Disclosing Conflicts of Interest to Potential Research Participants: Good for Nothing?

Abstract

The growing commercialization of science has raised concerns about financial conflicts of interest (COIs). Evidence suggests that such conflicts threaten the integrity of research and the well-being of research participants. Trying to minimize these negative effects, federal agencies, academic institutions, and publishers have developed conflict-of-interest policies. Among such policies, recommendations or requirements to disclose financial COIs to potential research participants and patients have become commonplace. Here, I argue that disclosing conflicts of interest to potential research participants fails to achieve the weighty moral goals that presumably ground such policies. This is so either because disclosure is simply a wrong means for achieving some of the goals in question or because, although disclosure could be an appropriate means for some of those goals, the way in which it is implemented prevents fulfillment of the desirable moral aim.

The commercialization of science in general and biomedical science in particular is now common.¹ Collaborations of pharmaceutical, device, and biotechnology industries with physicians and scientists in academic medical centers are widespread. Researchers can receive funding from industry, serve as consultants for the companies that fund their research, belong to company advisory boards and speaker bureaus, and own stocks in those companies. In the biomedical sciences, intellectual property rights—particularly in the form of patents and licensing agreements—and academic start-ups have been growing.² These activities have become important to producing advances in biomedical and clinical research and relevant to bringing such advances to clinical practice.³ Such collaborations, intellectual property rights, and start-ups can also encourage innovative approaches necessary to address complex health problems or can provide financial resources to the scientific enterprise in the face of increasing economic pressure and global competition.

These benefits notwithstanding, industry-academia collaborations come with some costs. Among a variety of the concerns that these relationships create,⁴ the presence of financial conflicts of interest (COIs) for individual researchers, clinicians, and institutions have been of particular interest.⁵ COIs have the potential to bias scientists’ judgments, leading them to act in ways that are inconsistent with their duties to ensure the soundness of science and protect the welfare of research participants. Financial COIs can thus undermine the integrity of the research and harm the interests of participants in research studies.⁶

Calculating the extent of financial COIs is difficult, but studies have found that these conflicts are common and becoming more so. For instance, recent studies have found that between 29% and 69% of published clinical trial reports include disclosures of COIs.⁷ Evidence also shows that about one-third of life sciences faculty members engage in industry consulting and are paid to advise companies whose activities are related to the faculty members’ areas of expertise.⁸ A 2015 survey of 446 publicly traded U.S. health care companies found that academic leaders held near 10% of company directorships and that 41% of companies had at least one academic director on their board.⁹ Moreover, undisclosed COIs appear to also be common. A recent cross-sectional study, for instance, assessing the accuracy of self-reported financial COIs in two high-impact academic journals found that over 80% of physician-authors failed to disclose relevant financial compensation.¹⁰ Similarly, recent evidence shows that reporting of financial COIs in meta-analyses published in high-impact journals is far from optimal.¹¹

Although it is challenging to measure the impacts of COIs on science and medicine, various studies have found that financial COIs can affect the objectivity and integrity of research and can compromise the clinical care of patients. For instance, a variety of studies have found that researchers whose studies receive funding from commercial parties are significantly more likely to report positive outcomes than are researchers whose studies are funded by not-for-profit organizations.¹² COIs can also negatively affect the quality and reliability of systematic reviews,¹³ as well as recommendations in clinical practice guidelines.¹⁴ Arguably, the increasing presence of financial COIs can also undermine warranted public trust in the scientific community and the medical profession.¹⁵

In an attempt to minimize these negative effects, federal agencies, academic institutions, and publishers have developed policies regarding COIs. Usually, such polices involve disclosure of financial COIs to institutional officials and agencies and management of conflicts in various ways.¹⁶ Among such policies, recommendations to disclose financial COIs to potential research participants and patients have gained importance.¹⁷ In this article, I argue that disclosing conflicts to potential participants fails to achieve the weighty moral goals that presumably ground such policies. This is so either because disclosure is simply a wrong means for achieving some of the goals in question or because, although disclosure could be an appropriate means for some of those goals, the way in which it is implemented prevents fulfillment of the desirable moral aim. Before making my case, I discuss what are taken to be some of the most significant moral goals that disclosing conflicts to potential participants presumably achieves. I conclude with some reflections on how to move forward.

