Clinical research involving human subjects and potentially marketable products carries with it unique ethical considerations. Human research subjects, the medical profession, and the public rely on clinical investigators to make decisions based solely on professional judgment, without regard for personal gain.1,2 However, growing evidence suggests that close financial ties between industry sponsors and clinical investigators may influence the quality and outcome of clinical studies.3,4 Furthermore, these relationships may undermine the public's trust of clinical research.5,6
As industry sponsorship of clinical research grows,7–9 investigators' personal financial relationships with their industry sponsors have come under increasing scrutiny. Although there is variability among institutions and across states,10 a range of policies exists to monitor and regulate investigators' financial relationships, including state and federal conflict-of-interest policies and policies at individual institutions. Recent studies show that more than 70% of academic research institutions have policies related to financial conflicts of interest. However, these policies may vary considerably in their substantive focus and reporting requirements.10,11 For instance, the University of California, San Francisco (UCSF) and Stanford University, two institutions that share comparable rates of clinical research activity, have divergent conflict-of-interest policies in effect.
At UCSF, faculty are required to disclose financial ties with companies sponsoring their research each time they submit a grant, contract, or gift agreement. UCSF has a campus-specific policy stating that faculty who are principal investigators in a sponsored clinical study may not concurrently receive any compensation from the sponsor, including honoraria and consulting fees, during the course of the clinical study. Nor may the faculty member have any investment or decision-making relationship with the sponsor (e.g., service on the board of directors or management committee) or be an officer or an employee of the company sponsoring the study. Clinical investigators must sign a statement agreeing to this policy before they begin a clinical trial.
At Stanford University, faculty disclose financial ties with each grant, contract, or gift submission and are also required to submit disclosures to their school's dean of related financial interests on an annual basis and when new relationships arise. Related financial interests include, for example, any consulting relationship, any employment relationship, and any significant financial interest in the sponsoring company (i.e., current or pending ownership interests in an entity amounting to at least 0.5% of the company's equity or at least $100,000 in ownership interests). Disclosures are reviewed and managed on a case-by-case basis by a conflict-of-interest committee.
To begin to explore the possible implications of such divergent policies for clinical researchers, we interviewed active clinical investigators at both Stanford University and UCSF. Our interviews were designed to elicit information on investigators' financial relationships with industry, their awareness and opinion of conflict-of-interest policies at their institutions, and their general attitudes about academic–industry relationships in clinical research. Participants from UCSF were randomly selected from a list of 106 principal investigators on the medical school faculty who were receiving industry funding for clinical research. At Stanford University, we randomly selected participants from a department of medicine list of 133 faculty members who appeared to be engaged in clinical research (based on the description of research on the Stanford University Web site in February 2001, available at 〈http://frd.stanford.edu〉). The interviews (19 at UCSF and 15 at Stanford) were conducted by telephone, and each lasted approximately 30 minutes. The UCSF Committee on Human Research and the Stanford University Human Subjects Panel approved the interviews.
The investigators we interviewed worked in a variety of fields within medicine, including cardiovascular medicine, gastroenterology, oncology, and pediatrics. Most had received industry funding for their recent clinical research projects (26 of 34). Of the faculty funded by industry sponsors, 19% had received consulting income or honoraria from the sponsor of the clinical study and 12% held equity in the company sponsoring their most recent clinical study. None held executive or management positions in the sponsoring company.
AWARENESS AND ACCURACY OF UNDERSTANDING OF CONFLICT-OF-INTEREST POLICIES
For institutional conflict-of-interest policies to be effective, faculty must be aware of and understand the requirements of the policy. Although all of the investigators we interviewed said that they were aware of a conflict-of-interest policy at their institution, more than half were unable to accurately describe that policy.
At UCSF, where the conflict-of-interest policy prohibits investigators from concurrently receiving compensation from the sponsor of the trial or holding a decision-making position within the company, 58% (11/19) of the investigators we interviewed inaccurately described the policy. The most common misunderstanding was that the policy does not prohibit personal compensation, but only requires its disclosure. As one UCSF investigator explained, “you are supposed to disclose if you are receiving personal money from a sponsoring company, such as for giving talks.” Another common misunderstanding was the belief that there is a limit to the amount of money investigators could receive, rather than a total ban on any form of concurrent compensation or service in a management capacity during the course of the study.
At Stanford University, where the conflict-of-interest policy related to clinical research is less restrictive, 27% (4/15) of the investigators we interviewed gave inaccurate descriptions of the policy, and 20% (3/15) had only a partial understanding of the policy. Most commonly, investigators misunderstood when they were required to disclose to the university (disclosure is required annually and when new relationships arise).
