Patient Outcomes Research Teams (PORTS): Managing Conflict of Interest (1991)

This chapter reviews definitions of conflict of interest used by other policy groups and institutions and provides a working definition. It also examines the categories of financial involvement that may give rise to such conflicts with particular attention given to university-industry relationships. Finally, it provides a brief overview of recent activities pertaining to conflict of interest in academic research and describes efforts in other sectors to deal with this issue.

Conflict of interest arises in a situation in which (1) one is in a fiduciary relationship with certain others, and (2) one's financial or professional self-interest substantially differs from the interests of those others. The concern with conflicts of interest arises in biomedical research because of the possibility that such conflicts, both real and perceived, may erode scientific objectivity and engender the loss of public trust (Barber, 1983, provides a general treatment; Relman, 1989).

Because PORTs conduct research for the public benefit, PORT researchers have fiduciary obligations to the public. Yet their interactions with other groups (e.g., specialty or medical societies, pharmaceutical and device manufacturers), although necessary and even desirable, in some instances risk creating conflicts of interest.

Financial conflicts are of greatest concern to the public and Congress. They are also easier to regulate and less subjective than nonfinancial conflicts. This committee does not, however, consider financial conflicts of interest to be the only, or even necessarily the most important, concern related to sources of bias in PORT research. Observers of scientific research note that conflicts of interest and conflicts

of commitment¹ occur ubiquitously and unavoidably in professional life where the ''coin of the realm'' is some admixture of influence and power as manifested in personal prestige and career advancement, commitment to teaching, and furtherance of the activities of a research program, a university department, or the university itself. These professional conflicts of interest need not be conscious decisions on the part of the investigator; nevertheless, intellectual attachment or commitment to a particular scientific theory may pose some possible conflicts. All such interests may affect professional judgment and actions and thus constitute broader concerns than simply financial conflicts of interest.

Several definitions of conflict of interest reviewed by the committee placed particular weight on the financial aspects of such conflicts. ² The committee took particular note of definitions of conflicts of interest that also addressed nonfinancial conflicts. The American Medical Association's Councils on Scientific Affairs and on Ethical and Judicial Affairs (1990) derived their definition of conflict of interest from Webster's Third New International Dictionary: a "conflict between private interests and official responsibilities of a person in a position of trust." This definition stresses private versus public interests where there is fiduciary responsibility. Similarly, the Association of Academic Health Centers (AAHC, 1990:7) offers the following: "A potential or actual conflict of interest exists when legal obligations or widely recognized professional norms are likely to be compromised by a person's other interests, particularly if those interests are not disclosed." This definition expands upon fiduciary responsibility by including legal and professional responsibilities, and it further implies that nondisclosure exacerbates the effect of a conflict of interest.

Concern about the professional judgment of researchers is evidenced in the guidelines published by the Association of American Medical Colleges (AAMC, 1990). They state that the term conflict of interest refers to situations in which "financial or other personal considerations may compromise, or have the appearance of compromising, an investigator's professional judgment in conducting or reporting research" (AAMC, 1990:6). The AAMC document acknowledges that the appearance of conflicting interests may be sufficient to discredit scientific results and undermine public trust.

The following sections of this chapter describe (1) financial conflicts of interest, and (2) academic and professional conflicts of interest.

Financial ties that may give rise to conflicts of interest take two typical forms for clinical researchers. These are (1) equity or other financial relationships with a company that owns a technology that the researcher is evaluating, and (2) industry support for university research.

Financial interests can take several forms:

• equity interests, including stock and stock options, in a company that markets drugs or devices that are held directly by the investigator or indirectly through various financial systems or holdings by relatives;

• other profit-sharing arrangements;

• management or executive positions with such companies;

• royalties from licensing of intellectual property rights, including patents on inventions, copyrights (e.g., computer software), and sale of other proprietary materials; and

• consultant relationships and lectures for which honoraria are received.

When medical researchers acquire financial interests in the new drugs or clinical devices they are studying, or when they receive support for their laboratories from the manufacturer of those products, they face possible conflicts of interest. The degree of potential conflict of interest varies with the arrangement. Consultant relationships are common among university scientists and are thought to pose little risk to scientists'

objectivity, although some institutions routinely review such relationships to ensure that they do not create conflicts of commitment.

The formation of companies in which investigators retain equity interest, however, may pose substantial risk (Louis et al., 1989). Even more problematic, in the view of some, is investigators who purchase equity interests in existing companies that do business in their field of practice. For example, one participant in the IOM workshop stated forcefully, "I see no socially redeeming value in equity interest. This is just something any private person does who wants to play the stock market. It would seem to me that anyone who is doing medical research ought to play the stock market in some field other than health care." Other observers, however, fear that such strictures would discourage capable clinicians from participating in investigation and innovation. They believe that most conflicts arising from equity interests can be mitigated by disclosure.

