm en t o f o c u la r hy pe rt en si ve s is i n pr og re ss - M .R .C . Co m m en ts E xt en si on o f ~i el d su rv e ys i s v e ry m u ch d ep en de nt o n th e de ve lo pm en ts o f s u it ab le h ae m at ol og ic al t ec hn iq ue s fo r u se in t he f ie ld . T he re a re c o n tr ad ic ti ng v ie w s o n th e v a lu e o f a dd it io n o f ir on ( an d fo r in st an ce v it am in s) t o f oo d fo r pr ev en ti on o f ir on d ef ic ie nc y a n a e m ia . Fi nd in g a be tt er i nd ex t o e a rl y c hr on ic gl au co m a th an t on om et ry w o u ld fa ci li ta te c a s e -f in di ng a s c o n tr as te d w it h s u rv e ys o f in tr a- oc ul ar t en si on . T he re i s a la ck o f m a te ri al o n ra n do m iz ed t ri a l o f m e di ca l tr ea tm en t o f c hr on ic g la uc om a a n d o f o c u la r hy pe rt en si ve s. 0 \ V I - 64 - Much of the work referred to here has been mentioned in the text7 where references have been cited. For permission to cite the work, not referred to elsewhere in the text and not yet published~ we acknowledge our gratitude to the authors. 5.4 Periodic ""ealth examinations 5.4.1 Introduction It will not have escaped notice that some of the conditions mentioned under the heading of early detection are detectable not by means of specific screening tests but by physical examination. Rectal cancer and, to a lesser extent, uterine cancer fall into this category9 so does examination for hernia and cateract. Other examinations, like blood-pressure and electrocardiography7 may or may not form part of a physical examination. The idea of the periodic health examination goes back a long way. Dobell published a monograph advocating "periodical examinations" in 1861. In 1925 the American Medical Association published a manual for physicians. The Commission on Chronic Illness41 was in favour of "all persons having a careful health examination including selected laboratory tests at appropriate intervals". The Commission advocated screening tests only as a substitute to personal medical examination 7 recognizing that shortage of medical manpower made universal routine medical examinations impossible. However, the difference between these two types of examination is quantitative rather than qualitative. Under periodic health examination the responder attends the doctor who examines him, determines \!hat (if any) laboratory tests are needed, and arranges for a second interview end re-examination if necessary, i.e. there are two physician interviews and one set of tests. With screening the responder undergoes a set of tests 7 then sees a doctor (only if necessary, the preliminary sorting of abnormal from normal having been already carried out by the screening tests)3 the doctor may then order more laboratory tests after examining the responder ~d the doctor sees him once again. Under periodic health examination the doctor sees all responders and himself acts as a selective screening agent ordering what laboratory tests he deems necessary. Finally, he sees the selected responders yet again, with the results of the laboratory tests in front of him. This can be represented in model form (Figure 5.1.). The advantage of the periodic health examination is, of course, the introduction of the trained analysing brain of the doctor into each examination, which is most probably a more efficient selector mechanism than the selection for further examination Fig. 5-l HEALTH EXAMINATION D • Doctor L • Labora- tory HE+ • Positive Examination HE- • Negative Examination NUMERALS • Number ot physician Contacts SCREENING EXAMINATION S + • Positive ~crecning S - • Negative Screenin~ WHO 70038 by a battery of tests, rigidly (however ingeniously) for~.~chosen on general probability grounds. 'l:'he: c;l;i~advantagf? is equally clearly the impracticability of routine physical examina.:tions for all. There is a second disadvantage as well7 the blunting of clinical acumen that. may occur in certain circumstances of repetitive physical examination, particularly where the yield of abnormal results is expected to be low. School physical examination,.without selection, is an example. On the other hand, the yield of abnormalities reported from the examination of middle-aged male business executives is .· . . 4"' . .. high (e.g. ~anco S.C. et al., ~; see also Periodic Health Examinations- Abstracts ... . . . . 43 from the Literature).· The subject of periodic health examination has been dealt within some detail in the "Prevention of Chronic Illness" 44 and it is not intended to deal with the , matter in any way exhaustively in this paper. However, it is worth noting two differing aspects of this kind of examination in relation to medical care. Relationship to general practice The first aspect of interest is in relation to the development of general practice. Where general practice has become relatively highly oli'ganized, whether by concentration on health centres or into group practices, or by voluntary limitation of the number of patients cared for (facilit~ted by a high,per capita firi~ncia~ reward), there has been a tendency towards the elaboration of routine medical exam;ination. Apart from actual physical examination, there may be a full examination of the blood, X-ray of the chest, E.C.G., procto-sigmoidoscopy, as well as testing of the urine and examination of the faeces for occult blood. There is special sco.P~ in tpis kind of environment for the dev~lopment of relatively sophisticated tes~s. The various biochemical estimations possible with the Gray Wedge photometer are a case in point: the recent development of an electronic apparatus for recording the knee jerk (as an index of thyroid function) is another. While information is available on the number and type of abnormalities revealed by this kind of search for early disease we do not have comparisons with the results of finding and treating disease by more conventional methods. Perhaps all that need be said is that it is clearly better to diagnose developed disease at the earliest possible stage and that the only comparison ethically:feasible would be the proportion ()f patients under treatment, by condition, in otherwise comparable general practices. - 66 - The second aspect of interest is the commercial or industrial periodic health examination. In this context we are considering examinations .for the general health of the worker and not statutory examinations for industrial hazards such as silicosis, lead poisoning, industrial X-ray exposure or extensive noise. General medical examinations have, of course, been demanded for a very long time by bodies which set their own terms of entry, for example insurance companies and the armed forces and railways and air lines. In this case there is either a commercial interent in the medical findings, based on actuarial calculations, or an interest in the public or other safety, or both. Thus it is perfectly legitimate, and desirable, that a man seelcing tn become an air pilot should be refUsed on the grounds of equivocal abnormali- ties in his E.C.G., though these changes would not be considered adequate causes on which to base a prognosis in civilian life. Again, if changes develop and are found at routine examination of a trained pilot this may well be a reason for taking him off flying duties. In industry, too, it is necessary that specific medical examinations be made on persons at special risk. For example, the periodic routine cytological screening of the urine in persons who have in the past been exposed to beta-naphtylamine is obviously needed? and other industrial examples can be thought of. It is when, as often happens, industrial firms offer medical examination as a "fringe benefit 11 to their employees, that doubts must arise. The legal term cui bono is applicable in these instances and it is well to be clear that the employee who submits to the examination derives at least as much benefit from it as does the employee. That both sides can banefit from periodic staff medical examination is probably true; it enables persons with early disability sometimes to be found more suitable occupations but there can also be the risk that employees may incur a loss of earnings and responsibility because of a risk that never materializes. It is salutary to remember that the Olympic 100 metre record was once held by a man with an aortic regurgitation that led to his rejection for military service. In some countries, periodic health examinations are cnrried out on all employees every year. In other countries, such examinations are carried out only on selected groups, such as those exposed to occupational hazards, groups with special work demands, such as crane operators, truck drivers, foremen and executives; or groups with higher .morbidity, such as middle aged and old personnel (above 45) and young workers (below 18); or thJse with high, long-term and short-term sickness absences; or those who, according to their foremen, etc., deviate from normal behaviour, absence record or productivity. - 67 - While medical care arrangements v~ry greatly from co1:1:~try.to country there can be special advantages in industrial health examinations. The industrial pbysician is often well placed to follow the same individuals for a long time and thus can detect early deviations from health because he is able to compare the results of the physicial examina- tion and screening with the findings of pre-employment health examinations. Industrial health services offer certain particular advantages for studying early disease detection in that it is possible to ensure the regular attendance of a relatively large population and that continuity of attendance is usually of a high order. The interest and full support of both management and employees has been achieved in several countries through national or local agreements on industrial health services or through legislation. The ILO Recommendation on this subject (No. 1121 1959) must also be mentioned in this connexion. 5-4-4 Routine examinations through life We should not forget~ of course, that there is in most developed countries already a well-established pattern of routine medical exe~inations 1 extending through life. If these are wisely spaced the individual can probably be given the maximum protection with the minimum interference vri th the normal pattern of his life. Starting with ante and post~natal and infant welfare examinations the child is nextexa~ined at or soon after school entry and again before leevJ.ng s2hool. After school there are pre- end youth employroent examinations as vrell as examinations before entering certain specified careers. In industry it would seem reasonable to ask for medical examination at key points in a person's career as part of the selection for prom~tion, considering the high· degree of investment in the individual necessary for senior appointments. However, the results of these examinations need interpreting with latitude~ taking into full account all that has been said aboutborder-lineconditions and giving the individual the benefit of any doubt there may be in a particular case. Finally, as vJilliamson, 45 among others, has shown there is a fruitful harvest of remediable defects to be found in the elderly and aged by routine physical examination. Defects of the special senses and of locomotion constitute some of the greatest handicaps of old age and these tend, in the elderly, to fall short of a level of clinical urgency necessary to compel the patient to seek medical advice. There seems to be considerable scope for clinics for the elderly where special attention would be paid to their medical needs. Certainly some of these clinics appear to be filling a real need. - 68 - 5. 5 Tne pla_c:;c' _ _?_f__:o;creenil'}0 j_n ti1e :')Povision of medical care 5. 5.1 ~ust_:1f_i.c_a.t),.:.C'!t witi-.in a country At the start of this paper we gave some space to a consideration of the aims of the early detection of illness. These aims need considering along i-Jith the policy aims which are central to any system for providing medical care. With greater prosperity more money will become available for personal health services. This will mean that greater efforts can be made to extend the disease-free period of life by all possible means and this will include the detection and correction of early departures from normal health: while poorer countries must needs manage with the minimal medical service necessary to prevent the major epidemic disasters and to maintain the working population in as fit a state as possible. The specific aims of early disease detection are firstly, the control of communicable diseases like pulmonary tuberculosis or bilharziasis; and secondly, the promotion of better health and less human suffering as the result of disease. Both aims are governed by the economic nexus. In the first case, it may well be in the national interest to control a disease or group of diseases and a national campaign may be mounted. In the second case, the advance to better health may be marginal only and may be left to local effort where that is forthcoming. Ideally, and in areas where the economic aspect is all-important, the cost to the community of early detection and treatment of e. condition should be less than the cost of treatment at a later stage. For example, it has been calculated that the cost of diagnosing and treating one patient with pre-symptomatic pulmonary tuberculosis costs about~ 400 a sum which, in relation to the cost of treating a developed case 5 represents a worthwhile saving even without taking into account the probable saving of a human family from anxiety and suffering. On the other hand, serological screening of certain communities for venereal disease, which used to be considered economically worthwhile, may no longer be so considered. Similarly, the returns on screening patients for lung cancer are so poor (vide 6.6.1.) costly as they are, that this procedure is probably neither acceptable on economic nor medical grounds. In working out cost a large number of factors may need to be taken into account. One important factor that comes into the calculation is the possible saving of the time of highly-trained people, representing the cost in time and money of their education and experience. If a screening procedure can, by automation or by employing e. less highly trained person, act as a substitute for some of the time of a highly-skilled person, there will then be a saving. However, it must be remembered that the overall cost to the - 69 - community is ustially greater than before because something is being done 1.,rhich was not being done previously. For instance, the detection of unascertained disability among the elderly and its treatment and rehabilitation undoubtedly costs the community extra even though the ascertainment may be carried out by questionary or ancillary workers rather than doctors. An example in which the overall cost may be lowered can be found where the treatment of a condition is in any case essential. Thus the detection and treatment of carcinoma-in-situ of the cervix uteri should cost considerably less than does the diagnosis and treatment of established invasive ce.ncer. While the diagnosis of carc~noma-in-situ may cost more than the diagnosis of the invasive lesion the cost of treating carcinoma-in-situ is nmch less. But again it should be remembered that there is a period when the overall cost is increased~ from the admission to hospital both of the newly.-detected carcinoma-in-situ patients together with the normal quota of invasive cancer patients during the time when it is still too early to see a fall in the.incidence rate of invasive lesions. Cost of multiple screening The actual cost of multiple screening programmes naturally varies with the examina- tions made. A typical· one, that of the San Francisco longshoreman, is given by Breslow 6 as $5.04 per person screened. In this case, twelve tests were ca-rried out (as cited above). He also quotes two other schemes in which chest X-ray and blood sugar only were done and that the cost lay between $1.50 and $1.75 per person. Other costs are given in the A.N.A. 11 Study of Multiple Screeningn, already referred to. What is difficult, however, is to relate costs like this to what benefits would have derived from spending the same sums in other forms of medical care. It is, perhaps, of interest to consider these services in relation to what :i.s spent generally on medical care. In USA dollar equivalent terms 9 an uverage British general practitioner earns from all sources about $8 000 per year and he looks after, on an average, 2 350 pa tiEmts. This allows between $3.00 and $3.50 only per head for all general medical services. The cost of the whole National Health Service for England and Wales in 1962-63 was £605 million, which allows an expenditure of £13 (or $36) per head of the population over hospital, public health and general medical. California, Union had a total health budget (Federal and State) of over the year for all services, the wealthiest State in the $43 million in 1960-61; 46 this allows expenditure on health and welfare services of about $3 per head in the year, (of which 36 per cent was spent on hospital construction and seven per cent on preventive medical services, about $0.20 per head). Thus in - 70-. relation to total health expenditure multiple screening is costly and its use would need to be judged on the benefits to. health 1.-1hich it could provide. (It should, of course, be remembered that this health budget is for public health only (apart from the contribution hospital construction) and personal medical services are supplied under private contract.) 5.5.3. Inter-country differences Besides justifying itself on "within-country11 economic grounds screening procedures need to fit into existing systems of medical care and the variation of these between countries may influence the pattern of screening. Unlike the usual sequence in clinical medicine when the patient, feeling unwell, consults a doctor and thus submits to diag- nostic examination, early detection entails an appeal to the public to come and be examined. An appeal of this kind is most easily organized through the public health services, and not through the clinical services, (though it may well be to clinicians that the public go for their actual tests as, for example, cervical smears). Where there is no unity of the medical services and where they may be financed in different ways the efficient practice of screening becomes difficult and there is the twinfold danger of a failure in communication and a failure to follow-up, dangers which have been discussed earlier in this paper. On the other hand, where there is a unified medical service it is relatively easy to carry out the whole screening operation, beginning with an appeal to the public, through screening tests, definitive diagnosis, treatment and follow-up, without breakdown in communications at any point. Figure 5.2 represents the two dimensions of economic development and unification of the services for medical care and the straight line passing through the origin indicates an average axis of development of the two variables. Near the point of origin of the curve is the area of poor countries with little in the way of organized medical services. In this area is indicated a medical aid team which might visit the country at its government's request for some special screening operation, (for example, the control of bilharziasis). The other end of the curve is the area of wealthy countries with highly-organized systems of medical care where integrated screening operations carried outunder national arrange- ments may.be expected. At the opposite extremes are countries HXn and "Y". nxn is wealthy but has non-integrated medical services and expensive sporadic screening exercises may take place here, with poor communication and follow-up. Country "Y" on the other hand, is poor economically but has a unified system of medical ce.re. It would comm£.nd ·ideal conditions for the control of the kind of disease which is a problem still in the less prosperous parts of the world 5 for. instance pulmonary tuberculosis. Unfortunately, Fig. 5-2 RELATIONSHIP OF MEDICAL CARE SYSTEM TO ECONOMIC DEVELOPMENT Econ~ic Develop~ent 3 2 1 1 2 3 WHO 70039 Integration of Medical Care Services - '71 - there are·few economically underdeveloped countries which heve achieved integrated m8dical services since this is in itself costly. 5.5.4 Example of gynaecological cytology Soms lessons on the medical care problems and economics of screening can perhaps be learned from experience with initiating service gynaecological cytology in England and Wales, .where. there is a tripartite, decentralized, national health service with semi-independent general practitioner~ public health and hospital services. Here~ the responsibility for the examination is regarded as primarily belonging to the general practitioner~ as he who practices personal preventive medicine. Either he, a public health authority doctor or a doctor from a voluntary family planning organization may, in tt'e case of nwell" women, actually take the cytological material and it is generally regarded as best that this examination should be carried out by a doctor. This poses one problem of screening~ ideally examinations should be carried out by a doctor since women have considerable confidence in the examination and are likely to believe that a negative test indicates that they are not only free from cancer but unlikely to develop it. Whilst a trained assistant~ of nurse or similar standing 9 can easily learn to take cytological material she is not so well trained in observing local pathological changes. Yet 9 if doctors only were to carry out this examination~ not only would a heavy economic burden need to be borne but the service would be likely to· remain limited due to a shortage of medical manpower. If screening (taking material) is to be carried out for uterine cancer by non-medically qualified people, therefore, as to some extent it almost certainly has to be, it is ex-tremely important that the public should understand that this form of medical care is provisional and not definitive. A problem, in a way similar to the above, has to be faced in the examination of cytological specimens. It is of the first importance that those responsible for deciding whether specimens are negative or positive, both pathologists m1d technicians, shall be well trained. Inadequate training can easily bring a whole service into disrepute;, wrong decisions on smears and biopsy material will have far-reaching effects since the definitive treatment for intra-epithelial (carcinoma-in-situ) cervical lesions is normally total hysterectomy. However~ insistence on training in cytology being limited to laboratory technicians, while admittedly ensui'ing a high standard of technical service, is liable to clog the rate of development of a service. The USA and some British Commonwealth countries (among others) have succ·essfully adopted the procedure of ad hoc training in cyto-technology for persons of US university graduate or similar education. This is important because, with all advanced countries, competition for man and women- - 72 - power from all sides has become extremely keen and an over-insistance on formal qualifica- tions prior to training in cyto-technology could be another factor slowing up the development of a service. The number of technicians needed for carrying out not only primary examinations, but also regularly spaced re-examinations, is considerable; as a rough calculation will show~ there e.re about 50 million persons in England and Wales with~ approximately 15 million a.dul t women. Supposing cytologica.l examination were ce.rried out at an average of three-yearly intervals by techiiicians who could examine 10 000 women/specimens per year, a total of 500 technicians would be needed. The problem of the future may however be that of persuading those women at greatest risk to attend for examination (and,. incidentally, perhaps to give advice on health measures likely to diminish the chances of developing cervical cancer); and it·is likely that a rather large proportion of the population may fail to submit to regular examina- tion. Even with smaller numbers the recording and efficient recall, with follow-up if necessary, of all these women presents formidable problems. A possible means of recall that is being explored is the registration centrally of all cytological examinations followed by automatic data processing and the issue of recall appointments. Such a central registration system clearly also has great epidemiological potentialities. By reason of the man-power shortage the search for automated techniques for cytological screening is clearly of great interest and would be acceptable even with a high false positive rate. Provided that the cost per examination could be kept low, automation could be combined with a high degree of centralization of screening services 9 with a consequent saving in overhead costs. However, it would need to be remembered that the lengthened lines of co~~unication would be more vulnerable and it seems likely that on-the-spot re-examination of reported positive specimens would still be necessary. It still remains to be seen whether cytological screening will be accompanied by a significant reduction in deaths from uterine cancer 1 but it appears that already the incidence of invasive cervical cancer can be lowered by this technique. INhere this is economically possible it is no longer therefore a question of whether one form of treat- ment is economically more desirable than another since it can hardly be argued that women might wait until they have developed early invasive cancer, rather than have it diagnosed and treated in the pre-invasive stage. It is, however, of some interest to examine briefly the consequences in hospital bed usage and the following simple calcula- tion may give some idea of the effects of widespread population screening for uterine cancer: - 73 - ; At present there are in England and We.les about 10 500 discharges from hospital annually for carcinoma of the cervix and the average duration of stay is 21 days~ giving . 47 a total of 220 000 hospital days. These discharges refer to admission and not persons. The approximate total time spent in hospital as an in-patient by a woman with carcinoma of the cervix during her lifetime is in the region of 50 days. The average length of admission for cone diopsy for a patient with carcinoma-in-situ is in the neighbourhood of seven days or roughly one seventh of 50 days. On the other hand, perhaps twice the number of women that would ultimately have developed invasive cancer, would be subjected to cone biopsy. Supposing cervical screening to be nearly 100 per cent effective, this would still cut down total hospital bed usage by a factor of three or four. However, admission for biopsy during the early years of a screening prograinrne would occur simultaneously >vi th the same number of admissions for invasive cancer as previously experienced. From the Hospital In-Patient Enquiry reports of England and Wales it cen b8 estimated that in a population of 250 000 there wo11ld be !oughly 35 admissions of cancer of the cervix per year. If one-fifth of the female population over 20 years of age were screened annually, on the l'1eophis, Tennessee rates 50-60 carcinomas-in-situ would be discovered, thus nearly trebling the existing admission rate. At 10 days in hospital for 'each case this would need, on average, e.n extra two beds for a population of 250 ooo, over and above the two to three that would be needed in any case. 5.5.5 Conclusion Clearly some way is needed for measuring the value to a community of the early detection and treatment of both pre-symptomatic and declared disease. The 11value" may be either of the 11cost-benefit'1 kind (e.g. productive of better health or comfort), or of the "cost-effective" variety (i.e., of economic value to the community). Obviously, in terms of health 5 the two sorts of value cannot be fully separated - the man or woman who enjoys better physical or mental health can work better, though it may be impossible to measure the difference in output between loss and more healthy people. The important point, probably in determining whether a particular measure for early disease detection is, or is not, worthwhile, is the degree to which the positive results from screening has been established. For example, it seems doubtful whether it is of value to screen a whole community for high blood·pressure, other than what is already done by the clinical services, in the absence of good criteria and a proved and acceptable form of early treatment. On the other hand, as we have seen 5 it does seem possible that .the early·detection and treatment of cancer of the cervix may lead to its virtual - 74 - eradication5 and this, in terms of benefit to the individual woman, is well worthwhile. However, the economic cost of a campaign to eradicate cervical cancer must, if it is to succeed fully, inevitably be relatively high. The cost must therefore be considered by any community in terms of its own particular economy. If5 as is usual, there is a limit to the health budget, the benefit and effectiveness of the proposed programme needs to be compared with other desirable objectives. For example, an intensive campaign to prevent young people starting to smoke, or a nation-wide attempt to prevent overweight? could improve the life and health of a far lo.rger number of persons in the active span of life than the cervical cancer programme. In the event communities often decide that every feasible form of early disease detection end prevention must be offered to the population with the result that the overall cost cannot satisfactorily be met, and a number of programmes is carried out in a less than effective way, instead of a few that are well-planned and well-executed. 5.6 The place of education in the ea:clv detection of diseqscs 5.6.1 .Lch'_<?2-tion of the medical profession Undergraduate medical education is traditionally disease-orientated. Education in public health has in the past been largely concerned with environmental health and the primary prevention of communicable disease. The teaching of hygiene and public health in Europe has been reviewed in a WHO monograph by Grundy and Meckintosh. 