Can transnational research be ethical in the developing world?
23 December 2002 (The Lancet)
by Paul Farmer (Harvard Medical School and Partners In Health)
Researchers are increasingly aware of infectious diseases that mainly affect poor people. Some diseases, like AIDS, were unknown only a generation ago; others, including tuberculosis and malaria, are NOW resurgent in new and more difficult-to-treat forms. These three pathogens will probably kill 6 million people this year alone , and social inequalities are central to the emergence of these pandemics . Identification of new methods of diagnosis, drugs, and control strategies is of the utmost importance if medical and public-health workers hope to control these diseases in the settings in which they take their greatest toll.
It is not surprising, then, that thousands of research projects link affluent universities in developed countries with slums and villages in the less-developed world. Nor is it surprising that these collaborations are beset with difficulties and controversies: the social context is one, after all, of deepening inequalities between rich and poor. The past decade has been marked by many debates, often buried deep in the columns of our professional journals, about the ethics of transnational research projects. Are institutional review boards, irrespective of their constitution, capable of monitoring research across such steep grades of inequality? Is truly ethical research possible in poor communities?
In today’s Lancet, Daniel Fitzgerald and colleagues, writing from Haiti, take up the debate from a starting point that few would dispute: “Unfortunately, there are few practical guidelines on how best to inform research volunteers or how to ensure their understanding of the consent form.��? The data they present stem in part from a questionnaire, a series of true or false questions relevant to HIV-1 transmission, the topic of the study for which volunteers were being recruited. Although detailed social characteristics of the cohort are not included, most participants were poor and illiterate city-dwellers; all were negative for HIV-1, and had had sexual contact with someone infected with HIV-1 (GHESKIO, the site of the survey, is urban Haiti’s largest source of free care for sexually transmitted infections and for some complications of HIV; the centre has offered many of these services for 20 years).
Fitzgerald and colleagues found that only three (20%, 95% CI 0–40%) of the 15 volunteers passed the test. They showed that counselling sessions with a non-physician psychologist could substantially increase the proportion of volunteers able to pass the qualifying exam. This small study suggests that modest investments in better communication can improve participants’ understanding of the process in which they are being engaged.
If the goal of such an exercise is merely to increase the number of volunteers who can pass an exam, the debate might end there. But although improvement of tools such as those for informed consent is a fine goal, sights should not be set too low. Focusing on only the process can lead to an overemphasis on the consent form (or institutional review board) as the key to rendering research in such settings ethical.
The fact that so few volunteers could pass the simple test is a reminder that, increasingly, researcher and subject are living in two different worlds. Counselling sessions before signing a consent form do not change the social conditions that structure the growing gap, cognitive and social, between those who do research and those who are participants. These social gaps underpin the growing “outcome gap��? that characterises transnational research projects . Empirical research on the process of informed consent across this gap is urgently needed.
Fitzgerald and colleagues’ study allows a frank discussion of what is at stake as inequalities of risk grow for certain diseases. Disparities grow with social inequalities, and social inequalities are little understood by either medical ethicists or medical researchers; thus, this and other sincere attempts to address social context are welcome .
Social context is not merely local, nor are standards of care. In studies linking developed and less-developed countries, context is transnational, and such research is a reminder that some populations are not really developing, but rather being left behind by the same global economic processes that enable powerful universities to do research in poor countries. Does this disparity mean that research should not be done at all in such settings? Such Luddite responses need to be anticipated and prepared for to defend research in less-developed countries if researchers can answer local demands for equity and criticisms of the uses to which findings are put. This process will be best helped by in-depth social analysis of inequalities between countries and plans to remedy them. In studies of AIDS, research might have to be linked to access to the best treatment available, rather than to local standards of care.
All medical and public-health researchers would like a magic bullet that would make research undeniably ethical. But there is no magic bullet. There is only the complex and difficult process of linking research in resource-poor settings to the services demanded by poor people. The alternative prospect—a world in which medical research is dedicated wholly to the diseases of the affluent—is too painful to contemplate.
Lancet 2002; 360: 1266
 World Health Report 2001. In 2000, tuberculosis, HIV/AIDS, and malaria killed an estimated 5.7 million people. Geneva: WHO, 2001: 144.
 Farmer, P. Social inequalities and emerging infectious diseases. Emerg Infect Dis 1996; 2: 259-69.
 Farmer, P. Infections and inequalities: the modern plagues. 2nd edn. Berkeley: University of California Press, 2001.
 Churchill, LR. 1997. Bioethics in social context. In: The social medicine reader. Henderson, GE, King, NMP, Strauss, RP, et al., eds. Durham: Duke University Press, 1997: 310-320.