Ethics in occupational health research in South Africa

S Naidoo1, MH Ross1,2, A Dhai1, 


1 Faculty of Health Sciences, University of the Witwatersrand, Johannesburg


 2 National Institute for Occupational Health,  National Health Laboratory Service, Johannesburg


Corresponding author: S Naidoo                  

e-mail:  shan.naidoo@wits.





Occupational health research encompasses the same ethical issues that apply to all research on human subjects, together with unique challenges related to the workplace setting. This article covers the main principles of research ethics relating to the occupational environment, illustrated by the authors’ experiences related to ethical challenges in conducting occupational research, and reports an analysis of submissions made to the Research Ethics Committee of the University of the Witwatersrand during the period 2004 to 2006 and some lessons learnt in that regard. It also provides an overview of the inter­national and national ethical guidelines pertinent to occupational health research with a view to assisting researchers in doing ethically acceptable research.



Occupational health research is essential in the establishment of evidence-based practice in the workplace and is integral to improving health and safety at work.1 It comprises a combination of clinical occupational medical research and public health oriented research and thus encounters the same ethical issues as all health and other research on human subjects. In addition, the workplace setting necessitates the consideration of specific ethical issues. These are related to the needs, interests and vulnerability of the workforce being studied, the individual versus the collective benefits and the potential health care relationship between the researchers and the research participants. In essence, ethics in research is primarily protecting vulnerable participants from possible harm. Respect for the dignity, safety and well-being of participants should be of main concern, and culture, language, beliefs, perceptions and customs must all be taken into consideration.2,3

Occupational health research must, like all other health research, comply with national and international ethical norms and standards and should be approved by a Research Ethics Committee (REC) or Institutional Review Board (IRB) prior to being conducted. The imperative for all health or human related research to be ethically reviewed grew from the experience of the holocaust during the ­Second World War, when it was found that researchers ignored the interests of research subjects. As a consequence, the Nuremberg Code was formulated in 1947.4 The fundamental principles were that research participants should be informed about the nature and outcome of the research, be free to decide whether or not to participate without any effect on the care they received, and the benefits should outweigh the risks of the research. This was further refined, other principles were added and the Declaration of Helsinki was formulated in 1964. This Declaration has been revised five times with the latest revision being in 2004. It is today regarded as the international benchmark for ethics in research.5 South Africa has its own set of guidelines, ‘Ethics in ­Health Research: Principles, Structures and Processes’. It draws extensively from the international documents and was launched by the ­Department of Health (DoH) in 2005.2  

The four main principles that guide ethical research are: respect for persons (autonomy), beneficence, non-maleficence and distributive justice.6 For occupational health research these hold particular importance because workers are often vulnerable or potentially coercible populations. In addition, researcher and scientific integrity are further considerations to ensuring ethical research. The dis­cussion that follows addresses these principles as they pertain to occupational research ­specifically in the South African context.


Respect for persons

This principle is applied through the appropriate use of informed consent. For the research participant to give informed consent, it implies that the person has agreed to participate through their own volition and with a reasonable knowledge of the possible risks and benefits. In addition, participant responsibilities towards their involvement should also be explained.

Occupational health research is primarily conducted on workers in the workplace and care should be taken with regard to the power relationships and the workers' vulnerability. The potential for misuse of power in the research setting cannot be ignored. This is particularly so if the research is perceived to have come from management. Another important aspect is the need to involve the worker representatives, sometimes at both a national and local level of organized labour movements. Researchers should attempt to involve the workforce and management at all stages of the research from planning to feedback of the results and recommendations. This principle of involvement and collaborative partnerships with the community is of particular importance in the South African context, given its history. However, permission to conduct the research by worker representatives does not imply or replace informed consent from individual ­participants. Such informed consent by participants is essential in prospective research with direct participant involvement. It should also be considered when there is indirect contact such as record reviews, even though these may be retrospective. Where participants are ­subordinate in a hierarchical relationship with the researcher, or where they are patients of the researcher, it is advisable for consent to be obtained by an independent ­person.

In the mining industry, deceased miners are entitled to an autopsy under the Occupational Diseases in Mines and Works Act (ODMWA),7 primarily for potential compensation payments to be made to the surviving dependants. For occupational health research to be conducted on autopsy data, appropriate informed consent needs to be obtained for pathological specimens (either archival or prospective). Thus prospective post mortem studies using non-ODMWA controls or ODMWA post mortems for tissue other than those for which consent normally applies (heart and lungs), require an information sheet on the proposed study for the person consenting to the post mortem to give ‘informed consent’ for the procedure or specimen use.

