Letter Criticizing Draft Report on Ethics of Research in Developing Countries

December 6, 2000

Harold T. Shapiro, Chairman
National Bioethics Advisory Commission
6705 Rockledge Drive  Suite 700
Bethesda, MD 20892-7979

Dear Dr. Shapiro:

We are dismayed and deeply disappointed that the National Bioethics Advisory Commission (NBAC) has seen fit to radically alter its draft report on the ethics of research in developing countries at the last minute to no longer require review by a U.S. Institutional Review Board (IRB) of U.S. government-funded research in developing countries. This ill-considered proposal would effectively remove the requirement for American ethical review of some American research and frequently leave participant protection solely to often inexperienced and unregulated foreign IRBs.

It is not as if current review practices are adequate. Research commissioned by the NBAC suggests that, according to U.S. researchers, only 22% of pharmaceutical/biotech studies in developing countries receive review from a U.S. IRB.[1] This is in part because the U.S. Food and Drug Administration (FDA) does not require such review. Offered the opportunity to improve this lamentable state of affairs by extending the requirement for U.S. IRB approval to studies subject to FDA regulation, the NBAC has opted instead to make recommendations that could actually weaken review procedures.

At the last NBAC meeting on November 22, the Commission for the first time circulated the following draft recommendation:

Clinical trials conducted in developing countries and subject to U.S. research regulations must, as a minimum, be given prior approval by an independent ethics review committee in the country where the research will take place. Researchers should include in the research protocol plans for facilitating communication between or among IRBs in the United States and with ethics review committees in the collaborating countries.[2]

Importantly, the first sentence of the previous draft of this section read: "Clinical trials must meet (sic) prior approval by an independent ethics review committee in the country where the research will take place as well as by a U.S. IRB (emphasis added)."[2] The italicized part of the sentence has now been omitted, meaning that developing country IRB approval alone could suffice, a dramatic departure from the current practices of U.S. government-funded researchers.

Under the new NBAC language, U.S. researchers from universities and government who now generally receive IRB approval from both the U.S. and foreign IRBs, could ethically conduct research in developing countries without any U.S. IRB approval. The onus of participant protection in such studies could be shifted from the U.S. to the developing world, where the ethical review infrastructure, as the NBAC documents clearly in its report, is often grossly inadequate. Yet it is to this infrastructure that the NBAC would entrust participant protection, even though it is seen as so inadequate that the NBAC report recommends the need for capacity building. Can we really expect countries with no IRBs or IRBs with little experience to turn down research, particularly with the prospect of millions of dollars of hard currency in the offing? Even the better-funded, longer-established, more-regulated U.S. IRBs, have been heavily criticized by the government's own Inspector General for inadequately protecting participants.[3] The most appropriate standard is that no researcher should ever perform a project that was not approved in his or her country of origin.

We are mystified as to the reason this dangerous change has been proposed. We have examined all three comments related to this recommendation that the NBAC has described as being "of particular interest," and none call for the removal of the requirement for U.S. IRB approval of research subject to U.S. regulation.

Information from a Previous Draft NBAC Report Deleted

The June 1, 2000 draft of Chapter 5 of the main report (Volume 1),[4] now changed to contain the recommendation that U.S. IRB approval no longer be required, contained evidence of the limitations of IRBs in developing countries. However, all of the material cited below has, for unclear reasons, been excised from the current draft:

In Morocco, no IRB exists and the Ministry of Health does not feel that it is necessary. Thus our approval came in the form of a letter from the Director of the Division of Epidemiology and no IRB review process was undertaken.

... the idea of creating an ethics review board within foreign countries that do not have such an institution is a very complicated issue, and I have severe reservations about such an undertaking ... Furthermore, from my experiences in Turkey, to force the creation of such an institution would only add to the already torturous bureaucratic maze one must go through in order to get a research permit.

Dr. Jean Pape, the Haitian researcher, said that when he began his research in collaboration with Cornell University, a review committee had to be created in Haiti because there was no committee at the institutional level anywhere. The national IRB in Haiti took a long time to create, but it was finally put in place in 1999.

Dr. Sana Loue told us that the notion of an unbiased, uninvested review committee is still something that is quite new to Uganda.