THE MORAL GOALS THAT DISCLOSURE TO POTENTIAL PARTICIPANTS AIMS TO ACHIEVE

Disclosure of COIs to potential research participants has been defended on various weighty moral grounds.¹⁸ Three of them seem particularly worthy and have received a significant amount of attention: promoting warranted trust,¹⁹ fostering informed decision-making, and protecting research participants’ welfare.²⁰

As to the first of these, trust, both ethical and epistemic, is necessary for the clinical research enterprise to realize its epistemic and social aims.²¹ Lack of trust can lead the public to show skepticism with regard to the funding of science. Moreover, clinical research needs people to enroll in clinical trials. If potential participants fail to trust researchers and research institutions, they are unlikely to volunteer for studies. Lack of participation impedes the progress of science and hinders the development of beneficial treatment and interventions. The transparency exemplified by disclosure of COIs to potential participants can help to persuade them that researchers are honest, which denotes trustworthiness.²² To the extent that participants would feel betrayed by learning about financial COIs that have been hidden, disclosure can also safeguard trust by eliminating accidental discovery. Arguably, the openness and transparency of disclosure also encourages warranted trust by assuring those recruited to enroll in research that institutions are actually attending to financial COIs and are putting strategies in place to manage the risks that such conflicts create.

Although there are some disagreements regarding how much and what information should be provided to potential research participants in order to facilitate informed decision-making, there is a general agreement that they should receive information material to the decision in question.²³ Clearly, information involving risks is material to someone’s decision to participate in a research study, and, as indicated earlier, financial COIs pose risks to research participants. This is the case for various reasons. Because conflicts can intentionally or unintentionally bias researchers’ judgments, this might lead researchers to enroll participants who do not meet inclusion criteria or fail to monitor appropriately the health of those enrolled in studies or those being recruited to enroll,²⁴ to provide insufficient or false information during the consent process and enroll participants in trials based on disclosure of fabricated data,²⁵ or to use invalid methodologies and fail to report or monitor adverse events.²⁶ Hence, researchers’ COIs can increase the risk that participants’ health and life will be harmed, that their rights will be undermined, and that the integrity of the research will be threatened. When the scientific results are invalid, risks to participants cannot be balanced against potential benefits. Disclosing COIs to potential participants, then, provides them with information necessary to determine whether they want to participate in a particular research study.

If, as pointed out, researchers’ COIs increase risks of harm to participants and threaten the integrity of research, disclosing such conflicts could contribute to safeguarding the well-being of potential participants. Presumably, disclosure could do so in several ways. First, disclosing conflicts could lead potential participants to recognize the increased risks of participation. As a result, they could either choose not to participate or be more attentive to the particulars of the research and how it is being conducted. Second, researchers could be apprehensive about having to disclose COIs to potential participants—some of whom could be their patients. This could deter some researchers from entering into relationships that create financial COIs.²⁷ Alternatively, the need to disclose could make researchers more attentive to possible bias so that they could prevent it.

DOES DISCLOSURE ACHIEVE ITS INTENDED MORAL GOALS?

Few would deny that promoting warranted trust in scientists and the research enterprise, facilitating informed decision-making, and protecting participants’ welfare are worthy goals. Unfortunately, it is not clear that disclosure of financial COIs to potential research participants constitutes an appropriate means to achieve them. This is either because disclosure is simply the wrong means to realize some of these desirable aims or because, although it might be an appropriate means in principle, the way in which financial COI disclosure is implemented prevents their fulfillment.

Let’s begin with the promotion of warranted trust. Trust is a complex phenomenon.²⁸ In general, people do not trust others completely. Usually, we trust others to do certain things. For example, potential research participants may trust investigators to conduct research according to appropriate scientific standards, but not to fix their cars. They may trust investigators to protect them from unnecessary harm, but not to take care of their children. In trusting, we presuppose that the person trusted is competent in some regard, such as conducting research, giving advice, or flying a plane. There is also a presupposition that the person trusted will be rightly motivated to do what we are trusting them to do. Both competency and willingness or motivation are central elements of warranted trust—even if there is disagreement about the exact nature of such motivation.²⁹

Although transparency has often been proposed as a condition for warranted public trust in the context of institutions—such as corporations and governments—what transparency in that context entails and what benefits it offers are controversial matters.³⁰ But transparency has rarely, if ever, been understood as a condition of interpersonal trust. Indeed, transparency is neither necessary nor sufficient to foster interpersonal trust, and disclosure of at least some information, for example, of behavior that shows questionable character, can actually undermine trust.³¹ In some cases, disclosure requirements are indications that trust is lacking, and the disclosure works as a way to show that one is actually trustworthy.

