The Role of Inclusion Benefits in Ethics Committee Assessment of Research Studies

Abstract

The relationship between risks and benefits is central to the ethics of research involving human participants. Traditionally, to be ethically justifiable, risks should be reasonable in relation to anticipated benefits (if any) to subjects and to the potential social benefits resulting from research. This calculus is being further complicated by findings from an increasing number of social science studies that reveal that prospective and actual research participants frequently describe various types of inclusion benefits (for example, personal benefits such as access to or improved health care, increased knowledge about their disease or condition, and greater solidarity with the local community) as important to them. What is the ethical significance of such inclusion benefits, particularly those nonmedical in nature, for research ethics committees’ risk-benefit assessment of research studies? We argue that, unless participants are clearly mistaken in their perceptions, ethics committees should take these types of inclusion benefits into account, and we suggest a few ways this might look in practice.

The relationship between risks and benefits is central to the assessment of human research studies. According to U.S. federal regulations governing research with humans, to be ethically justifiable, risks should be reasonable in relation to anticipated benefits (if any) to subjects and to the importance of the knowledge that may reasonably be expected to result.¹ Furthermore, The Belmont Report encourages research ethics committees and ethicists to take a broad view of the kinds of research-related benefits (medical, psychological, economic, legal, social) that should enter the risk-benefit calculus and help determine the reasonableness of the risks.² These views have been influential far beyond the U.S. context.³

In this article, we examine whether and to what extent research ethics committees’ risk-benefit assessments of research studies should consider inclusion benefits identified from social science research with prospective and actual research participants. By “inclusion benefits,” we mean benefits that result from participating in a research study, regardless of whether the participant receives an experimental intervention.⁴ Such benefits could conceivably play a role in ethics committees’ risk-benefit assessments of proposed research studies.

Our starting point for reflection comes from the fact that there may be substantial differences between how researchers and ethics committees perceive research-related benefits and how such benefits are perceived by prospective research participants and by individuals who are enrolled in research studies. Social scientists often identify these differences in perspectives when they conduct studies with potential and actual research participants. Two examples derived from recent social science research illustrate these differences.

For a research project in Kenya, our research team conducted focus group discussions (using vignettes) with parents, adolescents, and health and other professionals to solicit opinions about appropriate ways to conduct HIV research with adolescents. Our data revealed that many focus group participants had a profoundly positive conception of participation in health research, including studies conferring seemingly few personal benefits from the research team’s perspective. More specifically, we identified a rich array of benefit types as perceived by participants in the focus group discussions: informational, existential, emotional, health, and material benefits.⁵ Our findings were consistent with similar research in Kenya related to participant perceptions of benefit.⁶

In Thailand, early phase (first-in-human) HIV cure or remission studies are underway. These studies involve significant and unknown risk and seemingly little personal benefit for participants, and therefore their ethical justification would seem to rely heavily on the importance of societal benefits in terms of the scientific knowledge reasonably expected to result. Our social science research on participant experiences in the trials, however, revealed a different story about personal benefit.⁷ For example, some of these studies involved analytic treatment interruption, that is, cessation of antiretroviral treatment in order to test the effects of a cure or remission intervention. Since participants were doing well on treatment, analytic treatment interruption was identified by researchers and ethics committees as a significant medical risk. However, our interviews revealed that many participants regarded analytic treatment interruption also as a benefit: while acknowledging the risks, they valued the return to a “sense of normality” in not having to take their daily antiretroviral medication.

Similar experiences of benefit by research participants are not at all uncommon.⁸ In hindsight, it seems obvious that inclusion benefits can become more apparent after a study has begun. Designing and ethically assessing a research study, including identifying potential benefits, is one thing; implementing research in real-world circumstances is another. These sorts of inclusion benefits can also be expected to come to light when participants’ study experiences become the subject of social science research. In addition, as social science research delves further into this area, uncovering patterns and trends, more potential inclusion benefits will come to be known in advance. Such information could reveal researchers’ and ethics committees’ assumptions and blind spots about what may be beneficial for participants.