48 Tne dominant role in teaching medical students is still to too large an extent filled by the clinician whose treining, in turn, has not concerned him to any great extent with preventive medicine. At present there remains a tendency for a self-perpetuating system of teaching which allows too little emphasis on new thinking in preventive methods. In recent years there havo been signs of a new epproach, by teaching both the preventive and clinical aspects of medicine at the same time. The preventive side of medical practice in thE:. community needs to be learned both by the undergraduate student and in the pre-registration and post-graduation periods. Thus a ltJHO Expert Committee49 considered that the preventive aspects of medicine should be -taugnt at all stages in the curriculum, including the basic medical sciences and pathology courses. Interest in the earlier detection of disease must largely stem from an understanding of the epidemiological approach. As the WHO Expert Committee report points out, a number of new and special subjects are needed in training for preventive medical services, of which the epidemiological method takes first place, and which should include medical statistics, the social sciences, genetics and the organization of health and welfare services. These subjects are best taught by participation (involving some degree of personal - 75 - respohsibfli ty) in laboratory work, field surveys, out-patient .md peripheral services, domiciliary care and rural health programmes. For fostering the idea of early disease detectidn it is important to introduce the student to e type of general medical practice organized in such a way as to make this work feasible, e.g. well-organized group practice or health centre. Similar views have been set out in a WHO study "The Teaching of the Medical Student for Comprehensive Medical Practice".so This study particulerly E;lmp~asizes the.value of the health centre for this type of teaching. It is, in fact, now realized that it must be for the doctor practising in the community to undertake the first-line prevention of disease, whether at the primary or secondary stage. He is best placed to see the beginnings of illness and to be aware of local genetic and environmental factors. University faculties of general practice are now springing up and their role in relation to early disease detection can fruitfully be exploited in a.ssociation with their fellow departments of epidemiology, social medicine and medical statistics. The place of the general practitioner in the prevention of disease was recently considered by the College of General Practitioners.51 In addition to doctors it is important to remember that there are other workers in the practice of medicine who also need educating towards more positive attitudes to the early detection and treatment of illness, for example, nurses, health visitors, chiropodists and pharmacists. 5.6.2 Education of the public In order to successfully carry out the early detection of disease it is not only essential to have an alert and prevent'ion-orientated medical profession; the public needs also to co-operate in the operation. We know from experience that, for example, in the reporting of symptomatic cancers, the public tends to defer seeing a doct::>r until a condition is relctively far advanced. Knowledge about the value of early diagnosis and treatment needs to be spread to the community (and this, in its turn, needs. to be based on sound scientific studies). Thus it is accepted that education of the public is essential7 what is not always known is the form which this health education should take, both in its contentand in its application.· Early disease detection as a means of prevention in the community needs, in order to succeed, to reach all, or nearly all, those at risk regardless of cultural and economic differences between the various parts of the population. A common experience is that persons with the best general education make use of facilities of which they may perhaps, in reality, be in little need? while others, at a much higher risk of disease but with a less good education, fa.il to utilize a service out of ignorance of their specie.l need (the example of high-risk groups in - 76 - cancer of the cervix is briefly referred to in sections 5.5.4 and 6.6.2 ). Planning health education to reach all kinds of people is therefore essential and~ in order to do this? special studies of public attitudes to disease may be needed in the first place. Studies of ~his kind? have, for example? been carried out for gynaecological cytology in California, 52 nationally in the USA; 53 and one is in progress in Manchester.54 No doubt much still needs.to be learned about the best ways of reaching populations in providing health education. One point of importance is the need for proper organiza- tion. "Delivering the message 11 ? whether in print, by direct speech, radio or television, is a specialized technique and if carried out in an amateurish way 5 can lose most of its potential impact. There is a place in public health administration for specialist health education officers to assist the medical and nursing staff. In addition to local officers the central planning and teaching of health education me.y be valuable? as suggested in Great Britain by the Committee on Health Education. 55 As well as staff employed by public health authorities spreading education on health, other specially-trained persons have an important part to play in educating the public. General practitioners are well placed to influence their patients }n particular high- risk groups to attend for examination) and nurses and health visitors, for example, can influence the public beneficially during the course of their work. 6. ILLUSTRATIVE EXAMPLES OF SCREENING DISEASES The previous sections have dealt almost entirely with the geJ:ler?-1 conE3ide_rations of early disease detection. It now seems useful to consider some specific conditions in rather more detail. In the following section we have therefore chosen certain examples to illustrate specific points made in the general soctions. 6.1 Diabetes mellitus 6.1.1 General Diabetes detection, according to Joslin and others 56 goes back to 1909 when Barringer reported the findings on over 70 000 persons examined for life insurance purposes. However, the concept of population screening for diabetes dates from the survey carried out by Wilkerson and Krall in 194757 on the people of Oxford, Massachusetts. Post-prandial blood sugar estimation together with glycosuria testing was carried out. Since that time an increasing amount of case-finding, both by urine and blood examination, or both together? has been practised in the advanced countries of the world. For typical examples see 1.Jalker and Kerridg8.5B Despite all this work it is still difficult to evaluate the results in terms of benefit to the populations - 77 - screened. Some of the criteria for case-finding discussed above remain unsatisfied. This question will be dealt with later in this section. There is little question that the best documented case-finding has been carried out during the past decade in the USA. Under the Diabetes and Arthritis PrograJn of the Chronic Diseases Division of the US PHS information has been gathered together and published. Thus NcDonald and his co-workers 59 give figures for 1958-1963; during 1963 nearly 530 000 persons were screened by blood sugar estimation, as compared with 190 000 in 1959, an incree.se from one per 1 000 total population to three per 1 000. Of 338 500 screened in 1962, 10 300 tests were positive~ 9 000 (2.7 per cent) of these were referred to their physician, and of these 2 500 (0.7 per cent) wore known to be diagnosed as diabetic. The US National Health Survey finds a rate of nine per 1 000 population of all ages for known cases of diabetes, while thG estimate of unsuspected 60 diabetes is at a rate of e,ight per thousand for all ages. This would yield a total in the USA of 1 500 000 known and 1 400 000 unknown diabetics, so that screening, even in a country where a relatively large amount h&s b0en done, has still a great way to go. In general, blood sugar examination has superseded urine testing, rapid screening being pos:;lible using either the 11Clini tron" apparatus or the simpler slowe.r but much cheaper Glover-Edwards kit, both employing the Wilkerson-Heftmann screening test. 61 Remein and Wilkerson 15 show that, on a. study of 580 persons of varying age and race, the following specificity and sensitivity was obtained for blood sugar and urine screen~ng respectively: :Bi~'Oct- ·(s~~~-gyi-Nel son) 160 mg/100 ml, 1 hour after meal ~- - ... ~~ . . . ~ . Urine (Dreypack, glucose oxidase) 2 + TABIE 6.1.1 test Sensitivity %positive .. 52.9 45.9 Specificity % negative 99.4 90.3 I It is c::lear. that_:ne;ither :ti:Jchnic::ue is .. good .a.t.picking. up a .. high.pr~portion of persons with diabetes though the blood sugar method has the advantage of finding very few false positives. As a practical example of tho application of the above blood sugar sensitivity and specificity, in a. population of 10 000 with a true prevalence of 150 diabetics, there - 78 - would be 70 false negative results (i.e. 70 missed cases) and 60 false positive tests. The practice of case-finding is discussed by the C.C.I. in "Chronic Illness in the United States 11 , Vol. IV, under methodology (Appendix D). The Baltimore screening survey 62 may be of interest~ as an example. The urine was first tested 30 to 50 minutes after 50 g of glucose given as a drink. The vt:nous blood sugar was screened at 160 mg per 100 ml using the Wilkerson-Heftman technique? 45 to 75 minutes after glucose. If positive, a two hour specimen of blood was screened by the same technique at 130 mg per 100 mi blood. Out of 1 916 persons of all ages 33 were found with positive one and two hour values (or not done at two hours)? 14 males and 19 females preponderantly in the 45-64 age-group. In nine persons only was follow-up completed) five were confirmed DS diabetics, of whom four were previously undetected. The rate in which both tests were abnormal (33 in all) was eight per thousand persons examined. This is fairly typicel of the results of case-finding. The cost of this type of operation is estimated by the US PHS as being in the neighbourhood of $0.84 per person screened~ providing this is carried out by an adequately staffed loca.l health department and not including overheads and publicity. 61 At a pick-up rate of eight per thousand population the cost per diabetic discovered would be about $105. A difficulty in comparing case-finding progre~es has been that similar criteria for diagnosing diabetes have not been used from place to place. Tho US PHS "Diabetes Program Guide" lists nine different s~ts of criteria. The Committee to the Public Health Servi·ce. Diagnostic Study recommends the following criteria? with a point value allotted to each reading:- TABlE 6.1.2 Time Minimum Point positive B.S. value of positive Fasting 110 1 1 hour 170 1/2 ... 2 hours 120 1/2 3 hours 110 1 A total of two or more points is considered to indicate a diagnosis of diabetes. - 79 - Despite these attempts at defining diagnostic criteria it is still very difficult to know in practice which are being used in a particular case-finding programme. Where separate forms of medical care co-exist side by side~ follow-up of screened positive persons can be fruitless and it is not very uncommon for this to consist only of a urine exrunina tion following on a positive blood test. Reid's report to WHO 63 makes a number of cogent points: (a) Differences of opinion about the value of generalised~ versus selective~ screening. (b) In the case of selective screening~ the criteria which should be used for selection. (c) The relative merits of urine versus blood screening, although opinion seems to have hardened in favour of the latter. (d) Varying conditions under which screening is carried out. Thus it is sometimes undertaken on the basis of random blood samples) sometimes on tests carried out at special times in relation to meals; sometimes with prior carbohydrate loading; and sometimes with a substantial period of dietary preparation. (e) There is a. wide divergence of opinion about the interpretation of screening levelsj and there are, of course, varying standards according to whether a true glucose or other type of blood sugar estimation is carried out. (f) Follow-up arrangements for th_ose who screen positive e,re variable and many a.re not adequately investigated. (g) Even if full tolerance tests are arranged, there are again variations in such metters as dietary preparation, carbohydrate loading~ and the interpretation of results. (h) There is further confusion (and this applies not only to the USA) because many of the medical publications reporting diabetic screening campaigns do not give enough information to purmit judgements to be formed~ e.g. no details may be given of the preparation of the patient for testing~ of the type of blood used; or of the variety of quantitative estimation employed. (i) Pre.ctically all the surveys which have been carried out point to· the .. conclusion that diabetes mellitus is difficult to define in terms of blood chemistry as any series of tolerance tests will show, at one end of the scale, - 80 - those who are clearly not diabetic, at the other, those who are definitely diabetic 7 but in the middle, a substantic..l group who fall into neither category. In many ways the position resembles that which applies to the diagnosis of hypertension, and greater knowledge is required of the natural history of diabetes bGfore anything of a more definite nature can be said about the intermediate group. (j) This last point is, in turn 1 related to the question of the preventability of diabetic cardiovascular complications, which is again a subject which has not yet been finally answered and which is 5 indeed, unlikely to be decided for another ten or twenty years? by which time various longitudinal studies will have been completed. Before advocating e. general policy for diabetes screening, therefore, more needs to be learned about the disease and the results of treatment. This will be discussed in the next part of this section. However, the detection of overt diabetics can well be encouraged, and high blood levels for screening could be agreed as an interim measure, so as to exclude"border-line 11 cases. Selective screening of the high-risk groups could be more economical since the yield is greater, but a real difficulty, and a potentially expensive one 9 is to make; contact with these groups. What is also needed is the better education of the public and the medical and nursing profession. As Reid 64 has pointed out, enquiry shows that public knowledge of the symptoms of diabetes is extremely vague:, and a proportion of diabetics diagnosed in surveys have already had symptoms for some time, according to Redhead 7 65 Wilkerson and Kral1, 57 College of GGneral Practitioners. 66 In the USA emphasis is now being given to educating the medical profession itself in the diagnosis and control of diabetes. Ancillary workers are trained as "public health representatives" to organize local detection work. They are based on local health departments and visit local doctors or other professional person- nel. 63 L1 some US projects self selectior:. plays a part a.nd c:.s much as an eight per cent yield of dia.bctics may be obtained in this way. Another high-risk group is that of patients attending hospital. The Dundee workers 67 have shown a surprisingly high proportion of undetected diabetics in hospital wards and there is scope for increasing the awareness of doctors and nurses to the sad fact that admission to hospital is in itself no guarantee that the asymptomatic diabetic will be diagnosed. The evidence of the value of screening In screening populations for diabetes it is important to QSk whether the criteria discussed under "Principles" have been fulfilled. Two of these principles merit further discussion: - 81 - (1) the question., whether it can be accepted that ea~ly treatment is valuable, and (ii) the question of the criteriafor diagnosis. 6.1.2 Value of early treatment Early treatment cau be considered at two stages; firstly its effect on early clinical diabetes and secondly, its effect at a later stage of the condition. It is not possible to obtain irrefutable evidence for the value of early (or late) treatment of developed clinical diabetes because of the ethical veto on withholding treatment from a control group. However, clinical evidence has accumulated that care- fully-controlled diabetics experience less complications than badly-controlled patients. Rundles68 has noted that neuropathy is particularly prone to develop in poorly- controlled patients. Garlanc169 states that amyotrophy is totally reversible with full diabetic control. In Ashton's view70 11 good control" of diabetes over many years 71 72 may have some beneficial. effect on retinal microaneurysms. Both Dunlop . and Marble have reported that diabe:t.~cs under good control are much less likely to develop nephropathy: than poor;Ly-controlled patients; while Johnsson73 found significantly less. nephropathy and severe retinopathy in a series of patients treated by strict dieting., with .an att,E;lmpt to keep the urine sugar-free., compared with. a series allowed a morel ··-' : . liberal diet, i:o. whicn only control of polylJ!ia and ketonuria was attempted. 11>/olff Salt 74 found, in acute e:Jeperiments, that hyperlipaemia returns to normal in children when diabetes is controlled, thus suggesting a possible causal relationship between diabetes and early arteriosclerotic changes; and Keen75 has -reported that cataract is more common in poorly-controlled diabetics and >'lhen arteriosclerosis is als<> present. Newburgh and Gonn76 have pointed out, among others, that weight reduction in the obese middle-aged diabetic may result in the glucose tolerance test returning to normal. The relationship between diabetes and obasity has recently been briefly reviewed.77 Finally; as a general finding, patients still living free of complications after a 25 year history of diabetes have been found by Joslin 1 s clinic to be those vrho had controlled· their diabetes meticulously.78 The subjeet has been ably reviewed by Beckett.,79 who urges the need to reduce delay· in the diagnosis of diabetes. There is, therefore., a considerable body of evidence in favour of the benefits of treatments .in minimizing diabetic. complications. However, it must also be allowed .that there is also some evidence to the contrary., particularly on the progress of retinopathy 8o 81 82 and renal changes. " . " When treatment at the earlier, pre-clinical stage is considered the matter becomes in on'1 sense more complicated because of the diffj.culty in deciding where to draw the arbitro.ry line between 11 health11 and 11 disease11 ; though on the other hand, investigation is e:ade simple by the real uncertainty which exists about the value· o:f'"'"fre·atment, thus and t H t 83 . . ethical~y allo;-ring randomization into treated con rol groups. oe nas snown that "pre-d:lt.betic" women put on insulin during pregnancy have significantly more live- birt~.s follo\'rtng t::..~ee.tment than before; and Wilkerson 84 finds insulin in 11 pre-diabetic" wo:::,:m p::>e·vents overweight in the infant, with a probable reduction in perinatal mortality: Reid63, 64 reports on a visit to the US PHS Diabetes Field Research F?,cjlity in Boston, where this work is being carried out, that more members of the control group of 11 pre-diabetica women (not treated with insulin) have developed clinical di.:1botes than have women in the treated group, though the numbers are not statistically 11 Prc·-diabetes" in women is designated by Vvilkerson as a group which includes "a high inc~.dc:;:we of big babies, stillbirths, neonatal deaths, spontaneous abortions, premature del:.:.veries, toxaemia and congenital abnormalities", as well as showing "transient 85 ~0no~rr.alities of carbohydrate metabolism during pregnancy". Fajans and Conn regard the 11 potential" diabetic as a person with a normal oral glucose tolerance curve but shcwl.ng .::. positive response to the cortisone-glucose tolerance test. They advocate· apply5 .. ng th:t.s test to those with an increased risk of diabetes, i.e. the women in the cc::cegories just cited as well as the relatives of diabetics. 6 , "· • ..J...,..,. Tnis question ~.s connected with that of the value of early treatment in that criter:l::> adopted det8rmine the s·Gage of blood sugar abnormality at which the diagnosis of diabetes j.s made and thus the stage at which treatment might be started. A.n irr:vm.~tant factor in considerj_ng the nosology of diabetes mellitus is the part heredity plays. The familial aggregation of diabetics has been attritbuted to the ir.he:ritnnce of a Mendelian recessive gene86, 87 For9. and Glen88 found the prevalence of unrecogn:i.sed diabetes was five times as high among relatives of known diabetics as in . th~ general population. Other surveys have confirmed these findings but it is not: of cslwse~ necessar,y to invoke a single gene to account for this. HarrisS9,90 considers thsrG ls insufficient evidence to determine the type of transmission. Recent blood 21~, 91 sugar surveys have demonstrated a continuous distribution of this index with no visible tendency to bimodality, though this could be present but concealed. On balance, present evidence probably favours a multifactorial inheritance. The question has been - 83 - 92 revievved recently in the British f'iledical Journal. 1'lh8ther single gene or multi- factoriably transmitted, the index blood sugar fails to segregate a discrete group of 11 diabetics11 from "non-diabetics" and therefore the problem of the "border-line" case of Fig. 4.5 (lower drRwing) occurs. This is the problem that lies at the root of discussion of diagnostic crj.tc::-ia and of "pre-diabetes". The best diagnostic criteria are based on the experience of those treating diabetic patients and one such group has advised the US PHS on the set of criteria already cited.61 However, it should be noted that these criteria are not ~;he same as those a¢1.opted by the American Diabetes Association and there ~s no goneral agreement as yet. International agreement on a set of criteria could be a v<:?.J.uable contribution to the comparability of survey work. Recently, a 1tlli0 Expert Commit~ee ras reviewed current knowledge about diabetes mellitus and has - ~ recommended work:illg criteria for single blood sugar levels after glucose loading./ Whilst it. must .be agreed that a diagnosis .. based on a glucose tolerance curve is arbitrary, dividin::.~ "diabetics" from 11 non-diabetics 11 by means of a single blood sugar (or~ even more 8o, urine test) examination is even more arbitrary arrl subject to doubt. Apart from patients actually found with the symptoms of diabetes$ there will be at one end of the scale those with such high blood sugar levels that there can be little if any doubt of the diagnosis; whilst at the other end there will be those persons with lew blood sugar levels, in whom there can be no suspicion of di.abetes. The intermediate group, as studies Jn Birmingnam66 and Bedforo24 and by the US National Health Survey, 91 have Fhown, comprise a formidable proportion. In the US National Health Survey, for example, 15.5 per cent. of persons had a blood sugar level of 160 mg per cent. or more one hour after 50 g of glucose by mouth { se.e. Table. E. 1.3), a level which Remein 15 and Wilkerson showed to be 99.4 per cent. specific for diabetes by the accepted criteria of their advisory panel. This proportion increased greatly with increasing age as the table shows~ thus, nearly half of the women over 65 could be classified as diabetic on this criterion. Similar findings have been reported from Birmingham and Bedford. The US National Health Survey report suggests that current standards for a normal blood glucose level are unrealistically low. As the report says, it is outside the scope of the survey to answer this question but within the scope to raise it. A follow-up of -~-he border-line group found in the Bedford survey is in progress, with randomization of treatment, so that in due course information should become available on the criteria to be adopted for diabetics in need of treatment. Until that time comes it might be legitimate to screen at a higher blood sugar level of, say, 18o-200 mg>; at two hours after 50 g glucose, in order to detect only unequivocal diabetics. There are hints, however, that relati.vely mild hyperglycaemia is associated with cardiovascular - 84 - changes; 94,95 though it remains to be seen whether measures aimed at lowering the blood sugar level are able to prevent or delay these changes. Blood suga.r screening could be greatly facilitated in practice if the newly available glucose oxidase strip paper technique for blood proves its worth. This technique would make much more simple the screening by individual practitioners of the high-risk patients known to them. A large scale field trial of this method has yet to be published~ at least one is in progress in the United Kingdom. There has been hope that other and better diagnostic indices than the blood sugar might be found 5 for exe~ple non-esterified fatty acids (NEFA). But~ so far~ nothing more relie. ble than blood sugar as a clinical index has been demonstrated" TABLE 6.1.3 I Percentage of Adults having Blood Sugar Level of 160 mg% i or more one hour after 50 g of Glucose by Mouth I ~ ! I I t ; I j Age group~ i Total I ! i I ! j years ! 18-79 18-24 I 25-34 35-44 45-54 t 55-64 I 65-74 I i i i I I I I ! I I I I ! Males 1L8 1.3 5.0 11.0 ! 13.2 I 17.7 27.4 I I i ; I 1 Females 18.8 5.1 7.3 11.0 I 19.6 34.5 43.0 i I l Persons 15.5 J.4 6.2 11.0 16.5 26.4 36.0 ; ! i , I t Reference: US National Health Survey. 6.2 Heart disease 6.2.1 Rheumatic and congenital i·1eart disease 75-79 24.7 58.2 4L5 In developed countries rheumatic fever and its consequences appears to be diminishing though convincing figures are not available. Convincing proof can only be provided by incidence figures which in turn depend on the condition being notifiable; and notifi- cation has only been carried out in localized areas" However, deaths from chronic rheumatic heart disease provide some support; and death-rates are diminishing. In England and Wales for example, the crude death-rate fell from 255 to 130 per million persons between 1950 and 1964.96 , 97 Some of this fall is, of course, likely to be due to improved treatment. In developed countries 1 again~ the prevalence rate for chronic rheumatic heart disease is relatively lm.,r. We do not know what it may be in developing -parts of the world. The British general practice survey 98 gives a patient consulting rate per - 85 - annum of 1.4 per thousand persons of all ages (2.J per thousand aged 45-:-64). The US National Health Survey 99 on; examinations carried out on a national sample between 1961-62, found a prevalence rate of 1.1 per cent of persons aged 18-79 with rheumatic heart disease, the highest rate (3.8 per cent) occurring in men aged 75-79. In women 100, 101 .. of 45-54 the rate was 1.8 per cent. \AJHO set enqu~r~es as to world prevalence on foot in 1963 through the cardiological societies of member countries but data are not yet available. Preventive steps can be taken at two stages: firstly by treating prophylactically all persons who have hE:.d at least one attack of rheumatic fever:; and secondly, by administering continuing prophylaxis to persons and who have developed rheumatic heart disease. In areas where the prevalence of sequelae is not high it is probably not economic to attempt mass detection~ (by measuring anti-streptolysing titres, by taping heart sounds and by electrocardiography) and detection and prophylaxis can safely be left to the personal doctor~ An example of detection by taped heart sounds is the study of heart disease in tho Chicago public school child population. 102 In this case the yield proved so small that the cost per case found would surely be prohibitive. The heart sounds of 27 000 elementary school children were recorded on m~~!lE:tic tape a..11d these were interpreted by two physicians. A rate of two per 1 000 children were discovered to have organic heart disease, of whom one per 1 000 were previously unrecognized. Of these previously unrecognized cases 0.2 per 1 000 were due to rheumatic fever and 0.8 per 1 000 to congenital heart disease. An exception to.this principle may be the case where many young people are living together, as in schools and the armed forces. Trials of mass prophylaxis have, for example, been carried out in the US armed forces where·it was found extremely difficult to maintain continuous prophylaxis. It may well be, however, that much higher rates for acute rheumatism and rheumatic heart disease prevail in Africa and Asia, for example, and that prophyla~s in these area.s should be tried, despite the difficulties of dispersion and supervision that would be met. We understand that a trial of case-finding for rhe~atic heart disease is in progress in Russia, based on the National Rheumatological Institute in Moscow. 101 Little more need be said about congenital heart disee.se. Serious cases will normally be discovered at an early stage in life at infant welfare examinations, whether carried out by the family doctor or at health authority clinics. At a later age the school health service takes on this screening role. \~ether so much medical man- and woman-power is best used in carrying out repeated routine health examinations is a - 86 - matter of current discussion. The answer may perhaps li0 partially in streamlining the examination end partially in cutting the number to the minimum, which would be carried out at key times in the life of the schoolchild~ e.g. at the start, at puberty and at school leaving. 6.2.2 Ischaemic heart disease 6.2.2.1 Mortality Of all diseases in developed p~rts of the world arteriosclerotic and d0generative hee.rt disease accounts for the greatest number of deaths, by far the greatest number having specified involvement of the coronary arteries (I.C.D. 420.1.). The death rate has been steadily climbing. For example, 103 in England and \.Jale s, the crude death rate per million living from this assigned cause rose, for men, from 808 in 1940 to 2 731 in 1962 and from women, from 374 to 1 613. There is thus not only about e. three- fold rise for both sexes but an approximately two to one male to female ratio, (with the ratio diminishing the later the year). The evidence that this recorded increase is real is reasonably firm. 103 6.2.2.2 Morbidity For ischaemic heart disease (I.H.D. ), too, reasonably good measures of prevalence and incidence are provided by various prospective surveys. Various studies give mean incidence figures, for men (which of course rise with age) from two to 12 per thousand which seven per thousand as the mode? while prevalence estima.tes vary even more widely from overe.ll rates of 24 per thousand men aged 30-59 (Framingham) 3 to 24 per cent in men aged 55-64 (Annandale). 104 A recent publica- tion of the National Centre for Health Statistics99 finds an overall prevalence of definite I.H.D. in a probability sample of white males aged 18-79 years of 38 per thousand. Epstein and persons over the age of ischaemic heart disease thousand females. 105 his co-workers, in a survey totalling 90 per cent (8 641) of all 16 in the town of Tecumseh, Michigan, measured a prevalence of (including probE,ble disease) of 49 per thousand males and 33 per The prevalence for white men between 65 and 74 years was 12.2 per cent with definite I.H.D. and a further 5.1 per cent with suspected I.H.D. The diagnosis was based on a medical history and cardiovascular exa~ination, including a 12 lead ECG and a full scale chest X-ray. The subject has recently been well reviewed by Epstein. 