University-industry ties are formed when industry enters into an agreement with an academic institution to support the work of an investigator. There are several types of arrangements for research and technology evaluation. Most arrangements that involve clinical trials of experimental diagnostics and therapeutics have one of the forms described below (MacCordy, 1988; American Medical Association Councils on Scientific Affairs and on Ethical and Judicial Affairs, 1990):

• Single projects in which a sole principal investigator is supported by one company. The research protocol may be devised by a drug company or other firm, designed by the investigator, or developed jointly;

• Program consortia involving multiple projects in a broad field of interest, many participating researchers, and both industrial sponsors and state or federal agency sponsors;

• Programs involving multiple projects and investigators but sponsored by a single company and without involvement by state or federal agencies; and

• Technology transfer cooperative agreements between industry and a university to promote rapid commercialization of new inventions such as those in biotechnology.

These last are of special interest to observers of PORTs because as PORTs develop special software, assessment instruments, and methods of dissemination of results, they may also engage in spinoff ventures.

To concentrate on the extent to which people are motivated by money and to attribute any and all bias to personal greed ignores other powerful, noneconomic sources of bias that may be of special salience for professional groups. For researchers, desire for public recognition, publication, grant renewal, career advancement, or tenure may exert strong inducements to produce positive results. These outcomes may eventually lead to financial remuneration and increased financial security at some point, but this potential connection does not fully explain the intellectual satisfaction and similar psychological rewards not directly tied to academic advancement that can motivate researchers. Thus, professional conflicts of interest need not be conscious decisions on the part of the investigator. Subtle and unrecognized intellectual attachment or commitment to a particular theory can shape the design and interpretation of an investigation.

Although not unique to PORTs, the complexities of academic and professional conflicts of interest posed by the PORT approach to health services research raise special problems. PORTs' evaluations of clinical conditions depend on data from, and acceptance of their results by, professional groups and communities whose livelihood or professional behavior may be directly challenged by the research results. Wennberg has called attention to the effect of PORT research findings on the careers of both PORT investigators and those whose practices are examined (Wennberg, 1990b). If PORT members try either directly or indirectly to protect the subjects of their work, their colleagues, or themselves, such strong conflicting motives could cause bias in study design, data collection, data analysis, and dissemination of results. In addition to any bias in the collection and use of data, loss or distortion might occur when the results of PORT research are communicated to, and interpreted by, practitioners and patients. Dissemination of results by third-party payers or their consultants may add another source of bias (Wennberg, 1990b). Publication bias occurs when there is a greater likelihood of submitting and having accepted for publication positive results. Still other sources of bias stem from a desire to downplay the uncertainty that abounds in medical practice and which can result in the presentation of findings with greater implied certainty than is warranted (Thomas, 1983; Eddy, 1984; Wennberg, 1984).

Professional conflicts of interest arising from nonfinancial motivations are more difficult to identify, much less evaluate, than financial conflicts, and hence less likely to be amenable to direct university or federal regulations. The final section of this chapter reviews ways in which conflicts of interest have been addressed by academic institutions,

regulatory agencies and federal officials, and others with fiduciary responsibilities.

Until the 1980s conflicts of interest were usually managed by voluntary disclosure occurring within academic institutions, typically to the chairman of the department or other academic unit and, if appropriate, to a dean or other supervisor. Some state laws, however, prohibited public university employees from receiving compensation or engaging in other employment related to the subject of their research. During the late 1980s attention focused on several egregious cases involving clinical trials in which investigators appeared to have substantial conflicts of interest (Booth, 1988; Marshall, 1990).

In 1988 and 1989 Representative Ted Weiss (Democrat, New York) held hearings before the Human Resources and Intergovernmental Relations Subcommittee of the Committee on Government Operations. Although the 1988 hearings focused on fraud and scientific misconduct and particularly on delays and mishandling of such investigations (U.S. Congress, House Committee on Government Operations, 1989a), the 1989 hearings ("Is Science for Sale? Conflicts of Interest vs. the Public Interest," see U.S. Congress, House Committee on Government Operations, 1989b) focused on conflicts of interest and, in particular, conflicts deriving from industry-academic relationships in the field of biomedical technology. Other recent events related to conflict-of-interest issues in medical research are summarized by McNeil and Roberts (1990). Table 3.1 is a synopsis of the major events between 1960 and 1990.