What should one think about these discrepancies in what is considered beneficial? Moreover, what is the ethical significance of perceived inclusion benefits? We will argue that such benefits exist and that their exclusion from risk-benefit assessments by research ethics committees is unjustified. We will start by describing what we call the “standard view” on participant experiences of benefit, as well as the arguments supporting it. Much of the standard view is reasonable, but the arguments underpinning it have limitations. Taking these limitations as a point of departure, and after providing some key conceptual distinctions, we will make an appeal for permitting participant experiences of inclusion benefits to play some role in ethics committees’ risk-benefit assessments of research studies. We will conclude by responding to some important objections to this position.

THE STANDARD VIEW

King and Churchill map out the basic ingredients of the standard view.⁹ They start from the assertion that research ethics committees, not research participants, are the ones who should determine the ethical acceptability of research studies. How individuals perceive benefits and weigh the risks and benefits when deciding whether to enroll or remain in a study (whether participation is “worth it” to them) is to be distinguished sharply from third-party, external ethical assessments of benefits and risks. According to King and Churchill, the potential participant’s perspective is “logically and chronologically” secondary to the ethical assessment: only once an ethics committee has determined that potential research risks are reasonable in the light of anticipated benefits is it justified to invite individuals to enroll in a given study. Another way of marking this distinction is to imagine individuals willing to enroll in a study that has extravagantly excessive research-related risks without any countervailing benefits; King and Churchill argue that even if these individuals think the risks are acceptable, the risks can be unreasonable and the study ethically suspect.¹⁰ Alternatively, a study might be designed in such a way that its risks are deemed reasonable but prospective participants do not think participation is worth it. According to this view, the fact that individuals unfavorably view the risk-benefit relationship does not mean that the study (in this respect) is unethical.¹¹

King and Churchill are at pains to argue against an alternative view that permits the prospective participants’ perspective to determine benefits to be weighed in the ethical assessment of research, particularly concerning inclusion benefits. They mention three arguments, and we will add a fourth that is often implicit in the literature.

Problematic manipulation of the risk-benefit relationship.

Inclusion benefits, King and Churchill write, are largely in the investigator’s control. That being the case, researchers could deploy inclusion benefits to further scientific goals in ways at odds with ethical values. There are two main concerns: first, researchers could essentially justify any level of risk in a study by simply offering participants more inclusion benefits; second, there is a threat of undue inducement when inclusion benefits are medical and participants have limited access to health care outside the research context. Furthermore, they write, permitting these sorts of inclusion benefits in the risk-benefit assessment by ethics committees would legitimize and entrench such practices. Instead of increasing inclusion benefits and giving them ethical import, they argue, investigators and ethics committees should find ways to minimize research-related risks.

Benefits and injustice.

Increasing inclusion benefits creates a problematic division between standards of care within a study and outside of it. This permits investigators “to hold out high-quality care as a quid pro quo of research participation, and to allow or excuse deterioration in access to and quality of off-study treatment.”¹² According to King and Churchill, these are reasons both not to have inclusion benefits in excess of what would be offered to patients in standard clinical settings and not to include such benefits in the ethical assessment of risks and benefits.

Variable character of inclusion benefits.

King and Churchill point out that subjectivity enters into the experience of inclusion benefits. Some individuals may experience an aspect of research participation (such as extra monitoring) as a benefit, while others may see it as a burden. Some may experience psychological benefits (for example, pride in helping future patients), and others may not, despite following the same protocol, at the same site, with the same investigators. Here the objection to inclusion benefits is not so much ethical as practical: how can subjectively variable benefits be integrated into an overall ethical assessment of the risks and benefits of a study? These kinds of benefits are to be contrasted, according to King and Churchill, with direct medical benefits, which are objectively beneficial.

Genuineness of benefit claims.

Benefits are not just variable; they are also fallible. Participants may experience phantom benefits, that is, benefits that are not really there and are based on false beliefs. For example, Appiah-Poku et al. describe in-depth interviews conducted with participants in a genetic epidemiological study in Ghana. The epidemiological study focused on genetic factors associated with contracting tuberculosis (TB). Treatment for TB was provided as a benefit to those identified by the study as having the disease. However, interviews revealed that participants associated the research with a great many benefits, some of them not reflecting study procedures or aims as stated in consent processes. These included production of new knowledge, protection of those without TB, education about TB, and the discovery of a new, shorter TB treatment regimen.¹³ Additionally, in some cases, participants’ perceptions of benefits may also be determined or magnified by background poverty and need: the worse off one is, the more benefits in research one may tend to perceive. While it is not always obvious how to determine whether participants are mistaken in their beliefs, there are sometimes clear-cut cases of error, and at least in this range of cases, there are good reasons not to take participant experiences of benefit seriously (with respect to ethics considerations).