[Dr. Christopher Plowe of the University of Maryland] testified that 'the local [developing country] review process is really inscrutable for U.S. investigators.' ... Still, Plowe said, because he does not speak the local language and did not attend the committee meetings, he remains ignorant of how the actual process of review works, and if it works the way it is supposed to work. For that, he has to rely on his collaborators.

All of the material above was at one time deemed important enough to be included in Volume 1 of the report, but has now, for no discernible reason, been at best consigned to Volume 2.

Recommendations of NBAC Consultants Not Heeded

As part of writing its report, the NBAC commissioned researchers at Johns Hopkins University to conduct a survey of developing country and U.S. researchers.1 In our letter to you of November 13, 2000, we pointed out that the NBAC had ignored the recommendations of these consultants, who had recommended that tests of understanding be incorporated into the informed consent process.

With respect to the need for U.S. IRB approval, the NBAC has continued its practice of ignoring the recommendations of its own consultants. The Johns Hopkins researchers concluded: "Studies involving international collaborations need to be reviewed in both/all countries." Apparently this recommendation was too inconvenient to include or to even to be referred to in Volume 1 of the NBAC report.

Data from Developing Country Researchers Collected for NBAC Not Heeded

The NBAC report has tended to heavily downplay the useful data collected for it by Johns Hopkins, even though these data could place the NBAC's report on a more credible research-based footing. In fact, the last-minute revision to the recommendation on IRB review flies in the face of much of the data collected for NBAC. The following are direct quotes from the Johns Hopkins survey of developing country researchers. None are included in Volume 1 of the report, but all are contained in Volume 2 of the report, which includes the Johns Hopkins report and which has not been made widely available for public review.

Once we got [the IRB] set up in the university then we moved on to the Ministry (of Health) and it took a while ... I'm not saying that all Ministry people are bad, but some people ... (the government) took money from, these researchers, and turned their backs, so whatever happens to the patients, they don't care too much.

In [African country] there was no ethics or research committee by the time I got there and when there were a lot of researchers coming from abroad and calling themselves researchers who just came to the country and they did what they wanted to do and left. It took a while for us to push the government to the point (of addressing the situation) ... This has been addressed in most African developing countries by now because the patient is extremely important.

[The Ministry of Health is] More concerned about the money than (whether the study is) Ok for the people or not.

... Institutions that have universities attached to it and have a lot of people that have been outside and are (also) used to the developing world system, its more probable that you will have something comparable to what you have in the developed nations ... moving from those cities that have universities and just set up a collaboration with some private medical practitioner or some small group somewhere ... you might find ... that they have no control at all, and that is a huge problem for the government.

They (local IRBs) are not really concerned about ethical issues, they are looking (at) technical (issues). And you know, and who (is) giving you money, how much are you getting ... But now (we need to look at) the ethical aspects, what people are doing, is it right.

... it is a political approval. It is not an approval that is about ethics. It was more about whether we would be spies or we would be real researchers that would benefit (Asian country).

In resource poor countries the issues of corruption and 'kickbacks' for government officials was a concern in establishing and, particularly, in enforcing standards for research ethics at a national level. These concerns were relevant for all respondents. One respondent expressed their concern that external organizations took advantage of the resource poor countries in that they could dictate what they wanted by finding government officials that they could control." (Quote is from the text of the Johns Hopkins report, not a respondent.)

... they (pharmaceutical companies) get an institution somewhere that has a half-controlled person that would be willing to just take them in and do ... whatever kinds of studies they want to do.

... the companies when they want to market the new drug in a new market ... they would promote these trials, and for them to get into the market was to get the approval of the Ministry of Health and then find collaborators at the universities and then move on.

... but in terms of who is running these bodies and who is controlling what's really happening, you will be amazed. It is mostly people who have no idea about this. They just know it [ethics] is a word.

The biggest problem in developing countries is that our poverty puts us in a situation (where) the beggar has no choice.

The Johns Hopkins researchers also collected useful quantitative data in a field not hitherto characterized by a surfeit of such data. Developing country researchers who responded to the survey (approximately 200 of 500 or 40%) indicated that U.S. IRBs were more likely than developing country IRBs to indicate that confidentiality protections in the index study were inadequate (42% [U.S.] vs. 18% [developing country]; p=0.005), to comment on the need for a local language consent form (84% vs. 58%; p=0.0065) and to raise concerns about the complexity of the informed consent form (64% vs. 45%; p=0.070). Twenty-five percent of developing country respondents indicated that their studies were reviewed by neither the Ministry or Health nor by an IRB or equivalent. The NBAC consultants calculated that this could represent 15% of studies, about one-third of which were funded by the U.S.