Furthermore, in the context of research, the disclosure of a COI is presumably a way to call attention to the fact that the researcher might be affected by bias—conscious or unconscious—and thus that the participants should consider how such bias could affect their own welfare and the conduct of research. Hence, the appropriate role of disclosure would be to increase warranted distrust or at least to encourage one to question whether one should place one’s trust in researchers. Notice that this need not involve a negative judgment regarding the researcher’s moral character but simply a recognition of the role of unconscious bias in the researcher’s actions or decisions.

It is true that, in some contexts, disclosing COIs can provide evidence of honesty and integrity and thus of trustworthiness.³² But the disclosure of COIs to research participants does not depend on researchers’ motivations. Such disclosure is mandated by institutional policies. Hence, disclosure itself does not constitute evidence of integrity or honesty. It seems, then, that despite the common claim that disclosure is an appropriate means to promote the warranted trust of potential research participants, this is not obvious from a conceptual point of view.

The empirical evidence regarding the effects of disclosure on trust is not very helpful either. In fact, it is hard to determine the value, particularly as a guide for policy, of much of the empirical evidence on the relationship between disclosure of COIs and trust. Although studies provide conclusions about effects on trust, often those are inferred indirectly from some other factor, such as peoples’ inclination to accept advice or to decide to enroll in research studies. Furthermore, while some studies report that disclosing financial COIs increases trust,³³ others conclude that it can actually decrease it,³⁴ and still others show no particular effect.³⁵ These mixed results are, at least in part, attributable to the complexity of the phenomenon of trust and the lack of conceptual clarity that often vex empirical studies on this topic. The empirical literature on disclosure and trust rarely provides definitions of trust and its various dimensions, does not indicate what exactly it is that research participants and patients should trust when conflicts are disclosed to them, or even fails to discuss who or what is supposed to be the object of trust, for instance, the investigator, the research institution, or the research produced. More important still, the literature often does not consider whether the trust affected by disclosure is warranted or not.

These are not small problems when attempting to use empirical evidence as grounds for policy-making. After all, if disclosure increases trust but such trust is unwarranted, we should hardly conclude that disclosure is an appropriate means of promoting a worthy moral goal. For example, some studies evaluating the effects of disclosure on trust in various contexts, such as medical and financial-investment arenas, have found that disclosure of financial COIs can increase trust because the disclosure is interpreted as an indication of honesty or evidence of professional standing.³⁶ But, as mentioned earlier, these interpretations are incorrect. Although, of course, some researchers might be willing to disclose financial COIs even if not required to do so, the majority do so because their institutions mandate it. Hence, such disclosures need not track honesty.³⁷ Similarly, even though financial COIs can indicate the professional standing of researchers, what disclosures are supposed to be doing is calling attention to the possibility of unconscious bias. Reputation or standing is not correlated with less likelihood of unconscious—or conscious—bias. Moreover, and showing again the complexity of the phenomenon of trust, some studies have shown that although after disclosure people might feel less willing to participate in research or to take the advice of an investor—because people do not trust such advice—they are also more inclined to participate or take the advice because they do not want to signal that they now distrust the researcher or advisor.³⁸ In these cases, the conclusion that disclosure increases trust would be unsound. What this evidence shows is people’s beliefs about when it is appropriate or not to show even warranted distrust.

Conceptually, then, it is not obvious that disclosing financial COIs is an appropriate means of promoting potential participants’ warranted trust in researchers’ judgments, and no arguments have been put forward of how this would be so. The empirical evidence is muddled or shows problematic effects of disclosure on trust. At a minimum, it provides the wrong kind of policy guidance regarding the effects of disclosure on warranted trust.

One might argue that disclosing conflicts can help promote public trust by showing that institutions are putting in place the necessary mechanisms to prevent the effects of possible bias in research resulting from financial COIs.³⁹ Certainly, strong regulations and enforcing mechanisms can contribute to people’s warranted trust in research institutions. Institutional and structural factors that are likely to encourage appropriate research practices and that make interest in research integrity and participants’ welfare primary can strengthen scientists’ resolve to be honest. They can make the development of character traits implicated in trustworthiness, such as honesty and integrity and benevolence, easier. But it is not obvious that disclosing conflicts to research participants can have these effects or that it is the best means of achieving warranted public trust. After all, few people not participating in research would be aware of the requirement to disclose COIs in consent documents. More importantly, other forms of disclosure to the public—researchers’ institutional website, federal government databases, journals, and so forth—are likely to be more effective in informing people that institutions are working toward protecting the integrity of scientific research and the welfare of research participants.