All the above seemingly converges on the following conclusion: whatever social scientists may discover in their studies of participant experiences concerning benefits, none of it has (or should have) any bearing on ethics committees’ risk-benefit assessments of research studies. Researchers might be interested in such information in their pragmatic bid to recruit and retain participants, and social scientists might find such experiences fascinating in their own right, but ethics committees need not—or, more strongly, should not—take them into consideration in evaluating the ethics of research studies.

However, King and Churchill do not draw such a strong conclusion. “IRBs,” they say, “should consider carefully how to address inclusion benefits, both in particular clinical trials and as a matter of policy.”¹⁴ But they provide no further guidance. To make room for addressing inclusion benefits, the four arguments above need to be examined critically. But to facilitate the discussion, we need to first make some important distinctions concerning the standard view.

PRELIMINARY DISTINCTIONS

First, the focus of the discussion is experiences of personal benefit that research participants can have, beyond just experiences of medical benefit or improved health care derived from research participation. The list (increased knowledge about their disease or condition, greater solidarity with others, improved interpersonal relationships, acquisition of life skills, positive behavioral change, enhanced sense of purpose, and bolstered self-esteem) is not exhaustive; social scientists continue to explore the landscape of participant experiences in particular studies globally. These inclusion benefits are to be distinguished from direct benefits (resulting from an intervention being studied) or aspirational benefits (accruing to future patients or society).¹⁵ These kinds of benefits should also be distinguished from (beneficial) incentives—such as in cash-transfer studies—provided to research participants as interventions to study their effects on health behavior.¹⁶

Second, not all inclusion benefits are created equal. Some of them are certain (as King and Churchill put it, “built in” simply by being provided), while others are merely probable (“as a participant in this study, you may benefit …”). Some of them are to some extent foreseeable (and foreseen), while others may be surprising and unexpected, at least for nonparticipants. As indicated below, these distinctions can make an ethical difference.

Third, as Bernabe et al. propose, there are useful distinctions to be made between separable activities of risk-benefit analysis, risk-benefit assessment, risk treatment, and ethical decision-making.¹⁷ Risk-benefit analysis refers to the empirical study of what events, causes, and consequences related to a study can be risks or benefits for participants and society. Risk-benefit assessment incorporates and makes use of risk-benefit analysis to determine the significance of the risks and benefits, that is, to “weigh” the given risks against the benefits to participants and society. “Risk treatment” refers to the use of strategies to reduce risks to participants and society in ways that do not unduly compromise the scientific validity and social value of the research. Ethical decision-making brings all the preceding activities to a head and concerns the final discussion of the research ethics committee on whether benefits outweigh the (modified) risks to participants and, if they do not, whether the social value of the research justifies an unfavorable risk-benefit ratio for participants. Our question is where inclusion benefits ought to enter in this process, if at all.

EXAMINING THE STANDARD VIEW

Problematic manipulation of the risk-benefit relationship.

In regard to this first concern, researchers indeed could add all sorts of inclusion benefits in an attempt to justify high research risks.¹⁸ This is a genuine concern, though little is known about the prevalence of such manipulation. The question here is whether making inclusion benefits inadmissible in the ethical review of research is the most effective way to counteract this possibility. An alternative approach would be to allow inclusion benefits to be an input in the risk-benefit analysis, leaving it to ethics committees in their assessments to decide whether researchers have or have not “spiked” their studies with inclusion benefits simply to facilitate approval.

