It is widely agreed that medical researchers who conduct studies in low- and middle-income countries (LMICs) are morally required to ensure that their research benefits the broader host community, not only the subjects. The justification for this moral requirement has not been adequately examined. Most attempts to justify this requirement focus on researchers' interaction with the community as a whole, not on their relationship with their subjects. This paper argues that in some cases, research must benefit the broader host community for researchers to treat subjects and prospective subjects ethically. If research presents substantial net risks to subjects, researchers can ethically ask LMIC citizens to participate only if people in LMICs, normally including people in the host community, stand to benefit.
- Research ethics
- international affairs
- distributive justice
Statistics from Altmetric.com
Some medical research in low- and middle-income countries (LMICs) aims to develop new interventions primarily for use in high-income countries. Research sponsors, both public and private, have various reasons for engaging in such research.1 ,2 Sometimes it is easier to recruit subjects either because a disease is more prevalent or because prospective subjects have fewer treatment options. Sometimes researchers want data from subjects who have not previously been treated for a disease. Often, research sponsors choose to conduct research in LMICs simply because it is cheaper.
Many bioethicists believe that research sponsored by organisations in high-income countries but conducted in LMICs—‘international research,’ for short—unethically ‘exploits communities’ unless host communities benefit adequately. On this view, whether international research is ethical does not depend only on how subjects and prospective subjects are treated. It also depends on how research affects others in the host community. Though it is widely agreed that international researchers are morally required to benefit people in host communities other than the research subjects, the justification for this requirement has not been adequately examined. To know precisely what this requirement demands, it is necessary to understand its grounds and scope.
This paper offers a partial remedy for this gap in justification. It explains why, when research involves substantial net risk to subjects, individual risk and community benefit are ethically linked. Though many studies do not present substantial net risk to subjects, those that do are those that raise the most serious subject-oriented ethical issues. Researchers can ethically ask LMIC citizens to take on significant net risks only if there is sufficient prospect of benefit to others in LMICs, typically including people in the host community.
Community benefit requirements
The belief that ethical research must benefit host communities lies behind two widely held views about the requirements for ethical international research. The first is the view that the topics of research done in resource-poor settings must be responsive to the host community's needs. There is widespread consensus that a responsiveness requirement applies to all international research, though there is disagreement about how this requirement is to be interpreted.3–5
The other widely held view is that if the fruits of a successful trial would not ordinarily be locally available on reasonable terms, research sponsors are required to provide benefits to the community. Bioethicists have extensively debated what benefits must be provided. Some hold that research sponsors must make any new interventions developed reasonably available to the host community.3 Others hold that researchers are instead required to provide a fair package of benefits to the community that need not include the intervention under study. The benefits could include, for example, building a new clinic or training healthcare workers.6 Nonetheless, it is widely if not universally agreed that benefits of some kind are required. This requirement is distinct from the responsiveness requirement, and it is possible to satisfy one requirement without satisfying the other. The responsiveness requirement concerns the topic of research. A requirement to provide benefits concerns an economic transaction between researchers or research sponsors and the host community.
There has been extensive discussion of the concept of community in the context of genetic research that specifically targets groups.7 ,8 Those who advocate more broadly applicable community benefit requirements generally do not explicitly specify what they mean by ‘community.’ There seems to be an agreement that it is defined at least partly by geography. Concern for the host community is concern for the ways in which research affects people other than subjects in the village, neighbourhood, city, district or nation from which subjects are drawn. It may be a concern about how research will affect people generally in this geographic area, or it may be a concern more specifically about how research affects people in the area who are also in the subjects' socioeconomic class. My argument presupposes that community is to be understood geographically, but it does not presuppose a more specific view.
Though it is widely believed that there is a responsiveness requirement and a requirement that researchers provide benefits to resource-poor host communities, it is not obvious what justifies these requirements. Some argue that sponsors of international research must ensure that their research benefits host communities because they have a general moral duty to address pressing health needs in LMICs.5 ,9 This argument does not explain why research sponsors who recruit subjects in LMICs have a special obligation to help LMICs. The pressing needs of LMICs presumably give all research sponsors a moral reason to support research that helps to address these needs, regardless of where these research sponsors recruit their subjects.
