WORLD HEALTH ORGANIZATION
ORGANISATION MONDIALE DE LA SANT~
REGIONAL OFFICE FOR
THE WESTERN PACIFIC
BUREAU REGIONAL DU PACIFIQUE OCCIDENTAL
REGIONAL CXlvlMIT.l'EE
TECHNICAL DISCUSSIONS
Fifteenth Session Manila
WP/RC15/TD7
11 September 1964
17-22 September 1964
ORIGINAL: ENGLISH
From. the Unit of Development of Health statistical Services
Division of Health statistics W.H.O., Geneva
, .
~)..
i
1.
TYPES OF ST.A!rISTICS USED
1 1 1
1.1 statistics concerrd..ng the health status of the population ••••••••••••••••••••••••••••••••••••••• 1.2 1.3 2.
statistics concerning' health services and medical
care •••••••••••••••••••••••••••••••••••••••••••••• statistical data about the conditions influencing health and health s~v1clj!s ........................ . ~A
2
SOURCES OF THE
••••••••••••••••••••••••••••••••••••
2 2
2.1
Official records ••••••••••••••••••••••••••••••••••
2.2 2.3 2.4 3.
Records in health departmerrts and medical institutions •••••••••••••••••••••••••••••••••••••• Records of special surveys or investigations •••••• Records of other departments and programmes •••••••
3 3 4
AVAILABILITY AND USEFULNESS OF THE STATISTICS ••••••••••
4 4 4 5 5
3.1 statistics reflecting the health status of the population •••••••••••••••••••••••••••••••••••••••• 3.1.1 Vital statistics ••••••••••••••••••••••••••• statistics of notifiable diseases ••••••••• General morbidity statistics •••••••••••••• Health service statistics •••••••••••••••••••••••••
6
.....
wp/RCl5/Trr( page 1
The variety of data which are of use to health administration, the number of sources from which they are derived, the many WB\Vs they can be utilized, the differential availability from one place to another and variations in their quality preclude an exhaustive global treatment of the subject in a paper of this size. The following reviews briefly the type of statistics used and their sources with more space given to the possibilities of obtaining them and their relative usefulness in health administration. 1. TYPE OF STATISTICS USED
statistics which are used, or of use, to health administrations for the establishment of health pOlicy, fom.ulation of programmes, control of operations and evaluation of results are generally considered under three broad headings: a) b) c) 1.1
statistics concerning the health status of the population. statistics dealing with health facilities, personnel and services. statistical data about the conditions influencing health and the health services.
Statistics concerning the health status of the population
This group comprises: vital statistics, including statistics of causes of death; general morbidity statistics; morbidity statistics in respect of one disease or group(s) of diseases; statistics of impairments (e.g., deafness, blindness, skeletal infirmities). Other data which may be included here are, for instance, anthropometric measurements, or the level of immunity in respect of a particular disease. 1.2
statistics concerning health services and medical care
The subjects of these statistics are the need for health services; availability and distribution of health facilities and health personnel; utilization of resources and manpower; evaluation of health programmes; and assessment of expenditure connected with health services and medical care. They apply to institutions and establishments providing medical and health services, such as hospitals, dispensaries, health units, diagnostic laboratories, treatment centres. They deal with a variety of health and medical care programmes, such as school hygiene, health education of the public, mass immunization, maternal and child care, case-finding programmes, social-security schemes, food and drug control.
WP/Rms/'IDT page 2
3..3
Statistica.l. dat§ about the COtIditiol?§ innuencing healtb e.nd health flervices •
These relate to the wide range of conditions which influence or are influenced by health, such as soeio-economic conditions, nutrition, education, occupation, housing, water SUlmly J climate and other environmental conditions, as well as living habits and genetic factors. Health is affected in various ways by these factors acting either singly or in an interrelated pattern, for instance, low economic status being responsible for undernutrition, poor housing, unhygienic living habits, insufficient education, this in turn reducing the economic possibilities of the individual. The t~e of health services in the country are directly influenced by these conditions, in terms of the health measures needed (e.g., sanitation of the environment); the limitations fixed by the financial resources of the country; the availability of professional and auxiliary health personne1, which is dependent on the educational level of the country; and indirectly by the health status of the population resulting from these factors. Population statistics may be included here, considering that the structure of the population according to age, sex, occupation, residence (urban, rural) and other attributes presents different health problems calling for different health measures, e.g., preponderance of old persons. However, apart from suggesting certain health problems, population data as derived from census enumeration and annually adjusted according to population movement have a much wider significance. They provide the basis for programmes of development and supply the denominator of indices and rates of various kinds which are applicable to all types of statistics. 2. SOURCES OF THE DATA
The type and range of statistics used by health administration is reflected in the variety of sources from which they are obtained. They can be grouped under the following headings: Records required by law Records in health departments and medical institutions Records of special surveys and investigations Records of other departments or programmes containing medical data ar~ items relating to health. 2.1 Official records
Here belong in the first pla.ce registration records of births, deaths and other vital events which are legal documents providing evidence of the fact for use by the individual. and for a variety of other purposes.
wp/RC15/'IDl page 3 Records of population censuses are also of'f'icial. documents periodical~ established according to the law of' the country. Conu>letion of the medical certificate of cause of death is a legal obligation in a number of countries. Depending on the health legislation of the country,· certain health measures or actions may be governed by law or regulation, althOUgh some of the resulting records may be of temporary value, for instance, notification of communicable or occupational diseases, licensing of hospitals, pharmacies, physiCians, laboratories. 2.2 Records in health departments and medical institutions
j
...,..