With respect to the second concern—the familiar question of undue inducement by medical benefit—it is important to analyze the risks and benefits of a study as a whole before deeming particular medical inclusion benefits ethically problematic. For example, even if an individual would be “crazy not to join” a study because it includes medical benefits otherwise inaccessible to the person, inclusion of these medical benefits could be permissible if the study risks are sufficiently low.¹⁹ It is also worth noting that empirical evidence for the prevalence of undue inducement by medical inclusion benefits is lacking and that research stakeholders often disagree on the definition of undue inducement, frequently make inconsistent judgments on its basis, and regularly confound it with coercion.²⁰ Reluctance to include medical inclusion benefits in risk-benefit assessments may reflect an assumption that reducing or eliminating potential medical benefits—out of fear for undue inducement—is always the ethically right thing to do.

In addition, the assumption that inclusion benefits are in the investigator’s control seems only partly true, and perhaps stems from a preoccupation with the clinical trial context and medical inclusion benefits. Clearly, the intentional provision of inclusion benefits has been advocated for and does happen, motivated often by efforts to minimize exploitation²¹ or to fulfill obligations to provide ancillary care.²² In other cases, however, the relationship between study design and participant benefit may be looser and more complicated. From the perspective of the researcher, we can distinguish between unforeseeable and foreseeable participant benefits. Unforeseeable participant benefits are inclusion benefits that participants experience but that become discernable only after the study is underway. Participants in the Thai study seem to have experienced such benefits. With unforeseeable inclusion benefits, worries about intentional undue inducement by researchers are moot. In addition, if these sorts of benefits tend to come into view when a study is implemented, rather than before, they would have little influence on willingness to participate in a study and, for the same reason, could not figure in ethical review. These sorts of inclusion benefits might influence study retention or willingness to participate in multiple studies over time, but these effects of inclusion benefits have not been studied extensively.²³ However, some potential inclusion benefits may be to a greater or lesser extent foreseeable. In Kenya, social scientists have previously studied and published about inclusion benefits in this context in regard to a number of different research projects.²⁴ In principle, researchers and research ethics committee members for the Kenya study could have made use of such data in determining potential participant benefits, just as would be expected if there was available, relevant data about potential risks of social or psychological harms.

Benefits and injustice.

The standard view claims that if inclusion benefits (particularly medical ones) are allowed to play a role in risk-benefit assessments, there is potential for injustice, if it means that researchers “hold out high quality care as a quid pro quo of research participation, and to allow or excuse deterioration in access to and quality of off-study treatment.”²⁵ While this is a legitimate concern, the argument needs further empirical support. To what extent do researchers in fact offer medical standards of care that are significantly superior to off-study standards of care, rather than occasional advantages such as, say, blood pressure monitoring? To what extent do funder restrictions place limits on what standards of care researchers can offer participants? What do we know about the specific impact research standards of care have on off-study standards of care, in terms of access and quality, as opposed to other factors such as lack of government investment? Certainly, ethical red flags should be raised in cases in which researchers provide extensive medical benefits to participants out of keeping with usual standards of care, particularly when those benefits aim at boosting enrollment rather than helping answer the research question. However, it does not follow that the provision of any medical inclusion benefit necessarily involves injustice, much less that inclusion benefits as such are unjust.

Variable character of inclusion benefits.

This consideration poses problems for integrating inclusion benefits in risk-benefit assessments. But King and Churchill overstate their case when claiming that direct medical benefits are universal and certain, whereas inclusion benefits are subjective and variable. This framing seems to underplay how much uncertainty there already is in determining direct benefits (not to mention risks) in the practice of assessment by research ethics committees.²⁶ The objective risks and benefits of research for participants or society are rarely known with certainty: this is reflected in the language of research regulations, with their references to “potential” benefits and “probable” risks. Given variations among individuals, groups, and populations, the possibility that even medical research-related risks and benefits are uniform for all participants is unlikely. So if risk-benefit assessments are always decisions under uncertainty, and study risks and benefits typically subject to variability, it is hard to say why on this basis these sorts of inclusion benefits should be a priori ruled out of such assessments. They would merely add complexity to an already complex process.

Genuineness of benefit claims.

It should be conceded that some experiences of benefit on the part of participants are simply mistaken, including but going beyond cases of therapeutic misconception. Clearly, just because participants think they are benefiting or will likely benefit does not itself make the benefits real. Nevertheless, just as participants might be wrong, they might be right. They might benefit from studies in ways that researchers might not imagine when conceiving their studies or that research ethics committees might overlook in their assessments—a likely scenario given that the ethical import of inclusion benefits is typically discounted.