Others argue that research tends to place burdens on host communities' public resources, and that researchers must provide communities fair compensation for these burdens and for the benefits they gain from relying on these resources.10 ,11 This argument explains why researchers sometimes have an obligation to benefit host communities, and it gives a clear answer to the question how broadly community benefit must extend: the community that bears the burdens of research ought to receive benefits. But this argument does not always support a robust obligation to benefit host communities. If researchers rely only minimally on the host community's public resources, and if they do not impose significant burdens, this argument supports at most a weak obligation to benefit the host community.
The conclusion that researchers have only a minimal obligation to the host community when they make minimal use of the community's public resources may be acceptable if the research does not present significant net risk to subjects. I will argue that when research does present significant net risks to subjects, however, researchers have obligations to host communities even if their direct interaction with host communities is limited.
Research involves net risk to subjects when the risk of the intervention or interventions is greater than the potential for benefit to subjects.12 Prospective subjects thus could not rationally choose to receive this intervention for their own medical benefit. The term ‘risk’ refers to any potential harm to subjects, including death, physical injury, pain, psychological harm, and negative social or economic consequences for participation. It includes both the potential for harm that may or may not materialise and burdens that are entirely predictable, such as the pain of a blood draw or the expected side effects of chemotherapy.
Some studies involve net risk that is slight either because the risks are minimal or because the potential benefits are nearly great enough to outweigh the risks. The studies that raise subject-oriented ethical issues in the absence of community benefit are those that involve substantial net risk. The clearest examples of studies that involve substantial net risk are those with no prospect of medical benefit to subjects, such as phase I studies of new drugs and vaccines in healthy volunteers. Studies of stigmatised populations (such as sex workers or, in some countries, men who have sex with men) also involve significant net risk if there is no prospect of benefit to subjects and there is a significant chance that participants' confidentiality could be compromised.
Studies may present substantial net risk even if they involve experimental interventions that may benefit subjects. (When they do, there is reason to be especially attentive to the possibility of defects in consent.) For example, many psychiatric trials involve a placebo arm and require all subjects to go off their existing medications. For patients with serious psychiatric conditions that are at least somewhat well managed with existing drugs, participating in such a trial would involve substantial net risk to their mental health. More broadly, early phase trials on patients can involve substantial net risk, depending on what risks the interventions pose, how subjects might benefit and how likely they are to benefit. Though these trials do have the potential to benefit subjects, the evidence of interventions' effectiveness is speculative, and their safety has yet to be established.
A study that would present net benefit or a slight net risk to subjects in high-income countries may present a substantial net risk to subjects in LMICs. This can happen, for instance, if there is a substantial chance that an experimental intervention has side effects that will not surface until well after the trial is over (and thus would not be treated by the researchers). If these side effects are easily treatable but serious if untreated, the study presents a risk to people who lack access to healthcare that it does not present to people who have reliable access to good healthcare.
It is unclear how many studies involving substantial net risk have taken place in LMICs, in part because many early phase trials are not registered.2 But early phase trials in LMICs are not rare. Among clinical trials that have been registered at clinicaltrials.gov as of 12 April 2008, 15.8% of clinical trials in Latin America and 17.9% of clinical trials in Africa were phase I or phase II, compared with 23.9% of clinical trials in North America.13 Though it may be that most studies in LMICs do not involve substantial net risk to subjects, those that do warrant close attention because they present special ethical concerns.
Justifying substantial net risk
For an invitation to participate in research to be ethical, researchers need adequate justification for inviting people to take on the risks. It is relatively easy to justify asking people to participate in research that involves net medical benefit to subjects. It is also relatively easy to justify asking people to participate in research that presents only a slight net risk either because the risks are minimal or because the potential benefits nearly outweigh the risks. It is more difficult to justify asking people to participate in research that presents a substantial net risk.
One way to justify asking people to enrol in net risk studies is to provide some benefit to subjects that would make enrolment in the interest of prospective subjects. This could involve medical benefit, as when researchers provide ancillary care or screening that is not readily available outside of the study. There is controversy about whether screening or ancillary care can ever justify research risks; at least one LMIC's research guidelines hold that they never do.14 It is clear that they cannot justify some of the larger risks of research, as in a first-in-human phase I study. Though particular subjects may benefit greatly from ancillary care, researchers in such studies cannot expect that the potential benefits of ancillary care will be greater than the risks for a typical prospective subject.