106 - 87 - In 1963 the WHO, Regional Qffice for Europe held a Technical Heeting on Surveys of - . 1~ the Prevalence of I.H.D. in certain European countries. At this meeting Rose reported on I.H.D. prevalence in working men in England aged 35-59. An overall prevalence of 10 per cent was found, using his questionary 108 and ECG' s read by the Minnesota code. If diagnostically weaker ECG items (ST depression and T inversion) were included as indicating I.H.D. the prevalence increased to 20 per cent. This variability in ECG criteria may account for much of the difference between the prevalence rates cited above. At the same meeting Hostmann and Thomson reported findings for I.H.D. prevalence from Odensa. They found an overall prevalence of six per cent for men aged 40-59, with f, rate of 10 per cent in the 55-59 age group. 6.2.2.3 Screening One of the primary needs for screening, a high prevalence rate, is therefore satisfied. The next important question is, whether a recognizable pre-symptomatic stage exists. This question will be examined in two ways; firstly, is there a latent stage of actual myocardial ischaemia; secondly, can a satisfactory high risk group be defined. (This distinction is referred to in the WHO Technical Report on the Preventive Aspects in Arterial Hypertension and Ischaemic Heart Disease. 109 (a) The latent stage of myocardial ischaemia. The development of the cardio- vascular questionary 6.2-15, 16 has led to the recognition that many people actually have symptoms of I.H.D. without being under medical care. Rose found four per cent of the men aged 35-59 who were examined by questionary had symptoms of angina pectoris and 4.5 per cent h~d symptoms of possible myocardial infarctium. About three-quarters of this angina had been previou~ly undiagnosed. 112 Thus in the symptomatic group alone there is scope for screening in this way. Electrocardiography is the other chief screening technique. Routine examination in this way shows that in a high proportion of people there are changes consistent with I.H.D. in the absence of symptoms, quite apart from the appreciable proportion of those developing actual myocardial infarction without pain - 21 per cent in the Framingham series. l13 I th An d 1 . 1 d f d t 104 . t t f th t t n e nan a e ser1es, a rea y re erre o, 1n en ou o e wen y four men aged 55.64 found to have I.H.D. there was ECG evidence only. In the Ach~son' s 114 gr~up of 53 men aged 65-85 with I. H. D. 27 had ECG changes only. The US National Health Survey,found five per cent of persons aged 18-79 years with either definite or suspected I.H.D. In addition there was a further - 88 - 6.4 per cent who had electrocardiographic abnormalities falling "just short of the rather severe survey criteria for myocardial infarction". 99 In the series, quoted by Rose, 108 of 1 848 English working men aged 35-59, two per cent had definite EGG changes in the absence of symptoms (in answer to the questionary) and a further 10 per cent had the less definite ST depression and/or T inversion. It should be noted here that Hinkle and l] 5 others, · in studies on the variation of the EGG under conditions of daily life, report that changes in the ST segment and T wave occur so frequently in people of all ages in association with ordinary activities that they believe it would be hazardous to assume that these changes necessarily indicate a pathological process. The prognostic significance of some of these EGG changes in the absence of other signs of I.H.D. can be accepted from the results of prospective studies.ll6 We may ask how successful is the ECG in only detecting disease when it is present, i.e. what is its specificity? what proportion of false positives mt=ty we expect? This question cannot properly be answered because we do not know enough about series of persons having abnormal EGG's who later died and were found not to have myocardial disease post mortem. Regarding sensitivity (indicating the proportion of false negatives) of electrocardiography, there is more evidence. The six-year follow-up of the Framingham Survey revealed that 88 men had developed myocardial infarctionj of these, 15 (all of whom died suddenly) had had no previous evidence of I.H.D. 117 An earlier report from Framingham showed that, out of 44 EGG's on patients with I.H.D. 22 were normal. llB Findings of this kind indicate ti1at the val.ue of the ECG by itself as a screening method must be strictly limited. Moreover, as with other tests, the ECG is subject to the drawback of variabi- lity in interpretation between observers. ll9 The Minnesota Code 120 is a help in .improving observer agreement. A further help will be the training of observers (who need not be medically trained) on the set of reference tracings being prepared by WHO Headquarters, Geneva. 107 Inter-observer variation amongst selected observers can be rewardingly low, while inter- observer variation is of the order of twice the variation in reading of the ll('\ 111 same observer. 1 We can conclude, then, that the EGG though highly (b) - 89 - significant when positive 9 is far from being a certain diagnostic technique when used by itself. A third method of screening for I.H.D. (as well as other forms of heart disease) is by mass radiography. Mass radiography is sometimes usE::d in case-finding for heart disease, including arteriosclerotic disease. A good e~mple is the study by Thompson and his colleagues, carried out in Oklahoma. 121 When screening for pulmonary disease the searci:1 for heart dis- ease in addition is undoubtedly vwrtnwhile and it is perhaps best viewed in this wo.y. C8.rried out by itself ci1.e yield for all i1eart disease· is small ~nd coi;l)C.:i.'ed with other methods of finding I.H.D. ~ very small. Nor is radiography likely to find early I.H.D. or hypertensive disease. In Thompson's Oklahoma project 8 126 persons over the ege of 15 were screened by MMR, about one quarter of the whole population of Carter County in this age group. There were 917 (11 per cent) 11 suspect 11 films of which 302 (four per ·cent) were confirmed by 2_ physician. Fifty-two of these cases proved to be previously undiagnosed (60 per cent). Of these 52 cases 45 were either arteriosclerotic, hypertensive~ or both. Lastly, screening for I.H.D. is sometimes carried out by estimation of the serum lipids (usually serum cholesterol level). The predictive value of a high cholesterol, while highly correlated with I.H.D. for groups of the population, is not of a high order in the individual and is therefore not of high value, by itself, as e sc;reening test. There is indeed danger thEJ.t, used unexpectedly, thE: serum cholesterol might be harmfully used by leading to rest:dction of life in persons vri th rele.tively high levels but no other signs of I. H. D. Meanwhile,. (;pidemiological studies of the significance or rise in the se1~ lipid ·1eyels in different populations are in progress. A WHO stande.rdization programme is centred at Atlante, Georgia, JJSA, under :the direction of G.R. Cooper. 122 A review of cardiovascular survey methods is at present being prepared for vJHO by Dr Henry Blackburn of the Laboratory of Physiological Hygiene) University of Minnesota~ and Dr G.A. Rose of the London School of Hygiene and Tropical Medicine. High-risk groups. A number of definite risk factors can now bE: recognized as a result of the various studies, prospective and otherwise, of the epidemiology of I.H.D. These factors will be discussed briefly:. - 90 - (1) Age: The death rate from I.H.D. increases with age and so does the evidence of disease. For example, in the Framingham survey, 117 the six-year incidence for men aged 30-44 was 24.9 per thousand population while it was 90.6 per thousand for men aged 45-62. (2) Sex: Men, as is well recognized, suffer a higher incidence of I.H.D. than women. In the Framingham study the male to female ratio of incidence at ages 30-44 was 25:2 whilst at 45-62 it was 91:45. The tendency for this ratio to approach unity with age is well-recognized end is considered to be related to the hormonal changes at the menopause. 123, 124 (3) Physical activity: Morris and his co-workers 125, 126 have shown thc:tt it is likely that I.H.D. develops more frequently in those Hhose work is physically inactive than'in those Hith physically active occupations. (4) Occupation: There is a gradient amongst men (not amongst married HOmen) in mortality by social class. The Registrar General's Decennial Supplement for England and tvales (1958), 127 gives a Standardised MortPlity Ratio (SMR) gradient from 147 for males on Social Class I to 89 for males Social Class V and the rBnge is even greater for specific occupations (registered medical practitioners, 159); farmers, 62). (5) Serum lipids: The Framingham survey, amongst much other evidence, clearly relates group susceptibility to I.H.D. to the serum cholesterol level. There is no evidence of bimodality in the distribution, except perhaps for the very small group of familial hypercholosterol-aemias. The risk of I.H.D. has been shown to increase with cholesterol level. 117 Even more, perhaps, than with other continuously distributed variables the predictive value for myocardial .infarction of the serum cholesterol in the individual is low. (6) Blood pressure: There is a firm relationship between high blood pressure and I.H.D. The Framingham Survey shows that the. risk of developing I.H.D. increases as the level of the blood pressure (both systolic and diastolic) increases) the observed incidence of I.H.D. being not far short of twice the expected in persons with definite hypertension (systolic B.P. over 159 mm Hg or diastolic B.P. over 95 mm Hg). 117 When left ventricular hypertrophy is present the ratio was more than trebled. (7) Body .1<reigi1t: The independent contribution of overweight as a factor in I. H. D. is not supported by the Framingha~ Survey. 117 The Framingham work and - 91 - also that of Doyle128 2nd others - but not that of Oglesby Pau1116 - does show a positive correlation between I.H.D. and gross overweight; but it appears that this contribution to morbidity is mainly due to the cross- correlation of overweight with serum cholesterol level and blood pressure. Excess weight is also associated with diabetes mellitus which in turn increases the liability to I.H.D. Again, the association may be indirect. However, since overweight is more readily identified than blood pressure or cholesterol level, weight remains an important correlate (even though indirect) for I.H.D. Apart from body weight itself there is certain evidence that body-build may be directly correlated with I.H.D., in that mesormorphic persons are at higher risk than exo- or endormorphic. 129 Spain found a higher prevalence rate for I.H.D •. in endomesomorphic persons which was not altered by excluding h t . . d" "d 1 130 p 116 yper ens1ve 1n 1v1 ua s. aul while not:i.ng n.:-' significant difference .in weight between I.H.D. patients and controls, did find a signifi- cant difference in skinfold thickness, the I.H.D. patients having a higher fat content. (8) Smoking: There is ample evidence that cigarette smoking is in some way correlated with an increased risk of I.H.D. The study of Doll and Hilll3l of British doctors and the prospective surveys of Hammond and Horn# 132 Dawber and his associates133 A.t Framingham, and of Doyle and others.134 as well as the follow-up of the San Francisco 1ongshoremen28 all support this conclusion. The important conclusions are: (a) that the excess risk of I.H.D. increases with the number of cigarettes smoked5 (b) no relationship between risk and the duration of exposure to cigarette smoking has been shown; (c) there appears to be no excess risk from pipe and cigar smoking? (d) there appears to be a rapid reversion of the excess risk of I.H.D. to normal in persons who have given up cigarette smoking. No causal factor in cigarette smoke has been identified but it appears probable from these findings that the excess risk of I.H.D. may pe due to the repeated adrenergic effect of the nicotine content of cigarettes. The Framingham workers estimate that, if cigarette smoking ceased, mortality and morbidity from I.H.D. might be reduced by almost one half. 135 6.2.3 - 92 - Towards the end of 1965 a WHO Working Group met in Prague. 136 The group considered the present possibilities for preventive measures in I.H.D. and outlined the high-risk factus, in order of importance, as: Conclusions high serum lipids high blood-pressure diabetes mellitus cigarette smokjng overweight stress We have seen that more studies have been made into the epidemiology of I.H.D. than into most other non-communicable chronic diseases and that a considerable amount is now known about· the early natural history of the condition (though tantalizing gaps in our knowledge of causation still exist). Much is also known about the available early diagnostic tests~ (though we do not yet know the sensitivity and specificity of the cardiac questionary). We should now ask whether, in the light of this knowledge, enough is known with sufficient accuracy about the natural history of early I.H.D., the tests for its detection in the individual, and its treatment, to justify case-finding programmes. i.Ve have noted above that the practical possibilities for early detection and treat- ment have been recently reviewed by vmo. 136 At present we must perhaps regard the value of early diagnosis in I.H.D. in a tentative manner in view of our ignorance of causation and of methods of prevention. However, there appears to be at least three risk factors which can, through individual effort be reduced~ cigarette smoking, overweight and sedentary habits. It is possible that, with the development of a simple cardiovascular questionary, combined with electro- cardiography, blood-pressure recording, chest X-ray and weighing in older persons, a particularly high-risk group of the population could be identified for which there would be a special incentive to stop smoking, reduce weight and increase physical activity. At least one trial on these lines is in progress. The other line of prevention which has received much attention is the reduction of serum lipid levels by dieting. At present firm information is lacking on the value of diets low in saturated fats and/or with added poly-unsaturated fats in preventing myocardial infarction. However, this is regarded as a reasonable clinical preventive . - 93 - measure in persons at particular risk and it has been recommended on a large scale (e.g. in Norway~ in Sweden by the Swedish Board of Health; and in the USA by the American Heart Association). 1Jhether or not it is justified on present knowledge as a general recommendation 7 alteration (where necessary) of the diet in persons identified in the population through screening measures as being at particular risk may well appear reasonable. In conclusion~ the matter of communication with the public 1 and how this is to be handled 7 is of grea.t importance. A health education policy that presents the public with a reasonable and balanced view of the risks of I.H.D. and what can be done about them needs to be evolved at the same time as the epidemiological work is carried out and not left until actual case-finding is being widely practised. It could be disastrous if persons, knowing themselves to belong to a high-risk group were to take too gloomy a view of what could be done about this situation) it would be equally if not more 5 disastrous if peop~e tried radically to alter their way of living on the basis of inadequate epidemiological evidence. Though a jest9 Myers' "Thumbnail Sketch of the Man least likely to have Coronary Heart Disease" does point a lesson. 137 6.3 High blood-pressure Sphygmomanometry is usually carried out as part of a screening programme. It is of course of value to diagnose severe degrees of hypertension as early as possible. Undoubtedly cases of severe hypertension occur with absence of minimal symptoms though this must be the exception rather than the rule. Perhaps the most valuable reason for taking the blood-pressure is in the case of the young person with a high pressure in whom a. remediable cause may be present 7 like coarctation of the aorta or phaeochromocytoma. Even when the cause lies in the kidney remediable measures may well improve the outlook. There is evidence 5 too? that treating severe essential hypertension improves the prognosis and may reverse retinopathy:3S, 139 There is, :i10wever, the large popu.:..ation of persons havin;; mo~erately "raised" blood-pressure for their age and sex in which the value of treatment has not been established. This is work 1-1hich needs to be done. It could well turn out that drugs 5 while lowering blood-pressure and marginally improving prognosis, were so unpleasant to take for the rest of a patient's life 5 · and induced such a sense of invalidism, that the overall benefit could not be considered worthwhile, despite the incontrovertible evidence that life expectation in the actuarial sense diminishes 1.-ri th rising blood-pressure. 140 The Rostural hypotension and sexual impotence induced by some modern drugs are both severe disabilities in themselves. - 94 - Apart from our present ignorance of the value of treating the early essential hypertensive (and of what criteria to adopt)~ there is the methodological difficulty of the blood-pressure recording itself. Two factors contribute to this difficulty~ (a) interpreting the level in relation to the circumstances under which the reading was taken~ (casual, resting, circumstances of the examination): recent work relating average casual pressures to average continuous recorded pressure in normal and hypertensive persons carrying out normal duties? indicates that there are wide departures from the average relationship in individual instances, and that the average recorded pressure is considerably lower than the average casual 16,141 pressure. (b) the observer variation in recording blood-pressure: To obviate this? sphygmomanometers are now becoming available which remove this 142 source of error. These points are clearly important in practtce since they can seriously affect the validtty of blood-pressure examination. Recently~ the subject of measuring blood-pressure has been well reviewed in the Lancet. 17 We do know that "high blood-pressure" is com:.10n~ increasing with age. There is argument as to whether transmission might be by a single gene or whether it is multi- factorial. Population surveys by Boe et al., 143 Miall and Oldham, 144 Hamilton et 145 146 J47 al,, Kagan et al., and the US National Health Survey- indicate a skewed normal distribution of blood pressure with no sign of the segregation of a "disease" group of hypertensives. The Framingham workers arbitrarily define "high blood-pressurerr as readings by two examiners of 160 mm Hg or over systolic or 95 mm Hg or over diastolic. By these criteria 391 out of 2 024 men aged 29-62 were hypertensive (19 per cent) and 410 out of 2 445 women in the same age-group (17 per cent). Using the same criteria the US National Health Survey found a total of 15 per cent males and 16.7 per cent females between the ages of 18 and 79 years had blood-pressure at or above these levels. The numbers with high blood pressure rose with age in the proportions shown in the Table 6.1.1 (from 147). 105 Epstein and his collaborators in the Tecumseh Survey, found overall a. prevalence of hypertensions heart disease (HHD) of seven per thousand males and nine per thousand females (defining HHD as a systolic blood-pressure of 160 mm Hg or more or diastolic pressure of 96 mm Hg or over, or both together with suggestive evidence of left ventricular hypertrophy). As would be expected, the prevalence rises steeply with age and was higher in females than in males; for example, in the decade 60-69 it was 39 per thousand males and 71 per thousand females. - 95 - We do not yet know whether treatment would be effective and worthwhile in improving prognosis. It is difficult, as Pickering148 has pointed out, to compare the prognosis in hypertensive patients with that expected for the whole population because of the almost inevitable selection in the case of the hypertensive series. Bechgaard,149 however, has compared a group of persons witn raised arterial blood-pressure with the population of Denmark and shows that the mortality for men in the group 40-49 with systolic BP over 180 and less than 200 is nearly five times that of the whole population. (For women it is only 1-1/2 times as high as normal.) The records of insurance companies, too? show the relationship between levels of blood-pressure and mortality. There ·are, therefore, good grounds for treating mild essential hypertension provided it can be demonstrated that this is effective at the lower levels of blood-pressure and does not interfere unacceptably with the normal enjoyment of living. What is evidently needed is a controlled trial of early therapy to answer these questions, including that of the acceptability of long-term drug administration in the absence of symptoms. 6.4 Overweight 6.4.1 In multiple screening programmes it has been usualto record the weight, height e..nd determine, by reference to standard weight tables, whether a person is 11 overweightn. If so, he is told of the risks of overweight and advised about weight reduction. This advice is normally based on the judgement that the excess.weight is due to obesity, since above-average muscle or bone mass may not be associated with a reduced life expectation. Life tables, like those published by the .American Society of Actuaries, show that expectation of life falls with increasing degrees of overweight when weight is standardized for height, build, age and sex and that expectation can in fact be improved by weight reduction. 15° The main risks associated with overw_eigh.t are .cardiovascular disease (only with gross obesity, accordinng to Doyle et al., 129) diabetes, chronic ·- .. -~' respiratory disease and degenerative arthritis. But, as with other continuous distribution conditions, there is difficulty in deciding what should be regarded as overweight~ anything from 10-30 per cent above the mean has been regarded as the lower 151 clinical limit in different studies. Kemsley and his co-workers suggest the upper and lower quartiles as reasonable practical limits, for overweight and underweight. An important factor, not fully allowed for in most standard tables, is the relative adiposity .. Measuring this adds to the complication of the examination, though it may be done relatively simply with an instrument like the Harpenden Calipers. 152 - 96 - Another difficulty about using tables is that normal weight (and~ of course 7 height) for men and women varies from country to country and within the same country at different periods of time. For example, comparison shows that the average height and weight of persons in the USAl53 has been increasing during the present century and US weights (for heights) were greater than those for some other countries (Canada and Norway). In the CCI Baltimore study, 154 in the physician evaluation of a sample by physician interview, the prevalence of obesity as judged by observation was 129 per 1 000 popula.tion~ or nearly one in eight persons. Of the sample submitted to screening 40 per cent of males were 10 per cent or more above the Canadian normals used 9 18 per cent were 20 per cent above and eight per cent were 30 per cent above. As the report says? 11 the high prevalence of obesity •.•.•• points up the need for further research to clarify the relationship between obesity and health ..•..• The size of the problem suggests that the approach may need to be directed to the public at large as well as to the individual obese patients". Strong motivation is indeed needed to induce obese people to reduce their weight by dieting, as those who have had clinical experience of this particular problem know. To some extent this recommendation, made in 1956, is beginning to be translated into action. Norway, for example~ has recently issued a public statement, through the Norwegian Board of Health, advising moderate dieting~ primarily aimed at lowering the intake of polyunsaturated facts but also recommending keeping to a moderate total calorie intake. The problems of definitions and of weight control in obesity were recently reviewed in the British Medical Journal, l55 when the need for further longitudinal studies of the effects of treatment was pointed out. There is no doubt that weighing is one of the simplest and most precise screening tools available and that the information it gives is important for health~ what we still need to learn is how best to act on the information it gives. 6.5 Respiratory diseases 6.5.1 Pulmonary tuberculosis 6.5.1.1 Whilst deaths from pulmonexy tuberculosis began to fall dramatically with the introduction of streptomycin and chemotheraphyl56 notifications have fallen at nothing like the same rate and indeed~ in the older age-groups have even tended to rise. This effect may be attributable to improvement in methods of detection and especially to MMR. Having reached low levels in developed countries there is a tendency for the notification rate to level off. In England and Wales at the end of 1963 there were 340 000 persons under supervision for pulmonary tuberculosis, a ra.te for the population of about 0.75 per cent. The Director-GenerRl of WHO has pointed out that not a single country had reached - 97 - the poi11t qf c_o1;11:,rol wher@, there is less. than one per cent prevale11ce of nfJ,j:;ural. reactors .:,,~ ' ~· '> ' . "- '. - . . to tuberculosis a.moi1g cll.ildrE?n in the 14 year age~group, a cri:ter;i.pn~et by 1;1. previous 1.<JHO expert comm:l_·ctse as indicating that tuberculosis is no longer a serious _public health problem. 6.5.1.2 Pulmonary tuberculosis is perhaps the classical condition for early, often pre-symptomatic, detection (as well as prima.ry prevention with BCG) and it .meets the screening criteria well in that - 1. ~lmonary tuberculosis is an important public health problem. 2. · Facilities for diagnosis and treatment should be ava.ilable, otherwise there is . J:lO benefit only harm, from screening. : 3. The natural history of the precursor stage of the disease he.srbeen' elucidated? early_ infiltration does lead -to. overt disease. 4. There is a recognizable lat~ntstage (positive tuberculin reaction and · infiltration). 5. Suitable tests are available, the tuberculin reaction and mass miniature radib~faphy··· (MMR). 6. The tests are acceptable to the population . .. :·.; .' 7. There is .an accepted (and effective) treatment. . - .. "·'· . '' . . ,~·, .. P~}i',~,0ns with9ut rec_ognised disease (small f~~:r:_osed infiltrations) are :qot treated .as pat~ents. ·:.i..''.' 9"'. ,Th~ cost is t.olerable where MMR is used, by the saving in cost of treating a flqrid. case of tuberculosis. 10. Long-term·follow-up is a built-in part of schemes for the control of tuberculosis.- 6.5.1.3. This brief summary makes no .attempt to review the bulky lite'rature on ·the early detection of tuberculosis. · The subject has been reviewed recently by a World Health Orgai1~Za,1;.ion Expert Committee.~57 Only a few salient points will be discussed: . . 1. Unlike cancer there ic evidence that small tuberculous lesions in the lung equate with early lesions. 158 In a five ·year ·p;ospective survey of civil servants and other workers in the London area the size of lesion'at first X.;..:rayhas proved of the greatest prognostic significance. - 98 - 2. Small infiltrative lesions can only be ignored at peril; persons with these lesions are in fact in need of treatment and between 70 per cent. and 80 per cent. will have further trouble if the lesion is ignored. 3· Proof of the effect of early treatment: this is very difficult to demonstrate since it would be necessary to show that the rate of fall in registrations was influenced by earlier detection and that a significant change in the slope of the curve in a downward direction had occurred. What has in fact occurred as a result of introducing M.M.R. (and this has been seen best in intensive surveys such as those of some Scottish cities) is an early rise in registrations at the time of introduction~ followed by a compensatory fall in registrations afterwards; and with a third relatively rapid swing back to the level of the falling base-line after which the fall in registration continues at its "original" rate (Figure 6.5-1). Thus in England and Wales in 1938-39 the mean annual notification rate per 100 000 persons was 88; in 1947-50, following the impact of M.M.R. it rose to 100 per 100 000 and this was followed by a l5ri fall in 1954-55 to 80 per 100 000 persons • ·· In the comparison carried out in two Welsh mining valleys the fall in registration rate was practically the same after the completion of the M.M.R. Campaign. 4. As with other tests radiological screening sets the problem of sensitivity and specificity~ linked with that of observer variation. Yerushalmyl59 found a false negative rate (a missed lesion of some kind) of nearly one third in addition to a false positive rate of about 2 per cent. By dual reading one third of the lesions missed by a single reader could be picked up~ but this need necessarily adds to the cost of case-finding. 5. As radiological screening progresses the number of persons needing con- tinuing supervision increases cumulatively; few are discharged but additions are always being made. This need for arrangements to keep the intermediate type of patient under surveillance of course adds to the expense. The experience of the Danish Tuberculosis Index160 is, that for every one person needing treatment between five and 10 will need to be kept under observation. Similarly, Styblo 161 has some 20 000 - 30 000 persons under supervision in Czechoslovakia~ in whom the natural Vl c: 0 -~ -Vl C'l ~ ....... 0 ..... Q) ..0 E ::;:) z Fig. 6.5-1 t Time----. Introduction of M.M.R. WHO 70C'IO history of the border-line lesion is being followed (thus critari6n "3" · · above, has not been completely met,; there are early pulnlonary stages · · which require watching rather than treatment). 6. Economics: (a) In a developed country the yield from M.M.R. rapidly declines and it then becomes a matter for decision as to whethe.r to make it . . . . .. .. ~ ; ' more and more selective; or to continue offering it to the whole a,dult population in the expectation that the rare missed case may otherwise go '. . . . longer undetected tharl had M.M.R. nev;er been introduced, (beqause .of the ._.:. very rarity of the disease)#. thereby prolonging the danger of contact cases occurring. One alternative is to make M.M.R. selective for high' risk . ·. -~ , ;.'. --~-c;·~~:· ... · .. :.::c.·:.