The large proportion of the statistics are based on records established in the operation and administration of health and medical services in order to preserve evidence of health action taken and/or to initiate appropriate health action. For instance, notification of communicable diseases serves the primary purpose of applying various control measures, such as isolation of the case, investigation of the source of infection, sanitation of the environment, examination of contacts, which actions in turn become the subject of a record. An in-patient case history records the clinical observations and treatment giveu during the current illness and is used as reference for any fUture medical attention. Records are kept of vaccinatiOns, sanitary inspections, home nursing visits, pre~natal care, case-finding programmes and of the wide range of curative and preventive services extended to the individual or COmmunity and intended to protect, promote and restore health. These records primari~ inform on the service given, but many of them contain morbidity data related to illness or disease as such or to a specific condition. For instance, records of mass screening programmes contain not o~ data for reporting on the work and progress of the programme in terms of the individuals covered, type and number of examinations made, number of positi','"e cases of illness found, but also allowing conclusions as to the morbidity, general and speCific, of the surveyed population. All these records can serve as the basis for statistical reporting, either already constituting the report (e.g., notification record of communicable diseases) or forming the source of data for completion of the statistical report. 2.3 Records of special surveys or investigations
J.
Many countries find it necessary to resort to special surveys to obtain information not available in routine sources. The objectives of a survey, its scope, whether single-subject (morbidity survey) or multiplesubject (health survey with eeveral,cOnu>onents), its advantages (e.g., flexibility of deeign, available population baSis, person-centred information) find its methodological aspects have been discussed in detail in the Seventh Report of the Expert Committee on Health statistics (Wld Hlth OrIS. +,echn. Rep. Ser., l~l, 218),
WP/R(Jl5/TIr7 page 4 2.4 Records of other deper!;ments and programmes
There are various records outside those of health departments and medical institutions which contain data on conditions related to health. They are maintained by other governmental departments dealing with subjects such as food production, hOUSing, economic conditions, communications, education,occupation, metereology a.o. In addition, a number of records which are not the responsibility of health administration contain morbidity data of various kind. The Third Report of the Ex;pert Committee on Health Statistics (Wld Hlth Org •.techn. Rep. Ser., 1952, .2J.) lists a whole series of records containing morbidity data, regardless of their source, whether health administration or other, such as: Notifiable disease records Registers of certain diseases (e.g., cancer) Census enumeration of impairments Records of medical institutions and health centres Records of other medical and nursing services (e.g., practitioner's records, records of home visits) Records of medical care systems (e.g., social security, sickness insurance, hospitalization insurance plans) Records of physical examinations, sickness and absenteeism of specific population groups (e.g., armed forces, veterans' records, civil servants, industrial employees) Records of mass screening programmes Records of morbidity or health surveys other records (e.g., life insurance records, records of road accidents)
3.
AVAILABILITY .AND USEFULNESS OF THE! STATISTICS
The type of statistics of use to health administrations differ in their availability depending on the social and administrative development of the country and the level, coverage and co-ordination of medical and health services. 3.1 3.1.1 statistics reflecting the health status of the population Vital statistics
Vital statistics are affected by the lack of or gross incompleteness of registration of births and deaths. Mortality statistics by cause are further affected by scarcity of physicians to certify the cause of death
page 5
wp/RCJ..5/TfT(
on the death certificate. These deficiencies are common in devel.oping countries and as establisblllent of registration systems and extension of medical services are slow processes, mortality statistics, general and by cause, of a reasonable coverage and completeness cannot be expected in these countries in the immediate future. 3.1.2 Statistics of notifiable diseases
Notification of communicable diseases is a traditional health measure existing in a large number of countries, incl.uding developing countries. These statistics are again hampered by the lack of physicians, apart from the general defect of under-reporting and in addition population data for computing incidence rates may not be available in the absence of vital statistics. However, even with these defects statistics of communicable diseases can serve for general orientation as to epidemiological. trends.
3.1..3
General morbidity statistics
As already indicated, morbidity data are availabl.e in a variety of records maintained by individuals and institutions providing health and medical services or as part of the administration of heal.th, wel.fare and soci~security programmes. However, these morbidity data, characterizing the service or benefit rendered in terms of speCific conditions, are not directly suitable for assessment of the morbidity in the community. In the first place, the morbidity data are fragmentary, covering only conditions which are the subject of a particular health measure or benefit (e.g., notification of infectious diseases, hospitalization, incapacity benefit) and omitting cases without medical attention. The data are selective, either relating to specific diseases or groups of them (e.g., registered cases of cancer) or having a composition different from the morbidity in the general po:pu.lation (e.g., hospital morbidity). Furthermore, a person may be the object of several services for the same condition, e.g., in the practitioner's office, in the dispensary, in the hospital, and the information concerning the individual. will be scattered in different record systems. Finally the popul.ation at ri~ will not always be definable and, if known, may represent only a segment of the population.