IMPLICATIONS

As social scientists examine the experiences of participants in research studies, they will learn more about the ways in which participants benefit from their participation. Some of the benefits participants experience will be inclusion benefits, raising the question of their ethical import for ethics committees in their risk-benefit assessment of future trials.²⁷ The standard view, reflected in many research regulations and some of the bioethical literature, argues against there being any meaningful place for inclusion benefits in ethics committees’ risk-benefit assessments. We have argued that, on closer inspection, the arguments in support of the standard view have significant limitations. These limitations suggest that there might be an ethical justification for ethics committees’ formal consideration of the participant benefits that social scientists are uncovering. How would that work?

To make some headway, let us return to the framework of Bernabe and colleagues and their model of risk-benefit analysis, risk-benefit assessment, risk treatment, and ethical decision-making. At the stage of risk-benefit analysis, the concern is to describe, using the best empirical information available, what the probable risks and benefits of the research actually are.²⁸ Following this model, inclusion benefits experienced by participants—at least those not based on false beliefs—could and should enter into the process at the stage of risk-benefit analysis. Bernabe and colleagues inserted certain and probable inclusion benefits into their model of risk-benefit analysis without question, with certain and probable inclusion benefits treated as inputs to the descriptive analysis of risks and benefits on a par with any other kind of benefit.²⁹ The next stage of risk-benefit assessment is an evaluative, deliberative process of comparing the described risks and benefits to determine the significance of the risks and the benefits in their relationship with each other. In their model, inclusion benefits also appear at the level of assessment, where they are “weighed” along with other benefits and risks. In principle, inclusion benefits could make a difference to the final ethical decision-making regarding a research study, though this would depend on how they were treated in the assessment. For our purposes, the important point is that, rather than sidelining inclusion benefits on the grounds of the standard view, this model permits such benefits to be at least taken into consideration within a process where evaluators are expected to make judgments about their significance explicit.

To be more concrete, let us return to the studies conducted in Kenya and Thailand. What if the respective ethics committees had information about how participants experienced benefits from previous studies? In the Kenya case, ethics committees could have consulted the rich social science literature on participant and community experiences in Kenyan biomedical research.³⁰ They could then have related that to the protocol and the approved consent form to inform judgments of likely benefits. Note that if there were prior Kenyan studies about risks in research, the appropriateness of taking such information into consideration in risk-benefit assessments would go without saying. At the same time, given that the Kenyan research was originally considered low risk, any inclusion benefit determined to be significant would only improve the risk-benefit relationship.

With respect to the study in Thailand, the ethics committees could regard “feeling normal again” through ceasing HIV treatment as a potential inclusion benefit to be weighed against study risks or could consider changing consent forms to include this as a potential study benefit. Considering the unknown and potentially serious risks involved in HIV cure research involving treatment interruption, the first option would unlikely survive ethical assessment. When “weighed,” it would likely be found very light. The second option may be worthy of more serious consideration. Given that prospective participants are supposed to be informed of all potential risks and benefits, as well as provided information that might affect their continued participation once they are enrolled in a study, what would justify withholding this information from them? The findings from the Kenyan and Thai studies suggest that a role for inclusion benefits in ethical assessments could promote the ethical conduct of research rather than undermine it. Again, if social science research revealed hitherto unknown social risks related to research in a certain setting (such as stigma or loss of employment), that information would clearly be considered admissible in ethical assessments. Why not the same with inclusion benefits, since such assessments are, after all, of risks and benefits?

Whether and to what extent incorporating inclusion benefits into ethical risk-benefit assessments is reasonable is likely to depend on a number of context-sensitive factors, the most prominent of which may be the level of study-related risk. When a study poses little risk to participants, there is little reason to be concerned about whether inclusion benefits have been duly considered. However, when a study poses significant risks, whether inclusion benefits have been taken into consideration in the risk-benefit analysis becomes more important. One may object that such benefits could serve to justify high-risk studies and that inclusion benefits are hard to anticipate before a study starts. But the “weight” those benefits have in ethical assessment depends on the judgment of the ethics committee, and many risks that play a role in ethical assessments are speculative in nature. Note again that there is little opposition to the idea that ethics committees should take inclusion risks (such as stigma, loss of employment, legal jeopardy) into consideration. Why not also bring in inclusion benefits, if a comprehensive and critical assessment of risks and benefits is what it is all about?