Researchers can also appeal to subjects' self-interest by paying them, either in currency or in kind. The dominant view in research ethics is that payment is not a benefit that can justify risks to subjects.15 ,16 Although some have questioned this view, I will assume that payment cannot justify asking prospective subjects to undergo risks that researchers would not otherwise be justified in asking people to undergo.17 This does not mean that paying subjects is impermissible. It means that if study participation would involve a substantial net medical risk, and if ancillary care associated with study participation is not sufficient to make up for the risks of research interventions, researchers must be justified in asking prospective subjects to participate at least partly out of altruism.
Asking for help
Although we sometimes say, ‘It can’t hurt to ask,' there are ethical limits on asking for others' help even when they have the opportunity to refuse. One limit is that it is unreasonable to ask people to take on too much risk for too little benefit to others. But it also makes a difference what relationship there is between the would-be benefactor and the beneficiaries. I shall argue that the relationship between citizens of LMICs, collectively, and citizens of high-income countries, collectively, is such that LMIC citizens cannot reasonably be asked to take substantial risks to their health primarily to benefit people in high-income countries.
A breach in a relationship or a positively bad relationship can make it unreasonable to ask for help that would otherwise be reasonable to request. Suppose that A has wronged B without making amends. A now needs B's help, and the sort of help A needs is burdensome and not morally mandatory for B to give. It may have been reasonable for A to ask B for help before the wrong, and it might be reasonable for B to forgive A and to offer help unprompted. But A should not ask B's help now, at least not without first seeking B's forgiveness for the past wrong.
Even if A has not wronged B, there can be other negative features of the relationship between A and B that could make it unreasonable for A to ask B's help. Suppose, for instance, that in the past, B has asked for A's help in an important matter, and A declined. Declining to help B may not have been morally wrong. But having declined to help B, it would typically be disrespectful for A to ask B for help that is more burdensome than the help A declined to give and that would address a less important need.
Asking LMIC subjects for help
The considerations just discussed speak against asking people in LMICs to take substantial health risks primarily in order to benefit people in high-income countries. Research primarily benefits people in high-income countries if any new interventions introduced as a result will be widely available only in high-income countries for the near future. If new interventions would be available only to the wealthy in LMICs, or if they will eventually be generally available in LMICs but only in the fairly distant future, asking prospective subjects' help is asking them primarily to help people in rich countries.
This is objectionable because of at least one negative feature of the relationship between citizens of LMICs and citizens of richer countries. At least arguably, LMICs' limited resources are partly a result of injustice for which high-income countries and their citizens are responsible. If this is true, the unjust distribution of resources among nations makes it unreasonable to ask LMIC citizens to take substantial net risks for the benefit of people in rich countries. Even if international economic inequality is not unjust, rich countries' lack of generosity towards LMICs makes such requests for help unreasonable.
There are two reasons one might think that LMICs' relative lack of resources is an injustice. On the cosmopolitan view of international justice, the very fact that some countries have far more limited resources than others is itself an injustice, which high-income countries have a moral duty to remedy.18 Some in the international justice literature deny this. They argue that norms of distributive justice normally apply only within states, not between or among them.19 ,20 But one could reject cosmopolitanism while still holding that high-income countries have wrongfully caused or contributed to international economic inequality. Arguably, high-income countries have taken active steps that intentionally or negligently perpetuate international inequality.21
If either of these views is correct, then the relative poverty of less well-off nations is an injustice, and citizens of high-income countries bear blame for it. Responsibility for the unjust inequality does not only fall on rich countries' leaders. At least in democracies and arguably in all states, citizens bear blame for the wrongs their countries do. If the citizens of rich countries collectively wrong LMIC citizens by failing to address or by actively perpetuating unjust international inequality, there are some forms of help that LMIC citizens cannot reasonably be asked to give to the citizens of rich countries.