The use or abuse of incentives offered to potential study participants could also lead to an erosion of the voluntariness of informed consent. This could constitute an undue influence, have a coercive effect and could compromise the potential participants’ ability to make an ­independent informed decision. However, it is unethical for participants to bear unreasonable costs of their research involvement. Hence, reimbursement for travel, food, inconvenience or any ‘out of pocket’ expenses may be appropriate.

A major concern in South Africa in every instance of obtaining informed consent in occupational health research is the ability of the potential study participant to understand the risks and benefits i.e. are they really informed? Researchers need to focus attention on ­creating an understanding of the nature of the scientific enquiry and they should indicate this to the ethics committee as clearly as possible. It is important that potential participants are given information based on a full risk/benefit analysis. Oral ­explanation should always be provided, and it is essential for there to be a written document for the participants containing a full, objective and comprehensible description of the purpose, design and pro­cedures of the research plus its risks, discomforts and inconveniences and details of an independent person to consult. Moreover, informed consent is a process and not just a signature on a document at a particular moment in time.

Respect for persons also entails respecting their ­privacy and confidentiality. In its simplest form privacy is concerned with access to personal records, while confidentiality refers to the use of personal information once it has been disclosed. The researcher should ensure that where personal information about participants or a community is ­collected, stored, used or destroyed, this is done in ways that respect the privacy and confidentiality of participants or the community and any agreements made with them.2 Data of a confidential nature (e.g. ­patient records, personal questionnaires) must not be disclosed except where consent has been obtained to do so or there is a legal or professional obligation to disclose (e.g. notification of a notifiable disease under occupational health legislation). If there is any like­lihood of legal obligation to disclose data obtained through the research pro­cess, the participants should be informed before consent to participate in the research is obtained. Participants who belong to small groups are identifiable cohorts and require additional protection. The confidentiality of information is critical to occupational health research. Researchers must design the research to avoid situations where they may be forced to divulge confidential information. This is contrasted with anonymous data in which even the researchers are unable to identify which persons contributed to what information. However, in contrast to the obligation to keep information confidential, researchers also have an obligation to share the findings of the research. Research findings are typically presented as aggregate data and do not reveal the identities of individual participants. There is an ethical obligation to share the findings with the greater society, particularly negative findings, because the information is potentially useful but often goes unpublished. Publication in reputable journals usually requires ethics clearance of the research by an ethics committee of ­standing.6


Risks and benefits of research

The principles of beneficence (doing good) and non-­maleficence (preventing harm) are also critical to ethical occupational health research. Benefits should preferably be for the participants but are frequently more for the ‘greater good’ of the workforce in the future. The risks to participants of the study vary according to the type of research. This could include physical harm from a pro­cedure; illness from medication, intervention or exposure; psychological harm or emotional trauma; workplace disruption; loss of morale, negative consequences on relationships or even on employment because of detection of disease or failure to keep information confidential; and social stigma or embarrassment from loss of privacy. A risk/benefit analysis should precede the study and should take full consideration of the benefits and harms beyond the ­duration of the research, particularly in the case of stigma or job loss. It is an ethical imperative that participants who might be placed at undue risk be excluded.3


Distributive justice

It is important to remember that the participants are not the only stakeholders whose interest should be con­sidered. The benefits and risks of research should be distributed equitably in society. There should be a reasonable likelihood that the population on whom research is to be carried out will benefit from the research and its results. Where participants are selected to be in the study but are unlikely to be beneficiaries of subsequent applications of the research, adequate motivation and justification for this must be provided to the REC. Research proposals should indicate whether long-term interventions would be provided to participants after completion of the study.2


Oversight of research

Ensuring that a researcher adheres to these principles of ethical research is entrusted to the REC or IRB. The ­primary role of such committees is to protect the safety and well-being of human participants involved in research, i.e. to facilitate ethical research. Additional duties are to advance knowledge for the common good, and protect researchers from undue criticism, and ensure that researchers are treated fairly. The functions of these committees or boards are to review study proposals, inform investigators when their study protocols fall short of conventional ethical standards, approve ethically sound protocols and monitor studies over their duration to ensure that ethical standards are adhered to throughout the course of the study.

