Epidemiological methods for the study of onchocerciasis

iEXPERT COMMITIEE ON EPIDEMIOLOGY OF ONCHOCERCIASIS Geneva, 10-18 Novenb er L975 Draft asenda item: Te chnical Annex I UONCA/ WORLD HEALTH ORGANIZATION ORGANISATION MONDIALE DE LA SANTE oNcHo/r,IPl 75. t5 ORIGINAL: ENGLISH Ce document ne constrtue pas une publrcatron. ll ne doit farre l'objet d'aucun compte rendu ou rdsum€ ni d'aucune crtatron sans l'autorrsatron de l'Organisation Mondiale de la Santi. Les opinions exprim6es dans les articles sign6s n'engagent que leurs auteurs. 'l IDEMOLOGICAL }IETIIODS EOR IIIE SI]T'DY OF ONCHOCERCIASIS by aE. Sartwell-l P :_n ), tI 1. 10n OUJ Any d SS ion of epideoiological methods must be prefaced by a stateEent of the purpose of the study. There are several broad general purPoses that may be served by an epidemio- logical approach. Although a particular study may on occasion serve more than one PurPose, it should not be relied on to do so. It is also necessary to know someEhing of the clinical characteristics, natural history and nethods of diagnosis of the disease in order to decide what methods of investigation are appropriate and practicable. Onchocerciasis occurs mainly in undeveloped areas where nedical care is largely unavailable and access is often difficult. The clinical disease evolves very slowIy over a period of years in Ehe human host. Ihese two facts make the determination of incidence, as opposed to prevalence, impractical, and the reliance on official notifications hopeless. There are, however, a number of clinical and laboratory- aided procedures such that reasonably accurate indices of prevalence, based on physical inspection for subcutaneous nodules, skin lesions, hanging groin, or defective vision may be utilized; and for a highly reliable, though perhaPs insensitive, measure of infection, Ehe skin snip is available. In onchocerciasis studies, attention will usually need to be given, along with measure- ment of the frequency of huoan infection, to Ehe insect vector. Such issues as its density, breeding habits, flight range, feeding habits, zoophily, and extent of parasitization will be explored. These matters will not, however, be considered in this Paper. 2. Programme purpose gne major purpose of epideniological study may be termed the .g1]EIame purPose. One wishes to know the frequency of a disease, its severity, chronicity, and its social and econoEic cost to the individual and the conrmunity in such terms as lost manPoTrer, crippling or reproductive wastage. This inforrnation is needed so that a rational decision can be rnade as to the amount of effort, resources and funds to devote Eo furEher study and attempts at prevention of the disease, relaEive to other probl-ems. I I'orr".ly of the Department of Epidemiology, TheJohnsHopkins University School of Hygiene and Public Health, Baltimore, Mary,land, United SEates of America. I I I The issue of this document does not constrtute formal publication. lt should not be reviewed, abstracted or quoted without the agreement of the World Health Organizatron. Authors alone are responsible for views expressed in stgned articles. oNcHo/wP/75.15 Page 2 Ihe programme goal requires a knowledge not only of the frequency of the disease but of its natural hisEory, clinical feaEures and other rnatters, including some that appear unrelated to medicine. As pointed out intronchocerciasis ConErol in the Volta River Basinrrrl the avoidance of areas of highest prevalence has had adverse economic consequences for entire countries, since it has meant the abandonment of the best agricultural 1and. Ttris is a striking example of the fact that analysis directed toward the programne goal cannot be limited to a mere counting of casesl a broader approach must be taken. The i.ssue includes but is not limlted to the securing and use of epidemiological data. 3. Surveillance A second purpose, related to the programme purpose, is surveillance. this aims to keep track of the occurrence of a disease with special attention to its distribution in time and place, so Ehat one IIEIy know where it is most prevalent, and whether it is increasing or diminishing. In onchocerciasis, prevalence surveys done by medical teams have been the means of getting data for both these purposes. Estirnates of total infection rates and of blindness have been two of the most often used indices. For the surveillance purpose, the two consideraEions of