~·-· ·groups in the population (the adolescent, the elderly, the economically depressed# person's. exposed to special' industrial or professiona.i 'riskS:. persons with cough); or to link M.M.R. w~th ~antoux tes_tinE. ~<? .. that all contacts of Mantoux positive children are X-rayed. ;-;, ·~. ' . In countJ;>ies .. where the co:;;t of a. oountrywide service might be prohibi:. ·•···· ·. . • • . . ·' . • ~ . .l... . • . • . • • . tive and where m~dical services woul,<i not be adequate to carry-out extensive follow-up of border-line cases. In the first place~ it goes without saying that the prerequisite for a detection programme is an., a.deqlla.te mecii~al service for the definitive dle.gnosis and treatment ancl.afte;t'- .. . .. . . : ·' care of the. patient. There is clearly no point in diagnosin_g if facilities for treatment and the prevention of contacts are not available .• Secondly, the stage at which tuberculosis is diagnosedmay be varied according to· .· :' :··resources. M.M.R. is expensive and, as we have noted, produc~s ~"large number of border..:li~~ patient~ f~r costly follow-up. There is a case, now that chemother~py.''k{cl.. antibiotic treatment is so. effective, for pas.sing over the earliest stag~~ of the dise~eand endeavouring only to detect. the f) • .•. ··.. 162 condition when the sputum has become positive. Clearly. there are .. '".·,-:· -· .:.c-~ risks of contact c~es associated with this technique but it is. perhaps the •• r -.. •', ··,• ' t, . . . . .. ' ~· ,.i~ ... . ~ . -- - - most economical secondary preventive step available and therefore capabl:.e of. being widely used. ·,-, ... '·' ;.·•' :".;. i_.: ... t . ~. ' . - 100 -· 6.5.1.4 In sunimary, the words of a memorandum of the Chief, Tuberculosis Unit, WHO, Geneva (1/2/65) :put the world position on screening into perspective: "·••••• symptoms provide a much more selective criterion for achieving high yields than other epidemiological screening criteria, single or combined. ·. 6.5.2 ••••••• where pre-symptomatic case-finding is considered justifiable and feasible, screening procedures dictated by the epidemiological situation can increase the efficiency of a community examination, but the latter's contribution to the total case yield will always remain marginal." And" ••••••• follow-up (after screening) may well be the most important aim from the application of screening procedures, as there are indications that it yields, in the long-run, quite substantial returns." Non-specific r.espiratory disease Chronic non-tuberculous disease of the chest presents rather different problems from those discussed under tuberculosis. By far the largest contributor to this category of illnesses is chronic bronchitis and it is with this that this section will be concerned. There is now overwhelming evidence that the prevalence of chronic br<?nchi tis is associated with the twin factors of atmospheric pollution and cigarette smoking; macro- and micro-air pollution. 163 Wide international differences in mortality and prevalence are found and much of this appears to be related to the factors first mentioned. Undoubtedly, also, the diagnostic customs in the countries concerned play their part. For example, the mortality attributed to bronchitis in England and Wales in 1958 for all :Persons was 65 per 100 000 compared with a rate of only 2.3 per 160 000 persons in the. United States of America (white). However, it is doubtful whether differences in diagnostic habit or coding of the cause of death could account for such large differences. Comparative studies of pathology are now in progress~ in Chicago and London and these should help to resolve these ··,( doubts. As an example of the type of epidemiological work needed in order to understand the real meaning of international differences, the study of Mark of comparative prevalence of respiratory disease in England and Wales and in Norway should be - 101- mentioned. 164 Mork foun<i that, while the prevalence of minor symptoms was nearly the same in the two countries, the prevalence of severe symptoms of chronic respiratory illness was considerably higher in England and Wales. Studies of similar working groups of postal and telegraph workers in London, rural England 165 and United States cities have been carried out by Holland, with comparable results to those of Mork. In England and Wales chronic bronchitis appears to be more prevalent than in 166 any other country. A survey by the College of General Practitioners showed a prevalence of 17 per cent. in men aged 40-64; and in an earlier survey bron- chi tis was found to be the commonest re.ason for consulting a doctor in general practice (260 consultati.ons per 1000 patients per year), second only to the common cold in accounting for numbers of patients consulting their doctor.97 Though much has still to be learned about the aetiology of chronic bronchitis, a great deal of epidemiological work has by now been carried out which has enabled us to identifY some at least of the precipitating factors. Use of this knowledge, largely by abating atmospheric pollution and by persuading people not to smoke cigarettes, should go far towards preventing the disease. However, it will take a long time to achieve these aims and in the meanwhile many young persons are insidiously developing chronic bronchitis. We should try to discover what might be done to arrest the course of this pathological process and, if possible, to reverse it. In the absence of a specific aetiological agent or of specific pathology the assessment of chronic bronchitis has to depend on recording symptoms and on measuring functions like the forced expiratory volume (F. E. V.) or peak expiratory flow rate (P.E.F.) and on sputum volume. Records have proved their worth in cross-sectional epidemiological studies. Work is now in progress on prospective studies aimed at identifYing personal factors of susceptibility to chronic bronchi tis. Preliminary evidence 167 seems to incriminate atmosph~ric pollution at a much earlier stage of life than has been previously recognised. If it proves a practical proposition, by using a short respiratory questionary and a simple measurement of lung function, to identify a high risk group of individuals at a very early stage of respiratory function abnormality, the next stage would be to conduct a trial of treatment. In this particular case the most effective treatment we know is stopping smoking and a - 102- comp~rLson between early chronic bronchi tics still smoking and h'lving stopped smoldng shou.l<i give helpful information abbut the reversibility of the condition. A d'i.fficu}ty wc.u.1..d certainly be thCJ.t c.f cbta:lning test g:!:"'oups compe.re.ble in other respect-s. At the same tj_me, nabn"al~.y, it wcJt.~ld be useful to study the effect of g:t.ving up :::moking on the incid.anee of ischaem:i.c lnart d:Lsease, as referrE:d. to in the sect.ion of this paper which deals "tv·ith that condJtion. Studies of this kind a:re now getting vnder :·u:;;y but it will be a con::::;ide:t'ai)le tim0 before we can pt>rso:::.s c~n l:e identified who would speciaJJ.y benefit from ::nt·ms:i.ve preventive measures, over cmd above those normaJ.J_y <;.dvif>ed. li'inaJ.ly_, industrial risks should also be considered. Th8 P>~·0cise relation:'lhip b::;,t~'!ecn ccPtain industrial hazards~ for exrunple in tlJc stGeJ. i.nc.:ustry_, is not fuJJ_y u·.!.derstood. Her.: again work is needed to discovt>r h'"hcther a high-risk group exists Khich coFld b0 identified and advised en other em:;:>loym'?.nt~ A.t loa;:;t two studies of -:-::1.is kind are in progress in -~:1e United Ki.ngdcm and nv doubt 0+.hers are in progress Th0 need for internatic;J.ally compar£:ble nomenclature and meth~:io].ogy ih chronic non-specific Ju."'lg disease, nnd for further epidemiologic:;.}. r.:;tucl5.cs..., ·.was discusos:i at. .-- 168 a. Hl:IO European Symposium held in Moscow in 19o2, We do not have accur·atc morb:i.di ty data on lung c.:.ncer ~·Jhich indicate the :2:tzE: of the problem. Ravertheles:::. c>.n idea can be. obtained from the general prnctic8 Durvey carried ou·:~ by the College of G~mera::_ Pract:i.t:ioners -:tnd ·c-,~ Gcn8ral Regl;;;.ccr Office ir~ Engl3.!1cl '"'nd Wale:::; in 1955·~56, s:i.nce i·~ i3 likely that all. paticmts Hi th lung c~cE-~r consulted ·::.heir family doctor ;J.t least once d.ur).n::; the observation yca.r. The rate for 11 pn/~ients consulting11 for.- neoplasm of tho lung., bronchus &nd tr·achea is Oo5 pe:::> t..lJ.ousund pe;rsons (1 .. 0 per thotJ..5c-.!ld fO.i" n:::t.lf".l3 &'1<'!. O.l per thousand fm.' fe;;12.les); for men betiJeen the age.:; of 45 and 64 :l t JB 2.1 per.- i.;hou;.:;r.nd. In thi:::; high-risk group luns cancer is found, thc:J:C!fc:::'e, at a rather lo;•J8r prevalence ro.tc ths.n is ca."t'cinoma-il1-si tu of the ut;::rina c0rvix in r,lJ. adul +. v;ome:::J. (3 per thous~.nd)" Hc:·rdver, lung cancer is so lethal. that deA.tbs frcm this nonditicn ['.ro more· than fi•re - 103- time.s as .. frequent as those :from cancer of the cervix. Moreover~ deaths have been increasing annu~ly at an alarming rate in western countries~ associated with the earlier increase in cigaJ."ette smoking; in England and Wales deaths in males have increased by nearly two thirds between 1954 and 1964~ having risen from 14 000 to 24 500 malignant neoplasms of the trachea, bronchus and lung (6.6-.1). Such is the nature of lung cancer~ therefore, that if early detection can be made effective~ mass screening of the adult population in highly-developed countries would be indicated. Unlike the accessible cancers, (cervix~ lip and skin) lung cancer is customarily at an ad.V:i:Uloed ·stage when diagnosed with the help of X-rays and t..,_e prognosis is nearly always bad.:_ The corrected survival rate at five years for males in England and Wales regi5.tered during 1945-1947 was only 14 per cent. for. early cases sub- mitte<;ltoradical treatment {6.6.2). Forall early cases) whether treated or not, the fiye . .,.year. survival· rate was only two per cent. Seven out of eight cases were in men and of these only. 13 per cent. were classified as early at diagnosis while nearly one quarter had metastases. The median duration of· symptoms Nas about six months for the early cases but only four and a half months for those already with metastases. The five-year survival rate of neither early nor late cases treated radically was affected by the duration of the symptomatic history~ being for the early cases 12 per cent~. with symptoms of from 0 to 2 months duration and 10 per cent. for personswith a 12 months and over history. This is a gloomy picture but it is necessary to remember that it is based on persons developing the disease twenty years ago~ Since then mass miniature radio- graphy (M.M.R.) has been practiced widely and it is necessary to study its effect in combination with advances in thoracic surgery. M.M.R. was, of cOurse, developed for case-finding in the epidemiological control of pulmonary tuberculosis. With the decline.of tuberculosis in many countries attention has turned towards the possibility of using M .. M.R. more for diagnosing other lung conditions, of which cancer is the most important~ rather than just for seeking out tuberculosis. We need therefore to see as clearly as possible what is the evidence for benefit so that a sensible policy may be evolved. Four surveys will be considered; that of Posner, McDowell and Cross in Birmingham; 171 Cuthbert 1 s · review of the Glasgow X-ray c~paign; 172 - lo4 - \vaddington 1 s comparison of the Liverpool survey with his own and Gifford'' s routine hospital admissions; 173, 174 and Boucot, Cooper and Weiss' experience with the Philadelphia Pulmonary Neoplasm Research Project. 175 6.6.1.1 (1) Birmingham: Posner and his colleagues analysed all cases of lung cancer diagnosed by M.M.R. units in the Birmingham Hospital Region during one year, 1955-56, 238 in all being investigated. As with the Glasgow series patients diagnosed by 11 conventional11 M.M.R. {"routine M.ri.R. cases") were compared with patients referred to the units by general practitioners. There were rather more older patients in the general practitioner referred group. At one year the survival of the routine M.M.R. group was better than that of the general practitioner referred group - 50 per cent. compared with 36 per cent., though this difference is, not significant. More routine M.M.R. cases proved to be resectable than were the g.p. referred - 44 per cent. compared with 30 per cent._; also there was a higher proportion of lobectomy as opposed t~ pneumonectomy in the routine M.M.R. group, which is usually considered to equate with a b~tter prognosis. As with the Glasgow series 85 per cent. of the routine M.M.R. group had had symptoms at the time of presenting for examination. These workers concluded that the smallest cancers were very easily missed and recommended the selective screening of men over 35 referred by their own doctors, irrespective of symptoms, to static 100 mm camera units. They considered it would be a major mistake if a relatively costly programme of this kind were to be allowed to deflect attention from the importance of primary prevention. 6.6.1.2 (2) Glasgow: 48 patients with proved bronchogenic carcinoma were found through the operations of one of the city chest clinics,and these patients.were compared with 48 consecutive patients referred to the same clinic by general practitioners. The average age of the two series was about the same with a similar range, most of the patients being between 50 and 65. Of the 48 M.M.R., cases 36 were found to have had one or more of the cardinal symptoms of chest ill- ness; but on several counts the disease in the patients referred by general practitioners was more advanced. As might be expected, the patients in both series who were considered suitable for surge~ had tl}e better survival at 18 months; but the M.M~R. group did better than the g.p. referral group, 13 of the M.M.R. group surviving at 18 months compared with only eight of the g.p. group. - 105- 6.6.1.3 (3) Liverpool: Out of more than 450 000 persons over the age of 15 X-rayed in the Liverpool Campaign of February/March, 1959, 235 were admitted to surgical wards for investigation. Of these, 163 were suspected of suffering from bronchial neoplasm, a rate of 0.36 per thousand persons examined (0.5 per thousand males) which is somewhat lower than the general practice survey rate quoted above. Of the 163 suspects there were 118 proved to have a primary bronchogenic carcinoma and Waddington compared these with his earlier series of patients admitted to the Liverpool Thoracic Unit. 80 out of the 118 (68 per cent.-) from the Liverpool Campaign were resected, 90 per cent. of all those surgically·explored~· This is a rather higher rate than was found in the hospital series where o_nly 70 per cent. of those explored c-oUld be resected. They were also able to treat a higher proportion: of the Liverpool Campaign patients by lobectomy (as opposed to pneu- monectoiny) than were so treated in the hosp:ltcil series. HoWever, at the time of the report, while only. 61 per cent. of all the 'operated patients fwho had survived the first two ~onths) had survived one year 67 per cent. of the hospital· series had so survived. There 'seems therefore to nave been no advantage .. at the time of writing, for those -diagnose-d presumptively earlier by M.M.R. 6.6.1.4 (4) Philadelphia: In this well-known project 6137 men of 45 years of age or more were enrolled in an experimental prospective survey with the aim of following them by means of six monthly 70 mm chest X-rays and a short medical history. A previous M.M.R. campaign in :Philadelphia had shown a prev'a.lence rate for lung cancer of 2. 7 per thousand men over 45 which is rather more than the Coll~ge of General Practitioners su.rVey :found in a similar age group. During the course of the study, 26 men developed lung cancer, in whom no radio- logical evidence of neoplasia had been discovered on entry to the project. Only five were without symptoms up to the time of the first positive X-ray and only two were asymptomatic actually at the time of the examination. The other finding of direct interest is the survival: only two out of the 26 had survived to the time . _, ·- of writing the paper, which could have given a maximum survival time of seven and a half years, though there is no statement as to actual survival times. From these surveys it is clear that the prognosis of lung cancer is little influenced by detection by routine radiography and that, in fact, most of those who are detected in this way are symptomatic at the time of their X-ray. It seems - lo6- probable that~ as at present carried out~ routine chest X-ray at any interval greater than six monthly would be of little use. More frequent x~ray examination would not only be uneconomic but would also pose problems of persuasion to attend and of possible harm from the frequency of examination. The best present use of routine radiology is likely to be for selectively screening the middle-aged person~ particularly males, with a persisting cough. The examination of sputum by exfoliative cytology has also been employed as a means of detecting early cancer of the lung. Compared with exfoliative cytology for cervical cancer there are drawbacks. The bronchus is not accessible in the way the cervix is; and examining the sputum is considerably more time-consuming than examining cervical smears if it is to be at all acceptably reliable. At ,Tohns Hopkins Hospital 176 the positive sputum rate in patients with bronchial carcinoma rose from only 20 per cent. when one sputum was examined to 56 per cent. when five specimens (3 smears of each~ i.e. 15 slides) were examined (or from 42 per cent. to 95 per cent. when suspicious reports were included). In another series of 144 177 patients with suspected lung cancer at St. Bartholomew's Hospital,- · London, there were 10 per cent. false· negative results from examining 3 smears from one specimen of sputum. The time needed to achieve this degree of accuracy was unacceptably long. A rapid ·sputum cell concentration technique is needed to shorten examination time and work on these lines is proceeding. Lilienfeld 178 has reported on a comparison between sputum cytology and radiology in the early detection of lung cancer in persons living in u.s. l~terans Administration homes. Up to 1960 over 12 000 persons, aged 45 and over, had been subrrdtted to a 6 monthly X-ray and sputum cytological examination at least once and some 4000 had been screened between three and four times. Of 43 cases of lung cancer diagnosed by follow-up, cytological screening contributed to the diagnosis in 15 and this diagnosis would have been missed if X-ray screening alone had been used. On the other hand, if cytology alone had been used, 21 out of the 43 would have been missed. Unfortunately survival at six months was no better than previous experience where repeated screening was not carried out. It appears, therefore, that sputum cytology~ while it can add to the diagnostic accuracy of X-ray examination alone, is not atcpresent able to offer an improvement to the prognosis of lung cancer and would not be economically justifiable. - 1-07 - .Clearly -the prevent:ton ·of lting cancer would be far better than early detection and .it is interesting that" the American Cancer Society and Veterans Administration Study just refepred to showed that one per cent. of-present smokers had positive or suspect sputum cytology as compared with 0.35 per cent. of those who had never smoked and 0.47 per cent. of past smokers. 6.6.2 Cancer of- the cervix The evidence for the value of detecting cervical cancer early is relatively strong. The earlier the stage at diagnosis the better is the survival rate. It is reasonable to suppose that diagnosis at the pre-invasive, carcinoma-in-situ, stage would very greatly improve on existing sUrV-ival though it is as yet too soon for . . .. ·. statistical data on this point to have accumulated, nor though mortali~y from cervicar ~~6e~- is in general decreasing at a slow but steady rate in advanced countries, is there evidence of a reduction in the death-rate directly attributable to diagrl.osis and treatment of the pre-invasive lesion. It is, indeed, probably too soon to expect to see this effect, since intensive commUnity screening for can.cer of the cervix has only been practised for a few years; and that only in a relatively few centres. A significant fall in the incidence of invasive cervical cancer has however been observed in British Columbia since the introduction of widespread screening by exfoliative cytology. Unlike other pre-symptomatic conditions- the natural histo:cy. of pre-invasive - .. cerv.ical canc~r has been reasonably closely studied. Firstly, there is relatively good evidence that the in-situ lesion pre~-~des and turns into invasive cancer of the-cervix. The direct evidence for this is both retrospective179 and pros~dtive~lSO, 181; 182 >. 183 There is also indirect evidence based ori the age- -.. . . 184, 185 specificdistl:-ibution of in-situ and invasive cancers. Secondly, estimates are available of the proportion of in-situ lesions which. b . . 180, 181, 182, 185, 186 t t . l f ecome ~nvas~ve though these es ima es vary w~de y, rom one quarter to two-thirds. Another important feature of the pre-symptomatic lesion about which there is admittedly incomplete information is its duration. Dunn187 has done useful work on this point, based on his study of the U.S.P.H.S., Memphis, Tennesee, survey. : ' . . 188 He and his colleagues now estimate __ an average duration of aboy.t 10 years - 108- (calculated from the age-specific prevalence and incidence rates). This accords reasonably well with the estimate of Boyes, Fidler & Lock, 185 of 12 to l3 years, based on mean age of onset of both in-situ and clinical invasive carcinoma. Some of the picture presented by pre-invasive cancer of the cervix has therefore been filled in, but there are still important gaps in our knowledge. We need to know about the effect on mortality as quickly as possible since it is just possible that incidence could fall without a reduction in the death rate. To obtain this information mortality data related to numbers of examinations in populations screened heed to be collated. It is also important, for practical screening, to discover how frequently cytological examinations need to be carried out. To learn this it is necessary to note the time elapsing between the last negative examination and the first positive in as large a number of reported screenings as possible, the screening interval being varied in different groups of women. In this way it should be possible to construct a frequency distribution of the rate of progression from negative, through in-situ, to invasive cancer in the small population of women in whom this progress is relatively fast. Depending on the shape of this distri- bution the optimum screening interval could be selected. Registration of cyto- logical examination and linkage with invasive cancer registration is one way in which this might be done. There are other practical problems associated with population screening, the solution of which depends on epidemiological knowledge. One of the problems is that of ensuring that examination is offered to women at greatest risk. The evidence shows that incidence of cervical cancer increases with age; that it affects parous rather than multiparous women; that early coitus is an aetiological factor rather than parity; that there is a sharp s0cial class gradient with the higher incidence in the lower social classes; and that there are marked cultural and geographical differences. 189,l90 It is important therefore, in organizing population screening, to take steps to ensure as far as possible that parous women in the lower socio-economic groups are not only offered, but accept, examination. This means a 11 free" service and organized health education along lines planned from prior attitude studies. Without this approach it may be doubted whether merely providing facilities will in practice get to the root of the matter. In the United States of America, - 109- California State Health Department has adopted this approach in Alameda County52 and, more recently, a National United States Sample Study has been carried out. 53 In the United Kingdom an attitude study is in progress in Manchester.Slf A recent development, at present still in the evaluation stage, shows signs of overcoming the difficulty of persuading women at high risk to submit to cytological examination. This is the irrigation-pipette, 25 a plastic pipette which can be inserted by the woman herself into the posterior fornix region of the vagina. The pipette contains fixative which is eXPelled into the vaginal pool and sucked back together with exfoliated material. The pipette is then placed in a container and mailed to the laboratory where the material is centrifuged before making smear preparations. Davis has reported as much as a 90 per cent. acceptance of this 2") technique among 11 semi-indigent" women in King County, Maryland; ·· and a similar acceptance rate is being found in general practices in /1_ ·)srdeen, to give only 101 two examples.- J-· With the general acceptance of the value of cytological screening for uterine cancer attention has naturally turned to the possibility of developing automated techniques. Two may be mentioned. 6.6.2.1 A number of workers are examining the value of estimating the 6-phosphogluconate dehydrogenase {6-PGD) level in vaginal aspirate as an index of the presence of malignant cellso So far this technique has been proving too unreliable for case-finding. Though apparently reliable as regards sensitivity as a test for invasive cancer of.the cervix, it has a false negative rate in the region of 50 per 192 cent. for carcinoma-in-situ. · The false positive rate is also high, at between 20 and 40 per cent. 192,193 The value of varying the technique is now being examined. · 6.6.2.2 The development of an automatic electr9nic scanner of preparations of vaginal cells has been in progress since the 1950's. The "Cytoanalyser" has not so far been developed into a practicable instrument but work on these lines is advancing. There are numerous possible uses for an instrument of this type and more than medical interests are involved. The cost of developing a prototype instrument is considerable and there might well be a case for co-operative effort in this field. More recently, the possibility of detecting cervical cancer cells in vaginal - 11'0 - aspirate, using the Coulter Counter, has been reported. 194 being o~itically investigated. The technique is still 6.6.3 Breast Cancer 6.6.3.1 In England and Hales carcinoma of the breast in 1964 accounted for 20 per cent. of all female cancer deaths and for nearly four per cent. of all female deaths. 195 It is easily the commonest cause of death from cancer in the female and mortality has remained steady from year to year regardless of treatment. Thus the death rate for all females in England and vJales per million living (standardized) was 158 in the decade 1901-10 and in 1950-54 was 182. This trend is carried through all age-groups, as rates for 1936-1954 demonstrate. 196 Nulliparous women are more prone to the disease than women who have lactated and there is evidence that some tumours are hormone-dependent; in one group of tumours the peak registration age is reached at the ·time of the menopause and tl1is is followed by a temporary fall, though another group show a steady increase in the registration rate throughout life.197 6.6.3.2 There is no question, therefore, about there being a group at higll. risk involving large numbers of middle-aged women, at least in certain countries (some countries, Japan for example, have an extremely low prevalence of breast cancer).198 Unfortunately the results of treatment are discouraging; the corrected five-year survival rate for all stages in Encland and l"Jales, 1945-47, was only 37 per cent., though it was 67 per cent. for the earliest stage when radically treated.199 This form of cancer appears to a large extent to run its own course and to be not' greatly 200 influenced by treatment, whether surgical or radiotherapeutic. Park & Lees have estimated that, at most, treatment accounts for cure in no more than five to 10 per cent. of women. Lewison, 201 from the breast clinic at John Hopkins Hospital found little difference in survival when differing forms of surgical treatment, carried out t d .ff t t· . d d Be " R bb. 202 . 20 f 11 a ~ eren ~me per~o s, were compare • rg ~ o ~ns ~n a -year o ow-up of breast cancer at the Memorial Hospital, New York, considered the cure rate by surgery for anaplastic duct carcinomas was 12 per cent. 6.6.3.3 For a highly prevalent and lethal condition that tends to run its course whatever the treatment the question of early diagnosis, with a view to more complete - lll- eradication of the tumour and consequent improvement in the prognosis, assumes high importance. We should enquire wnether, if breast cancers can be brought to treat- ment at an earlier stage than is the present practice, by health education, frequent self-examination of the breast or by scre~ning by soft X-rays (or possibly by infra- red or ultra-sonic scanning), the prognosis is likely to be improved. Delay may occur for two main reasons: firstly, fear; ru1d secondly, failure to be aware of a small lump in the breast. The first of these reasons for delay is gradually diminishing pari passu with a rising level in general education. Better general education enables people to reason more clearly and. to plan mtional steps to meet a situation. More specifi- cally, ignorance and fear go cheek by jowl and health education, in seeking to over- come ignorance, at the same time aims to dispel the fear which can prevent a woman from consulting her doctor as soon as she notices something wrong. The second reason for delay, the fact that the lesion is smail enoUgh to escape attention at all unless looked for in some special way, needs highly organized mass scr~ening to overcome it. The methods are- costly, use up valuable resdurces~-both human and material, and involve the female public in a co-operative effort which cou*q._ .have bad, as well as good, effects on morale. Therefore, before advocating case-finding of this kind, it is necessary to examine the case for detecting breast cancer when the lesion is as small as possible. The question of the value of detecting the lesion when small may be examined in two ways. Firstly,· it is worth looking at some of the evidence for the benefits of di~osing bre~st cancer at an early stage in its course (which is linked with the grade of malignancy). Secondly, the evidence for the relationship between size of lesion and stage and degree of malignancy needs to be examined. 1. Prognosis by Stage: (a) There is a considerable correlation between stage at registration of breast cancer in the female and duration as reported by the patient (6.6-36). Nevertheless in this series of registrations in England and Wales in nearly 20 per cent. of women with a history of more .than two years the growth had neither spread nor apparently invaded the lymphatics; while in over 25 per cent. of those with a declared duration of less than one month the growth had invaded extra-mammary tissue - 112 - or produced distant metastases. Therefore length of history alone will not account for the facts; there must also be a wide variation in degree and nature of malig- 207 nancy between tumours. ~ (b) Survival: the average prognosis for breast cancer is poor: the Registrar General's Supplement finds, as we have seen, a corrected five-year survival rate of 37 per cent. and concludes that the chances of survival depend almost entirely on the clinical stage when treatment is begun and that, independently of this, neither the duration of the t~our before diagnosis nor the age of the patient ... .. 