J.-
All these aspects are obstacl.es to a successful. integration of the partial data into an estimate of morbidity relating to the population as a whole. However, in certain cases these records can serve for the assessment of morbidity in the COmmunity. An example is the health record combining all information of the various episodes of il.lness of the individual regardless of the health service received, as is :possibl.e in countries with integrated medical and health services covering the whole popul.atlon where the type of system favours the maintenance of an individual health record. other examples are health service records which by their nature are centred on the person, such as notification records of communicable diseases or records of mass screening progr8l!Jl1les to detect cases of illness from certain diseases. Valid conclusions as to the distribution of diseases in. the general population can be drawn fran hospital records in respect of morbid conditions requiring nearly always hospitalization, such as fracture of the skull, tetanus, cerebral haemorrhage.
WPjRC15/TIYl
page 6
Nevertheless, on the 'Whole the primary usef'ulness of morbidity data in health and related records will be for the admi pi stration of the particular programme with which they are connected. Their contribution to morbidity statistics will depend on a number of factors, such as the type of record, the purpose for which it is kept, the degree of co-ordination of medical, hospital and other health services, the . possibilities of linking records in various systems. Only a morbidity survey covering the whole population in an area or s~les of it can provide a complete picture of the distribution of diseases in the c am!l],1 mi ty. 3.2 Health service statistics
It is in the field of health service statistics where the ~hasis should be placed considering the direct need of health administrations for operational data and the possibilities of obtaining them from the records of health services. Regardless of the level of development of medical and health services in a country, it is essential to know the number and distribution of health facilities of whatever kind; the number and type of personnel attached to them; the amount and nature of services rendered; the utilization of resources and personnel; the efficacy of the work performed; and the cost of such services. These data are needed not only on the local level for operation of the individual establishment but also on higher administrative levels for the purposes of comparisons in order to detect maldistribution of resources and their inefficient utilization and to obtain a basis for corrective action. This type of analysis is needed and feasible at any stage of development of the health services and may range from conclusions based on purely descriptive statistics to sophisticated studies qualii'ying under the terms of operational research. Relati vely simple statistical procedures and tabulations can yiel.d an amount of useful information and hospital statistics may be quoted as an example. The Expert Committee on Health Statistics in its Eighth Report (Wld Hlth Org. techno Rep. Ser., 1963, g§l) established a minimum general content of a hospital statistics programme which all countries could carry out, differentiating between two types of statistics, namely, those based on t~e tlhospitalll and those relating to the "patient". In respect of the "hospital" the EJQ;>ert Committee recommend collection of data on:
a)
resources of the hospital, including beds and special. diagnostic and therapeutic facilities (e.g., laboratory, radiology, operat:l.ng room);
b ) utilization of bed facHi ties in terms of patient-movement and days of care;
WP/RCl5/TIJ7 page 7 c} :personnel, diutinguislrlng prof'ession.eJ.. medicalj nursing; other prof'essional and technical; and other, 1ncluding administrative and general serv1ce staff; d} general financial. aspects of hospital. services.
In respect of statistics reJ.e.ting to the patient (hospital lllOrb!dity statistics) the Committee recommended as a min:l.mum account of patients discharged and of their ho~italization days, by diagnosis and sex.
. .,..
Already such a basic type of statistiCS, especia.l.ly through crosstabuJ.e.tions of various items, is able to suppl¥ va1ua.ble data for effective administration of the individual hospital to prov1de proper care for its patients; for organization, co. ordination and pJ.e.nn1ng ho~ital. services in an a.dministrative area; and for economic utilization of hospital facilities within the general health programme of the Community, region or country. For instance, statistics of discharged patients by diagnosis and length of st8¥ are useful in bringing out variations between hospitals warranting further studies as to the possible reasons for the differences, such as insufficient bed facilities allowing on1¥ admission of serious cases; differential. attitudes of ho~ital physicians; economic factors; J.e.ck of home nursing and other services outside the ho~ita1. All these a~ects require careful consideration and ~ call for correct!ve measures. Applied to the range of other health services, current operational data provide the health administrator with a pool of information of direct use to him and wh;1.ch he can obtain within the framework of the d.a;y-to-d8¥ health activ1ties. This emphasis on health service statistics sh01,1ld not detract from the usefulness of v1tal statistics and mortality statistics by cause, but in their absence there remains a wide range of statistics which can be developed, illlproved ar.d utilized for the plll1loses of health administration. In recognizing the imJ?ortance of health service statistics to countries lacking registration systems, several other aspects should be stressed, namel¥: the pOl'!sibility of the country of establishing a usei'ul. health statistics system without having to wait for the introduction of registration; the cormected need for improving the basic records from which the statistics are abstracted, thus contributing to the efficiency of the particular health service (e.g., reorganization of medical and other records in ho~ita1s); the possibility of training personnel in record keeping and simple statistical procedures.