Although promoting warranted trust is not a goal that disclosure of financial COIs can achieve, it seems that disclosing such COIs can contribute to good decision-making. As mentioned earlier, information about risks to one’s well-being is material to making an informed decision about whether to enroll in a study, and financial COIs increase risks to research participants.

But can the information provided about such conflicts facilitate good decision-making on the part of individuals considering enrolling in a clinical trial? There is no federal policy requiring disclosure of COIs to research participants, and thus institutional policies that have those requirements can vary in the information they mandate or recommend. Nonetheless, usually, the disclosure is short and indicates the particular type of financial relationship affecting the researcher or team, for example, company funding, equity ownership, royalty payments, or patent ownership. For instance, if the research is sponsored by company X, the consent form can include that the researcher(s) receives money from company X or is a consultant to or a member of the advisory board for company X. Potential research participants might also be given information about whom to contact—the relevant IRB, conflicts office, researcher, or so on—if they wish to receive more information. More often than not, that the relationship in question presents a COI is not mentioned. No information is provided about why the disclosure is relevant or how the financial relationship disclosed might affect a research particiant’s welfare or the conduct of the research.

This minimal information might be sufficiently informative in a context where potential participants are well aware of what financial COIs are and of the ways in which they could negatively impact their welfare and the integrity of the research. Unfortunately, evidence suggests that potential participants have low awareness of the extent of financial relationships in biomedical research and medical care as well as of the implications of such relationships.⁴⁰ To the extent that this is the case, then, current disclosures are unlikely to promote informed decision-making.

If fostering informed decision-making is the goal of disclosing financial COIs, then such disclosure should provide appropriate information about the potential negative effects of financial COIs. Importantly, the revised Common Rule strengthens the informed consent requirement and calls for presentation of information in sufficient detail to facilitate understanding of what participation in a clinical trial can involve.⁴¹ This, however, presents some challenges. A significant amount of evidence shows that consent documents usually fail to facilitate decision-making because of their length and complexity.⁴² Adding sufficient information to make sense of the implications of financial COIs would only lengthen already-long consent documents. This is unlikely to facilitate decision-making. Moreover, whatever the relevance of financial COIs for potential participants’ welfare, it is clear that basic facts about research methodology, such as randomization and dosing, as well as about known and quantifiable risks, are at least as relevant to people. However, if, despite ample evidence indicating that many participants do not understand key facts about the studies in which they are participating, the consent process has not improved, it is unrealistic to expect that providing more information would be more successful in the case of financial COIs.

Perhaps information about financial COIs and their relevance could be succinctly provided on the newly required key-information portion of informed consent documents. According to the revised Common Rule, the consent document must include a concise presentation of key information that is most likely to assist prospective participants in the decision-making process. Yet it is difficult to see how this would solve the problems. Understanding how exactly financial COIs can negatively affect participants’ welfare and the integrity of research requires some knowledge of how unconscious bias can play a role in various aspects of the research. Given that people already have difficulties understanding important components of clinical research, it is not clear that they would be able to make much sense of the role that financial COIs can play, or at least that they could do so without a detailed description. And such a description would defeat the purpose of the key-information section, which must be concise.

It seems then that, although disclosing financial COIs could facilitate the moral goal of promoting informed decision-making, it is unlikely to do so in practice. This is because the information provided is perfunctory. However, providing adequate information for an individual to be able to make an informed decision would only compound the many problems already making it difficult for individuals to understand aspects of research necessary to make informed decisions.

Whatever the concerns with disclosure as an appropriate means to promote warranted trust or to facilitate informed decision-making, perhaps such policies are still an appropriate means of protecting research participants’ welfare. How would they do so? As mentioned earlier, disclosure could make people more reluctant to participate in research, and, thus, given that financial COIs could increase the risk of harms to participants in various ways, declining participation would protect individuals from those possible harms. But this seems problematic for various reasons. If we believed that financial COIs are sufficiently risky that there ought to be policies whose goals are to deter people from participating in research, it would seem that the right way to protect research participants would be to prohibit such conflicts rather than simply providing information to people considering whether to enroll in a research study.