Among the many financial arrangements described earlier in this chapter, those concerning technology transfer have aroused particular fears about compromised objectivity, enhanced secrecy, and possible subversion of the mission of the university. These concerns center on equity holdings and spin-off ventures as components of biotechnology agreements with universities (Frankel, 1988).

The genesis of such agreements can be traced to legislative inducements in the early 1980s. In 1980, recognizing the increasing competition from foreign industrial nations and their government-supported incentives for technology transfer, Congress amended the patent laws to establish a presumption of ownership by universities of patentable

inventions produced by them under government-sponsored research. In addition, revisions of the tax code in the early 1980s created research and development limited partnerships; these offered tax shelters and high-investment income and may have attracted private capital to university research (Krimsky, 1988). At the same time, the potential rate of growth of biomedical research outpaced the federal money available for research stimulating researchers and their universities to seek funding elsewhere (MacCordy, 1988; Culliton, 1989; DHHS, 1989b; Harvard University Faculty of Medicine, 1990).

University-industry collaboration grew rapidly during the 1980s. The propensity toward commercialization of faculty inventions through more numerous university-industry ties became evident throughout biomedical research. This increasing number of collaborative arrangements between industry and academia raised questions in public and professional circles about their effect on the scientific enterprise.

PORTs examine the effectiveness of alternative treatments for clinical conditions. Because of their collected expertise, and the cost and time benefits of using existing patient cohorts, data, and investigative methodologies, drug and device manufacturers may offer financial support to PORTs that are willing to become involved in technology evaluations of the manufacturers' products. Such relationships may encounter the full spectrum of risks and benefits described above. In trying to anticipate them it is reasonable to review what is known about these risks and benefits.

To determine the effects of university-industry collaboration Blumenthal and his colleagues (Blumenthal et al., 1986a, 1986b; Gluck et al., 1987; Louis et al., 1989) conducted surveys in 1985 of faculty, postdoctoral students, and fellows whose work involved the new biotechnologies. According to the authors, faculty with industry support published more, had more patents, earned more, and served in more administrative roles than those without such support. Both groups devoted the same amount of time to teaching.

Compared to faculty with federal funding, faculty with industry support reported less "red tape" in conducting their research, increased rates of commercial application of their basic research, increased availability of resources, and enhanced career opportunities for students. These researchers reported, however, that their research involved more trade secrets (information kept secret to protect its proprietary value), although this only rarely affected their ability to publish; that they shared less

information with colleagues when it contained proprietary information; and that they were more likely than their colleagues without industry support to consider commercial applications in choosing research projects. Students and fellows with industry support occasionally reported fewer or delayed publications, some inhibition of scientific communication, and some restrictions on research. A few students reported that their advisors held equity in the companies for which the students were doing research.

In the last decade university-industry relationships have increasingly included the biomedical sciences; yet, as noted earlier, very few comparative assessments have been published on the risks and benefits of the variety of relationships adopted by universities and academic centers (on the one hand) and industries and for-profit and nonprofit external organizations (on the other). Policymakers do not have rigorous data about risks and benefits or data assessing the result of policies meant to affect these risks and benefits. There are many impressions, anecdotes, and a few well-publicized cases, and these tend to drive policy formulation (Hanna, 1989). For this reason, there is no single, proven, applicable model (much less an egregious case) to guide AHCPR in the development of industry-PORT relationships. Rather, investigators and AHCPR must proceed using their good judgment, knowledge of the attendant risks, and experience with other efforts to deal with conflicts of interest. The efforts described in this section may provide useful models for PORTs.

The federal government enacted ethical standards for all full-time and special employees through the Ethics in Government Act of 1978 and an explanatory memorandum in 1982. The act requires workers in executive, legislative, and judicial branches to make information about their financial interests publicly available through the Office of Government Ethics. All federal employees at grade GS-16 or higher must file an annual Executive Personnel Financial Disclosure Report (SF 278). In addition, DHHS issues standards of conduct for its employees that prohibit behavior regarded as a conflict of interest (see current standards in DHHS, 1989a). The Food and Drug Administration (FDA) has issued its own, even more stringent, standards for control employees—those who make regulatory decisions—in the FDA supplement to the HEW [Health

Education and Welfare] Standards of Conduct (45 CFR, Part 73a, February 24, 1978).

The FDA uses a specific disclosure form (FD 2637) and guide (FDA Staff Manual Guide 3118.2) for members of federal advisory committees to the Center for Drug Evaluation and Research. The form requests information on financial interests, current and anticipated contracts and grants, and professional employment and activities. The last include compensated or uncompensated consulting, lecturing, writing, teaching, and committee membership. The form also asks for service on behalf of corporations, state and local governments, societies, and so forth.

