Exploitation in International Research
Tom Tomlinson
In this short essay, I want to better understand the danger of exploiting research subjects in developing countries.
Concerns about exploitation lie behind a number of ethical principles applied to international research. The 2002 CIOMS Guideline 10, for example, mandates that
“Before undertaking research in a population or community with limited resources, the sponsor and the investigator must make every effort to ensure that:
The commentary on the guideline explains that this requirement is necessary to avoid exploitation.
“Even when a product to be tested in a particular country is much cheaper than the standard treatment in some other countries, the government or individuals in that country may still be unable to afford it. If the knowledge gained from the researchin such a country is used primarily for the benefit of populations that can afford the tested product, the research may rightly be characterized as exploitative and, therefore, unethical.” (Found at http://www.cioms.ch/guidelines_nov_2002_blurb.htm, 01/02/06)
The requirement is motivated by a number of studies in developing countries in which the principle beneficiaries of a successful clinical research trial appeared to be persons in the developed world who have the means to purchase any resulting treatment. One possible example is a trial of adjuvant breast cancer treatment following mastectomy in Vietnam, in which the control group received mastectomy alone. Since adjuvant therapy following mastectomy is standard practice in the US, such a trial would not be permitted here. Yet, since Vietnam could not afford to provide adjuvant therapy to all mastectomy patients, whatever knowledge might be gained about the usefulness of this therapy is likely to mostly benefit women (and possibly their insurers) in the US and other developed countries.<1> It might be said, then, that the Vietnamese women in the trial suffered all the risks of the research, but will not enjoy all of its benefits. CIOMS Guideline 10 would remedy this unfair exchange by pledging to make beneficial treatment available to the subjects or to similarly situated patients.
The idea that post-trial benefits help to remedy exploitation raises a number of questions. What form of post-trial benefits best addresses the sorts of exploitation produced by research? Are there situations in which post-trial benefits might increase the risk of exploitation? And, when is research in a developing country exploitative in the first place?
In this essay, I will only have space to discuss this last question. In my discussion, I will be drawing on, and at times departing from, an article on post-trial benefits produced by participants in the 2001 Conference on Ethical Aspects of Research in Developing Countries, and published in the Hastings Center Report.<2> (Henceforth, the “Conference Report.”) Readers who want to pursue some of the other questions raised are referred there.
Let’s start by asking what counts as “exploitation.” Note first that it is not exploitative merely to “use” another person for my benefit. I do that with my auto mechanic whenever I take my car in to be serviced. Of course, he gets something out of it as well. Since he’s “using” me to make his living, there is at least some rough parity of benefits.
This leads to the idea that exploitation occurs when there is an inequality of benefits (or harms). This might indeed be a necessary condition, but it is not sufficient. Imagine a company which hires a management consultant, who collects her standard fee, while the company goes on to save millions using the consultant’s recommendations. There’s a great inequality of benefits, but we wouldn’t say the consultant had been exploited without having more information about the conditions under which the deal was struck.
We also should recognize that if inequality of harms or benefits were a sufficient condition for exploitation, much human subjects research would be exploitative by its nature. Research is not therapy; it is not conducted primarily to benefit its subjects. It’s conducted to gain knowledge that will often be of benefit only to others, while the subjects endure the risks. There is thus an inherent inequality in the distribution of harms and benefits produced by medical research, but it would go too far to condemn it all as “exploitative.”
We might attempt to address this by incorporating a requirement of informed, voluntary consent into our conception of exploitation. Neither the consultant nor the research subject is exploited so long as they freely agreed to the “bargain.” But this isn’t quite right either. Together with inequality of harms/benefits, the lack of consent may contribute to exploitation. But it’s not a necessary condition. To borrow an example from the Conference Report, if I run into a snow bank and find myself at the mercy of a tow truck operator who demands three times the usual rate to tow me out, I’m a victim of exploitation even as I freely pull the cash out of my wallet.
So what then is “exploitation?” Borrowing from Alan Wertheimer <3>, the authors of the Conference Report claim that “Party A exploits party B when B receives an unfair level of benefits as a result of B’s interactions with A” (Conference Report 19). But what makes a different level of benefits an unfair and hence exploitative one? According to the Conference Report (and Wertheimer), “Fairness in individual interactions, which is the concern of exploitation, is based on ideal market transactions. Thus a fair distribution of benefits at the micro-level is based on the level of benefits that would occur in a market transaction devoid of fraud, deception, or force” (Conference Report 20).
This is an unsatisfactory answer. It can’t easily make sense of the tow truck example, which is a market transaction involving neither fraud, deception, nor force. Neither is it clear how to translate this market model into the context of clinical research, which doesn’t involve markets of the usual kind and so doesn’t offer paradigm examples of what counts as a fair market transaction.