accuracy and consistency have different relative priorities than for the purposes to be discussed later. Consistency should be considered of equal or even greater importance than accuracy, since the purpose of the surveys is to compare prevalences in differenE places and at different tioes. A large inaccuracy of diagnosis of course could not be counEenanced, buE noncomparable measurements would vitiate the inEended cornparisons. Why cannot we demand a maximum of both qualities? Ihe answer must consider the great effort in examining and reexamining neny thousands of people at hundreds or thousands of Iocations. Expenditure of time and money must be kept aE a practical level, in balance with other aspects of the prograume. What are needed are indices, fairly siqle to obtain, noE too costly, and comparatively error-free, by which the disease burden can be reliably compared between villages and districts. They should be such that different trained observers can approxinately replicate one anotherrs findings, and the same observer can on repeated Eries approximately replicate his own. Observations by Hamilton and his colleagues2 in the United Cameroon Republic indicate that conventional current methods fall short of meeting this goal, even when used by skilled and experienced workers. Upon reexamination of villagers a year after the initial examination, they found sizable differences in the findings of ocular, skin and subcutaneous nodule abnormalities. They suggested that agreed-upon definitions should be published for the various skin manifestations of onchocerciasis. Skilled ophthalmologisEs cannot always be relied upon to participate in surveys, and standardized methods for scoring ocular and visual defects need to be used. Ihere seems to be a need for further work on these problems. 4. Establishment of prevalence and incidence to evaluate intervent,ions A third purpose of epidemiological investigation, related to the last two, is to establish the prevalence or incidence of a condition so as to deterrnine the effects of a {lqeage preven- tion or eradication programme. For this purpose a series of measurements must be made beforer perhaps during, and after the operation of Lhe programme. Furthermore, in order to really evaluate the prograrmne there should be trcontrolrtvillages, areas or subjects, si.n-ilar in all important resPects to those in which the preventive progranuoe is put in operation, which have precisely similar measurements of prevalence taken, at the same times, preferably by the same personnel. If these stePs are not taken, the programme is in danger of being a tot.aI failure from the point of view of assessment of the value of the procedures used, so that in other areas or in fuEure programmes one must start from zero kncrr^rledge of what will work. The considerations concerning repeatability of measurements, discussed under surveillance, are of equal or greater importance here. Ihe design of conErolled trials of preventive programres will not be discussed further in this paper, except to say that the parEicipation of an epidemiologist is almost mandatory. oNcHo/wP/75.15 Page 3 5. The determinant of the disease A fourth seE of goals, which have come into focus especially since chronic diseases of unknown etiology have been under study, are Eo learn causes or detetminants of disease, including both those intrinsic and extrinsic to Ehe huoan host. Ttrese goals are pursued by sEudying the distribution of the disease in the population in relation to what have come to be termed ttrisk factorstr, and thus measuring the rrrelative riskrt of the disease anong those having a particular risk factor. Ihe methods for doing this are classified on a dynamic basis as the case-control or retrospective, and the population-based or prospective method. There has been no great tendency to adopt these methods in the study of infectious di.seases of known etiology, but they could doubtless be of value in sme situations. They require that some question or hypothesis be formulated regarding the relation of a possible risk factor t,o the disease. Thus, cigarette smoking was long ago identified as a risk factor in lung cancer and myocardial infarctionl hypercholesterolaemia as a risk factor in myocardial infarcEion; and occupation in industries involving exposure to certaln chemicals in bladder cancer. The association of thrombosis and use of oral contracepti.ves was