Investigators from both campuses who had difficulty describing their institution's conflict-of-interest policy felt that this was proof of their distance from potential conflicts. According to one Stanford University investigator, “Since I've never come close to a conflict, I can't cite the policies.” Similarly, a researcher at UCSF said, “It's never been an issue for me, so I'm not sure what the rule is.”
ATTITUDES TOWARD CONFLICT-OF-INTEREST POLICIES FOR CLINICAL TRIALS
Although there is speculation that restrictive policies covering conflict of interest for clinical trials could inhibit research or that nonrestrictive policies could erode patients' trust, little is known about the effect of such policies. At both UCSF and Stanford University, the majority of investigators we interviewed assessed their campus' policies as “reasonable,” “necessary,” and exhibiting an “appropriate level of restrictiveness.” However, some investigators at each institution found their campus' policy too restrictive. According to one Stanford University investigator:
I can't breathe. It has absolutely hindered my research. The bureaucratic delays. The cover-my-behind things I have to spend my time doing. I have faced major impediments. … I do have stock in the [sponsoring] company, but I don't think it is appropriate for Stanford to oversee me in that way.
Investigators from both campuses were critical of blanket bans on personal ties to industry (such as the UCSF policy), and some suggested that close financial ties to industry might be necessary. One Stanford University investigator commented:
For phase I and phase II trials of new and innovative drugs and procedures, often it is essential to have an investigator with [a financial] interest involved in the study because this is the person who has the most knowledge – the person with prior hands-on experience. Total barriers to personal financial ties with industry can be bad for the patient.
On the other hand, some investigators characterized the policies as quite lenient. According to one Stanford University investigator, “From my experience, Stanford's policies are less stringent than any institution I know of. This is probably because Stanford encourages entrepreneurship and industry interaction.”
Almost half of the investigators who responded would recommend changes to their campus' policies. At Stanford University, investigators called for more enforcement of the existing policies and a “more stringent” approach to financial relationships; in contrast, at UCSF, investigators called for a less stringent policy that would allow certain relationships during the study, such as consulting. One investigator recommended, “getting rid of the nickel and dime stuff. One hundred dollars here or there isn't going to influence investigators. Go after the big stuff—that's what matters.”
Investigators from both institutions who participated in our study expressed concern that, despite the policies, institutions might not be effective in monitoring all the relationships between faculty and industry and that some of their fellow investigators were involved in inappropriate relationships. A Stanford University investigator said, “The policy is reasonable, but I've always wondered how all of my colleagues can do what they do and still not be violating the policy.” A UCSF investigator expressed similar concerns, “I think [the policy is] reasonable, although I have my personal doubts about a lot of investigators at UCSF disclosing what I think they should be disclosing.”
When asked directly how their research had been affected, most (76%) investigators said that the policies had no impact on their research. A majority of investigators (74%) also felt that conflict-of-interest policies were neither an incentive nor a disincentive to work at their institution. Those investigators who felt discouraged from conducting clinical trials generally felt that the conflict-of-interest policy had little to do with their discouragement. Faculty members who felt discouraged about conducting their research reported that their discouragement had much more to do with bureaucratic delays. According to one Stanford University researcher, “There are layers of inhibiting levels of bureaucracy,” and a UCSF investigator, also stressing the multiplicity of problems involved in conducting clinical research, said, “Conflict-of-interest policies are only a minor cog in a very large wheel of abuse.”
SITUATING REGULATORY RESPONSIBILITY
When asked where the responsibility for the ethical conduct of research lies, many investigators suggested that professional societies, the Food and Drug Administration (FDA), the general public, and even the investigators themselves were appropriate arbiters of ethical research behavior.
Several investigators argued that professional societies were active and well suited for monitoring and regulating investigators' financial relationships. According to one Stanford University investigator:
What you should understand is that a lot of conflict-of-interest rules have less to do with university rules than with outside professional associations. … Every time I present a paper or publish, I need to declare my financial conflicts. … These professional organizations have rules that are more restrictive and affect me more than Stanford's rules. Because everyone knows each other and each other's research, people in these organizations are well aware if you are being up front about conflict-of-interest issues.
Another investigator, calling the FDA a “very unforgiving authority,” felt that the FDA's monitoring of clinical trials was sufficient. Other investigators suggested that the public might be the best judge of disclosed conflicts. As a Stanford University investigator said, “The best solution would be to have an open book at Stanford of all faculty and what kind of money they receive from industry. That way, public scrutiny could decide who is acting inappropriately, and those people could be prosecuted.”