SOME OBJECTIONS

Before concluding, we should acknowledge that there are at least two other arguments against the idea that inclusion benefits should play a role in risk-benefit assessments besides those of the standard view explored above.³¹ According to the first argument, incorporating inclusion benefits involves a misunderstanding of research as a public good. Research aims to contribute social value by conducting high-quality science, not by satisfying participant preferences; pursuing the latter could in fact detract from the former. Because of its value as a public good, research can be ethically justified even if it provides no benefits to participants, as long as risks are minimized and are necessary to produce knowledge that is sufficiently socially valuable.³² However, in arguing that inclusion benefits ought sometimes to be taken into account in ethical assessment, we are not arguing that these benefits should trump other values, such as social value or scientific validity. The “weight” of inclusion benefits could be a matter of committee discernment, including the judgment that they are relatively unimportant.

A second, related objection claims that incorporation of inclusion benefits involves a misunderstanding of the nature of risk-benefit evaluations in research ethics review. According to a leading conception of risk-benefit evaluation (component analysis), the review should proceed by looking at each study intervention individually and asking if the risks are favorable in relation to the benefits of the intervention.³³ This approach is partly motivated by the desire to avoid “package-deal fallacies,” a situation in which researchers “buy off” the risks of some study interventions by reference to benefits associated with other ones. In our view, this criticism highlights something important about inclusion benefits: some of them (like the benefits of analytic treatment interruption in the Thai study) are related to individual study interventions, and some are not. The traditional approach is to exclude all inclusion benefits from risk-benefit assessments in an attempt to avoid package-deal fallacies altogether. But there could be other options. If research ethics committees are—as they should be in general—attentive to the dangers of package-deal fallacies, they could bring inclusion benefits (especially those associated with particular interventions) into component analyses. Alternatively, inclusion benefits could come in after component analyses are done. For example, in first-in-human research, concerns are sometimes raised that potential risks are exclusively being weighed against the potential future benefits for society, seeing as there are no study-related benefits for participants. While the social value of the research may be regarded as sufficient ethical justification, worries about possible exploitation may be alleviated to some extent by the recognition that participants may find value in such “no-benefit” studies after all.

SPACE FOR PARTICIPANT EXPERIENCES OF BENEFIT IN ETHICAL ASSESSMENTS OF RESEARCH

Our conclusion is relatively modest. We are not saying that inclusion benefits experienced by participants and revealed by social science should be driving forces in ethical assessment and decision-making in research. Instead, against the standard view, we have argued that they can and sometimes should play a role in deliberations and practices, and we have provided a few concrete examples of how this might look. Clearly, more empirical and conceptual work in this area is needed. For one thing, how ethics committees should interpret social science data about the benefits of research from the participant perspective needs further investigation. In addition, in some low- and middle-income countries, there has been little social science research about inclusion benefits, and hence, there is little for ethics committees to draw on.

A more robust account would also include an explanation as to why the standard view has been so dominant and why inclusion benefits have been so devalued. We close with some speculation on the latter point. As is known, research regulations and research ethics committees were largely developed in response to questionable and unethical research perpetrated on vulnerable populations. Among populations deemed especially vulnerable are children and those with cognitive disabilities who cannot provide informed consent. To justify research with the latter, particularly research with greater than minimal risk, the participants would need to have a likelihood of personal benefit—but not just any kind of benefit, and this is where a strong evaluative distinction between direct and inclusion benefits comes in: the only kinds of benefit with justificatory power for research with these vulnerable populations are those directly related to the study intervention and the health condition associated with it. For these vulnerable populations, offsetting research risks with inclusion benefits (food, extra medical services, the sense of being useful to society, and so forth) was considered insufficient, even offensive. This seems reasonable. But what seems to have happened more recently is less reasonable: a general and indiscriminate devaluation of inclusion benefits within ethical assessment regardless of the study population, with inclusion benefits regarded as dangers to the ethical assessment process rather than potentially positive contributors to it.