Data from U.S. Researchers Collected for NBAC Not Heeded

Moreover, the NBAC has also ignored the data collected by its own consultants from U.S. researchers. The Johns Hopkins survey includes the following observations:

Some of [the developing country IRBs] do really quite a decent job, just as you would want them to be. And there are others that are completely rubber stamps, and nothing else ... Yes, there's an IRB, [but] I don't have any faith that there was any real review.

[Local IRB members] may be people that are not all equipped intellectually, culturally, scientifically to deal with the issues you are asked to deal with. And therefore, you introduce a false sense of security and conformity with the rule, with the letter, when the spirit is actually vacant.

In some cases, the developing country ethical review is actually a process of seeking permission to conduct research, and no ethical questions are raised at all. Developing country review boards are often more concerned about the financial aspects of the study than about ethics.

Useful quantitative data collected from U.S. researchers are also largely ignored by the NBAC in Volume 1. For example, according to the 333 U.S. researchers (response rate: 35%) who responded to the NBAC consultants' survey, 91% of university studies (characterized by the researcher's employer, not the funding source) and 100% of U.S. government studies underwent review by a U.S. IRB, compared to 22% of pharmaceutical/biotech studies. As we have suggested elsewhere, more general efforts to weaken ethical codes are most likely to be exploited by pharmaceutical companies. Moreover, according to the Johns Hopkins report, "79% of [U.S.] researchers believed that developing country IRBs sometimes or always are more concerned with politics than they are with protecting the interests of research subjects."

The data collected for the NBAC also confirm that U.S. IRBs make particular contributions to the ethical review process. Very frequently, they are more rigorous than the developing country IRBs. Among the 12% of U.S. respondents who had engaged in research that represented more than a minimal risk to participants, the U.S. IRB was more likely to raise questions about whether the intervention was too risky (18% vs. 2%), procedures for the control group (35% vs. 15%), use of placebos (32% vs. 7%), whether benefits offered might compromise voluntariness (25% vs. 8%), relevance of the research to the developing country (59% vs. 27%), post-trial availability (52% vs. 17%), complexity of the informed consent form (66% vs. 42%), and need for a local language consent form (84% vs. 63%). All findings listed here were statistically significant at p=0.02 or less.

This systematic failure to include relevant information in Volume 1 is deeply troubling. While honest disagreements over policy may be expected, it is not acceptable to sanitize the report by relegating inconvenient findings to a volume not made widely available for public review and which most observers will never read.

It is clear that Volume 2 of the NBAC report also contains some criticisms of the U.S. ethical requirements. In particular, the requirements were criticized for being inflexible, particularly with regard to the composition of IRBs. Some respondents did have positive things to say about developing country IRBs and indicated that the ethical review situation in some developing counties is improving. But the data collected here also make it clear that in numerous instances, the ethical review process in developing countries is deeply flawed and that U.S. review adds a critical additional layer of protection for research participants.

U.S. Advisory Committee Legal Procedures Violated by the NBAC

We are mystified as to why so fundamental a change could be introduced into the document at so late a stage in the document revision process, particularly because the meeting at which this change was made public was illegally constituted under the Federal Advisory Committee Act, which requires the provision of 15 days notice before a public meeting. When we brought the NBAC's failure to provide timely notice to the attention of the Commission, you reacted on the day of the meeting by canceling the official meeting, although the Commission continued to meet as if there was no cancellation. The meeting was scheduled for the day before the four-day Thanksgiving holiday weekend, ensuring poor attendance from both the public (Public Citizen was the only group that provided public testimony) and even from the commissioners themselves (only two of the 21 commissioners were present in person). Incidentally, the NBAC's next meeting, on December 7 and 8 in Washington, D.C., is also illegal because it, too, was not properly announced 15 or more days ahead of time.