Second, attempting to deter people from participating in research with the goal of protecting their welfare is an incongruous strategy for scientists and scientific institutions to promote. Research with humans is essential to knowledge production, and the ability to develop new treatments and interventions to cure and treat diseases depends on such research. Deterring people from enrolling in research studies might protect them from possible harms involved in the research but only at the cost of preventing the development of beneficial knowledge and interventions.

Third, the evidence—admittedly not very extensive—on the impact of disclosure on potential participants’ willingness to participate shows that it has little deterrent effect on the majority of people.⁴³ Worse still, some studies have actually shown that after disclosure of financial COIs, people are actually more, rather than less, willing to participate in research.⁴⁴ If this is correct, using disclosure of financial COIs as a means of protecting participants’ welfare by deterring them from participating is dubious. It is misguided conceptually, and if the evidence is any indication, it is also useless in practice.

Perhaps, then, disclosure would work to protect participants’ welfare by making them more attentive to what the research involves, how it is being conducted, and how the researcher’s possible bias (conscious or unconscious) could affect the research. Presumably, careful attention would allow participants to take the necessary measures to protect their welfare and the integrity of the research.

But again, this strategy is questionable. It would be morally wrong to use disclosure with this purpose. The burden of protection would fall directly on research participants. But research participants are those with the least ability to protect themselves in this way. They lack power and resources necessary to control the aspects of research participation relevant for their protection. Protecting participants’ welfare and safeguarding the integrity of research are the responsibilities of investigators, ethics committees, and research institutions, and such responsibilities should not be transferred to participants.

Furthermore, even if using disclosure for this purpose were appropriate, it is completely implausible that potential participants could use it in this way. It is not simply that disclosure of financial COIs in consent documents is usually perfunctory; it is that most people lack sufficient information about what clinical research involves, how bias operates, or how such conflicts can actually affect the research and participants’ welfare.⁴⁵ Indeed, determining the negative effects of financial COIs in research is difficult even for those with sufficient expertise, and it requires significant time and resources.⁴⁶

Perhaps, then, disclosing conflicts can help protect participants’ welfare because of the impact of disclosure on investigators rather than on research participants. This would appropriately put the burden of protection on researchers. Disclosure could deter researchers from entering into relationships that create conflicts. Contrary to the effect of dissuading people from participating in research, deterrence of relationships that create conflicts is not inconsistent with the aims of science. However, evidence indicates that most researchers do not find such relationships problematic.⁴⁷ Moreover, if the prevalence of reported conflicts of interest in research and the increasing funding from industry are any indication, disclosure does not seem to be limiting financial relationships.⁴⁸

Likewise, the idea that conflict disclosure could serve to make researchers more attentive to possible bias resulting from financial COIs is implausible and betrays a misunderstanding of how such conflicts might influence research conduct.⁴⁹ The main concern with financial COIs is that they can result in unconscious bias. Such bias might affect subject evaluation, how the evidence is interpreted, how trials are designed, what research questions to ask, and so on. Given that possible biases occur unintentionally, it is not clear how scientists themselves could be more attentive to them.⁵⁰

To make matters worse, at least some evidence suggests that, far from being an appropriate means to protect research participants’ welfare, conflict disclosure might actually increase risks to participants by making biases more likely to occur.⁵¹ This kind of evidence shows that disclosure can have perverse effects. Once disclosure has occurred, researchers can show less selfrestraint, perhaps because they believe that participants have been warned. They can also be inclined to emphasize certain aspects of research participation or to overstate potential benefits as a way to counteract the potential for disclosure to cause people not to participate in research.⁵²

Using disclosure of conflicts as a means to protect the welfare of participants is thus morally misguided, as it places the burden of protection on research participants, and conceptually confused, insofar as it presupposes that scientists can control unconscious bias on their own. The practice also seems to have no discernible effect on decreasing the prevalence of financial relationships between researchers and the companies that stand to benefit from the development and sale of drugs or other interventions studied, and it might actually increase bias and, with that, possible harms to participants and the integrity of research.

MOVING FORWARD

My claims here have several upshots. First, we should stop arguing that disclosing financial COIs to potential research participants can help achieve the moral goals of promoting warranted trust or the welfare of those who enroll in research studies. Those are worthy goals, but disclosing financial COIs cannot attain them. Indeed, in some ways, disclosure of financial COIs can actually undermine such goals.