What produces exploitation is not just inequality of harms or benefits, but also how that inequality comes about. Here the role of vulnerability becomes critical, because exploitation involves “taking advantage” of another person. John exploits Sam when
The authors of the Conference Report, however, brush aside the relevance of vulnerability: “Since exploitation involves the distribution of benefits and burdens, vulnerability is neither necessary nor sufficient for its occurrence” (Conference Report 20). Of course, by itself vulnerability is not sufficient for exploitation. There can be a parity of harms/benefits despite the vulnerability of one of the parties. But by itself, inequality of harms/benefits is not sufficient for exploitation either. It’s the relationship between inequality and vulnerability that produces exploitation.
Given this conception of vulnerability, when is research in developing countries exploitative? Perhaps less often than assumed. Certainly we can no longer conclude that research is exploitative merely because it promises much greater benefits for developed countries than it does for developing countries. It will depend as well on the reasons for that disparity. Compare two studies. One is a surveillance study of HIV-discordant heterosexual couples in Africa which seeks in part to determine the relationship between HIV viral load and seroconversion of the HIV negative partner.<4> Uganda could not afford the resources for viral load testing and for the antiretrovirals needed to reduce viral load in HIV positive patients. For these reasons, any results showing a strong relationship between viral load and infectivity would prove much more useful for HIV-positive persons in the US than in Uganda. The other is Phase I testing of a potential treatment for oral cancers, using patients recruited at a cancer center in India.<5> If successful, the drug, being developed by a commercial start-up company, would likely be beyond the reach of most Indian cancer patients.
In both these cases, the benefits of the research will go disproportionately to patients in developed countries. But the Ugandan study is less exploitative than the Indian study, for two reasons. The first has to do with the reasons for conducting the research in the developing country and the way these rely upon vulnerability. The question posed by the HIV seroconversion study would be much more difficult to answer in the US, since for scientific validity the study requires a high rate of heterosexual transmission, together with an incidence of HIV-discordant heterosexual couples sufficient to recruit the numbers required. By contrast, I will suppose, the questions posed by the Indian cancer trial could have just as readily been answered by trials conducted in the US. Arguably, they were conducted in India because there participants are more easily and cheaply recruited, due to poverty, illiteracy, and desperation. Of course, the high incidence of HIV transmission in Uganda is a disadvantage suffered by those who will be recruited into the study. But the trial relies only on there being a high incidence of HIV transmission, not on this high incidence being a disadvantage relative to other subject populations. Unlike the Indian trial, the Ugandan research does not rely on the subject’s being disadvantaged in order to further its aims.
The second reason is that the HIV study has a much more benign risk-benefit ratio for the subjects than the Indian study. The Ugandan subjects are exposed to little or no additional risk, and may benefit from the additional HIV education and counseling provided by the research project. The Indian subjects are at risk for whatever toxic side-effects may be caused by the agent being studied, with no prospects for any therapeutic benefit. This enlarges the disparity in harms/benefits between the subject population and the populations of patients who will benefit from the knowledge obtained. This implies that one way to reduce the danger of exploitation is to maximize the benefit-risk ratio for the study subjects. (And this can’t be accomplished by luring subjects into ignoring the likelihood or significance of real risks.)
This discussion has only begun to scratch the surface. Even if my analysis of exploitation is headed in the right direction, there are still plenty of questions. What counts as a “vulnerability?”<6> Under what circumstances has someone improperly relied on or used a vulnerability in gaining an advantage? And, of course, how do we translate the answers to these questions to the international research context?
References
1. See Love, RR, and NC Fost. 1997. Ethical and regulatory challenges in a randomized control trial of adjuvant treatment for breast cancer in Vietnam. Journal of Investigative Medicine 45(8):423-31.
2 . Participants. 2004. Moral standards for research in developing countries: from “reasonable availability” to “fair benefits.” Hastings Center Report 34(4): 17-27.
3 . Wertheimer, A. 1999. Exploitation. Princeton, NJ: Princeton University Press.
4 . For example, see Gray, RH, and MJ Wawer MJ, and R Brookmeyer, et al. 2001. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. The Lancet357(9263):1149-53.
5 . I’m thinking of the Phase I trial of M4N conducted by a Johns Hopkins researcher without approval of the Hopkins IRB. For more disturbing details, see Bagla, P and E. Marshall. 2001. Hopkins reviews investment in Indian cancer drug trial. Science 293, 1024.
6 . For more on the concept of “vulnerability,” see Kipnis, K. 2001. Vulnerability in research subjects: a bioethical taxonomy. In Ethical and Policy Issues in Research Involving Human Participants, Volume II: Commissioned Papers and Staff Analyses, National Bioethics Advisory Commission, pp G1-G13. (available at http://www.georgetown.edu/research/nrcbl/nbac/pubs.html)