Section 73 (1) of the National Health Act8 affirms that every institution, health agency and health establishment at which health research is conducted must establish or have access to a health research ethics committee which is registered with the Health Research Ethics Council. Most academic institutions already have such a body to review research proposals. Most professional journals also require that a manuscript submitted for publication of research study findings indicate REC or IRB approval. Occupational health research in South Africa is usually submitted to the local REC (where the researchers are affiliated or the workplace is situated).

Table 1 shows an analysis of the research protocol submissions to the Research Ethics Committee of the University of the Witwatersrand during the period January 2004 to August 2006. In keeping with occupational health research internationally, studies are usually epidemiological rather than clinical i.e. observational rather than intervention studies such as randomized control trials.1 The number of submissions reflects on average an 8% of total submissions to be of an occupational health nature and a somewhat surprising increase to 9,6% in 2006 (up to end of August 2006). There is an expected increase in overall occupational health research under the auspices of the Research Ethics Committee of the University of the Witwatersrand in this last year. This may relate to an increase in research activity in the occupational health research field as well as the now locally legal requirement to have research subjected to ­ethical review by RECs. The Research Ethics Committee of the University of the Witwatersrand has learned some ­important lessons in the past few years in its review of occupational health research and this is illustrated in ­Table 2.


Record reviews and epidemiological studies in occupational health

Epidemiology is the foundation of public health practice and research. However, regulations designed to protect the rights of individual clinical research subjects require modification in the face of public health focused research.9 The Council of International Organisations of Medical ­Sciences (CIOMS) noted that individual informed consent was not always practical in epidemiological studies and that consent could be waived under these special ­conditions.6 This paved the way for the use of retrospective record reviews without informed consent. The South African DoH Guidelines recommend that data collected in epidemiological studies should preferably be in the de-identified form. Data should be stored with personal identifiers (identified or potentially identified data) only if absolutely necessary and where approved by a research ­ethics committee with sound justification.2 ­Section 16 of the National Health Act affirms that a health care provider may examine a user’s (patient’s) health records for the purposes of study, teaching or research with the ­authorization of the user, head of the health establishment and the relevant health research ethics committee.8 ­However, where no information as to the identity of the user is reflected, it will not be necessary to obtain user authorization. In the South African occupational health research setting this poses particular challenges and also provides immense research opportunities. The challenge is to ensure that the risks to subjects are minimal, the recorded data is anonymous and research on the subject matter could not easily be otherwise conducted. An important caveat is that the review is of data already collected at the time of formulating the research proposal and those records are not going to be used for ongoing consultations and are to be archived. For example, occupational health records can be accessed for research as long as the participants have left employment or are no longer visiting the occupational health service on an ongoing basis. In the latter circumstances, informed consent is imperative and signed permission to use past and/or future data should be sought from individual workers. Another example, an occupational physician might wish to review clinical records retrospectively to identify workers with a specific clinical disorder in order to contact them and seek their participation in a study. Record reviews of patient admissions to hospitals or ­clinics are dealt with in the same manner. Some institutions, including the University of the Witwatersrand have pro forma information sheets and consent documents that can be easily accessed and used.



Surveillance has served as a critical element in the practice of public and occupational health. Surveillance has imposed on health care institutions and physicians, particularly, the duty to report cases to confidential registries.10 Almost invariably such reports have included the name of the afflicted. For most of the twentieth century, such practices went largely unchallenged until the AIDS epidemic raised some concerns. There was resistance to name-based reporting particularly as it pertained to HIV.9 However, at the end of 1990 the CDC declared its support for HIV reporting stating that it could “enhance the ability of local, state and national agencies to project the level of required resources” for care and prevention services.11 The American Medical Association endorsed the reporting of names as well. More recently, SARS (Severe Acute Respiratory Syndrome) and avian influenza have posed an occupational and public health threat and governments throughout the world have enforced stringent surveillance and quarantine measures to control epidemic spread and the urgency to publish precluded all the operational ­research from receiving formal ethical clearance prior to publication. As the threat from SARS receded, ­surveillance measures waned and became less stringent while the need for submission of further research involving ­human subjects for ethics approval increased.9,12 Governments can impose restrictions on individuals who represent a serious risk to others. Problems emerge when the risk is uncertain. In the occupational health setting, surveillance of certain conditions is a widely acceptable public health practice and the reporting of these, even if found through research, is not only acceptable but in many cases mandatory. What is suggested, especially if publication of findings as a form of research is anticipated, is that RECs are timeously informed. The DoH Guidelines2 state that public health surveillance should be distinguished from public health and epidemiological research. Its role is to monitor the health status of the community (workplace or non-workplace related), known risk factors and emerging threats to community health. Its purpose is to facilitate a prompt, effective and corrective response. There is often a requirement or authorization by law for these agencies to carry out health surveillance. Under certain circumstances, however, surveillance data may be used for research purposes and considerations pertaining to data privacy and record review should apply.