04 seriously affect survival. 2 Among others Bloom, in a recent report 205 has demonstrated, by careful grading of breas.t carcinoma according to histological type, that survival correlates well not only with stage but also with grade. The stage two cancers, for example, bear a f;ive..;year survival rate of 71 per cent. for grade I tumours down to a 26 per cent. survival rate for grade III tumours. 2. Relationship between size of lesion and stage and degree of malignancy: th . h b . 1 d 206 ~s question as een rev~ewed by Suther an • with small tumours experienc better survival rates. There is evidence that women Hat-fkins, 207 for example, in a series of over 3000 found a five-year survival of 86 per cent. when the lesion was less than 1.5 em in diameter; and Taylor end Wallace 208 have reported an inverse relationship between the size of the primary grotNth at operation and five year survival, varying from 89 per cent. when the tumour was less than 2 em in diameter to 18 per cent. when it was over 4 em. Small size, however, is not uniformly related to a good prognosis. I\reyberg and Chri~tiansen 209 considered, in a review of nearly 1000 patients, that small cancers did not have a specially favourable prognosis. They concluded that more than half of the patients who present with . ,_ tumours no bigger than 1 x 2 em are likely to die from their cancer. The evidencE! therefore supports the case for a degree of correlation, well short of absolute, betvfeen size of lesion and prognosis. Wh~t then can be said about the bverall value of early treatment? We have seen that, though littie affected by duration of symptoms, survival depends on ~e;' stage and tumour grade. The effect of age can be accounted for mainly by the later clinical stage at registration in these patients; and this in turh is associated· with the longer duration of history found in older women. There is also evidence that, to some ext·ent, the smaller the lesion the better the prognosis (though not to - 11.3 ... an extent that could be considered satisfactory). ' ' ' The conclusion is# therefore# that a proportion of patients would in fact be treated at an earlier stage if delay in duration of symptoms could be reduced; but this proportion may not be very high beci:i.tise' 56 many patients are found to have a later stage of cancer or a highly invasive type even when the history is very short. 6.6.3.lJ X-ray mammography For the above reasons it has been suggested that earlier diagnosis of breast carcinoma at a pre-clinical and perhaps impalpable stage should improve the prognosis. This is a reasonable hypothesis but one difficult to test. In addition to the development of diagnostic X-ray mammography a certain amount of screening of non- defined" pupui'ations.has been carried out. For example, Gershon - Cohen 210 and his cO-worke~s discovered 28 cancers in 1100 women over the age of 35 examined every six' months f'or· eight years. In a hosp~tal series of 2500 women with unr~lated breast pathology Egan detected .. 2ll 58 malignant growths. In order to determine the value of X-ray mammography in population screening a controlled trial on a defined population is needed, with a comparison between the results of standardized treatment in a group of women diagnosed by X-ray mammography and a second group in whom the diagnosis of breast cancer has been made in the usual way. A very large population is needed in order to provide adequately sized treatment groups. The annual incidence of carcinoma of the breast in England and Wales ~s of the order of 64 per 100 000 women aged between 35 and 74, so' that a popu:J.at_ion of this size, representing a total population of some quarter million persons would only yield individual treatment groups of 30 patients each, regardless "'' ?, .- of age, breakdown and histological type. Further difficulties in a therapeutic trial of this kind are its very long-term nature (follow-up fa~ a great many years is es::r~htial, with the consequent loss :to the trial from migration) and the large ainouht;of skill~dradiographic time and appari3.tus that needs to be available. IIdt-fever, a stil-vey o:f this kind is being Undertaken in the State of New York, in as~ocia£ion 'with the Hospital Itisu:r~ce Plart j~f Greater New York. 212 . - ·.· .. Until it can be discovered whether there is value in pre-symptomatic diagnosis in this way it wo~ld seem that X-ray mammography should be limited to its use as an adjunct to diagnosis. - ll4 .. - .Other techniques are being explored besides.· that of radiology. Infra-red photography is one possible method and it has been shown to be possible to produce outlines of tumours having an increased blood flow compared with the surrounding tissue (or vice_versa) by means of an infra-red scanning device. Another possibi- lity is the use of ultra-sonics and.this field~ too~ is being explored though it has not as yet reached a practicable stage. 6.6.4 Other cancers .6.6.4.1 · Screening for other cancers has been shown to be of value in certain groups in the population at special risk~ Of these~ perhaps the foremost are workers in certain industries, of which the rubber and electric cable industries are the most important. In the past benzidine and beta-naphthylamine or allied substance.s' ,were used in the manufacture of rubber articles or electric cable insula- ting material. These substances are now known to be highly carcinogenic, especially for the bladder. Pre-cancerous polyps and early cancers of the bladder can be accurately detected by means of exfoliative cytology. For those at risk routine cytological examination of the urine at six monthly intervals is recommended. 6.6.4.2 Exf<_:>liative cy-tology has also been shown to be valuable as a diagnostic aid in the early detection of cancers of the oro-pharynx. Dental inspection often shows small lesions of the tongue or cheek which would not normally be suspected of malignancy. However, by routine scraping of these lesions and cytological examina- tion of the material a proportion can be shown to be carcinomatous and radically exterminated. This is probably the best present use for oral exfoliative cytology. It is sometimes suggested that mouth washings or scrapings from the cheek should be routinally examined for all persons undergoing dental inspection~ but it is likely that the yield of unsuspected malignancy would be very low and the use of resources in this way uneconomic, 6.6.4.3 Cancer of the stomach is another form of malignancy where the methods of early de:tection can be employed. Unfortunately the prognosis for stomach cancer is appallingly bad when diagnosed by normal clinical means, at whatever stage. Indeed it is a paradox that. the shorter the history of symptoms the worse the prognosis in terms of survival. The best hope of improving the results of surgery is by diag- nosis at a pre-capcerous stage and this may be done either by gastric cytology or gastric photography, or both. However, the stomach is inaccessible and any kind of - ll5 .- mass screening is hardly practicable. The exfoliative cytological technique for examining the stomach requires particular skill and needs to be carried out under hospital conditions; its use is therefore virtually limited to aiding clinical di~1osis in patients with suspected lesions. It is used for screening high-risk populations, particularly elderly men, for example in Japan, where there is a high incidence of carcinoma of the stomach. Perhaps the particular use of these tech- niques will be found in a special high-risk group of persons who have already given positive results to a preliminary screening test. The haemoglobin level could be used as a preliminary test,. since persons with gastric atrophy tend to develop macrocytic anaemia. Another possibility is a tubeless gastric analysis for the presence of free hydrochloric acid, using an electrolyte-absorbing resin. For the future,. it may prove possible to develop a simple test for gastric parietal cell antibodies. 6.6.4:4 In ·England and Wales, in 1963, there were 5393 deaths from cancer of the rectum,. which accounted for 5.3 per cent. of all cancer deaths. In addition, there were over 9000 deaths from intestinal cancers, many of which will have ?..risen in the sigmoid colon. This comprises,. therefore, a very considerable cancerproblem. Although during the past 20years operative mortality for colon and ~ectal cance.r has fallen considerably, it is an unfortunate fact that survival is still about the 213 same as it was 20 years ago. The best hope, therefore, in improving the outlook may well lie in earlier deteetion (as indeed,. for other forms of cancer). Routine procto-sigmoidoscopy .. as part of any general medical examination, offers the possibility of detecting not only early invasive rectal and sigmoid cancers but also 1 . h 1 v .d 11 El d h" k. 214 f precancerous es~ons, sue as po yps. !~en a scm an ~s co-wor ers, or example,. at the University of Pennsylvania Diagnostic Clinic found, on routine procto- sigmoidoscopy of 10o6 persons,. 105 (10 per cent.) with rectal polyps. Carcinomatous change in the polyp was established by biopsy in three cases. Similarly, Hertz_ and 215 .. others reported from the Strang Cancer Prevention Clinic a finding of cancer in 2.2 per thousand out of 26 000 persons examined over the age of 45 by procto- sigmoidoscopy (i.B per 1000 in women and 3.1 per 1000 in men). 52 per cent .• of the patients were asymptomatic and 48 per cent. had only minimal symptoms (persons with ' . . '• definite preceding histories having been excluded) • More needs to be learned of the risk of ca~cer developing in polyps. The risk in familial generalized intestinal polyposis is k11own to be very considerable; the - ll6- conversion rate to invasive cancer of the various forms of sporadic polyp is at present not so well known. However, there are difficulties in offering procto-sigmoidoscopy to persons undergoing routine health examinations. In the first place there is difficulty in ensuring that faeces do not obscure the view at the time of examination; and, in the second place, there is reason to believe that the nature of the procedure may deter people from attending. Whatever the difficulties may be at routine health examinations there is little doubt there is good reason to advocate routine procto- sigmoidoscopy in persons at risk who are in any case undergoing medical examination at clinics or in hospital. 6.6.4.5 Lastly, the importance of searching for early pre-cancerous and cancerous skin lesions needs to be remembered. This is particularly true in countries where the sunlight is strong and the population largely of Caucasian stock and employed in agriculture. 6. 7 Diseases of the e're Chronic elaucoma It is only in the last two decades that chronic open angle glaucoma, or glaucoma simplex, has come to be recognised as a separate nosological entity from angle closure glaucoma. Chronic simple glaucoma, which for practical purposes only occurs after the age of forty, has been said to at tack 11 like a thief in the night" because of the long period during which physical signs gradually progress and the perhaps equally long period when subjective visual changes are minimal and only detected by 216 careful examination. For all glaucoma patients registered as blind,Sorsby found only 0.4 per cent. below the age of 40 at registration and 65 per cent. were between the ages of 60 and 79. Glaucoma (both acute and chronic) accounted in England and Wales, for the period 1951-54, for 13.6 per cent. of registered blindness, or 4200 persons. There is reason to believe that registration of blindness is incomplete and it is likely that there are far more persons in England and Wales either blind or with impaired eyesight due to glaucoma. There is a considerable literature on the prevalence of chronic glaucoma and its distribution is apparently world-wide. The accepted treatment of chronic glaucoma is by miotic drops administered daily - 117- for an indefinite period~ starting if possible in the early stage of the disease. If interference with vision becomes ser;tous an operation aimed at increasing the aqueous outflow is carried out. A commonly rec~mmended treatment consil?tS of pilocarpine hydrochloride (0.5 - 2 per cent. solution) administered two to four times daily. 217 One difficulty is that the value of this medical treatment has not been effectively evaluated; it is also an uncomfortable treatment which can be dangerous in that it severely limits night vision~ so that there must be real doubt about the degree to which patients follow treatment instructions. Established chronic glaucoma is diagnosed by finding characteristic visual field changes~ cupping of the optic disc and retardation of the normal outflow capacity of the aqueous fluid~ as measured by tonography. A rise in intra-ocular tension is regarded as the usual accompaniment of chronic wide angle glaucoma and this is the accepted sign by which early glaucoma is detected. Intra-ocular tension is measured by tonometry~ the Schiotz Tonometer being the most popular instrument used in screening. The principle followed in screening is that a rise in intra-ocular tension is the first sign of early glaucoma~ preceding the onset of other signs by some years (perhaps from 10 to 20 years). A population is examined by tonometry and perhaps six or seven per cent. are found to have pressures in the higher range (over 25 mm, of mercury). Of these~ about two pe.r cent. are considered to have probable glaucoma~ with an intra-ocular tension of over 30 mm of mercury (e.g. Glaucoma Program Guide~ pp 10-11). The U.S. P.H.S. Monograph No. 67218 on screening for disease illustrates the bimodal distribution of a disease attribute and takes the intra-ocular tension in glaucoma as a possible example~ showing separate normal distributions for diseased and non-diseased population. There is no good evidence that this is so and population samples of intra-ocular tension that have been recorded 20 .. 21, 219 do not support the bimodal model. Evidently, as with height and weight (and probably blood pressure and blood sugar) there is a continuous distribution of intra-ocular tension, with the probability of glaucoma increasing in the higher range of pressures. As with other conditions like diabetes and arterial hypertension, therefore, the "borderline" problem occurs~ posing the question: "does ocular hypertension indicate an early, pre-symptomatic, stage of chronic glaucoma?"· - ns- There is another question, too, which does not arise with diabetes or high blood-pressure: "does chronic glaucoma occur in the absence of a rise in intra- ocular tension? 11 • There is an authoritative body of ophthalmological opinion which 220 answers this question in ·the negative (e.g. Goldman). Howe rer, there are those who believe this does happen221 and a recent survey in the Welsh Rhondda20•21 supports this view (though admittedly the numbers are not great enough to be statistically significant). In the Ferndale, Rhondda survey, 13 cases of chronic simple glaucoma were found out of the population examined (between the ages of 40 and 74 years) of 4246 (92 per cent. of the total Ferndale population); this is a total glaucoma prevalence of 0.28 per· cent. Six of the 13 cases vrere already known glaucoma patients and only seven were n~wly discovered. All persons underwent not only applanation tonometry but also examination of the optic fundus and a one in three random sample visual field perimetry as well. Glaucoma-was diagnosed by finding a characteristic visual field defect with optic disc cupping. In addition, seven further suspected glaucoma patients had intra-ocular tensions of less than 21 mm of mercury and were .·,. diagnoseq on the basis of optic disc and visual field defects alone. Of these seven cases, two have since been observed to have raised pressure. Adding these seven low tension glaucoma cases to the 13 found by tonometry raises the prevalence rate for chronic simpl~ glaucoma in this population to 0.43 per cent. which is low compared with other recorded rates but similar to the findings of Stromberg. 219 If these findings are confirmed on a larger sample the inference is that the early detection of glaucoma by tonometry alone would need to accept a false negative rate of the order of more than 50 per cent. If this vvork is substantiated it will be necessary to search for other, more satisfactory, methods for early glaucoma detection. Visual field screening, which only detects patients when undoubted early clinical glaucoma is present (though its course, before subjective interference with vision begins to be noticed by the patient, !nay well be as long as 10 years) is an obviously attractive screening technique. Unfortunately there are two important drawbacks: firstly, the plotting of the visual fields, even with a new electronic flash apparatus (the "Globuck") takes up to five minutes for each examination; and secondly the number of false positive examinations found by this technique and needing full ophthalmological investigation is unacceptably high, - 119- though mapy non-glaucomatous defects were fonnd' ·by this technique (13 per cent. in the Ferndale· Survey though this was fonnd with the Freedman not the Globtick Screener) •.. Apart · frOm :'the possible need to look for other methods of diagnosing chronic glaucoma early there is the practical problem for a.riy commnnity of the ability of the lodal eye s~rvices to meet an increased load of examinations. 'Ihe Ferndale stl:f.ve;y- (the only. survey~ so far as the writers are aware~ carried out on either a complete defined population or a randomly selected popuiation sample~ and therefore representative of that population) has demonstrated a prevalence of 0.43 per cent. confirmed chr9llic sample glaucoma~ (prev:l,o:u~ly diagnosed.plus newly discovered)., a ... - :. ' .. co~~~e~ab~ )ower rate .. than that reported from other~ non-::representative ' :~: " . .. - ' . . i ·:· ~ : . . .. . ··' . • ·. . . • . : . • . ... . . . - . • population surveys. .Qne reason may well, be tl:la.t the rate is spuriously raised in .-. . ... . ~- .. most detection programmes by weighting with persons having an increased likelihood of glaucoma {e.g. relatives of glaucoma patients). Another~ artificial~ factor of defini tioh which may acc~nnt for the low prevale~ce fonnd at Ferndale is the use of the terni "ocular hypertensicin" by the authors of this survey. This in itself would tend to decrease the prevalence compared with other surveys. In contrast to this low true p~valence rate for glaucoma fonnd by tonometry., the false positive rate was high.,· 8.6 per cent. A rate of this order of persons needing further investi- gation by an .ophthalmologist would throw an impossible strain on any conn try's eye serJi6es. Perkins,. 222 for example, has calculated that in England and Wales this cbUid ,.inean a load ·of 4500 persons ·to be investigated by each consultant ophthalmologist Assuming· that by some means all the patients with early glaucoma in a population have been diagnosed, the que~tion of treatment next arises. We have. seen that the~e is doubt about the efficacy of medical treatment and that -'ther~ is ·a 'ha;rilf'l.ll ~lem~nt. Work l·s urgently needed to help decide on the two poirits: (~) ;ci6 pati~nt~ :i.n fact do as they are advised? and (b) if they do this~ (i) does it prevent them developing more advanced glaucoma? and (ii) has it deleterious effects, e.g. do those on miotic drug treatment suffer more accidents than the rest of ·the ·popUlation?. It might well be worth assembling all the evidence on the value of the medic.aL ··treatment of. early chronic glaucoma and deciding on the evidence whether a randomized trial of treatllient is justified. - 120- Lantly, there is the problem of' the "borderlines", those persons with intra- ocular tensions at the high end of' the distribution without symptoms or sign ·of' glaucoma. 222 In some surveys persons with a diminished aqueous outflow, as measured by tonography, are regarded as belonging to the "glaucoma" group and are excluded f'rom_the borderline group. In the Ferndale Survey only those with optic cupping and a field defect were regarded as glaucomatous. The "borderlines" com- prise a gro~p of' individuals to whom a randomized trial of' treatment can ethically be offered; there is real ignorance as to: (a) the risk to the individual of developing chronic glaucoma and (b) whether prophylactic medical treatment will lower what risk there is and, if so, to what degree. A trial of' this kind is in progress as part of' the Ferndale Survey but of' course it will be several years at least before significant results can be expectcdo Loolclng to .the future, it would be good to see work in hand to discover more about predisposing factors in chronic glaucoma Besides the relationship with a raised intra-ocular tension we know there is a familial trend; there is a possibility th~t there may be some correlation with iris colour. This is little enough to go on. Chronic glaucoma is one of the conditions in which useful correlations with other physiological variables or with other morbid processes might be revealed by a wide- f'ronted prospective study on the lines of' the TectUnseh experiment, where a complete J.. • popu1ation _is followed over time and its <Usease experience correlated with recordings of as many variables as possible. There is obvious scope for biochemical and digital automation in ma,k.ing leasible this form of large study. On ~he. tech,p.ical side, it was found in the Ferndale Survey that, while the coefficient of variation of the Goldman Applanation Tonometer was greater than that of thr::l Sch5.o-t.z instrtUne.nt, neither was considerable in terms of mean pressure levels. Applanation tonometry was more acceptable tq the population that the Schiotz technique~ The Mackay Marg tonometer is commercially available and it is claimed that this c~~ be used without anaesthetizing the cornea. If reasonably accurate readings can be obtained in this way it would clearly be a great advantage~ since using a local anaesthetic on the cornea inevitably adds some element of risk to the procedure. - 121 - Until the answers to some of the problems discussed have been provided by experimental work, what can usefully be done for the detection of chronic glaucoma? It seems that tonometry is the only practicable screening method we have at the present. In order to avoid an uneconomic and unbalanced use of the eye services available in most advanced countries· it would seem wise for the present to concent- rate on those at highest risk~ i.e. persons over 40 related to patients with known glaucoma,· since a familial tendency to glaucoma has been demonstrated. 223 What is done· in ·practice will depend a· great deal on the type of medical care prevailing in the area in qtiest{on.- · In the United States of America much glaucoma screening is at present befug carried out and efforts are being ·made to avoid the "drive" and to try to run a continuing ·:programme. In the San Jose, California., programme for 1961, for example 3286 persons were examined and, of these, 27, .were found to .be suffering from glaucoma at a cost of $ 175 per case found. This is compared with the estimated cost of $ 1200 per year needed to support one blind person under the United States "categorical'.'aid ·program :for the bifud". This report states that "it is apparent t~t eariy case di~ScQvery is of significant econOrilic benefit to the taxpayer". But . this is only true if . earf:Y detectiori arid treatment do in fact prevent or delay blindness and this still needs to be ascertained. In countries where health services are unified it is even more mandatory than in countries where me<tle.al. oar~ ,is _largely financed by the individual, that the facts .be ascertained befo:re ;·fo:rm\llating a country-wide policy and undertaking universal .glaucoma detection ... While glaucoma accounts for 13.6 per cent. of the registered blindness in England and Wales, cataract is responsible for 26.2 per cent. and senile macular lesions for 21.6 per cent.. "Congenital abnQrmalities" accotmts for a further 4.2 per cent. of blindness, 38 per cent. of it in children under four years. Much of this blindness, as well as impairment of vision short of blindness, is remediable. In infancy the condition of amblyopia, present from birth, can lead, unless corrected, to blindness in one eye. Detection of the condition needs to be carried out early in life, usually'before the child can read, so that orthoptic measures may be instituted. - J.22 -~ .. In California, for example, the public health nurse service is used to provide information and instruction to parents on screening young children at home :for amblyopia. At the other end of life impairment of vision from cataract, senile macular degener~ti.on or other cause can be detected by a visual acuity test and by examining the lens with a spot-light.. Cataract is of course. remediable by surgery and senile macular degeneration can be helped by pr~viding the patient with a visual aid. As · with glaucoma3 the medical care aspect needs examining carefully. The proportion of hospital beds available for eye surgery.is usually small and waiting lists for cataract operations long. Examination in the elderly.. as with the young, might well be carried out by the home nursing service after special.-:training .. 6.8 Diseases of the c.U'inarv tract 6.8.1 General Urine specimens are easy to obtain in screening projects and urinalysis is simple and can be indicative of a number of conditions.. of which diabetes and kidney .· ·· disease are the most important; though as an index of sub-clinical icterus in epidemic jaundice uribilinogen examination can be extrem3ly useful. paragraph is concerned with the urine in renal disease only. However, this rn t:he :Baltimore screening study a prevalence of albuminuria of 8. 4 per thousand parsons examined was found. In 4. 7 per thousand this was confirmed by diagnostic . tests~ and of these only 0.8 per thousand had been previously unknown to the patient. In his five year follow-up Wylie 38 found this test of high prognostic significa:n,ce !, . ·:·.,· .. , 6.8.2 ·-.Bacterillri:a , ·_, . _;._- '· ·· · In screening for urinary system disease the aetiology of the condition sought is of great importance from the preventive point of view.· While patients with chrtn1ic'o.g:tomerulo.-nephritis can indeed. be helped (particularly, perhaps_, by guarding against recurring streptococcal infectiQiiJ., ,, interest has in the past decade. turned to chronic pyelo.:..nephri tis and specially to its possible source in recurrent . attacks of urinar,r:Hifection, whether accompanied by symptoms or asymptomatic. Here agaJ.n is the classical screening situation: a potentially lethal. condition preceded by a. - 123 - possibly latent and.reversible stage. What is not yet certain is-the role played by asymptomatic_ bacteriuria infections, particularly in the female, both in child- hood and adult life. The natural history of these infections (which are certainly common) still need3 to be worked out in relation to the development of kidney damage, hypertensive disease and abnormalities of t.he urinary tract like vesico-ureteric reflux. Clearly the matter is potentially one of great public health .importance. Accurate prevalence figures for pyelo-nephritis are not available. The College of General Practiticners survey 5l showed a patient consulting rate of 13 per thousand persons for urinary tract infectj,ons. The possible precursor of much pyelo-nephritis, asymptomatic bacteriuria in pregnancy, has been found· in a number of surveys to be prevalent in about 5 per cent. of pregnant women; the inc1dence. of acute pyeio-nephritis in those women with . persistent bacteriuria· has be€m discovered to be ten times greater than in women without init±B:l. bacteriuria. 224 · Miail ·and Kass and theii- colleagues 22 ') have found a bacteriuria prevalence of 4.4 per cent. in the general female population aged over 15 years in Jamaica, but re;..examinations show that, while the prevalence remains the same, d~fferent women are affected at each survey.· ·-So far, the evidenc~ for an aetiological connection between the asymptomatic bacteriuria and later chronic pyelO- nephritis i:l,ppears to be incomplete. What is needed is a prospectj_ve- tri.al treat~ent; thoughthe·evidence for the immediate value of treatment in pregnancy, in preventing attacks of symptomatic urinary infection, prematurity and foetal loss, is now con- siderable~ suggests that untreated asymptomatic bacteriuria in pregnahcy i·s- accompanied by increased foetal loss and low birth weight; whilst . . -. . . ShouJ.d these findings be Sl:l.bstantj_ated thera are grobnds ':for -f1id.espread cai5e-finding, both in the female child and in pregnancy. Ttie'ineans fordoing this a::f.e already being worked out; the·triphenyl tetrazolium chloride (T.T~c"f screening test, described by Siriunons and Williams 227 shows promise· of p:i--oviding a useful screening test which can be used in the field, prcrviding facilities for refrigeration of urine specimens are available, Catheterization is· not required and indeed is contra-indicated ; 228 a "catch" mid-stre.ain :specimen of urine:. pasSed into a sterile container and immediately cooled by refrigeration; :is all that is required:, Doubtless .. n wider evaluation of the T.Jr. 0 •. test and the - 124 - possible development of improved tests is needed, since other workers have been unable to obtain comparable results in comparison with culture methods. 6.9 Rheumatic diseases 6.9.1 Rheumatoid arthritis "Arthritis and Rheumatism" accounts for a large part of the illness which presents to the general practitioner. In the British general practice enquiry51 it accounted for a patient consulting rate of 65 per thousand, (not including acute rheumatism) second only to acute nasopharyngitis in order of prevalence and of course causing far more disability. Population surveys for rheumatoid arthritis and for the presence of rheumatoid factor have been relatively extensive. 229 The prevalence of definite rheumatoid arthritis in different sample populations in Northern Europe varies from 2-3.5 per cent. for persons aged 55-64, with an additional 1-6 per cent. of persons probably affected. In North America the Tecumseh, Michigan, project finds a prevalence rate in all persons over the age of six years of 0.39 per cent. definite and 0. 85 per cent. probable rheumatoid g,rthritis. 