demonstrated by case-control studies. References to papers dealing with these methods may be found in textbooks of preventive medicirre or in Sartwell.3 Would it be useful to apply these meEhods in studies of onchocerciasis? It night be, if there are quesEions which they could appropriately explore. The case-control met,hod would be the more practical approach. Preferably choosing a cormunity with a lon, rather than high, prevalence of the disease, one would identify the infected persons and an approxioately age-matched sarnple of the uninfected as controls, and see how the Ewo groups coryared as to selected attributes, as for instance place of residence (proximiEy to a stream), duration of residence, or occupation. One could also coryare persons with ocular disease or those excreting microfilaria in the urine with other infected persons not havi.ng these characteris- tics; the association of presence of these conplications with intensity of infection as gauged by skin snips or by nunber and location of subcutaneous nodules could be examined. Such studies (several of which have been done) can lead to other, more original or rnore promising hypotheses that can be test,ed in a similar fashion. In order to interpret the associations found it is necessary to take account of so-ca11ed ttconfounding variablesri which may be responsible for spurious associaEions. Ttrere are tr.ro ways to do this: by oatching cases with controls on such variables, or by staEistical adjustment in the analysis phase of Ehe sEudy. 6. Conputer aided rnapping Since onchocerciasis is a highly focal disease, mrcthods of mapping prevalence data with the aid of computer programming as developed hy Hopps4 and used by him in studying the spatialdistribution of schistosomiasis in South America might be useful. Ihese methods could provide visr:aI evidence of the relationship of prevalence to various factors such as altitude, temPerature and rainfa11. It has been pointed out that people have deserted nany of the Eost heavily onchocerciasis-infested vilLages, but this should not greaEly confuse the picture. The geographical variati.on in prevalence of infection with Onchocerca volvulus is total ranging from practically one hundred per cent. by the age of twenty to coryleEe absence. This is presunably a reflection of the variable density of the vector flies. But other factors may be involved, and even the fly density nay be affected by ecologic factors which are not currently recognized and might be amenable to manipulation. oNcHo/wP/75.15 Page 4 7. l,latheoati 1 models The topic of mathematical models deserves consideration. Such models can be quite helpful in guiding and stimulaEing the investigatorrs thoughts about the dynamic Processes involved in an infectiouq disease. They are unlikely to lead to results of much immediate practical value. Mills5 has published a study that will be useful in the development of a model of onchocerciasis spread and maintenance in the fly and human hosts, although he did not carry it to the point of developing equations for the spread or dynaoic equilibrium of the disease. Ile estisuted the number of infective larvae of O. volvulus to which a population is exposed in unit Eime for the three bioclimatic zones of West Africa. Finally, there have been a number of WHO publications that can contribute in a general riray to this topic. I\uo of these, published recently in the Technical Report Seri-es, are nnmber 5LO, Lg72, entitled Statistigal Principles in Public ltealth Field SEudieso and nuuiber 5OL, Lg72, entitled Vector Ecology./ 8. Comunity surveys Investigations by two groups of workers have recently contributed heavily to our epideniological knowledge. One was by Anderso|,^Fuglsang and their collaborators from the London School of Hygiene and Tropical l.Iedicine.8,9 It was in the form of a cross-sectional survey of 3824 persons in 22 groups of villages, InElny of them examined on two occasions, a year apart. The roain purposes vrere comparison of populations having 0. volvulus infection and those without infection, and coryarison of characterisEics of the disease as seen in rain forest and savanna counEry. The extremely high prevalence of infection and of blindness were stressed. It was hypothesized that the differences observed between Ehe disease in rain forest and savanna night be due either Eo variations in host response resulting from differen- ces in the