Qualitative research

Ethics issues in qualitative occupational health research pose particular challenges with increasing use of activities such as unstructured interviews and focus group discussions. The limits of confidentiality must be stressed to the participants when recruiting them for research involving focus group discussions. Generally field investigators themselves are the measuring instruments and great care needs to be taken in this type of research. There are different approaches or stances regarding ethical issues in qualitative research. These include the absolutist stance, relativist stance, contextualist stance and deception model. Denzin and Lincoln caution against the deception stance where an investigator may use any method, including ­deliberate self misrepresentation, to do research. It is important that all qualitative research also be submitted for ethics approval.13


Avoidance of deceptive research

The DoH Guidelines state that as a general principle, deception of identifiable participants, concealment of the ­purposes of research or covert observation are not considered ethical because they are contrary to the principle of respect for persons and the obtaining of informed consent. However, there may be exceptional circumstances when studies cannot be conducted without concealment, deception or covert observation of participants.2 This usually occurs when the provision of detailed information will compromise the scientific validity of the research and there are no suitable alternative methods by which the desired information can be obtained. The Guidelines further state that there should be no increased risk of harm, there should be adequate and prompt disclosure including debriefing and allowing for ­participants to ­withdraw data and such activities should not erode the relationship between ­researchers and research in ­general and the community at large.   


International collaboration

There is a concern about international agencies conducting research in developing countries, particularly clinical research, and the application of standards of one country or not, as the case may be, in another.14 A classic example as experienced by the authors is that the ­National Institutes of Health (NIH) and the Centers for Disease Control and Prevention (CDC) do not offer insurance against injury due to their clinical trials done locally, ­although this is a requirement worldwide. It is the responsibility of local researchers to ensure that local participants are covered by insurance even where the international collaborator does not provide for this. Collaborative partnerships right at the outset, between researchers, sponsors, policy makers and communities in the developing countries help to minimize the possibility of exploitation. This would also serve to ensure that the developing country determines for itself whether the research is acceptable and responsive to the community’s health problems, will have a lasting impact, and will be likely to influence policy making, allocation of scarce health care resources and respect cultural dif­ferences.3 Moreover, determi­nations on issues around ­intellectual property rights would also be dealt with at the very beginning of the collaborative partnership by all ­respective parties concerned.3


Conflicts of interest

The researcher-participant association is a fiduciary ­relationship. Hence, when a set of circumstances e.g. researcher financial interests compromise or appear to compromise or have a potential to compromise the researcher’s professional judgment towards the participant, a conflict of interest situation in research could develop. The DoH Guidelines stipulate that the researcher discloses sources and extent of funding to the research participants, the research ethics committee and, where appropriate, to the relevant regulatory authority. Commercial ­affiliations and/or financial interests at the time of proposing and reporting the research must also be disclosed.2