230 "231 Tne world-wide distri~mtion of this dise2cse i12"s not j•et been defined by clinic2.l surveys and the ve.lue of international morcalit2r data is limited. However, serological testing for rheumatoid factor shows its distribution to be wide- spread in both Africa and Asia; though the exact relationship of the presence of positive tests (Latex fixation and Waaler-Rose) to the development of rheumatoid arthritis has still to be determined. There is therefore good evidence of a high degree of prevalence of rheumatoid arthritis (its diagnosis being based on the American Rheumatism Association criteria). This provides a strong prima facie case for early detection. However, the other criteria for justifying screening are not fulfilled, notably: (1) there is no clear pre-symptomatic stage (despite the somewhat obscure relationship with positive tests for rheumatoid factor): (2) there is.no specific treatment. The best accepted treatment for the early stages of rheumatoid artl1ritis is to protect the affected joints from excessive use. Case-finding before the patient presents with disability does not therefore appear to be at present justifiable. More needs yet to be learned about the aeti~ logy of the condition and then, perhaps, about its early treatment. Apart from -.125- basic work on a. possible aeti.ological factor {such as a virus;; as currently reported to be suspected by workers in Birmingham) continuing survey work can add to our . knowledge. of the rela,tionship of the various faqtors so far uncovered to the disease process itself. A possibly useful contribution to survey work might be a concomitant continuing population sur.ve:r of hiochcm:;.cal va:L'iables in blood. This would be feasible \'lith automated laboratory i'ncilitj.es. In this way early deviations from the "normal" biochemical pattern might be noted in relation to early rheumatoid cha.l1ges. Gout Although gout is not a highly pr·evalent condition {only 0.8 ·patients consulting per 1000 persons in the British general practice survey (though up to 4.5 per thousand in the prevalence studies quoted by Kellgren)229 .it is a condition {a) with a recog11i~}l,ble pre-clinical st,age (hyperuricaemia), (b) where there ia a;;high-risk group in the population (r.::latives of gouty probands) ancl. {c) for which an effective treatment exists (lo~i purine diet,. avoidance qf. exQess dietary fat and alcohol and the l1Se of uric~_suric dr.ugs). There is therefore a good case for screening for this condit~on, I:JOSsibly in a selective manner. Gout: e,ppen.rs to be one of the few clinical conditions which is with little doubt transmitted by one or, at most_, two genes and where a bi- or tri,..mooal distribution of se!'Uffi uric· .acid levels is found .. in n population. Kellgren 232 has suggested that an intermedia~e peak in the distribution curve_, at 6 mg per 100 ml_, may represent ··the heterozygotes in t.he population.. the gouty homozygotes having a peak distribution in the gouty. range-" at about 8.5 mg per 100 ml. A ·sub-·committee of the C.I.O.M.S. Symposium held in Rome i...'"l 1961 agreed that a serurn uric acid levels in males of over 7 mg per 100 ml and jJl females of above 6 mg per 100-ml should be regarded as one of four criteria fo:;." the G.iagnosis of gout • . It seems likely that .those at genetically high risk may be brought to the clinical level by variations in the habits of the population. A high purine diet, excessive intcl.l;:e of dietary fat and alcohol (as experienced by .well-to-do city dwelle~s as comp8~red with poorer country dv-rellers) and the use of drugs such as salicylates in lo>'l dosage_, _hypotensive &.gents <md diuretics may promote a higher prevalence of gout; whilst restriction in the diet and in the use of drugs (as - 126 - occurred during the 1939-45 war) tends to cause the prevalence to wane. Biochemical screening of the uric acid level will indicate those persons at high risk and the risk can then be lowered by alteration of the diet and advice on the use of drugs. In Great Britain a survey by Lawrence and his colleagues233 showed a prevalence of five and a half per cent. of males over the age of 15 in the rural area of Wensleydale with serum uric acid values of 6 mg per 100 ml or more. Among male inhabitants aged 55-64 of the town of Leigh, in Lancashire, the prevalence rate of men with serum uric acid levels above 6 mg per 100 ml was as high as 18 per cent. Thus selective screening would give a high yield of persons to whom useful advice could be given. 6.10 Mental illness On grounds of magnitude the early detection and treatment of mental illness in a community should have a high priority. In the British survey of general practice51 the r"'. te fo~' p::"tients consul tin;; their family doctor for psychoneurotic disorders ranked next only to acute nasopharyngitis, rheumatic disorders and bronchitis in frequency; the annual patient consulting rate was 46 per 1000 patients and the total consultation rate 166 per 1000, or more than 4 per cent. of the total consultation rate for all causes. In a survey directed specifically to mental 234 illness in general practice, Shepherd and his colleagues found, in a random sample of London practices, a total patient consulting rate for psychiatric morbidity of 140 per 1000 persons (176 per 1000 females and 98 per 1000 males) over a twelve month period. Of these, neuroses accounted for by far the greatest number (89 per lbOO) whilst'psychoses were diagnosed in six per 1000 persons. For women, psychiatric disorders ranked as the third commonest cause for consultation after respiratory and orthopaedic or traumatic conditions, in that order. A review of 50 surveys carried out in different parts of the world showed rates for all mental disorders from one to 370 per 1000 population.. The differences are mainly due to variations in the techniques employed in the surveys. It is clear therefore that mental ill health constitutes a great burden on society and it is reasonable to aim at early detection and treatment as a buttress to conventional methods. However, when examined more closely, difficulties appear. - 127 - ·. Ili;'contrast to most other branches of medicine it has not so far been possible to reach an aetiological classification of mental disease and for this reason there are very few specific treatments (neuro--vascular syphilis and phenylketonuria are examples of exceptions). Thus there is no clear agreement about the diagnosis and treatment of much overt mental disease. "Treatment" may consist of drug therapy, as in depressive illness, but it also includes the management of the total social situation of the patient and his immediate family group. In certain conditions it is not clear what should be the "best" treatment of the whole situation. It may be in the immediate interest of the patient himself to carry on as a member of the community, possibly remaining undiagnosed. On the other hand this situation may lead to breakdown of the family situat'ion, when diagnosis and removal for a time from th.e·sbcial.environment is the treatment of choice. It is therefore not necessarily beneficial for the psychoneurotic, say, to be the subject of early diagnosis; his best treatment may be that of continuing the struggle to integrate himself with society. After a fashion, this is the same problem of borderline diseas~ met with else- where; the problem of what to designate as "disease" in need of treatment, allied with the question of whether early treatment will be beneficial. When all this has bee8 said ;it still appears t_J;a.t there is a group of persons with relatively mild depre~sl:ve i:LlP:ess who could benefit from,,.~,a.rly diagnosis and treatment with, among other method~r modern ataractic drugs. T~ere is a shortage of evidence of the value of various treatments in depressive illness, especially in its earlier stages. The Medical Research Council has in progress a trial of treatments (including drugs and electroconvulsive therapy) of more advancied..depression; 235 but evidently a great deal of mild depressive illness is not refe:t'red for specialist advice and is treated by general practitioners with ataractic drugs. A trial of treatment of this milder illness might be valuable and could be combined with a survey of early, unreported, mental illness in the community. Work has recently been published on ascertained mental disease in general practice (as reported above, Shepherd and others,)234 but there is perhaps need for further work to identify persons who might be helped b;j early treatment (including social adjustment within their family group) and for a trial of such treatment. This would entail a trial, inter alia, of the benefits of using social workers in the pre-breakdown stage of psychiatric disease. - 128- Population surveys of psychiatric illness in the community are being undertaken (e.g. by the Medical Research Council in Camberwell and South Wales) but we do not kno\'r of trials of early treatment of the kind described though they may of course exist. 6.10.1 Mental retardation Examples of specific conditions where screening is carried out A detailed critical review has been made in 1964 by Gruenberg of a number of community wide surveys on mental retardation. 236 Remarkable variations are found between and within age-groups. At the age of peak-prevalence, about 14 years, Gruenberg found a 10-fold variation from about 1.0 per cent. to about 10 per cent. Such variations may be largely accounted for by differences in the limits and scope of studies, diversity of defL~itions and availability of measuring instruments. Some type of screening of school-age children has been widely carried out in count- ries with a well-developed school system, based largely on the criterion of ability to keep up with the classwork, but frequently supplemented by the use of IQ tests. Where suitable services exist, the children suspected of retardation are often referred to school psychological services for further testing and investigation of the possible cause of the retardation. Stress is increasingly being laid on the importance of using batteries of tests combined with socio-psychiatric investigation since failure to keep up with the school system has frequently been found to have causes other than low level of intelligence: moreover most specialists now doubt the prognostic value of IQ tests used alone. An example of the screening of a school-age population is that carried out by Jae~i and Jaeggi in Geneva237 as a preliminary to the re-organization of the Cantonal socio-medical educational service for the mentally retarded. The primary screening was carried out through consultation of institutional, educational and medical registers and reports. It was considered that practically all suspected cases could be found by this means, these reports being very thorough and kept up to date. (An indication of the reliability of the data is given by the fact that in 1964 10 per cent. of the total school population of the Canton was seen by the socio- medical service). A detailed questionary was then completed for each suspected case providing information on social, psychological and medical conditions as well as - 129 - educational achievemen.t. It .Has found that 0.8 per cent. of the children of school age l'fere· :in need of special social, medical or educational service. 6.lo.2 Phenylketonuria (PKU) PKU is one of the few disorders leading to mental retardation of which the cause is known - a genetically transmitted me~abolic abnormality. It. the defect is detected very early in life mental retardation can be prevented or favourably modified through a special diet low in phenylala~ine. A bacterial "inhibition assay" screening test for blood phenylalanine (Guthrie test) was used routinely by. State Health. Departments in the United States of America in 1962 and ·1963 on more than 400 000 newborn · infants born in about 500 hospitals. The frequency· of PKu was found to be· one per 10 347. Such routine testing has now become mandatory in some of the States (e.g. New York, Massachusetts). It is also used widely in the United Kingdom and.some.other countries. cannot be used in areas without a wide network of health services. Obviously it 6.10 .. 3 Galactosaemia This disorder is also caused by an inborn error of metabolism, but if untreated it leads to early death, unlike PKU. A diet loN in galactose prevents the develop- ment of the clinical condition which includes severe mental retardation. Promising attempts are being made.to attain simple screening methods, e .. g. paper strip tests. By quantitative biochemical screening procedures it seems possible to detect, with some measure of accuracy, the heterozygous state of galactosaemia thus making it possible to detect paren·ts who might produce galactosaemic children:38 . 6.11 Anaemia 6.11.1 Introduction There are, of course, many causes of anaemia but iron deficiency is by far the commonest ana'only this form of anaemia is· considered here. This deficiency is essentially ·~·imbalance between intake and absorption, and excretion or loss in other ways; it can be prevented by correcting either deficient intake or abnormal loss. Thus one of the priricipal criteria :for screening is met, in that the condition can be prevented by treatment (whether primary or secondary; early or late). - 130- A second feature of iron deficiency anaemia in connection with screening is the test for its detection. Measurement of the haemoglobin level is one of the few examinations which directly estimate the variable in question. There is therefore no doubt about the significance of what is being measured (as there is, for example, in measuring intra-ocular pressure as a test for pre-symptomatic glaucoma); though the other difficulties of definition and the variability of the method and the observer are the same as for other tests. 6.11.2 Prevalence Iron deficiency anaemia was the subject of a HEO Study Group in 1958.239 After considering the haematological data on apparently normal persons throughout the world the Study Group adopted the following criteria for haemoglobin values below which anaemia could be considered to exist: TABlE. 6.11.1. Age in Years Sex Hb g/loo ml 0.6 - 4 10.8 ) ) 5 - 9 11.5 ) 11.5 ) 10 - 14 12.5 ) Adults Male 14.0 Female 12.0 Pregnancy 10.0 i I The Group's Report reviewed the available studies of prevalence and considered that more of these studies were needed. It pointed out that in some parts of the 'I'Torld anaemia constituted a major public health problem. An investigation in Mauritius had shown that 50 per cent. or more of certain groups of the population were probably affected. In Western countries the prevalence is clearly much lower. 240 Using the above criteria, the survey by Berry and others of London housewives in 1951 showed nine per cent. of the women to have haemoglobin levels of less than 12 g - 131- per_lOO ml. Ki~patriok Cilld Hardisty 1 2·41 in a study in South Wales and North Englandc of men and women in age groups respectively of 35-64 and 55-64, found 14 per c~nt. of the women with haemoglobin levels of under 12 g per 100 ml and three per cent. of the men with levels of less than 12.5 g per 100 ml. Kilpatrick and Hardisty's female group were over the age of menopause. Iron deficiency anaemia is, of course, most prevalent in women who menstruate and it is likely that women in that age period would be found to be anaemic even more frequently. For this reason iron deficiency anaemia is a condition specially suited to selective screening of women between the age of, say, 20 and 44. More need·s to be learned, however, about the criteria constituting anaemia and the relationship of haemoglobin levels to symptoms. Though we may have a reasonable idea of what is clearly anaem:i.a .. we do not know much about optimum levels of haemo- _g:).obin. To this purpose there is need pf' further survey work: (a) to ascertain with the greatest possible accuracy the distribution of haemoglobin levels in a probability sample of a large population; (b) tq relat~ carefully, and without observer bias,· symptoms to haemoglobin leve],.s in iron de1"ici~ncy anaemia, as well as symptoms to the effects of trea"t;ment; (c) to discover the cheapest and most effective way of treating and preventing the recurrence of iron deficiency anaemia. Studies in this field are at present in progress under the auspices of the M.R.C. Epidemiological Research Unit (South Wales), and there may of course be other studies under way. 6.11.3 Methodology Where samples of venous blood are being withdrawn for other purposes, there is advantage in centralising the estimation of haemoglobin, so that it can be carried out on an accurate calibrated photo-electric instrument. On the other hand, collection of venous samples in tubes specially for this purpose is expensive and in many field circumstances it may be both cheaper and more convenient to estimate the haemoglobin on a sample of capillary blood. The simplest and cheapest way of screening for anaemia is the Phillips-Van Slyke specific gravity method (which is ---132 - used by-the British National Blood Transfusion Service for screening potential blood donors). However, its accuracy for use in general or selective screening is 6Peri 242 to question. A recent trial carried out by the M.R.C. Epidemiological Research Unit (South Wales) /18/ has shown the American Optical Company's Spencer haemo- globinometer to compare well with the M.R.C. Gray Wedge, the EEL photo-electric photometer and the Sahli method. The A.O. haemoglobinometer has the advantage that it works without· dilution of the blood sample; ·a drop of bloOd is simply placed on a special slide, haemolysed with a stick impregnated with saponin, and covered with an optical coverslip. The slide and coverslip are -optically prepared so as to present a standar4 depth of blood between them, th~ colour of which is compared with a st~dard. -:Another, and reasonably accurate, ·screening technique for capillary blood is the micro-haematocrit method. This method, of course, requires a centrifuge and is therefore best labor.~tory-based. 6.11.4 Screening Screening for anaemia is carried out, with highest yield, as we have seen, on women in the menstrual age ·groups. W.Omen in the· child-bearing age very largely attend maternity and child welfare clinics and this offers an excellent venue for screening. 'I'here are, however, the other causes of anaemia which may be picked up by routine haemoglo~inometry, such as other blood diseases, malignant.dis~ase, peptic ulceration and other gastro-intestinal disease, rheumatic and renal disease. Routine haemoglobin estimation is therefore probably one of the most profitable ·S:6fe·Eining tests f~r all ages in pointing to unsuspected disease; Jungner and Jungner243 found 1.4 per cent. of 30 000 persons in the ~rmland project were con- firmed as being ana.emic, the commonest previously undiagnosed condition. Other- than- part of a case-finding population screening programme, haemoglobino- metry is clearly a most useful addition to the normal examination ofpatients in a general practiitioner's work.· It can easily be ~arried out by an ancillary helper. Fry~_lj-~ has reported, for example, the results of routine haemoglobin estimation in an Outer London general practice. He took a 10 per cent. sample of all his adult patients and found 18 per cent. had haemoglobin values of less than 12 gjloo ml which wa.S ten timE3s the rate of "clinical1' anaemia in his practice. · - 133 - 6.11.5 Conclusions In conclusion, therefore, anaemia is probably one of the more acceptable con- ditions for screening that at present exist; it is highly prevalent~ can be sufficiently accurately detected· and, when due to primary iron deficiency, responds excellently to treatment. The haemoglobin level is also a sensitiv~ index to a number of other conditions, of which anaemia may be one of the earlier signs. Naturally, once anaemia is discovered, it is of the fil?st importance that a complete ~at.ological ~vestigation is then carried out, leading to a definiti~e diagnosis. 7. METHODOLOGICAL TRENDS IN SCREENING n1 the foliowing1 we will consider some common procedures and tests which have been applied in health screening progr~es, or in similar screening surveys in a population. There seem to be certain intE!rest.:i!}g t~nds in the tecpniques which may dominate in the futu~,.in contrast. with the wide variations in acting and think;Ln.g .. today .. 7.1 Clinical or technical screen~ Firstly, there is a strong trend to use automated methods, particularly for chemical tests. The. ~xtensive use of laboratory methods has, however~ ·been- criticized., and the value of impersonal and highly standardized methods may be questioned. Only if the resources are limited, can the medical value justify the cost aJ;ld,. effort P.f. automated or mechanized. tecl:miques with high capacity. ' -·· ' . - ... - '·" . .· . It can be noted that ~:volution may occur stepwise, ·when laboratory tests._ and certain . ,. . ··- -·' . . . . . ... ·· ' clinical examinatic;>Il$ may be applied in. "drives", but are later used as special facilities at the praotitioner's disposal for periodical health examin{ltion. Another obvious trend is the increasing use of data-processing machines. '!his matter will not be considered here. Health Services will adapt these principles used for sick eare, but in some respects - registration of data and storage - the work in- health examinations may lead the · evolution. 7.2 Participation by doctors The most essential and decisive factor for the extent and organization of health screening has been whether or not medical examination by a physician has been included in the survey. It is debatable whether a physical examination is to be regarded as a screening procedure. Depending on this., we can divide health investigations into two groups. One is a complete health examination., which includes a doctor's physical examination; the other group consists mainly of a ba~tery of clinical and laboratory tests, when the contribution of physicians is limited to evaluating certain tests., or is com- pletely lacking. Inclusion of the doctor's findings., ·as well as his interview and evaluation~ greatiy widen the scope of screening, and make it comparable to a general medical ex&nination. As an important part of the final medical conclusion, the examination by a physician is extremely valuable .. The signal factor is the time that is devoted to the check-up. Simple inspections and fast examinations, as well as c~prehensive physical examinations., have been practised. 7.2.1 The doctor's physical examination This examination is 'expected to be a complete physical examination, with inspection., palpation and auscultation, and should include measuring· the blood pressure and rectal palpation .. The_ physical examination is sometimes carried out as one of the first steps in a health screening programme, together with checking the questionary and ordering special tests, which are not carried out in all patients. made bytbe doctor after all tests J:lave b~en completed~ The final evaluation is Obviously, ·there is not always a sharp distinction between the technique used for physical examination in health and in sick care. Especiallywhen seeking· such conditions as malnutrition, the technique may be similar to methods-used for children'' s care in general, as seen for instance from the report by a WHO EXpert Committee on malnutrition. 245 It can be anticipated that some. of the work, which today. is done by the doctor:, may be taken over by technical personnel with special equipment. -By means of proper organization, many. clinical tests can be done more easily., and on a fairly large sc~e. This will make the goal easier to reach: always to have a complete medical examination. Cance.r det~ction is sometimes one of the main objects of health screening. This is despite the_ fact that no guarantee of freedom from malignancy can be given, . -', . .. . ., :· . .. ·~ ' . and that the poss_ib::ti.It~es-o:f detectinf!; ·cancer--- in an asymptomatic stage - are extremely limited. - 135 - Several suggestions·.· have been made about how to perform the cancer-detecting examination most effectively. Reference is made to,· for iilstance, the well known description byDay. 246 In a WHO Expert Committee's R;port, 247 the following !"" examination is mentioned: inspection of the entire skin area and all accessible body cavities, urine tests, chest X-rays and proctoscopy;. in males prostatic examination, in females cervical smears and palpation of the breast. Sputum cytology, gastrointestinal radiography, blood counts, colposcopy, and possibly mammography may be added. History-taking by doctor or specially trained medical staff • The history is very important, and can be obtained by proper questionaries. It has Qeen reported from many investigations that the medical history and the physi- cian's physical examination give the greatest contribution to the diagnosis. However;, most of the diagnoses are then !mown before the health screening. How much medical value is affoPded by the notation of earlier known disease remains to be seen. Obviously, the fuformation is most useful· the first time that an examination is undertaken. The ttalue of the history is tremendous and the advantage of questionaries is great. Questionaries · Many usef'l.ll questionari_es have been suggested. Best known is the Comell :r.1edical Index, a simple checklist medical history form (" self-screener"), which has guided many other attempts. The size and contents of the questionary depend on the 'purpose and the facilities available. Although the number of questions can be very small, a general survey of the state of health may require 200 - 500. questions, arranged in groups. Some workers use still more comprehensive systems, where positive findings, if any, are more carefully investigated. As a rule, the questionary is checked by a doctor or by specially trained hospital staff. The gain in time is al9o appreciated in connexion with the physical examination. 7.2.4 Clinical methods ~d laboratory tests used for screening .. Health examinations originate from·and resemble in many ways the traditional visits to a doctor's office. This was natural when health examinations started. There has, however, been a gradual change in some respects, depending on the lack of definite signs of a disease. - 136 - Looking at the procedures used, it is obvious that for a long time conventional diagnostic procedures or functional tests will be used. In Table 7.1,. an attempt is made to give a rapid idea of the examination programmes used. .A comparison is made between more conventional and traditional screening programmes {column 1) and some new attempts {columns 2 and 3). In the first column the frequency of the test is apparent from the numbers given. The material is taken from the study of 33 screening surveys 4 . in the United States during 1946-1954. The table shows that none of the tests was used in all 33 surveys but a few - chest X-ray or MMR, simple serology, and blood sugar determination - were most popular • . , . .. .. ~ ...... --.. ·-··-·----··· The other two colums in Table 7.1 give specific examples of more recent projects, according to somewhat different principles. The first (column 2) is a Swedish study in the county of ~land, where 100 000 people were offered health screening in connexion with a traditional survey for tuberculosis by MMR. 242 Apart from MMR, the basic examination consisted of blood pressure measurement, urine analysis and a multiple blo()d examination with a number of ·tests: · haemoglobin, haematocrit, serum analysis of serum iron, creatinine, the transaminases GC1I' and GPI' (e.g. the aminopherases glutamic - oxaloacetic and glutamic - pyruvic transaminase), cholesterol and beta-lipoprotein, zinc sulphate test for gamma-globulin, thymol rubidity test and, finally, the determination of total protein-bound hexoses and of sialic acid, both non-specific tests indicating many kinds of inflammatory reactions. The screening was carried out by a field group, who also took the blood samples. Analysis was done in a centrally located automated laboratory, and the results handled by computer. This kind of investigation was performed to find objective grounds for recommending examination by a doctor. In such an investigation, the methods chosen are obviously less diagnostic, but sensitive to various diseases. The blood sample taking - as well as supplementary clinical tests - is simple and can be done under primitive conditions, using disposable material (sterile needles, sample containers etc.). The last column (column 3) gives an ·example of- a periodic health examination in California by the Kaiser Foundation (see section 7.4). ( It gives a survey _of tfte selected methods with advanced techniques, which are used on a large scale. - 137 -· TABLE.7.1.1 .. SURVEY OF SCREENING PROCEDURES The examples chosen are taken from: 1. A summary of 33 studies in. the U.S.A. 4 24 2. A .?wedish pilot stud,y witt!. a b~tte:ry of chemical blood ~nsts; 3 3. A multiphasic study in Cal:l.fomia by Collen et al.248, 9 .. Physical examination History taking or questionary Oral and/or dental examination Intraocular tension, tdnometry Visual acuity Retinography Hearing tests Procto-sigmoidoscopy Biometric measurements,_ as height, weight, skinfold measurement Blood pressure ECG Lung function tests Cytology: vaginal smear Chest X-ray or MMR .. . .... X-ray mammography Serology: VDRL, etc. blood groups, Rh Haematology: haemoglobin haematocrit cell count, differentials count Blood in faeces Erythrocyte sedimentation rate 1. N'\lmber of stu:.;; dies using the test {from 33 surveys in USA, 1946-1954) 4 15 5 1 18 12 .. 2. 3. 'Ihe Vlmn1and. · Multiphasic study, in screening Sweden by·Collen et al. 1965 1962-1964 ·+ + + + + : + + + + + + + '· + ~ 7 , ... + .. + + + --- ... - ~···· + + + + + + - 13?- SURVEY OF SCREENING. PROCEDURES Chemistry Urine: sugar Urine: protein ... ~. · · ··· Urinary deposit· · ·· · · ··· Bacteriuria Blood sugar Blood sugar after test load Protein-bound hexoses, sialic acid Cholesterol Beta~ lipoproteins Serum albumin and/or total protein Gamma-globulin (Kunkel) Thymolturbidity or similar test Transaminases Creatinine and/or BUN Uric acid Calcium Serum iron and/or IBC 1. Number·of stu- dies using the test (from 33 ····surveys~ iri .. USA.,· 1946-1954) 16 16 3- .. . ..~ .... 30 2. The Varmland study, in Sweden 1962-1964 + + + + + + + + + + Multiphasic screening by Collen et al. 1965 + + ~ --. ' ·+ + + + + + + -- 139 - From the examples given can alsp-be :derived the possibilities of investigating only certain diseases. ·Although the variation of different projects is considerable, the main featur~~ in common are found in .. the:, 'table. One of the striking characteris- tics of the development is that the more .the ,eost .. can be. reduced per test, the more investigations are made • Whatever the real cause may be, there seems to be a wish . -'·' to increase such health studies... ;The willingness to follow up the investigations is not increasing to the same degree, although today some serious attempts can, fortunately, be noted. 7.3 Development of screening facilities Tec~ical evolution will have a drastic influence on all these f~ctors for screening. Generally speaking, the considerable efforts to improve _he~th screening ~Y technical means can be said to follow two _different lines of evolution. 