A quarter of the investigators pointed to public disclosure as the best way to mitigate the ethical risks associated with a researcher's personal financial ties to industry sponsors. “Disclosure gives us credibility,” said one UCSF researcher, and another said, “I believe that if these relationships are disclosed, then they are okay. There should be some limit to the amount investigators can receive, but generally, I think as long as they disclose, it's okay.”
Other investigators expressed the belief that the ultimate responsibility for regulating conflicts should lie with the investigator. According to one Stanford University researcher:
It's a delicate thing. You have to decide for yourself. For example, I'm getting money from [a large pharmaceutical company] for a study I'm working on. They also have me on speakers' bureau. I feel comfortable with this arrangement as long as the slides I use are my own, and I'm speaking about my own research and opinions. I don't think the information I present has anything to do with what [the company] wants me to say. This system can be, and is, abused. Some people do give canned talks prepared by the companies that are paying them.
Another Stanford University investigator stated, “Obviously there is the potential for bad science, but I think that exists regardless of whether or not industry is involved. The issue fundamentally boils down to the sense of responsibility of individual investigators.”
APPROPRIATE AND INAPPROPRIATE FINANCIAL RELATIONSHIPS
In addition to questions about their own institutions' policies, we also asked investigators about their attitudes, in general, toward personal financial relationships between investigators and their industry sponsors. Investigators from both institutions drew clear distinctions between holding equity in a company sponsoring one's research and other forms of financial ties. As one UCSF researcher said, “Holding equity is the biggest problem because then you have a vested financial interest in the success or failure of your research.”
Most investigators agreed that other activities, such as consulting, sitting on scientific advisory boards, and giving sponsored talks, were less problematic. Especially at UCSF, where campus policy prohibits faculty from concurrently receiving any form of compensation from the sponsor of the study, including stock, views were consistent that stock options are the primary problem. According to one researcher, being “a large stockholder or on the board of a company is obviously inappropriate. However, being prevented from giving a talk unrelated to your research is absurd.” Another UCSF researcher said that, “$10,000 here or there is not a big deal. Personal financial relationships with sponsors are necessary for growth. People have to look at the big picture and see the benefits that come from academic–industry relationships.” And another UCSF researcher was critical of those who would assume that, because researchers have been paid for providing consulting services to a company, they are automatically guilty of unethical behavior:
The biggest problem is stock ownership. I think there is a public misperception that if you consult or advise for a company you are in their pocket. Especially in the case of large companies, one advisor isn't that important. … These advisory board positions and consulting positions are usually scientific—they are not influential positions.
BENEFITS AND RISKS OF INDUSTRY RELATIONSHIPS
We also asked investigators about the benefits and risks they saw in their own relationships with industry sponsors. Investigators from both universities reported numerous benefits in having industry sponsors for research. The most commonly mentioned benefits were the extra funds that industry sponsorship provided for laboratory support, salaries for staff, other academic pursuits, and the access to new or otherwise unavailable drugs or devices. A Stanford University investigator said, “We have a drug available for testing that we otherwise wouldn't have. I also get a little salary support that helps me have time to do other research.” The intellectual stimulation of involvement with cutting-edge research was also a frequently mentioned benefit to industry relationships. A Stanford University investigator observed:
I see it as a two-way positive street. … Industry benefits because they get the best minds to help shape their research and also get the credibility that comes from having academics involved. Faculty benefit from industry relationships in that they get to have a voice in the cutting edge of what's going on in patient care and to shape the direction of research.
In general, investigators felt that their own relationships with industry posed little risk either to them personally or to the profession. Many investigators did not feel that their industry relationships posed any risk at all. Other investigators reported feeling slightly at risk, with a few worrying that they would be “used” by a sponsoring company. As one UCSF investigator said, “The only risk is that the company could take my ideas and run with them. The risk is to me, the consultant.”
In terms of risk to the profession, there was some concern that industry was shaping the direction of future research and that industry-supported research could be vulnerable to publication delays or suppression of results. A few investigators believed that there was a risk to scientific integrity. As one Stanford University investigator explained:
There is the risk that I become a complete whore and begin saying things that I don't believe. I'd hope I'd recognize this if it were happening to me, but it's hard to know. The risk to the public is one of fraud—scientists say something is true when it isn't. I don't think that conflict-of-interest rules can mitigate either of these risks—it is basically up to the individual investigators to act ethically.