South African requirements

According to Section 12 (2)(c) of the Bill of Rights of the Constitution of South Africa Act, No 108 of 1996: “Everyone has the right to bodily and psychological integrity which includes the right not to be subjected to medical or scientific experiments without their informed consent”.15 Chapter 9 of the National Health Act is unambiguous on the incorporating of ethics in health research.8 Accordingly, ethical research is now a statutory requirement in the country. There are very stringent requirements with regard to informed consent, with section 71 stipulating that written consent is requisite after the potential participant has been informed on the objectives and possible positive and negative consequences of the research. In the case of minors, consent from the parent or legal guardian together with minor assent will be necessary for therapeutic research. Where minor participants are involved in non-therapeutic research, additional permission from the Minister of Health will be necessary. As the National Health Act is not explicit as to the process of how the latter is to be obtained, nor on the exact age of the minor, it is hoped that the regulations to this section (still to be released) will clarify the situation. The National DoH Guidelines, ‘Ethics in Health Research: Principles, Structures and Processes’ are a response to Chapter 9 of the National Health Act and it is advisable that all researchers familiarize themselves with the regulations stipulated in both documents.2,8 (Also available at The Medicine Control Council approves the scientific aspect of clinical trials of new drugs.16 Usually, submissions for MCC and REC are done in parallel and final approval will only be issued once both provisional approvals have been obtained. An additional set of guidelines available in South Africa are those of the  Medical Research Council.3 Several international guidelines are available and easily accessible.6,17,18 The Inter­national Commission on Occupational Health (ICOH) also has an international code of ethics for occupational health professionals which covers research.19 Ultimately research protocols involving human participants need to be approved by an accredited Research Ethics Committee in South Africa for any local occupational health research to take place.2,20



The ethical review process for occupational health ­research is still in its infancy in South Africa. However, occupational health research by its very nature, i.e. research on workers as a vulnerable group, should strive to be as ethical as possible. It needs to meet international norms and practices and also local requirements. This paper attempts to inform prospective researchers in occupational health of the principles that guide ethical research, the ethical challenges of some types of research that they may consider and the role that the RECs play. It also makes reference to international and national guidelines that prospective researchers can easily access to improve the ethical standards of their research.


Conflict of interest statement

The authors, who are all members of the Human Research Ethics Committee (Medical) of the University of the Witwatersrand have no conflict of interest to declare.



The authors thank the Human Research Ethics Committee (Medical) of the University of the Witwatersrand for ­permission to analyse the committee agendas.



1. Faculty of Occupational Medicine. Guidance on ethics for occupational physicians. 6th edition. London; Faculty of Occupational Medicine of the Royal College of Physicians; 2006.

2. Department of Health. Ethics in health research: Principles, structures and processes. Government Printers. Pretoria; 2004.

3. South African Medical Research Council. Guidelines on ethics for medical research: General principles. Pretoria: South African Medical Research Council; 1999.

4. Nuremberg Military Tribunal. The Nuremberg Code. JAMA. 1996.

5.Williams J. World Medical Association Medical ethics manual. Geneva: World Medical Association; 2005.

6. CIOMS. International ethical guidelines for biomedical research involving human subjects. Geneva: CIOMS; 2002.

7. South Africa. The Occupational Diseases in Mines and Works Act, Act No. 78 of 1973. Pretoria: Government Printers; 1973.

8. South Africa. The National Health Act. Act No. 61 of 2003. Pretoria: Government Printers; 2003.

9. Bayer R, Fairchild AL. The genesis of public health ethics. Bioethics. 2004;18(6):473-92.

10. Koh D, Aw T-C. Surveillance in occupational health.  Occupational and Environmental Medicine. 2003; 60: 705-10.

11. Centers for Disease Control and Prevention. Update: Public health surveillance for HIV infection – United States, 1989 and 1990. MMWR; 1990.

12. Centers for Disease Control and Prevention. Efficiency of quarantine during an epidemic of severe acute respiratory syndrome – Beijing, China, 2003. MMWR; 2003.

13. Denzin NK, Lincoln YS. Handbook of qualitative research. California: Sage; 1994.

14. Nuffield Council on Bioethics. The ethics of clinical research in developing countries. London: Nuffield Council on Bioethics; 1999.

15. South Africa. The Constitution of the Republic of South ­Africa, Act 108 of 1996. Pretoria: Government Printers; 1996.

16. South Africa. The Medicines and Related Substances Control Amendment Act No. 90 of 1997. Pretoria: Government Printers; 1997.

17. The Belmont Report: Ethical principles and guidelines for the protection of human subjects of research. The National Commission for the Protection of Human Subjects of Biomedical and Behavioural Research. OPRR Reports; 1979.

18. International Conference on Harmonisation. Good Clinical Practice: Consolidated Guideline; May 1997.

19. International Commission on Occupational Health. International code of ethics for occupational health professionals. First update. Rome: International Commission of Occupational Health; 2002.

20. International Committee of Medical Journal Editors. 1993 Uniform requirements for manuscripts submitted to biomedical journals. Med Educ. 1999; 33(1):66-78.



Download this Article


Email address
Forgot password?