7.3-.1 Using simpler techniques for laboratory tests. 7.3.2 Laboratory automation. Simplification The first_ trend represents the use of tests that are extremely simplified, for instance, paper-strip tests. . ~.' ··. ··. . . and.~ill, ~crease in number. Such techniques will certainl:y: be greatly improved, Simplification of complicated method,s, such as those for determining sugar and urea in blood also seem promising. There are, however, definite limitations to such procedures, as well as certain difficulties in getting reliable registration, sample identification and reporting of results, etc. The analytical cost decreases when complicated procedures are done by extrem~ly simple methods, although by no means to the degree that might be expected. The manual handTing ana·· sorting is too time-consuming to be economical on a large scale. " So far, the cost of the material is not negligible. A good example of simplification in another r.espect is the laboratory routine worked out by s. Suchet in Paris.250 Using a well-planned routine, it is possible for one technician in a w·orking day" to perform 100 of each of the following analyses: erythJ:>ocyte sedimentation, rate, haema~~rit, lipoproteins, antistreptolysin titre, urin~ sugar, protein and blood, urea in serum and 300 serological tests !or syphilis. - _140 - The techniques can be nonsidered as semi-quantitative. In some instances, it may be permissible to sacrifice some accuracy in order to obtain a high capacity and low cost.- Although it remains to be decided how precise methods should be for screening purposes, it is certainly not necessary to have the same accuracy as in hospital-laboratory work or research. There are too many difficulties in sampling and handling specimens in large-scale investigations, so that a semi-quantitative level might well be justified. 7.3.2 Automation The second important trend is laboratory automation,; on which intense work has been done with promising results. The first clinically important analytical robot was the AutoAnalyzer. In health screening, the most common way of achieving sufficient capacity has been to have several AutoAnalyzers working in parallel. In such a way, large systems have been as.sembled, capable of carrying out a large battery of well-known tests. Many improvements that are of interest for health screening have been made on the AutoAnalyzers. Multi-channel equipment is now available, where several of the inconveniences have been overcome that 'were troublesome in a large-scale investi- gation. ·Examples are: patient identifidation with numberirig of sainples, peak detection and reading, calibration, analogue-digital conversion and auto~tic print- out. The analytical programme has gradually become directed towards health check-up needs, and for AutoAnalyzers there is now a large variety of procedures to choose from. In Sweden, experience of an automatic system for chemical mass analysis - partly based on AutoAnalyzers - led to a new equipment, called the AutoChemist, working with discrete samples and including a ~mall computer. 251 Being a machine for mass analysis, the A.ut6bhemist has a very ruih capacity, theoretically some 1/2 - 1 million samples per year,- with 20 or more analyses on each ·sample. In practice, other factor~ than th~- analytical work load are presumably decisive for normal operation, but with a fixed analytical programme - as for health screening - the conditions are especiall;i favourable. · The ma.xilnU.m efficiency, up to 150 specimens per hour - 141- and up to 40 different analyses on each sample1 will be used only periodioally1 but is advantageous for managing a ·temporB.ry high load... . The AutoChe!Jlist. has . 24 fixed anal~ical channels for different tests1 chemical as well as bacteriological and serol:oglcal. · · Handling of such an apparatus is extremely simple and personnel-saving. Loading is done on one side of the apparatus 1 and on the other side1 the samples are returned ··on a similar ·transport belt after analysis. The inclusion of a small, desk-size computer is very useful for screening purposes: e.g. editing the format 1 sorting and checking. Recording takes place "on line" by automatic print-out on a Teletypewriter with a paper-tape punch. It seems likely that technical evolution for automatic analytical apparatus will aim at an increasingly high speed of analysis. The essential feature will, however, presumably be an increase in the number of types of analysis. There seems to be a distinct tendency to efface the borderlines between chemical1 bacteriological1 serological1 haematological and other procedures. This implies that·. the equipment1 as an electro-mechanical device for automation of analyses 1 will be relatively similar for very different tests. The decisive factor for the apparatus system will be the choice of tests for health screening. Theoretically, almost any method could be automated. The analytical technique may be based on well-known procedures but has to be modified and adjusted to the automatic machine. In practice 1 however1 the difficulties and costs often become so great that the evolution seeks other ways1 tries completely different methQis to get the desired information. In order to give an idea of what might be of interest for screening purposes~ various methods are listed in Table 7.2. ~mportant advances are being made. The list is by no means complete, and In some cases, screening needs may promote automation. For instance, micro- biological determination of vitamin B 12 may be used in a large scale as a screening procedure. - 142 - TABLE 7.2 .1. SURVEY OF SCREENING PROCEDURES . . .l ~··. , .. ~ :. '. 1. ! 2. j .t 3· Manual method I Fully automatic i Comments ! on available for ; procedures automatic " i large-scale ! available with methods use : high speed ' BACTERIOLOGY ' I Growth (selected media) + + ; Turbimetric Bact. count i : (+) + Indirect counting CHEMISTRY ; : Urine l I ' Sugar .. I Automation + ' + Ketone bodies i ' economically I + I + I Protein I + + reasonable ( ! only i Urinary sediment. ! (+) ' - in a large i I j scale ' Bacteriuria l + + ' PKU (see .. also phenylalanine ' in blood) ' ; (-) + Fluorimetric I Blood: carbohydrates ' i I I SUgar I + + Galactose I (-) + Enzymatic Total protein-b. hexose I - l + I I Sialic acid ! (+) l + BloOd: lipids i : I ' Cholesterol + + ! Beta-lipoprotein + + Turbidimetric Total lipids·· I + + ! Phospnolip:i,ds ! - (+) ! ' Non-esterified fatty acids NEFA i - i (+) I Triglycerides ' (+) i (+) fluori-j '. i Evt. I i metric .. 1 (~) Denotes that by expensive means and special organization a fairly high effectiveness may be achieved. (+) Denotes that under certain circumstances large capacity can be reached. - 143- TABLE 7. 2. 2. SURVEY OF SCREENING PROCEDURES -·- ... '•• .. Serum protein r- Total protein Al'Puniin zinc sUlphate (gamma-globulin) Electrophoresis Haptoglobin Transferrin Thymol turbidity Ceruloplasmin Enzyme activities Transaininases Phosphatas~s Lactic acid dehydrogenase Non-protetn nitrogenous compounds Creatinine Urea, BUN NPN Uric acid Phenylalanine Electrolytes, etc. so~i~/Pdf~ssium Calcium/phosphorus Serum iron Iron-binding capacity l. Manual method Fully automatic available for procedures large-scale available with use high speed + (+) + ( +) ( +) (+) + - + ( +) (+) + ( +) (-) (-) ( +) (+) (-) ( -t-) (+) + + + - + + + + + + + + + + + + ( +) + + Comments on automatic methods Turbidimetric Non-specific test, still on research level Fluorimetric, better than urine test Of little value for screening As supplementary ~. haematological !i test r! (-) denote's that by exp€msive means and special organization· a fairly high effectiveness may be achieved (+) denotes that under certain circumstances large capacity can be reached - 144 - TABLE 7 .2.3. SD'RVEY OF SCREENING. PROCEDURES CYTOLOGY Vaginal smear Urine Cancer cells in blood Sputum P..AEMATOLOGY Haemoglobih · · Cell counts Differential count Erythrocyte sedimentation rate Haematocrit SEROLOGY WR VDRL .. etc. RA test ASrfA MISCElLANEOUS Bl-ood (faeces_, urine) PBI PoJ.ar!ography 6-phosphoglucose dehydrogenase (vaginal washing) 1. Manual method available for large-scale use (t) (-) (-) (-) + + (-) (+) ( +) ( +) + + + + - - (-) ·- 2. . Fully automatic procedures available with high speed + ( +) (-) { +) ( +) + ( +) - ( +) + + -. 3. Comments on automatic methods Results from electroscanners still experi- mental Can be replaced by protein-bounc carbohydrate tests Semi-automatic - electronic - methods available Large scale automatic methods may be available at high cuts Semi• automatic methods have been tried Non-specific test Still on a research level (-) denotes that by expensive means and special organization a fairly high effectiveness may be achieved (+) denotes that under certain circumstances large capacity can be reached - 145 - 7.4 Automated mUltitest laboratories ad modum Collen The most important methodological progress in this field has been made by Dr Morris ·F. Collen and his coilaborators. Automated multitest laboratories in.the KaiseJ?-Permanente medical centres have been arranged when advanced examination techniques have been ad~pted ·for periodic .he~th e~aminations on a large scale. 248 .. 249 In many ways., the organization and equipment indicate new trends., but are safely grounded on long experience. .The se~uP. is expensive, but inany details are of interest even when resources are limited. ·The examination progr~e includes the following: · · · 1. · ECG with 6 leads, combined with phonocardiogram. Results are recorded by mark.;.;sense cards, but the intention is to evaluate the ECG by computer analysis~ 2. · Glucose load test: 75 g of glucose in 240 ml of water; blood sugar level determined after 1 hour (evt. also after 2 hours). 3· Chest X-ray with 70 mm film., postero-anterior projection., read by a radiologist. 4. x-ray mammography on women over 45 years of age. Cephalocaudad and lateral views of· each breast are taken, mammographs read by a radiologist. Supine pulse rate and blood · pressllre. Recorded manually on mark-serise CardS• 6. ·Visual acuity by·reading a wall chart, as well as pupillary escape test. Mark~sertse card recording. Tonometry. The intraocular pressure is noted on mark~sense cards. "' ~ :. ~ (At the same time the left pupil is dilated for later retinal photography). ·-8. Vital capacity and one-second forced expiratory rate, recorded manually on mark-sense card. '. - 146 -~ 9. Hearing test by automated audiometer, with the graphed readings transferred to a mark-sense card. l,O. Questionar:i,es: (1) one medical, self-ac1ministe~ed in the form of 207 pre-punched cards, each with. a single question. The patients drop the cards in boxes with "yes" or "no", and: a card-reading machine records the results. (2) One psychological questionarY, also self-administered. 11. Blood tests: haemoglobin, white cell count, venereal disease research laboratories' test for syphilis (VDRLL rheumatoid factor (latex fixation slide test), bloo~ groups and eight chemical tests (serum glucose, creati- nine, albumin, .total protein~ cholesterol, uric acid, calcium and transa- minase) made by AutoAnalyzer and the results directly p~ched on cards. 12. Urine_ tests: for bacteriuria (chemically) and for pH, blood, glucose, and protein by paper-strip tests. 13. Retinal photograp~ read b~r a,:.--1 ophthalmologist. 14. Weight and skinfold thickness are measured manually • Height and trans- . verse body measurements are automatically recorded on punch cards, Routinely :I all patients above 40 years of age are recommended sigmoidoscopy.. and for women __ also a gynaecological examination w.ith cervical smear. The automated multitest laboratory has its own data centre. The ·computer plays an important role as an integrating part of the laboratory. Automatically, t.Q.e computer prints out the summary report for the phy:sician. The capacity is exceedingly high, and routine examinations for 4000 people a month are rep9rted. For detaiis, the reader is referred to the reports in the literature. Generally speaking, this is an indication of a possible evolution, at any rate in countries with good resources. One difficulty may be satisfactory co-operation with the physicians. In California, this has been the case, but in other parts of the world the conditions may differ. It is interesting to note that such a laboratory centre can adopt equipment and methods that otherwise would be impractical or too costly. This can help to develop and test methods that are specially suited to screening .. and for detecting diseases in a very early stage. Such possibilities for research are certainly most welcome. - 147- 8. Conclusions Reviewing the subject of early disease detection~ it is clear to us that, although for a number of years there has been an increasing interest both among the medical profession and the public, we are still at a very early and comparatively primitive stage in the systematic detection and treatment of early disease. For some conditions we have powerful methods of detection, but we do not yet know the effect of early treatment (e.g. diabetes mellitus); for others we are still experimenting to find satisfactory tests (e.g. chronic simple glaucoma); for only a relatively few condi- tions are there already established well-tested and successful means of pre-sympto- matic detection and treatment (e.g. cancer of the cervix, and even here there'are certairi qualifications). It is important therefore to ask ourselves in what particular respects is further investigation, promotion, or education needed; and, since these are big problems affecting whole populations and carrying the. implications of radical changes in emphasis in the practice of medicine in the direction of prevention, what part WHO might play as catalyst in this reaction. It may be helpful to consider some of the possible ways in which these aims may be realized, point by point. 8.1 The need for further epidemiological investigations and, allied .witb thi.$, · the need tp>,achieve standardizat;i,on ;(' Tooof:ten~ in,the past, work has been undertaken in different countri~s pn the ' \ . ' same ostensible condition, only for it to be found at a late stage th,at the workers •. . .. ' were using differing definitions,. This has happened, for instance~ over the condi- tions of carcinoma-in-situ of the cervi~ uteri and of glaucoma simplex. At the same time there is so much work to be done where there is need to examine large populations at a low risk for the condition under ~tudy, that there would be great advantage 'if' results could be pooled or at least intelligently compared. WHO has already played a large part in fostering co-operative international studies; we suggest there is yet room for extending this role. As we have seen~ WHO e~rt committees or seminars have considered many of the chronic diseases, notably tuberculosis, non-specific respiratory disease, anaemia, malnutrition, high blood-pressure~ ischaemic heart disease, diabetes mellitus, mental illness and the cancers. In some instances (e.g. anaemia, cardio-vascular disease and diabetes) special attention has been paid to reaching internationally acceptable definitions. TWo conditions, we believe, still needing internationally accepted definitions are carcinoma-in-situ of the cervix and glaucoma simplex. - 148- Ofcourse, rElaqhing acceptable definitions, diagnosing and then discovering the effect of treatment are lengthy processes entailing much painstaking work by teams of investigators. The needs of establishing the sensitivity and specificity of screening tests and the effect of early treatment are the reasons for the epidemiological surveys to which we have paid such attention in this paper. Surveys are, however, undertaken for a number of reasons, but the object of investigating the feasibility of pre-symptomatic diagnosis and treatment may not always be one of the reasons. Wr10 could, in our view, play a valuable part in both acting as a clearing house for information about epidemiological studies directed towards the development of screening techniques; and in acting as a watchdog to remind those considering studies that future screening possibilities might be borne in mind during the planning stages. While there is, in our opinion, everything to be said for the value of screening being developed by ~ hoc studies of individual conditions, so broad is the field of early disease detection that we consider there may be a place for WHO itself to keep the whole field under continuous review, stimulating and fertilizing here and there as the need appears. Conditions for which, in our view, more organized epidemiological work is· urgently needed are:- 8.1.1 Carcinoma of the uterus: more statistically controlled-trials of the effect of cervical cytology on mortality are badly needed. At present we are largely relying on British Columbia~ some United States cities like Memphis, San Diego and Louisville, Norwegian and British studies, all of iihich have drawbacks either of size of population or epidemiological acceptability. · 8.1.2 Glaucoma simplex: two or three small studies have suggested that ,previously accepted screening criteria are inadequate. are needed. More similar, and comparable, studies 8.1.3 Mental illness: is quantitatively a vast world problem yet little can at present be done about its early detection and treatment. Surveys aimed at defini- tions of early disease, acceptable diagnostic techniques, and incorporating controlled trials of treatment are badly needed. - 149 - 8.1.4 Asymptomatic bacteriuria: there is at least a suggestion that undetected ... urinary infections early in life may lie at the root of much crippling arterial hypertension in later life. Early diagnosis and prompt treatment could potentially have a very important impact. Surveys are ~eing carried out but, in view of the importance of the ultimate condition, the matter should perhaps be attacked on a , .... larger scale. 8.1..5 Cancer of the breast: we. are still ignorant of the effect of ma,king the effort to diagnose b.reast cancer before symptoms are reported, although advances in clinical treatment have had disappointing results and mortality remains little affected •. At least one large-scale mammography su_rvey is in progress. In view of th~ :l,()n& time that must elapse before an effect. on deaths can be seen, it. may perhaps be wise to ensure that valid results will be obtained by at least duplica;tipg the work elsewhere. The preventive value of breast self-examination has never b_een, adequately assessed, as far as we are aware, and~ since this is much cheaper, safer and generally more practicable than X-ray mammography, it seems most important that its value (or otherwise) should be known for ~el:'tain. If the technique were shown to be t'l:'tily beneficial it' should become much easier to propagate' its wider spread. ;::·l'-Lt':..:~ c:.". ; 8.1.6 Lung cancer~ the appalling prognosis for lung cancer even when detected at ·. <,;:, the earliest possible stage radiologically calls for more work directed towards • ' - i .. :~ ' .•. .. : . ·. better screening methods and, _if possible, prevention. ~ ,. ) : .. ~ :·· . ' '' ' ·' It seems likely that cyto- logical examination will show abnormal cells in the sputum at a pre-cancerous stage . ..... . .. , !.'· . \'<hich could lea~ to locali~ation of cancer in a bro~chu~ .. be~ore it appears radio- logically. There is need for more work on these lines, wi.th agreed standards of . •"'' " .;· • :•< ... . . cytological nomenclature, diagnostic procedures and follow-up arrangements. We consider WHO might usefully act in stimulating and co-ordinating work along these, and perhaps other, lines. 8.2 Recof'ds· For the ftelective screeniug of high-risk groups of the population a sine qua non is the ability to know who ··c6~stitutes these groups. It may, of course, be enough to issue a genera:! iri\riia:tion to a particular group (e.g. adult women for cervical cytology); but there :i::{ alWays a risk that those persons at highest risk may opt out. lt is 'there~iore often more successful (as well as better public relations) to - 150- ask people as individuals to attend. In the type of surveillance from a health centre or group general practice~ to which we have refer!'ed, individual invitations are~ essential. In order to draw particular-groups in large numbers from the population~ records capable of being dealt with by automatic data handling methods are needed. In practice this can be difficult, both from the point of view of the design of a suitable record and what should be put on it; and from the point of view of confidentiality. In general practice, for example, the patient's medical record is likely_only to be handled by doctors or those in his close c.onfidence. But for the future, it seems likely we must expect general practitioners' records to be handled at data processing centres. How is this to be done without . a breach of co!lfidentiality? It appears to us, therefore, that records the desigp and handling is an important and urgent problem and that WHO might usefully consider this question. 8.3 Economics Considering its importance surprisingly little is known about the economics of early disease detection. Most probably this is because screening has so far been largelyexperimental and the question of its economic cost has not arisen. Certainly, for screening for tuberculosis, a good deal is known about the cost since this was introduced as a service long ago. It would be valuable to study the cost of screening for· different conditions, either alone or in combination, tinder differing s~stems of medical care. Undoubtedly the arrangements for. medical' care in some. countries are at present more suited to early disease detection than those in other countries. Initiating comparative, and comparable, economic studies of this sort is something WHO might consider useful to undertake. 8.4 Education Practicable techniques for screening for early disease are now available but still, as a profession, we continue to see clinical conditions for the first time at a-late·· stage of development when treatment is less likely to be successful. Also clirticiaris,· in.ge~eral, are too oriented towards conventional diagposis and treatrri~nt to think readily in preventive terms. · Thus there is much room for improving both the attitude of the public and the attitude of the profession to the early detection of disease, as we have pointed out in Section 5.6. Hitherto - 151 - diagnosis has been the province of the clinician who has been concerned mainly with the individual patient. It has been possible to have a high standard of clinigal practice co-existing with poor overall medical care due to unequal distribution of ' -· . resource-s: ... rmf)iiciF_iii ... eariy~disease ~et~-<::~~-0.1?-:·.JcervicS;l cancer screening# for example) is the idea of extension of preventive clinical medical services to a whole community and this in turn postulates a new way of thinking in the-medical and ancill.a.:ry .. prQfess-:Lons ... The- medical pro-fession itself# and the public# may therefore be on the verge of a revolution in its attitude to clinical (as opposed to traditional preventive) medicine and there should be good opportunities for WHO to influence the course of medical and public education. -so as to further# to the best advantage# this changi.tlg approach. New departments of general practice# postgraduate institutes# • , I ~- " ' . ·.· . . . departments 'of soclar med.ic:fri.e and the like. are growing up. These should facilitate the training of doctors and ~issemination of new ideas through attachments and -~~J~ . . . .. feliowsnips~--iri. all of which 'WHO" cail-play a part both by facilitating interchanges and_ through the influence of its committees. - 152 - REFERENCES Chapters 1-5 1. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass.; Harvard University Press, Chapter 5, p. 45 2. World Health Organization, Regional Committee for Europe (1964) The Pre-symptomatic Diagnosis of Diseases by Organized Screening Procedures, EUR/ Rc 14/ techn. Disc., (unpublished working document) 3. Dawber, T. R., Moore, F. E. & Mann, G. V. (1957) Coronary Heart Diseases in the Framingham Study, Am. J. Publ. Health, 47, Suppl., p. 4 4. Council on Medical Service, American Medical Association (1955) A Study of Multiple Screening: Descriptive Data on Thirty-three Screening Surveys, (Revised) Chicago 5. Commission on Chronic Illness (1957-1959) Chronic Illness in the United States, Vol. I-IV, Cambridge, Mass., Harvard University Press 6. Breslow, L. (1955) Multiphasic Scree·ning in California, J. Chron. Dis., g, 375 7. American Public Health Association (196o) Chronic Disease and Rehabilitation: A Programme Guide for State and Local Health Authorities, The American Public Health Association, Inc., New York 8. Kurlander, A. B. & Carroll, B. E. (1953) Case-finding through Multiple Screening, Publ. Health Reports, 68, 1035 9. Wilson, J. M. G. (1962) Report on Multiphasic Screening, Wld Hlth Org. Rep. Travelling Fellowship to the USA, mimeographed paper, 62 R/UK - 13 10. Chronic Illness & Aging Unit, Bureau of Chronic Diseases, State Department of Public Health (1963) Bibliography on Disease Detection: Health Maintenance, Periodic Health Examination and Multiphasic Screening, Berkeley, Calif., mimeographed document 11. Chapman, A. L. (1949) The Concept of Multiphasic Screening, Publ. Health Reports, 64, 1311 12. Mountin, J. W. (1950) Multiple Screeening and Specialised Programmes, Publ. Health Reports, 65, 1359 13. Smillie, W. G. (1952) Multiple Screening, Am. J. Publ. Health, 42, 255 14. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 48 - 153 - 15. Remein~ Q. R. & Wilkerson,_ H. L .• (1961) The Efficiency of Screening Tests for . piabet~s, J-•. Chron •. Dis._,. :]..3, 6 16. Holland, W. W_. & Humerfelt, S. (1964) Comparison of Blood Pressure: Comparison of Intra-Arterial arid-·cuffValues, Brit. med. J., ii, 1241 17. Lanc-et· (T966) Measuring Blood-Pre-~sU:r-e_! Leadi~ J1.~~~-?.£.e~ Lancet; !· p. 414 18. Elwood, P. c. & Jacobs, A. (1966) Haemoglobin Estimation: A Comparison of Different Tecbrlique9, :Srit. med. J.,·i, 20 . -· ~--···-~--· ·-··~ 20. 21. 22. 24. 25. 26. 27. Thorner, R. M• & R~mein, Q. R. (1961) US Department of Health, Education and Welfare, Principles and Procedure§in the_Evaluationof Screening for Disease, Public Health Monograph No. 67, Public Health Service Publication No. 846, Washington, D.C. - Hollows, F. C. & Graham, __ P. A. (1966) The Ferndale Glaucoma Survey, in: criiiuc-oma, 'Ei:iideirifoiogy~- EarlY. Diagnosis and Some Aspect's of Treatment; Proceedings of a Symposium held at the Royal College of Surgeons of England,_ p. _ 24, Edinburgh, E. & ·S. Livingstone Graham, P. A. & Hollows, F. C. (1966) ibid, p. 103 Davis, H. J. (I965),The ·:rr:dgation Smear: Results in Population Screening for- Cervfca:r·c-a.ncer, --find International Congress of Exfoliative Cytology, Paris Davis~ _H. -J: (1962} -The ":trr"igation Smear - A Cytologic Method for Mass PopUiatlori-Screening by Mail, Am·~-;r. o'bstet~ GynecoL, 84, 1017 Butterfield, w. J. H., Keen, H. & Sharp, c. L. E. H._ (l964).Diabetes Survey in Bedford, l96~,.Proc. Roy. Soc. Med., 57, 193 United States Department of Health, Education and Welfare'(l964). Glucose Tolerance of Adu1ts, United States, 196o-1962, Diabetes Prevalence and Results-of ·a-Glucose Tolerance Test, by Age and Sex, Nat:'..onal Center for Health Statistics, Public Health Service Publication., No. 1000, Series 11, No. 2, Washington, D.C. St8.!lller., J., Berkson, ., Young, Q. D., Hall, Y. &-Miller, w. (1963) Approaches to the Primary Prevention of Clinical Hear:t T!.i~ease ~n :H:i,@-risk, Middle- aged Men, Ann. New York Acad. Med., 97, 932 Jolliffe, N., Rinzler, S. H. & Archer, M. (1959) The Anti-Coron,a.cy--_Cl_Ub: Including a Discussion of the Effects of a Prudent Diet on the Serum Ch~~~st:::-ol LeV:e~- _'?f_ Middle:..aged Men, Am. J. Clin. Nutr., l• 451 - 154 - 28. Christakis, G., Rinzler, s. H., Archer, M .• , Winslow, G., Jampel, S., Stephenson, J., Friedman, G., Fein, H., Kra._us~_ A;, _Gt._James, G. (1966) The Anti-Coronary Club: A Dietary Approach to the Prevention of Coronary Heart Disease - A Seven-Year Report, A.m. J. Publ. Health, 56, 299 . 29. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 65 )0. Ibid, Vol. IV, Chronic Illness in a Large City, p. 262, 448 31. Weinerman, E. R., Breslow, L., Belloc,.N. B., Waybur, A. & Milmore, B. K. (1952) Multiple Screening of Longshoremen with Organised Medical Follow-up, 1\·m:· i:- Pu"bi.' lleaith,· 42, 1552 - -- · --- · ·- · -- - · 32 New York State Department of Health, Bureau of Chronic Disease & Geriatrics, Annual Report (196o) Report on Cortland.County Screening Demonstration · -(~publisEe<l ·r-ep()~_~Y_ _ __ -····-33. Fa.no; R. M. & Corbato,- F • • Y. (1966) Time-sharing on Computers, Scientific American, 215:3, 129 34. Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. IV, Chronic Illness in a Large City,The Baltimore Study, Cambridge, -Mass"., ·Harvard University Press 35. Commission on Chronic Illness (1959) Chronic Illness in the United States, Vol. III, Chronic Illness in a Rural Area, The Hunterdon Study, Cambridge, Mass., _Harvard University Press 36. Beuchley, R. W., Robert, M. A., Dr~e, M. &.Breslow, .L. (l-958) Height, Weight and Mortality in a Population of Longsh9remen, J. Chron. Dis., 1, 363 37. Borhani, 0. N., Hechter, H. H. & Breslow, L. (1963) Report of a Ten-Year ·---Follow-up Study of ~h~--~~1:1 ~a.nci~c?_ IJ<:)~~~~o:s~~~~~--J. Chron Dis_., 16, 1251 38. Wylie, c. M. (1961) PartfciP"ation-I:n-·a: iviuitipie screerl.ing~-ciinic-witil Five- Year Follow-up, Publ. Health Reports, 76, 596 39. Wylie, c. M. (1961) Use of_Death Rat~s in Evaluating Multiple Screening, Publ. Health Reports, 76, llll 40. Stamler, J. (1962) Chronic Disease Control -bivision~-Chicago Board of Health, Ann. Report, p. 12, mimeographeq document. 41. Commission oq._Chron:i,c Illnes:;> (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p. 29 - 155 - 42. Franco, S. c., Gerl, A. J. & Murphy, ·a. T. (196l) ... P.eriodic Healtl:LExam,ip.~t:i,o_n~: a Long Term Study, 1949-1959, J. Occ. Med., L• 13 43. United States Department of Health, Education andWe.lfare (19q3) :P~:r:-iodic Health Examinations - Abstracts from the Literature, Public Health Service, Division of Occupational Health,Publication, No. 1010, Washington, D.C. 44. ·Commi--ssion on Chronic Illness (1957) Chronic I-llness in the United States, Vol. I, Secondary Prevention Tlwough Perio.dic Healt.h..E~~nations, Cambridge, Mass., Harvard University Press, p. 28 45. Williamson, J ., Stokoe, I. H., Gray, s., Fisher, M., Smith, A., McGhee, A. & Stephenson, E. (1964) Old People at Home, Lancet, _!, 1117 46. California's Health (1961) Biennial Report Edition, 19, No. 11 47~ · -1N1;1son;, J • M. G. (1965) Some Aspects· of the Epidemiology of Cervical Cancer, 'Month. Bull. Min. Health & Publ. Health Lab. Serv., 24, 72 _ . ·48' •. '6rundy;-- F ;· & Mackintosh, J. M. {1957) The Teaching of Hygiene and Public . . Health in Europe, Wld Hlth Org.,.:.Monogr. Ser., 34, .. 