These guidelines were developed by the Scientific Working Group on Epidemiology of the UNDP/World Bank/WHO Special Programme on Research and Training in Tropical Diseases, World Health Organization, Geneva.
These guidelines focus specifically upon ethical aspects that have posed particular problems in epidemiological studies of communities in developing countries. They are intended as a supplement to the Declaration of Helsinkia and the WHO/CIOMS guidelines on research involving human subjects,b and should not be viewed in isolation. They represent provisional guidelines for consideration, not rules for execution.
aWHO Chronicle, 30: 360-362 (1976).
b Since the first printing of this volume, CIOMS has published two works of particular relevance to Appendix 1: International guidelines for ethical review of epidemiological studies (Geneva, CIOMS, 1991); and the revised CIOMS/WHO International ethical guidelines for biomedical research involving human subjects (Geneva, CIOMS, 1993).
The international declaration concerning ethics as set forth in the Declaration of Helsinki, adopted by the 18th World Medical Assembly in 1964 and revised by the 29th World Medical Assembly in 1975, covers well most of the important ethical considerations concerned with clinical research in human subjects. The WHO/CIOMS proposed international guidelines for biomedical research involving human subjects were framed with special reference to developing countries and focus on problems of informed consent from certain disadvantaged subjects, such as children, the mentally ill or prisoners, and broadly ethical problems of developing countries and community-based research. Further, the guidelines discuss ethical review procedures at institutional and national levels and touch upon the need to inform communities in which research will be undertaken and upon the need for compensation from accidental personal injury.
The ethical concerns of controlled clinical trials, which are well outlined in these documents, include some that are also important to epidemiological studies. In particular the ethical issues surrounding three matters - informed consent, use of control or comparison (e.g. placebo) procedures, and trials of drugs or other materials that have not been approved for use in humans in the country of origin - all may be of great importance in epidemiological studies in developing countries, but are not specifically reviewed here.
Level of review
Proposals for research activities should undergo several levels of procedural review, as described below.
The fundamental level of review is the individual person; every participant in a study must provide informed consent to take part in the research. The many complex issues that informed consent may involve are generally well covered in the above-mentioned documents. Illiteracy and differing cultural concepts of health and disease do not alter the basic principles of informed consent. Written consent may be a legal requirement in some countries, but from the ethical viewpoint it is neither necessary nor sufficient. Written consent per se is not the requirement; the critical need is that the person consenting has an understanding of the procedures, including the benefits and hazards. Informed consent should not be considered as an inhibiting factor in research: indeed, a properly informed subject may be a much more useful participant.
The second level of review, which may be of particular importance for some types of epidemiological research, is a review by the community. It is not simply a matter of consent; it is a matter of understanding. For a research programme involving a community, the community and the researchers should be working in partnership since the purpose of the research should be mutually rewarding. A well informed community will not only be more cooperative, but also more useful in its collaboration.
The third level is that of the responsible institutional or national review body and is well covered in the WHO/CIOMS guidelines.
The final level of review for externally sponsored projects (i.e., those funded by an external national or international agency) should ensure that the research protocol conforms to the requirements of the Declaration of Helsinki and that the approaches to the three prior levels are documented.
Detailed procedures for obtaining individual consent or community understanding will vary considerably and must be left to the responsible institution and/or national review bodies and to the principal investigator.
Obtaining approval at all levels and, in particular, obtaining the informed consent of individuals in no way reduces the responsibility of the investigator to these individuals and their communities. A checklist of suggested areas for review at each level is given at the end of this appendix.
Issues of particular concern to epidemiological studies in developing countries
Health care expectations of the community
Epidemiological studies made in a community or other defined population in developing countries may involve important considerations that are not encountered in clinical research. In many areas, the only medical care available to the population is that of traditional practitioners, and modern health care with the infrastructure required for its delivery may be minimal or absent. The conduct of a health-related research programme brings with it the quite reasonable expectations on the part of the people that some kind of health care will be provided. One-shot, “bleed and fly” safari-type studies have often produced unfulfilled expectations in the people, leading to disappointment and reduced cooperation for future studies. Although one-shot studies have had their usefulness - and may not raise ethical issues when directed at solving a specific health problem such as an epidemic - they are generally unacceptable today. Such studies, usually carried out by outsiders, in addition to raising expectations among the people, make no long-term contribution to the health infrastructure or the research capabilities of the country. Almost always the same information can be better obtained through direct cooperation with institutions in the host country.