intensity and pattern of infection, or possibly to differences in the pathogenicity of the worms found in the two regions. The work of Buck, Anderscn, Thomas and their colleagues from the Johns Hopkins School of Hygiene and Public 11.61ghlo-14 is also noEeworthy. Their initial observations in Chad were part of a comprehensive health study. From these observations, suPPorted by later findings on a second visit and by observations in other areas, came imPortant new information on micro- filaruria, ocular manifestaEions, diurnal periodicity of microfilaria in the skin, the effecE of treatment with diethylcarbamazine on the abundance of microfilaria in the skin and secre- tions, and the question of the effectiveness of nodulectomy. The reports of these tr^ro groups, as well as oEhers, show how hypotheses arising from initial exploratory surveys can lead to further valuable discoveries dealing vrith Pathogenesis, therapy, and other clinical as well as epidemiologic aspects of the disease. The value of the surveys is enhanced if the team is able to return and repeat the observations, with appropriate modifications, at a later time, in order to examine the natural history of the disease and to test hyPoEheses suggested by the first visit. 9. Surnmary. Sore purposes of epidemiological study are programne; surveillance; evaluating the success of preventive campaigns; and measurement of risk factors. Appropriate methods to use toward Ehese goals in epidemiological research on onchocerciasis are considered. standardization of examination procedures is recorrmended, in order to increase the comparability of different st,udies. coryuter-aided mapping and nathematical models are sugSested as topics for study. pasE experience suggests, however, that continued reliance on conrnunity surveys' especially if observaEions are repeated and exEended in the sane cormunities at a later time, and the adoption of the case-control method, offer the greatest promise' Preparatory Assistance ltission: Geneva, L973 REFERENCES Onchocerciasis control in the Volta River Basin Area. 2. Hamilton, P. J. S., Marshall, T. F. deC, Anderson, J. & Fuglsang, H. (L974) Observer variation in onchocerciasis, Tlans. roy. Soc. trop I'led. Hyg., 68,187-189 1 3 5 oNcHo/wP/75.15 page 5 Ann. Int. LIed.Sartwe1l, P. E. (L974) Retrospective studies: A review for the clinician, 81, 381-386 4. Hopps, H. C. & Cannon, H. L., eds (L972) Geochenical environment in relation to health and disease, N Y Acad Sci ., 129, 32s-334 !1i11s, A. R. (1969) A quantitative approach to the epidemiology of onchocerciasis i-n West Africa, ltans. roy. Soc. trop. Med. Hyg., 63, 59L-6Oz 6. World Health Organization, Statistical principles in public health field studies (L972) Wld Hlth Ors. techn. Rep. Ser ., No. 51O World Health Organization, Vector ecology (L972) !.I1d lllth Oqg. techn. Rep. Ser. , No. 5O1 Anderson, J., Fuglsang, H., Hamilton, P. J. S. & I,larshall, I. F. deC (L974) Studies on onchocerciasis in the United Cameroon Republic: 1. Comparj.son of populations with and without O. volvulus Trans. roy. Soc. trop. !Ied. Hyg. ,68,190-208 7 8 t 10 9. Anderson, J. et al. (L974) Studies on onchocerciasis in the United Cameroon Republic 2. Coryarison of onchocerciasis in rain-forest and Sudan-savanna, Irans. roy. Soc. trop !Ied. Hve., 68, 209-222 Buck, A. A., Anderson, R. I., Kawata, K. & Hitchcock, J. C. (1969) Onchocerciasis: Some new epidemiologic and clinical findings. Results of an epidemiologic study in the Republic of Chad, Amer. J. trop. l,led. Hyg. , Y' 2L7-23O 11. Buck, A. A., Anderson, R. I. in Chad: Epidemiology, , Sasaki, T. T. & Kazuyoshi, K. (1970) Health and disease culture and environnent , Johns llopkins Press, Baltimore, United States of Arerica L2. Buck, A. A., Anderson, R. I., Colston, J. A. C., I{allace, C. K., Connor, D. H., llarrmn, L.8., Donner, M. W., & Ganley, J. P. (1971) }ticrofilaruria in onchocerciasis. A clinical and epidemiological follow-up study in the Republic of Chad,8u11. WId. Itlth 0rg. , E, 353-369 13. Anderson, R. I., Thomas, D. B., MacRae, A. A. & Buck, A. A. (L975) Onchocerciasis: Prevalence of microfilaruria and other manifestations in a village of Cameroon, Aroer. J. trop. l'fed. Hyg.,4,66-70 L4. Fazen, L.8., Anderson, R. I., Marroquin, II.F., Arthes, F. G. &Buck, A. A. (L975) Onchocerciasis in Guateoala. 1. Epiderniologic studies of microfilaruria, 44qf. J. trop. l.led. Hyg., 4, 52-57 I a!r