Though it is controversial whether the inequality in the international distribution of resources is unjust, it is indisputable that citizens of rich countries are collectively ungenerous towards citizens of LMICs. There are health needs in LMICs that charitable contributions or foreign aid from wealthier countries could do much to address.22 Many citizens of high-income countries earn quite a bit more than they need to live comfortably. Though people in high-income countries can help to address desperate needs in LMICs, and though they are regularly asked to do so, most have chosen not to make even modest financial sacrifices to save lives abroad. Some people in rich countries do give generously to save people in developing countries from preventable diseases, and many people in rich countries are not in a financial position to make substantial donations. As a group, however, people in rich countries are able but unwilling to make significant financial sacrifices to address the disease burden in LMICs. Even if this lack of generosity does not violate a moral duty, the collective refusal of high-income countries' citizens to make even modest financial sacrifices to save lives in LMICs makes it unreasonable to ask people in LMICs to risk their health for the sake of people in rich countries.
For both of these reasons, it is generally unreasonable to ask people in LMICs to risk their health with an altruistic motive if those who stand to benefit are primarily in rich countries. Since researchers should only ask people to enrol in high net risk research if it is reasonable to enrol at least partly out of altruism, researchers may not ask LMIC subjects to enrol in high net-risk research if the research will primarily benefit people in high-income countries.
These objections typically do not apply, however, to asking people in LMICs to take health risks for the benefit of others in their own communities, assuming that the benefits will be shared with people of the subjects' socioeconomic classes. It is not unreasonable to ask people to take risks with the motivation of helping people of their own socioeconomic class (or people who are worse off). There is typically no history of injustice that makes it unreasonable to ask people to help others in their own community who are economically no better off than they. Nor is there a typically a lack of reciprocal generosity that makes a request for help unreasonable. To be sure, if there is unjust economic inequality within an LMIC community, researchers may not ask the poor to risk their health for the benefit of rich members of their own community. But as long as benefit will extend to people of subjects' economic class, it is not unreasonable for researchers to ask subjects to take risks for the benefit of others in their communities.
For similar reasons, it is also not unreasonable for researchers to ask LMIC citizens to take risks for the benefit of people who live in other LMIC communities. There may be some cases in which researchers are justified in asking people in one LMIC to enrol in high net-risk research that will not benefit their own communities, but which will benefit people in other LMICs. Such cases are presumably unusual. Given that a study is going to benefit people in LMICs, it rarely makes sense to choose a study site that is in an LMIC community whose people are not among the potential beneficiaries. It is also easier to justify asking people to take altruistic risks if the beneficiaries may include people with whom they have existing relationships. Just as a breach in a relationship makes it harder to justify asking people for altruism, a positive relationship makes it easier to justify asking people for altruism. It is easier to justify asking people to take research risks if the beneficiaries could include their friends and family members, or at least their fellow citizens.
Whether researchers treat subjects and prospective subjects ethically sometimes depends on how research affects host communities. If research presents substantial net risk to subjects and other benefits such as screening and ancillary care do not make participation a net benefit to subjects, then researchers must ask prospective subjects to enrol at least partly out of altruism. But researchers cannot justifiably ask LMIC citizens to direct their altruism towards people in high-income countries. They may ask prospective subjects to take substantial risks out of altruism only if others in LMICs stand to benefit, normally including people in the host community.
This argument has two important consequences. First, it indicates which people in host communities must stand to benefit. If research involves significant net risk to subjects, the potential beneficiaries of research must include people who are of the subjects' socioeconomic class or who are less well-off. Second, the argument indicates how much benefit there must be to LMICs for high net risk research in LMICs to be ethical. The amount and type of benefit to people in LMICs should be such that LMIC citizens could reasonably choose to take on the risks of research to obtain these benefits for others. This means that, at a minimum, the prospect of benefit to people in LMICs must exceed the total risk. For purposes of risk-benefit assessment, benefit to people in high-income countries must be discounted.
I am grateful to Benjamin Chan, Joseph Millum, Seema Shah, Seana Shiffrin, David Wendler and Alan Wertheimer for helpful comments on previous drafts of this manuscript. This work was completed as part of the author's duties as a fellow of the U.S. National Institutes of Health and was supported by the Intramural Research Program of the NIH, Clinical Center. The views expressed here are the author's and do not reflect the policies and positions of the NIH, the U.S. Public Health Service, or the U.S. Department of Health and Human Services.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.