Among those researchers who did not see a risk associated with personal financial relationships with industry, several pointed to their belief in their own abilities to recognize when they were in conflict. As one Stanford University investigator explained, “I recognize that I am in conflict, but believe that I can handle it. If I couldn't handle the conflict, I wouldn't have gotten involved.” Another researcher argued that “data analysis is objective,” and therefore, he is not vulnerable to conflicts of interest.
The most striking feature of our interviews is the range of understandings and attitudes expressed by clinical investigators and their implications for administrators, professional societies, and policymakers concerned with conflicts of interest. Based on our interviews, it seems likely that faculty are poorly informed regarding their campus' conflict-of-interest policies. Fewer than half of the faculty we interviewed could accurately state their institution's policy, even though the policies are posted on universities' Web sites and staff at each institution are devoted to enforcing the policies. Because administrators depend on faculty to consistently disclose their relationships and because disclosure requires knowing when and what to disclose, these findings are of concern. If some investigators believe that they have no need to know the policy because they are not in conflict, then more must be done to educate investigators about both the specifics of relevant policies, as well as the nature of conflicts of interest. Furthermore, some faculty we interviewed perceived that the policies are inequitable because they are not consistently applied to all faculty. Such (mis)perceptions could also lead to noncompliance.
Many faculty in our study believed that public disclosure is the best way to mitigate conflicts of interest and monitor investigators' research behaviors. Indeed, investigators seem to treat disclosure as a panacea for managing all the problems associated with financial relationships with industry sponsors. However, there are no consistent standards for disclosure,12 many journals do not publish disclosed relationships,13,14 and little is known about disclosures made in professional meetings or other public settings. Reliance on disclosure as the best or primary solution to conflicts of interest seems problematic in light of these findings.
Faculty in our study also favored other nonregulatory ways of managing conflicts, such as voluntary compliance with standards set by professional societies or self-regulation (i.e., the investigator believing that she or he can “handle” the conflict). Recent guidelines from the American Association of Medical Colleges (AAMC) are one example of a conflict-of-interest policy for clinical trials promulgated by a professional society.15 The AAMC guidelines recommend that financial ties between investigators and the sponsors of their clinical research be evaluated on a case-by-case basis by an institutional conflict-of-interest committee. Financial ties between clinical investigators and their sponsors would be allowable under “compelling circumstances.”15 The circumstances that the conflict-of-interest committee should evaluate include the nature of the research, the magnitude of the financial interest and the degree to which it is related to the research, the extent to which the interest could be affected by the research, the degree of risk to the human subjects involved, and the extent to which the financial interest is amenable to oversight and management.15
The AAMC's guidelines raise the issue of how much money is considered an “acceptable” conflict of interest. The AAMC's guidelines, which are modeled after the Public Health Service financial disclosure regulations,16 could allow investigators to receive compensation of up to $10,000 from the companies that sponsor their clinical trials or hold up to 5% equity in the company. The investigators we interviewed would tend to agree with these financial limits on income as they suggested that “small” amounts of money do not pose conflicts of interest. However, health professionals and consumers (depending on their personal financial situations) might have different perspectives on what is considered an amount of money that poses no conflict. Little is known about the public's opinion of clinical investigators' financial ties to industry. However, even health professionals' decisions can be influenced by small gifts.17,18 Investigators agreed that owning equity in a company or having a decision-making capacity was a conflict of interest that should not be allowed.
The investigators we spoke with did not feel that the conflict-of-interest policies at their institutions hindered their research, although this argument is sometimes used against conflict-of-interest policies for clinical trials. The investigators we interviewed placed their clinical trial research into the larger context of conducting research at a major university and were equally troubled by other bureaucratic requirements.
Although most clinical investigators in our study recognized that financial relationships with industry sponsors pose possible conflicts of interest, many believed strongly in their own ability to recognize these conflicts, avoid bias, and regulate their own behavior. Investigators have this attitude despite the publicity that has been given to a few high-profile cases involving the suppression of research by investigators with financial ties to companies.19,20 The investigators' expressed belief that risks of conflict of interest do not apply to themselves, a viewpoint that is consistent with their support for self-regulation. However, the well-publicized risks, mentioned above, and the data on the association of funding and financial ties with outcomes of research21–23 suggest that investigators may be underestimating the risks to the integrity of their research. Furthermore, these views may ultimately undermine the effectiveness of institutional policies. Insofar as investigators believe in their own abilities to “handle” conflicts of interest, policies may be perceived (and perhaps treated) as irrelevant. Thus, a fundamental challenge facing administrators and policymakers is to demonstrate to all investigators, both clinical and nonclinical, that the potential for bias, pressure, and conflict is relevant to all investigators with industry relationships.