Second, if fostering informed decision-making is the goal, then current disclosure practices cannot accomplish it. Disclosure statements are at best uninformative and at worst confusing (insofar as potential participants might be unsure why such information is being provided to them).

What should be done then? Answering this question requires attending to various considerations. How important is this financial-relationship information in the context of making a decision about research participation? Evidence indicates that most potential participants are indifferent to information about financial COIs.⁵³ Even when they do want to have that information, evidence suggests that, for most people, it would make no difference in their decision to participate in the research in question. Of course, the significance of this evidence is unclear. Studies are not specific about how much potential participants know about the possible negative effects of financial COIs. Hence, that most of these individuals do not care about this information might simply be the result of a lack of understanding about its relevance. Nonetheless, it does not seem implausible that many would find information on financial COIs immaterial. After all, it is difficult to determine the magnitude of risks to participants’ welfare that can result from financial COIs, disclosure is not at all a mechanism to prevent those risks, and those risks are unlikely to have a greater effect on participants’ welfare than other research-related risks, such as those of experimental drugs.

Of course, that some information is not dispositive of a potential participant’s enrollment decision is not sufficient to determine that it should be withheld. Some research participants might feel morally wronged were they to discover such information by other means.⁵⁴ Still, this concern must be weighed against the burdens of disclosure. After all, even if one assumes that people have a right to know this information, it is implausible to argue that such a right is absolute.

Some of the costs or burdens of conflict disclosure are related to compliance and enforcement. Disclosing conflicts to potential participants requires institutional processes that ensure appropriate and accurate disclosures to institutional officials.⁵⁵ They involve educating research teams, IRB members and staff members, conflict committees, and so forth about rules; monitoring compliance; enforcing relevant policies; investigating cases when compliance fails; and administering sanctions for noncompliance. Researchers and research assistants must ensure that appropriate information about conflicts is disclosed in consent documents, and IRB members and staff members must check that the information in the consent form is accurate.

Disclosing conflicts also imposes burdens on potential participants. On the one hand, if the information provided is insufficient for potential participants to recognize its relevance, then adding information to consent documents would simply increase their length and the time involved in the consent process without offering any benefits. Furthermore, if, as some evidence indicates, information about conflicts might lead some potential participants to unjustifiably perceive researchers in a more favorable way—to be more competent, for instance—this would also create misconception burdens.⁵⁶

On the other hand, if more information was provided to make it meaningful enough, potential participants would have to decide how important it was and what weight to give it in relation to other choices regarding participation. Consent documents would increase in length, and more time would be needed for the consent process. Moreover, given the evidence showing that longer consent documents undermine good decisionmaking, providing more meaningful information might actually be counterproductive.

A third upshot of my claims is that more and better empirical research is needed to determine what people would want to know about researchers’ conflicts, why they would want that information, how such information is weighed, and what are the best ways to provide such information so as to actually foster good decisionmaking. Even if one believed that potential participants have a right to this information, empirical evidence regarding how to communicate it in effective ways would be of utmost importance.

CONCLUSION

Financial COIs in research are concerning. Disclosing such conflicts to, for example, public databases, research institutions, and publication venues has become a mandated practice. As has been acknowledged, disclosure of conflicts—in all these cases—is an inadequate strategy to address the concerns such conflicts raise.⁵⁷ At best, disclosure of conflicts is only a needed step in their management. Although not mandated by federal regulations, disclosure of financial COIs to potential research participants is becoming more common, with many research institutions mandating or highly recommending such disclosure.⁵⁸ Requirements or recommendations to disclose financial COIs to potential participants is said to promote several weighty moral goals: to foster warranted trust, protect the welfare of those who enroll in research, and facilitate informed decisions. I have shown here that using disclosure of financial COIs to achieve these goals is misguided. Such disclosures are simply unable to promote warranted trust or protect participants’ welfare. They could in principle serve to facilitate informed decision-making, but the way in which disclosure is commonly practiced fails to attain this goal. If that is indeed the aim of disclosure, then more research is necessary to determine how best to accomplish it. As currently implemented, disclosure of conflicts might protect against legal liability and might give research institutions the incorrect perception that they have discharged their responsibilities to potential participants. They have done nothing of the sort.