109' 49. World Health Organization (1964) Promotion of Medical Practitioners Interest in Preventive Medicine:-. rrwe·lfth Report of the WHO Expert ·Committee on . Professional and Technical Education of·Medical and Awdli~ryPersonnel, Wld Hl th Org. techn. Rep. Ser., 269 50. Brockington~ ·F.-,. Silver, G. & Vuletic, A. (1964) Teaching of the Medical ... · _Student for Comprehensive Medical Practice with emphasis on extramedical institutions) WHO Study, ~ffiO Educ 126, mimeographed document 51. College of General Practitioners (1965) Present State and Future Needs of General Practice, London, Council of the College of General Practitioners 52~ Br'eslow, L. & Hochstilll, J. R. (1964) Sociocultural Aspects·of·Cervical Cytology in Alameda County, Calif., Publ. Health Reports, 79, 107 53. Kegeles, s. S., Kirscht; "J.P., Haefner, D.P. & nosenstock, I .. M.:::.(1965) Survey of Beliefs About Cancer Detection and Taking Papaniqo::l:aou Tests., Publ. Health Reports, §£, 815 54. Wakefield, J. & Baric, L. (1965}= Public. and P:r:'ofessional Attitudes -to a Screening Programme for the Prevention of Cancer of the Uterine Cervix., Brl t.' J.. Prev·. ·Soc. Med •, 1:2, 151 55. Cebtl:-al Health Services Council (1964) Health Education: Report of a Joint Committe~ of the Central and Scottish Health Serv.iQes.Gouncils, p. 10, London,. H.M.s.o. - 156- Section 6~1 Diabetes Mellitus 56. Joslin, E. P., Root, H. F., Wb.fte,-- P~ & Marble, A. (1959) The Treatment of Diabetes Mellitus, (lOth Edition) London., Henry Kimpton, p. 19 57. Wilkerson, H. L. c. & Krall, ·1~ -P.- (1947) Diabetes in a.· New E~gland Town, J. Am. Med. Assoc., .ill,.. 209 58. Wa,.l~e.I', J ~ B. & Kerri~e, D.. (1961) Diabetes in an English Community - A Study of its Incidence an(~~t~ai fl1.~~ocy, Leicester UnivBrsity Press 59. McDonald, G. vJ., Fisher, G. F. & Pentz, P. C. (1965) Diabet"es Screening Activities, July 1958 to June 1963, Publ. Health Reports, 80, 163 60. United States Department of Health, Education & Welfare (1961) Diabetes Fact Book, Public Health Serv.i~~ Publication No. 890, Washington 61. . United States Department of Health, Education & Welfare (196o) Diabetes Program' Guide, Public .Health Serviqe Publication No. 506,· Wa~~gton, p. 51 --·· -~--. -· ····---·-----~-·-··--··--· •····· ··-·.. . . .. . 62. Conimission on Chronic Illness (1957) Chronic Illness in the United States., Vol. IV, Appendix D, Chronic.Illness in a Large Citl, Cambridge, Mass., Harvard University Press, p. 466 63. Reid, J. J. R. (1962) Diabetes Mellitus - Certain Aspects of Preventive, Detection and Treatment, Wld Hlth Org. Rep. Trav~lling Fellowship to the USA, mimeographed paper, 62 R/UK - 13 64, R~J_Q,,t __ ~. J .•.. ~~ (1960) __ Public Knowledge of Diabete~, Med. Off., 103,325 , .. 65. R~dll~-~d. .. ··-i. H~ · (1960}Incidence of ·aiycosuria and. Diabetes Mellitus in a General Practice, Brit. med. J., i, 695 66. Coll~g~ -~f- Gen~~al Pr~ctitioners (1962) A Diabetic Survey, Brit •. med. J ... i, 1947 67. Kirkland, J. A. & Morgan, H. G. (1961) An Assessment of Routine Hospital Urine Testing for. _!'rot_e_il?-__ ~ _ _q_!ucose, Scot. med. J., 6, 513 68. Rundles, R. w. (1945) Diabetic Neuropathy, General Review ~th Report of 125 cases, :Medicine, Baltimore, 24, 111 69. Garland, H. (196o) The Neurological Complications of Diabetes Mellitus, 'Clifiical Aspects., Proc. Roy. Soc. Med ... ..5.2_,. 137 ~ ... ' , . 70. Ashton, N. (1959) Diabetic Retinopathy, A New Approach, Lancet ... .li.,, 625 71. ..~.:.~l_o.p_, .P.• .. M. ( 1954) ·Are Diabetic ·Degenerative Complications Pr~:vep.table? Brit. ~d. J ~ 1 .:I.i~~~-~83---_- ·-- · _ ........ -_ .::.: .......... :. 72. Marble, A. (1955) Coronary Artery Disease in the Diabetic, Diabetes, 1 .. 290 74. 75. 76. 7.7· 78. 80. 81. 82. 8). 84. 85. 86. 88. - 157 - Johnssen, S. (196o) Retinopathy and Neuropathy in Diabetes Mellitus, Comparison of the Effects of Two Forms of Treatment, I)la~.!-~s ... ,.2 .. 1 Wo],ff, 0. lh ,& Sal:t;,_ H. B. (1958) Serum-Lipids and Blood Sugar .Levels in Childhood Diabetes, Lancet, i, 707 Keen, H. (1959) Paper read to Medical & Scientific Section, British Diabetic As:sociation Annua.":!, Clinical Meeting at Guy's Hoqpital, London 0 ~ H •''' • Oo •-• ---·-· ~'' ' '···--·-·• '0 '' 0 '' ••• '- Newburgh, ~. H. :& Conn;-·;r~. w: ( 1939) A New Interpretation of Hyperglycaemia in Obese, Middle-aged Persons, J. Am. Med. Assoc., ]lg, 7 ~cet (1965) Obesity & Diabetes, Leading Article, Lancet, _!, ~60 Root, H. F., ~lirsky, s. & Ditzel, J. (1959) Proliferative Retinopathy in Diabetes Mell;itu:s, :Rev;iew of Eigtlt Hundred & Forty-Seven Cases, J. Am. Med. Assoc., 169, 903 ...... __ _ ;Beckett, A. G. (1962) Ha.I;'111ful Effects of pelay inDiagn,osi.s .of Pi.abe.tes Mell,itus, The Prac.tit,ioner, 189, 57 . . .·r. Brown, I. K. & Jones, A •. T. (19q4) .Retinopathy .and Diabe~i~ .Con.tro:l, Brit. J. QPht~l., .. 48, 148 .· ' . . . Collyer, R. T. & Hazlett, B. E. (1961) Retinopathy and :NeuropathY· in One Hundred Growth-Onset Diabetic Patients, Canad. Med. Ass., I, ~, 1328 Daysog, A., jr, Dobson, H. L. & Brennan, J. c. (1961) Renal Glomerular & Vascular Lesions in Pre-Diabetes and in Diabetes Mellitus, A Study Based on Renal Biopsies, Ann. Int. Med., 2&, 672 Hoet, J. P. (1954) Carbohydrate Metabolism During Pregnancy, Diabetes, 2_, 1 Wilkerson, H. L. C. (1959) Pregnancy and the Prediabetic State, Ann. New York, Acad. Sci., 82, 219 Fajans, s. S. & Conn, J. W. (1954) An Approach to the Prediction of Diabetes Mellitus by Modification of the Glucose Tolerance Test with Cortisone, Diabetes, 2 .. 296 White, P., Joslin, E. P., Root, H. F. & Marble, A. (1959) Treatment of Diabetes Mellitus, (lOth Edition) London, Henry Kimpton, p. 48 Commission on Chronic Illness (1957) Chronic Illness in the United States, Vol. I, Prevention of Chronic Illness, Cambridge, Mass., Harvard University Press, p~ 166 Ford, N. J. & Glenn, B. (1951) Undetected Diabetes Among the Relatives of Diabetics, Sth. Med. J., 44, 239 - 158 - 89. Harris, H. (1947-1949) The Incidence of Parental Consangiiinity in Diabetes Mellitus~_ .Ann. Eugen., 14, 293 90. Harris, H •. (1950) The Familial Distribution of .Diabetes Mellitus, Ann. Eugen ... ~ .. 95 91. United States of America Department of Health, Education & Welfare (1964) Gluco::;e Tolerance of Adults,.· United States, 1960-62, Diabetes Prevalence and Results of a Gluccse Tolerance Test by Age & Sex, National Center for Health Statistics, Public Health Service, .Publication No. 1000, Series·ll, No. 2, Washington D~C. 92. British Medical Journal (1965) Inheritance of Diabetes Mellitus, Brit. med. ·J., _!, 940 93. World .. Health Organization (1965) Diabetes Mellitus.; Report of a WHO Expert Committee, Wld Hlth Org. tecl:m. Rep. Ser .... 310 ... · .......... - .. , .. 94. Keen, H.·_, Rose.. G., Pyke, D. A. .. Boyns, D. .. Chlouverakis, C. & Mistry.. S. ( 1965) Blood Sugar and Arterial Disease_, Lancet_, ii, 505 .. 95. Ostrander, L. D • ., Francis, T ... Hayner_, N. s ... .Kjelsberg, M. 0. & Epstein, F. H. · - (1965) The Relationship of Cardiovascular Disease to Hyperglycaemia, Ann. Int. Med., 62., 1188 - 159 - Section 6.2 Heart Disease .. 96. Registrar General (1966) The Registrar General's Statistical Review for the Year 1964, Part I, Tables, IVIedical, London, H.M.S.O. 97. Logan, W. P_. D. &_ Cus}:li()n, A~ .A •. {J95§) Studies on Medical and Population Sub,lects, No. 14, Vol. I, Morbidity Statistics from General Practice, London, H.M.S.O. - .. 98. Registrar General (1962) The Registrar General's Statistical Review for the Year 1960, Part I, Tables, Medical, London, H.M.S.O. 99. United States Department of Health, Education and Welfare (1964) Heart Disease in Adults, United States, 1960-1962, Public Health Service Publication, No. 6, Series 11, Washington, D.C. 100. World Health Organization (1965) The Work of the WHO in Cardicvascular Diseases, 1959:1964, (unpublished report) 101. World Health Organizat~on, Car~iovascular Diseases Unit, WHO, Geneva, to be published 102. Miller, R. A., Smith, J., Stamler, J., Hahneman, B., Paul, M. H., Abrams, J., Hai~, G., Edelman, J., Willard, J. & Stevens, w. (1962) The Detecting of Heart Disease in Children - Results of a Mass Field Trial with Use of Tape-Recorded Heart Sounds, Circulation, ~~ 85 103. Wilson, J. M. G. & Heasman, M. A. (1959) Coronary Artery Disease- An Epidemiological Review, Month. Bull. Hin. Hlth. & Publ. Hlth. Lab. Serv., 18, 94 104. · Kennedy, A. C. (1957) Observations on the Incidence of Coronary Heart Disease in a Rural Are.a in S. W. Scotland, Scot. med. J., .£, 420 105. Epstein, F. H., Ostrander, L. D., Johnson, B. C., Payne, M. vJ., Hayner, N. S., Keller, J. B. & Francis, T. (1965) Epidemiological Studies of Cardio- vascular Disease ip a Total Community - Tecumseh - Michigai.l., Ann. Int. Med., 62, 1170 106. Epstein, F. H. (1965) ':I'he Epidemiology of Coronary Heart Disease, J. Chron. Dis., 1§, 735 107. World Health OrganJ.iation (1963) Survey of the Prevalence of Ischaemic Heart Diseases in Certain European Countries - Report of a Technical Meeting convened. by the Regional Office -for Europe of the WHO, Copenhagen, 1963, EURO, 179.3, mimeographed document 108. Rose, G. A. (1962) The Diagnosis of Ischaemic Heart Pain and Intermittent Claudicat.ion~. in Field _Surveys, Bull. Wld Hlth Org., n_, 645 - 160 - 109. World Health Organization (1962) Arterial Hypertension and Ischaemic Heart. Disease, (Preventive & Therapeutic Measures) Wld Hlth Org. tecbn. Rep. Ser., 231.1! 19 . 110. Burgess, A. N., Fejfar, Z. & Kagan, A. (1963) Arterial Hypertension and Ischaemic Heart Disease - Comparison in Epidemiological Studies, WHO, Geneva 111. Kagan, A. (1965) Comparison of Electrocardiographic Data with Particular Reference to the WHO Minnesota Code Trial, Millbank Nemorial Quarterly, USA, .!2 112. Rose, G. A. Personal communication 113, Stokes, J. & Dawber, T. R. (1959) The Silent Coronary: The Frequency and Clinical Characteristics of Unrecognised Myocardial Infarction in the .. ____ "...:::. __ --~ra.nDJlgham Study, Ann. Int. Med., 2Q, 1359 114. Acheson, R. M. & Acheson, E. D. (1958) Coronary and Other Heart Disease in a Group of Males aged 65-85, Brit. J. Prev. Soc. Med., 12, 147 115. Hinkle, L. E., Carver, S., Benjamin, B., Christenson, W. & Strone, B. (1964) Studies in Ecology of Coronary Heart Disease; i, Variations in the Human Electrocardiogram U."1der Conditions of Daily Life, Archiv. Environ. Health, 9, 14 116. Paul, 0., Lepper, M. H., Phelan, W. H., Dupertnis, G. W., MacMillan, A., McKean, H. & Park, H. (1963) A Longitudinal Study of Coronary Heart Dis~~e,. Circulation,. 28, .20 117. Kannel, W. B., Dawber, T. R., Kagan, A., Revotskie, W. & Stokes, J. (1961) -__ , Factors of Risk in the Development of Coronary Heart Disease - Six Year -· Follow-up Experience ofthe Framingham .Study, Ann. Int. Med., _22, 33 118. Dawber, T. R., Kannel, W. B., Love, D. E. & Streeper, R. B. (-1952) The Electrocardiogram in Heart Disease Detection - A Comparison of the Multiple and Single Lead Procedures, Circulation, 2 .. 559 119. Thomas, A. J., Cochrane, A. L. & Higgins, I. T. T. (1958) The Measurement of the Prevalence of Ischaemic Heart Disease, Lancet, ii, 540 120. Blackburn, H., Keys, A., Sir.Jon.sor", E., Rantakarju, P. & Punsar, S. (l96o) . -~~_}!:lectrocardiogram in Population Studies, Circulation, 21, 1160 121. Thompson, W. B., Hudmet, H. B.,jr,Russo, P. E., Brown, F. 'R •. 'i:Mostl..ey, K. T. (1961) A Review and Study of Cardiovascular Disease Screen~ng with the Miniature Chest X-ray, J. Chron. Dis.,~ .. 148 122. Cooper, G. R. (1965) International Comparability of Blood Lipids, Millbank Memorial Fund Quarterly, ii, .!"2, 49 -161 - 123. Oliver, M. F. & Stuart-Harris, (1965) Experiments on Animals, Brit. med. J., ii, 1203 124. Oliver, M. F. & Boyd, G. s. (1959) Effect of Bilateral Ovariectomy on Coronary- Artery Disease and Serum-Lipid Levels, Lancet, ii, 690 125. Morris, J. N. & Heady_, J. A. (1953) Coronary Heart Disease and Physical Activity of Work, Brit. med. J., ii_, 10?5, 1111 126. Morris, J. N. & Crawford, M. D. (1958) Coronary Heart Disease and Physical Activity of Work, Brit. med. J., ii, 1485 127. Registrar General (1958) Registrar General's Statistical Review, Decennial Supplement on Occupational Nortality 1949-53, Part II, London, H.H.s.o. 128. Doyle, J. T., Heslin, A. s., Hillebal, H. E. & Formel, P. F. (1959) Early Diagnosis of Ischaemia Heart Disease, New Eng. J. Med., 261, 1096 129. Gertler, M. M. & White, P. D. (1954) Coronary Heart Disease in Young Adults, Cambridge, Mass., Harvard University Press 130. Spain, D. M., Nathan, D. J. & Gellis, M. ( 1963) Weight, Body Type and the Prevalence of Coronary Atherosclerotic Heart Disease in -Males, Am. J. Med. Sci. , 245, 63 131. Doll, W. R • .& Hill, A. B. (1956) Iung Cancer and Other Causes. of Death in Relation to Smoking, Brit. med. J., 11, 1071 -. 132. Hamnond, E. C. & Horn, D. (1958) Smoking and Death Rates - Report on Forty- four Months of Follow-up of 187 -783 Men, J. Am. Med. Assoc., 166, 1159, 1294 133. Dawber, T. R., Kannel, W. B., Revotskie_, N•, Stokes, J., Kag~_,A. ·&Gordon, T. (1959) Some Factors Associated with the Development of Coronary Heart Disease (Six Years Follow-Up Experience in the Framingham Study) Am. J. Publ. Health_, ~~ 1349 134. Doyle, J. T., Dawber, T. R._, Kannel, W. B., Kinch, S. H. & Kahn, H. A. (1964) The Relationship of Cigarette Smoking to Coronary Heart Disease, J. Am. Med. Assoc._, 190, 886 135. Kannel, W. B. (1964) Editorial: Cigarette Smoking and Coronary Heart Disease_, Ann. Int. Med., 60, 1103 136. World Health Organization (1965) Working Group an Studies of Preventive Measures in Ischaemia Heart Disease - Summary of Discussion_, Mimeographed document_, EURO- 179.3 137. Groom, D. (1961) Population Studies of Atherosclerosis_, Ann. Int. Med._, 221 51 - 162 - 138. Smirk, F. H. (1957) Hi.sh. Arterial Pressure, Oxford, Blackwill Scientific Publications, p. 687 139. Hodge, J. V. &Dollery, C. T. (1964) Retinal Soft E.xudates, .21!.~d-~., 2_2, (N,S.) 117 140. Pickering, G. W. (1965) Hyperpiesis: High Blood I1:•essure Without Evident Cause: Essential Hypertension, Brit. ~~d.~., Ji, 959, 1q21 141. Kain, H. K., Hinman, A. T. & Sokolow, M. (1964) Arterial Blood Pressure IJieasurement with a Portable Recorder in Hypertensive Patients; Iv variability and Correlation with "Casual 11 Pressures, Circulati..P~ )_Q, 882 142. Rose, G. A., Eollal1d, W. W. & Crowley,; E. A. (1964) A Sphygmomanometer for Epidemiologists, J.,ance,:t, l-_, 296 143. Boe, J. ~ -Hi..unerfelt, s. & Weden:~varig-:F. ·.(1957) The Blood Pressure in a Population, Acta Med. B_q§.l1d. Suppl., 321, Bergen . . 144. . Iii:i,all, W. E. & Oldham, P. D. (1958) The Inheritance of Arterial Blood Pressure, Clin. Sci., 11., 404 145. Hamilton, M., Piclcering, G. W., Roberts, J. A. F. &: Sowry, G. S. C. (1954) The Aetiology of Essential Hypertension I, Tne Arterial ~ressure in the General Population, p_lin. Sci.~- .J..l, 11 146. Kagan, A •. , (Jord_9n, T., K~el, W. B. & Dawber, T. R. (1959) Blood .tre~sure and its Relation -to_ .9.9}-"'.onary Hear~ Dis_Et~_ip._B~ Fr~ l'_9.Q11_l.a_lj_on,: HYperten~.s>_r:~,. Vol. III, American Heart Association 147. .United States Department of Health, Education & Uelfare (1964) ID_o_o.§_ Pr~S.§.ill'S, of Adults by __ jt~,.?1ld §_eb United_Si:;.?-.:t~e.s..,._J_$)6Q~-:- l_9..§g, National Centre for Health Statistics, Public Health Service ·Publication, No. 1000, Ser. 11, No. 4, ltJash:i,ngton, D.C. 148. Pickering, G. W. (1955) High Blood Pressure_, London, J .• &; A. Ohurc.hill, p. 299 149. Bechg('l.ard, P. (1946) Arterial Hypertension: a Follow-Up Study ,of One Thousand Hypertonies, Acta T1ed_~l?..9..§ll<!·, .Suppl., 172 .:.;:::·· .... - - 163 - 150. Ungerleider, H. E. & Gubner, R. S. (1958) 1J.J~_l}.§~~~p. MedicJ.ne ... Springfield, Illinois, Charles C. Thomas l5L Kemsley, W. F. F., Dillewicz, W. Z. & Thomson, A. M. (1962) A New· Height-for- Height Standard Based on British Anthropometric Data, Brit. J. Prev_. Soc. ~ ... 16, 189 152 ~ Edwards, D. A. w., Hammond, w. H... Healy I M. J. R., Tanner, J. 111. & vJhi tehouse, R. H. (1955) Desi0n and Accuracy of Oalipern for Measuring Subcutaneous Tissue Thic:~ess, f)r~_. _ __.;r_. Nutr., 2,, 133 153. United States Department of Agriculture, AGriculttu~al Research Service. Human Nutrition Research Division, (l96C) Heizj:l_ts and it/eights of_j\..Q...u..lJ..S. in the Unit_~l St_a_t_e.s ... Home Economics Research Report, No. 10, vlashington 154. Commission on Chronic Illness (1957) Chronic Illp~ss in the UniteSL_State~ Vol. I, Rr_eyep_t_i_op_of_ .. Qb.!'.9.J!.:i~:!Jle_9.§ .... CarJbridge, ll1ass ... Harvard University Press, p. 267 · 155. British Medic.al Journal (1964) Leading article, f3i"it. med. J., .;hi, 1543 - 164 - 156. General Register Office (1957) Tuberculos:h_s S1._ati_s.ti_(fs for ~~-'3ll.Clli~l.e§.. 12)8 .:.. ):95.5 .... studies on Medical and Population, Subjects No. 10, London, H.M.S.O. 157. \vorld .Health Organization (1964) W.H.O. Expert Committee on Tuberculosis, Wld Hl th Or_g_._j;_e..f!lli.1.-:..Rep. Ser ., 290 158. Springett, V. H. ( 1956) _t1inj.rnal Pulmon~'!'..u~""'.£.ll!.o.sJ..e_..£2und by_ rlla.§..lLR_a:_c:l.i..OU.§l..P~ London, Lewis 159. Yerushalmi, J. (1953) The Reliability of Chest Roentgenography and its Clinical Implications, P.J...§._. of the Ohe.§..,i, 24, 133 160. Horwitz, 0.'(1964) Epidemiological Basis of Tuberculosis E.:r'adiqation, 2, Dynamics of Tuberculosis, Morbidity f!: ~1ortali ty, Bull. \Vld Hl tl1....2l'Jl•, 2.Q., 609 - ... 161. Styblo, K. (1964) Identification of High Tl1berculosis Risk Groups in Connection with the Epidemiological and Clinical Study of Tuberculosis in Czec~oslovalda, in Co..~.laboration with HHO, J3_lJ.ll. Int. Un. _C2[lj:;ra. Tu.P.•, _22, 36)'. 162. Erin, L. (1960) Detection of Pulmonary Tu~erculosis by Sputum Survey, ~p~~~ London, 41, 363 163. Scottish Health Sei,.,.vices ColiDcil (1963) J?ronch.itis,., Jie.J2ort by a ~:.Pommittee of the St.anding Hedical Adv~ai tj;ee, Edinburgh 164. Horl;:, T. (1962) A Comparative Study of Respiratory Disease in Engla.'l"ld, Hales and Norway, 1\ct..?:._ £~_9._.__§_q.§llld., Sup pl., J8l~ 165. Holland, W. \l. (1965) Respiratory Disease in :Cn::;land and the United States, Ar.,ch. Envir'_?n_._]!_lth,., lQ., 338 166. College of General Practitioners (1961) ~nl~onic Bronchitis in Great Britain, Jki t. me~. , ).l, 973 167. Holland, W. W. Personal communication 168. World Health Organization, Regional Offic~~ f_ol: :::1..u."..?..P.e (1962j EJJ£..9..P_el!:ll Symposium on_lL~:?ll~~p-Specific LungDj~e~2~~ Moscow, EURO- 212, Mimeographed docuoent - 165 - Section 6.6.1 Lung Cancer 169. Registrar General (1966) Registrar General's Statistical Review of England f': Wales for the Year 1964, Part i, ·Tables, Medical, London, H.M.S .0. 170. Registrar General (1957) Registrar General·'s Statistical Review_of England f': Wales for the Year 1952, Supplement on Cancer, p. 114, London, H .. M .• S.O. 171. Posner, E., McDowell, L. A. f': Cross, K. W. (1959) Mass Radiography and and Cancer of the Lung, Drit. med. J., !, 1213 172. Cuthbert, J. (1959) Bronchogenic Carcinoma: A Mass Radiography Group Compared with a Practitioners Group, Brit. med. Dis. Chest., 53, 217 173. Gifford, .J. H. & Waddington, J. K. B. (1957) Review of 464 Cases of Carcinoma of. the .Lung. 'J:r.eated by Resection, Brit. med. J., --_!;,---7?~·-··-- . 174. Waddington, J. K. B. (1960) Surgical Aspects of the Mass X-ray Campaign, . Liverpoo~, 1959, Med.. . Off. , .10.4, 293 . . .. 175. Boucot, K. R., Cooper, D. A. 8: Weiss, W. (1961) The Philadelphia Pulmonary Neoplasm Research Project, An Interim Report, Ann. Int. Med., 54, 363 176. Frost, J. K. Cytology Laboratory, Department of Pathology., John Hopkins Hospital, Pal tim ore, rllary land, U.S • A. , Personal connnuni cation 177. Canti, G. (1964) Analysis of 100 Cases of Bronchial Carcinoma, Annual Meeting, British Society for Clinical Cytology, London 178. Lilienfeld, A. (1963) ACS-VA Co-operative Study for Evaluation of Radiologic and Cytologic Screening in the Early Detection of Lung Cancer, Progress Report, Unio~ Internat. Contra cancer. Acta, 19, 1330 - 166 - 6.6.2 Cervical Cancer 179. Jones, H. (1952) In Discussion of Paper by Hertig, A. T., Younge, P. A. & McKelvey,· J; L~~ ·entitled 'rA Debate, What is Cancer in situ of the Cervix? Is it the Pre-invasive Form of True Carcinoma?", Am. J. Obstet. Gynecol., 64, 807, 816 180. Petersen, 0. (i955) Precancerous Changes of the Cervical Epithelium in Relation to Manifest Cervical Carcinoma, D-~nish Sci. Press Ltd., Copenhagen 181. Lange, P. (1960) Clinical ~ Histological Studies on Cervical Carcinoma, Precancerosis, Early Metastases, and Tubular Structures in the Lymph-nodes, Acta Path. et Microb. Scand., Suppl. 143 182. Clernrnesen, J. (1962) On the Prognosis of Precancerous Conditions of the Uterine Cervix Proceedings of International Conference, The Morphological Precursors ·of Cancer, University of Perugia, Italy, p. 463 183. Koss, L. G. et al~(l961) A Long-Term Cyto-histologic Study of Untreated carcinoma-in-Situ and Related Abnormalities of the Uterine Cervix, Paper Presented at 1st. International Congress of Exfoliative Cytology, Vienna - - 184. Younge, P. A., Hertig, A. T. & Armstrong_, D. (1949) A Study of 135 Cases of carcinoma in Situ of the Cervix at the Free Hospital for Women, Am. J. Obstet. Synecol., 58, 867 185. Boyes, D. A., Ffd.l:er, H. K. i: Lock, D. R. (1962) Significance of In Situ carcinoma of the Uterine Cervix, Brit. med. J., !, 203 186. Dunn, J. E. (1962) The Use of Incidence and Prevalence in the Study of Disease Development in_a Population, Am. J. Publ. Health, 52, 1107 .. ,._ - 187. Dunn, J. E. (1958) Preliminary Findings of the Memphis-Shelby County Uterine Cancer Study and Their Interpretation, Am. J. Publ. Health, 48, 861 188. Kashgarian, M., Dunn, J. E., Erickson, C. C. r0 : Sprunt, D. H. paper Unpublished 189. Terris, M. (1960) Carcinoma of the Cervix: An Epidemiologic Study, J. Am. Assoc., 174, 1847 190. Boyd, J. T. ;": Doll, R. (1964) A Study of the Aetiology of Carcinoma of the Cervix Uteri, Brit. J. Cancer, 17, 419 - 167 - 191. r1acGregor, J. E., Fraser, H. E. & Mann, E. H. F. (1966) The Cytopipette in the Diagnosis of Early Cervical Carcinoma, Lancet, !, 252 192. Cameron, C. B. ~ Hussain, 0. A. N. (1965) 6-Phosphogluconate Dehydrogenase Activity in Vaginal Fluid: Limdtations as a Screening Test for G:~n:i,ta:L Cance.r "'· _Brit. med. J. , i.o_: __:!529 193. Labrum, A. H • .7: Gibbs, D. F. (1964) Clinical Significance of Levels of _6;:,Pho~phog1ucona:te Dehydrogenase in Vaginal Fluid, "Technicon" Conference: 11Automation in Analytical Chemistry", London .. - -- 194. Ladi,nsky,. J. L., Sarto, ''"• E. e: Peckham, B. M. {1964 ). Cell Size Distribution Patterns as a Means of Uterine Cancer Detection, J. Lab. C1in. Med., 64, 970 - 168 - 6.6.3 Breast Cancer 195. Registrar General·(l966) Statistical Review of England and Wales for the Year 1964; Part I, Tables, Medical, London, H.M.S.O. 196. Registrar General (1957) Studies on f!led~ical and Population Subjects, No. 13, Cancer Statistics for England and ~vales 1901-55, London, H.M.S .0. 197. Registrar General ( 1957) Statistical Review o.f England and Wales for 1952, Suppl. on Cancer; p. 11, Londott; ll.M.S. :J. 198. Lilienfeld, A. M. (1963) The EpidemiolovJ of Dreast Cancer, Cancer Res., ' 23, 1503 199. Registrar C'eneral (1957) Statistical Review of England and Wales for 1952, Suppl. on Cancer, p. 74, London, H.M.S.O. 200. Park, W. W. & Lees, J. c. (1951) The Absolute Curability of Cancer of the Breast, Surg. Gynaec. Obstetr., 22• 129 201. Lewison, E. F. (1963) An Appraisal of Long-term Results in the Treatment of Breast Cancer, Unio Internat. contra Cancra Acta, 19, 1547 202. Berg, J. W. & Robbins, G. F. (1963) TWenty Year Follow-up of Breast Cancer, Unio Internat. contra Cancra Acta, 12• 1575 203. Registrar General (1957) Statistical Review of England and Wales for 1952, Suppl. on Cancer, p. 12, London, H.M.S.O. 204. Registrar 'eneral (1957) Statistical l.eview of England and Wales for 1952, Suppl. on Cancer, Diagram P4 "· : '5, p. 17, London 205. I"-loom, H •. J .. G. (1965) The Influence of Delay on the Natural History and Prognosis of ''reast Cancer, '"ri t. J. Cancer, 19, 228 206. Sutherland, R. (1960) Cancer, The Significance of Delay, London, ,ezutterworth 207. Hawkins, J. H. (1944) Evaluation of ::"reast Cancer as a Cuide to Control ProQ;rammes, J. Nat. Cancer Inst., ~. 1~45 208. Taylor, r'. H. " Wallace, R. H. (1947) Carcinoma of the ~·reast: End Result, Massachusetts C'eneral Hospital, 1933-1935, New Eng. J. Med., 237, 475 209. 210. 211. 212. 213. ... 214~ 215. - 169 - Kreyberg~ L. -": Christiansen, T. (1953) The Prognostic Sig_n~_ficance of Small Size in Breast Cancer, Drit. J. Cancer, z, 37 -. Gershon.::c·ohe·n.; J. -;c !:':arden; .. (. G. };;·:--·(i964) Detection of Unsuspected Breast Cancer by Mammography~ Ann. New York Acad. Sci., 114, 782 -- . ·-·. --··· -··· Egan, R. L. (1962) Mammography, an Aid to Diagnosis of Breast Cancer, J. Am. Med. Assoc., 182, 839 Stapiso, S., Strax, P. 1'\:Venet, L. (1966) Evaluation of Periodic Breast cancer Screening with Mammography Methodology and Early Observation, J. Ain. Med. Assoc., 195, ni· · ·- · --· · · · ... ·· Eger, S. A. (1965) Early Diagnosis in Colon and Rectal Cancer, Ca., 15:6, 275 Clark, T. W., Schor, S. S., Elsom, K. 0., Hobbard, 0. B. ,.,; Elsom, K. A. (1961) Value of Periodic Examinations in Detecting Cancer of the Rectum and Colon, Postgrad. Med., 27, 290 Hertz, R. E., Deddish, M. R. & Day, E. (1960) The Periodic Examination. Evaluation of Routine Tests and Procedures, Ann. Int. Med., 54, 1209 -· -----· -----·------ . . .'• ·-- . --- --··-. --··-··· - 170 - 6.7 Eye Diseases .. ---~-. ----·-···· 216. Sorsby, A. (1956) Ministry of Health, Blindness in England 1951-1954, 217. London, H.M.S .0 •. . . ~--- New York State, Glaucoma Program Guide, P.O. Box 7283, Albany 1, New York, p. 10 Health Education Service, U.S.A. (unpublished pamphlet) 218. United States Department of Health, Education and Welfare. (1961) Principles and Procedures in the Evaluation of Screeninz for Disease, Public Health Service, Monograph, No. 67 _ ..... 219. StrBmberg, u. (1962) Ocular Hypertension: Frequency, Course and Relation to other Disorders occurring in Glaucoma, as seen from J1et~s Survey of all Inhabitants over Forty Years of Age in a Swedish Town, Acta Ophthalm, Suppl., 69. 220. Goldmann, H. (1959) Some Basic Problems of Simple Glaucoma, Am. J. Ophthalm, ~. 213 221. Duke-Elcier, S. (1957) The Bowman Lecture; the Aetiology of Simple Glaucoma, Trans. Ophthalm. Soc. of the_U.K., 77, 205 222. Perkins, E. S. (1965) Glaucoma Screening from a Public Health Clinic, Brit. med. J., i• 417 223. Paterson, G. D. (1966) The Value of Family Studies in the Detection of Glaucoma Simplex; in Glaucoma: Epidemiology, Early Diagnosis and Some Aspects of Treatment, Proceedings of a Symposium held at the Royal College of Surgeons of England, London, Edinburgh, E. & S. Livingstone, p. 51 - 171 - 6.8 Urinary Tract 224. Kaitz .. A~·-··L;;·& Hodder_, E. W. (1961) Bacteriuria and Pyelonephritis of Pregnancy (Prospective Study of 616-Pregnant Women) .. New Eng. J. Med., 265, 667 225. Mi.all_, W. E., Kass_, E. H., Ling, J. & Stuart, K. L. (1962) Factors Influencing Arterial-Pressures in the General Pqpulati.on in Jamaica;_ Brit.· m~d. J., ii, 497 226. Kass, E. H. (1962) Pyelonephritis and Bacteriuria, Ann. Int. Med., 56, 46 227. Simmons., N. A. & Williams., J;n. (1962) A Simple Test tor Significant Bacteriuria, Lancet_, l_, 1377 228. Brumfitt, w ... Davies_, B. I. & Rofiser_, E. ap i (1961) Uretheral Catheter as a Cause of Urinary Tract Infection in Pregnancy and Puerperium, Lancet., ii_, 1059 - 172 - ~ Rheumatic Disease 229. Kellgren, J. H. (Editor) (1963) The Epidemiology of Chronic Rheumatism a Symposium Arra.D.ged by the CIOMS, Vol. I, Blackwell, OXford 230. Mikkelsen, W. M., Dodge, H. J., Duff, I. F., Epstein, F. H. & Naiper, J. A. (1963) Clinical and Serological Estimates of the Prevalence of Rheumatoid Arthritis in the Population of Tecumseh, Michigan 1959-60, In the Epidemiology of Chronic Rheumatism, a Symposium arranged by the CIOMS, Vol. I, Blackwell, Oxford, p. · 239 231. Epstein, F. H., Francis, T., Hayner, N., Johnson, B. C., Kjelsberg, M. 0., Naiper, J. A., Ostrander, L. D., Payne~ M. w. & Dodge, H. J. (1965) Prevalence of Ch:i:-onic Diseases and Distribution of Selected Physiologic Variables in a Total Community, Tecumseh, Michigan, J. Chron. Dis., 81, 307 232. Kellgren, J. H. (1964) Heberden Oration, 1963, The Epidemiology of Rheumatic Diseases, Ann. Rheum. Dis., 23, 109 233. Lawrence, J. S., Hewitt, J. V. & Popert, A. J. (1963) Gout and Hyperuricaemia in the United Kingdom, Symposium arranged by the CIOMS, Vol. I, Blackwell, Oxford, p. 176 - 173 - 6.10 Mental Illness 235- 236. 237. Shepherd, M. (1964) Minor Mental. _Illness in London, Some Aspects of a General SurV'ey ,_ .Bri_t ~ · med. J., ii, 1359 ~., Medical Research Council: (1:965). Clinic~ Trial of Treatment of Depressive illness, Brit~ med •. J., 1, 881 Gruenberg, E. M. (1964) Mental retardat;ion: Epidemiology_, Chicago, Stevens, H. A. & Heber, R., p. 259 Jaeggi, A. & Jaeggi, F. (1965)-.Rimseignement des enfants et .adolescent reputes arrieres dans le canton de Geneve, Psych. 1 'enfant, Vol. VIII: 2, 453 238. Wein'berg,. A. N. (1961) .Detection of Congenital. Galactosemia . and the Carrier State Using Galactose-Cl4 andB1oQd cells·, Metaooiisffi;· ro; ·:p~-728 · -- · ·· --------·-- .. ·- · - 174 - 6 .11 Anaemia 239. World Health Organization (1959) Iron Deficiency Anaemia, Wld Hl th Org • techn. Rep. Ser • , 182 240. Berry, W. T. c., Cowin, P. J. & Mages, I{. E. (1952) Haemoglobin Levels in Adults and Children, Brit .. med. ;J., _!, 410 241. Kilpatrick, G. B. &: Hardisty, _R. M. (1991) The Prevalence of Anaemia in the Community, A Survey of a Random Sample of the Population, Brit. med. J., !' 773 242. Spooner, R. D. (1960) The incidence of Anaemia in General Practice in New South Wales, Med. J •. Australia, _g, 727 .. 243-. Jungner, G. & Jungner, L · (1966) The Health Screening· in Vctrmland, in SurveH.lahce and Early Diagnosis in General Practice, Office of Health Economies, London· 244. Fry, J. (1962) Minor Maladies, Practitioner, 189, 633 - 175 - Chapter 7 275. World Health Organization (1963) Expert Committee on Medical Assessment of Nutritional Status, Wld Hlth Org. techn. Rep • Ser ., 258 246. Day, E. (1960) What is an Adequate "Cancer Checkup"? Postgraduate Med ... 27, 274 247. World Health Organization (1964) Expert Committee on Prevention of Cancer, Wld Hlth Org. techn. Rep. Ser~ 276 248. Collen, M. F., Rubin, L., Neyman, J., Dantzig, G. B., Baer, R. M. & Siegelaub, A. B. (1964) Automated Multiphasic Screening and Diagnosis, Am. J. Publ. Health, 54, 741 249. Collen, M. F. (1966) Periodic Health Examinations Using an Automated Multitest Laboratory, J. Am. Med. Assoc., 195, 830 250. Suchet, A. s. (1963) Methode Active de Surveillance de la Sante des Grandes Collectivites Agricoles et Industrielles, Acta Medica et Sociologica, II, 237 251. Jungner, G. (1966) Data Processing in the Clinical Laboratory, in: Automated Data Processing in Hospitals, International Conference in Elsinor 1966 (in print)