Under current regulations, for research in a foreign country subject to U.S. regulations, "if a Department or Agency head determines that the procedures prescribed by the [foreign] institution afford protections that are at least equivalent to those provided in this policy, the Department or Agency head may approve the substitution of the foreign procedures in lieu of the procedural requirements provided in this policy."[5] The only agency known to have developed equivalence criteria is the U.S. Agency for International Development, but these have never been put into practice, to our knowledge. (Given the variability of the ethical review systems both between and within countries, it is difficult to imagine how such a determination could be made.) In effect, therefore, the equivalency language has never been utilized. Instead, foreign institutions have been required to obtain Single Project Assurances, which describe the process for local review for a particular study, but which, to our knowledge, do not supplant U.S. review.

The NBAC might have made an ethical argument that all U.S. research, whether or not it was subject to U.S. regulations, should have been subject to the two-IRB rule. (Whether the U.S. government has such authority legally is rather less certain.) Instead of improving ethical standards in this fashion, the NBAC has elected to lower standards below those currently recommended internationally. The NBAC proposal directly contradicts the Council for International Organizations of Medical Sciences (CIOMS) ethics document which requires: "After scientific and ethical approval in the country of the sponsoring agency, the appropriate authorities of the host country, including a national or local ethical review committee or its equivalent, should satisfy themselves that the proposed research meets their own ethical requirements."[6] It also is weaker than the Canadian requirements quoted in the NBAC report: "Research to be performed outside the jurisdiction or country of the institution which employs the researcher shall undergo prospective ethics review both (a) by the REB within the researcher's institution; and (b) by the REB, where such exists, within the legal responsibility and equivalent ethical and procedural safeguards in the country or jurisdiction where the research is to be done."[7]

We are certainly in favor of any effort to expand the capacity of developing countries to conduct ethical review and applaud recent efforts in this regard. Anything that converts international research studies into efforts more closely resembling true, equal collaborations in every aspect of the trial, including ethical review, is to be supported. But the uncomfortable fact, as documented clearly in Volume 2 of the NBAC report, is that there remain enormous deficiencies in developing and industrialized country review processes. To entrust the review of studies only to those in developing countries may comport with aspirations for future developing country ethical review capacities; it does not comport with the current NBAC-documented reality.

What started off as a promising attempt to address the challenging ethical issues in international research has now degenerated into a process whereby data collected for the Commission are essentially ignored and not made widely available for public review, the recommendations of those collecting data for the NBAC are discounted, findings that are uncomfortable to the research community and hence to some NBAC members are excised from draft chapters, meetings are constituted illegally, Commission consultants resign, report recommendations contradict the report's own ethical findings and major substantive changes are made at the last minute. The result is a document that could actually lower ethical standards below those currently accepted internationally.

Yours sincerely,

Peter Lurie, M.D., M.P.H.
Deputy Director
Public Citizen's Health Research Group

George Annas, J.D., M.P.H.
Global Lawyers & Physicians: Working Together for Human Rights
Health Law Department
Boston University School of Public Health

Troyen A. Brennan, M.D.
Human Subjects Committee
Harvard School of Public Health

Arthur Caplan, Ph.D.
Center for Bioethics
University of Pennsylvania

Dirceu Greco, M.D., Ph.D.
Professor, Internal Medicine
Coordinator of Infectious and Parasitic Disease Services
Chair, Federal University of Minas Gerais Ethics Review Committee
Belo Horizonte, Brazil

Michael Grodin, M.D.
Professor of Health Law, Psychiatry, Socio-Medical Sciences and Community
Medicine (Health and Human Rights)
Boston University Schools of Medicine and Public Health

Sidney M. Wolfe, M.D.
Public Citizen's Health Research Group


[1] Kass N, Hyder AA. Attitudes and experiences of U.S. and developing country investigators regarding U.S. human subjects regulations (draft report). Johns Hopkins University, June 15, 2000.

[2] Meslin EM. Memorandum to the National Bioethics Advisory Commission. Materials relating to public comments: international report. November 21, 2000.

[3] Office of the Inspector General. Institutional Review Boards: a time for reform. Department of Health and Human Services, June 1998.

[4] Macklin R, Page A. Chapter 5: Enhancing international collaborative research (draft report). National Bioethics Advisory Commission, June 1, 2000.

[5] 45 CFR 46.101(h)

[6] Council for International Organizations of Medical Sciences, World Health Organization. International ethical guidelines for biomedical research involving human subjects. Geneva, 1993.

[7] Tri-Council Policy Statement, Ethical Conduct for Research Involving Humans, 1998.