Longitudinal community-based studies inevitably raise the expectations of the people for improved health care. There are no clear answers to the vexing questions raised by these expectations. The first step is that a reasonably full understanding must be achieved between the community and the researchers before, during and after the research. Frequently, during a longitudinal, community-based study, the health infrastructure must be strengthened in order to obtain information about the health of the people. When this is the case, every attempt should be made to strengthen it on a continuing basis by such means as training, improved record-keeping, development of sampling procedures for future use. In general, it is understood that the actual provision of health care is the responsibility of the national health services and community: therefore, the researcher must work closely with both and resolve these issues as fully as possible.
Involvement of local personnel
Related to this is the need to strengthen the research infrastructure on a continuing basis whenever possible. At the least, this means as much involvement as possible of local personnel in the early stages of design of the protocol, its implementation, the collection of data, and the analysis and feedback of the information to the community and the government.
Control or placebo groups
In establishing the effectiveness of virtually any intervention procedure (vaccine trial, mass drug distribution, etc.) on a community basis, the need for control or placebo groups may create special problems. The basic ethical issues are the same as those in clinical trials, but the procedures used may be different. Important principles are (1) that the control group receive the best currently established form of intervention, if one exists, and (2) that if the new intervention procedure is demonstrated to be better (e.g. higher benefit, cost ratio) then it should be provided to the control population as soon as possible.
All the people concerned should be fully informed about the proposed research. They should be provided with a clear explanation of the investigation, the reasons for undertaking it and the possible implications. Any coercion is unacceptable.
Sometimes there may be practical problems in determining who truly represents the community; occasionally it is necessary to work through dual channels where there may be both legal and traditional authorities. Generally full discussions with community leaders - before, during and after the completion of the research - are both necessary and very helpful. Again to be emphasized is that the ultimate objective of research is to improve the health of those in the community; hence a cooperative partnership arrangement to ensure full understanding on the part of both the researchers and the community is vital.
Use of past medical records
Epidemiological studies carried out from a hospital or clinical base rarely raise ethical issues that are different from those of any other clinical research investigation. Although there has been some concern in the United Kingdom and the USA concerning the use of past medical records without specific individual consent of both the doctor and the patient as a potential “invasion of privacy”, it is no longer considered an issue provided that there is no identification of individual patients and that complete confidentiality is assured. As long as these principles are adhered to, the issue seems unlikely to become a problem in developing countries.
Anonymity of the community
The anonymity of the community, in general, should be protected with the same degree of concern as that given to the protection of the individual. If, as often happens, sensitive data emerge in the course of the studies of a community, it will be the responsibility of the investigator to use the utmost discretion in relating these data. However, sometimes the location and the circumstances are important to an understanding of the research that has taken place. Very often no purpose is served in not revealing the identity of the community. Indeed sometimes a community may take pride in being associated with the research. The best approach is to have continuing communication with the community as suggested above in the community review procedures (level 2 review).
It has sometimes been forgotten that informed consent may apply to observation of personal behaviour just as with any other research method affecting human subjects. The use of hidden observers and secret observation procedures (including photography) is rarely justifiable in behavioural research. When observers are required, as in research into human contact with water, their existence should be made known, but they can, of course, be stationed inconspicuously without being hidden. Experience has shown that such observers are eventually accepted - and even ignored.
In wide-scale application of measures to control vectors or intermediate hosts of disease organisms, the effect on the community and the environment of the methods used must be anticipated and carefully monitored.
Ethics in training
Training at all levels should include consideration of the ethical issues that will be involved in the trainees' future duties.
The following is a checklist of suggested areas for review at each level:
• Parent institution; funding organization
- scientific merit of the study
- consistency with ethical guidelines
- monitoring of results
- implementation of recommendations.
• National review committee
- scientific merit of the study
- consistency with ethical guideline
- use of controls
- confidentiality of records
- anonymity of subjects
- use of non-intervening observers
- cost to the community under study
- means of communication to the community of the nature of the research
- nature of informed consent
- acceptance of the research by the study community
- potential conflict and competition between proposals
- termination of controlled trials
- end points in the research process
- implementation of recommendations
- communication of results to the community.
• Local community
- nature and necessity for research
- possible harm and possible benefits to the community
- communication of issues to the people
- obtaining of consensus and the non-desirability of coercion
- monitoring of results
- implementation of recommendations
- end points.
- informed consent is required from all individuals.
These guidelines were originally drafted by Professor H. M. Gilles, Dean, Liverpool School of Hygiene and Tropical Medicine, with assistance from Professor F. Dunn, University of California, Berkeley. A number of useful suggestions from members of the Scientific Working Group on Epidemiology and the secretariat of the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases have been incorporated into the present version. Special thanks are due to Professor I. Riley, formerly Professor of Community Health, University of Papua New Guinea, for his comprehensive review and for the checklist of suggested areas for review and to Dr P. Rosenfield, formerly Secretary, Scientific Working Group on Social and Economic Research, for her compilation and analysis of the many useful suggestions submitted.