The United States Pharmacopeial Convention is an independent, nongovernmental, nonprofit organization. Its Committee of Revision, which is composed of volunteer members, establishes and revises drug standards published in the United States Pharmacopeia, the National Formulary, and drug information monographs. Its rules include both disclosure and conduct. Prospective members of the Committee of Revision must state their industrial or commercial affiliations, sources of research support, companies in which the member or his or her family have financial interests greater than $10,000, any other interests that might be affected by revisions in drug standards, and any other relevant professional or financial interests or responsibilities. The rules of the committee stipulate that members shall not vote on the approval of any item or have sole responsibility for work on a monograph about any item for which they have a conflict of interest or the appearance of a conflict of interest (U.S. Pharmacopeial Convention, 1990–1995).

Professional associations may also provide guidelines on conflict of interest in codes of ethics for their members. For example, the American Medical Association (AMA) has recently addressed conflicts of interest that may occur when practicing physicians are induced to prescribe a product in return for personal benefits. In December 1990, the AMA Council on Ethical and Judicial Affairs (American Medical Association,

Council on Ethical and Judicial Affairs, 1991) advised that the code of ethics for the medical profession would be violated by members accepting industry gifts including paid attendance at conferences and meetings. A similar and complementary report was adopted by the Pharmaceutical Manufacturers Association as a marketing code. The AMA report states that "gifts accepted by physicians individually should primarily entail a benefit to patients and should not be of substantial value…. Cash payments should not be accepted." The report also recommends that subsidies for conferences should be used to reduce the conference registration fee rather than being given directly to physicians, nor should subsidies be used to pay for travel or lodging or to compensate for physicians' time. Reasonable honoraria and travel expenses for faculty, however, are considered acceptable. ³

The National Research Council (NRC)⁴ has adopted a statement on potential sources of bias and conflicts of interest for use in its studies (NRC, 1989). These studies are conducted by its volunteer committees that are appointed for their special expertise in the area of study. The NRC organizations make extensive efforts to assure that its reports are, and are perceived to be, free of any significant conflict of interest and that its reports are not compromised by bias in committee appointment process or by circumstances that may occur during the course of the committees' work. The NRC statement describes several situations that might occur during the course of a study and that might give rise to concern about conflict of interest. These include: any actions taken on the basis of reports that might later result in economic benefit or loss to particular individuals or groups; access to proprietary information; potential bias

arising from public statements and positions on a given subject; the design or implementation of procurements (including work statements or request for proposals); access to sensitive government information that might confer unfair competitive advantage; and problems caused by review of one's own work during the course of a study.

Individuals participating in studies and other activities complete a "Potential Sources of Bias and Conflict of Interest" form that elicits only information that is relevant and that merits disclosure in the light of NRC policies and the task to be undertaken. In addition, committee members are asked to discuss with one another possible sources of bias or conflict of interest at their first committee meeting and annually thereafter. Information regarding conflict of interest is considered by the institution in the overall composition of the committees and in the appointment or reconsideration of appointment of individuals to committees. The NRC approach is often to appoint members representing a balance of potentially biasing backgrounds or professional or organizational perspectives.

The Public Health Service Grants Policy Statement entitled "Standards of Conduct for Employees" (Public Health Service, 1987:55) states that "recipient organizations must establish safeguards to prevent employees, consultants, or members of governing bodies from using their positions for purposes that are, or give the appearance of being, motivated by a desire for private financial gain." Each recipient institution must write policy guidelines on conflicts of interest that cover financial interests, gifts, gratuities, and favors. These rules must also indicate the conditions under which outside activities, relationships, or financial interests are proper or improper and provide for notification of a responsible, objective official within the grantee's institution, as well as notification of the grantee official, of any rule violations.

In September 1989, following a two-day conference sponsored by the National Institutes of Health (NIH) and the Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA), NIH and ADAMHA issued, as a special issue of the NIH Guide, a request for comments on a proposed rule (DHHS, 1989b). The rule would have required that grant recipients avoid any prospective conflicts of interest and prohibited "personal equity holdings or options in any company that would be affected by the outcome of the research or that produces a product or equipment being evaluated in the research project (DHHS, 1989b:4). The rule also stated that no honoraria or fees should be paid to grant recipients and that there should be no disclosure of results to a sponsoring company until

such results were publicly available. Following a storm of protest by investigators, institutional spokespersons, and industry, the proposed rule was withdrawn in December 1989. NIH and ADAMHA held a public hearing (''Conflicts of Interest in Clinical Evaluation of Commercial Products'') in late November 1990 to discuss approaches to rulemaking; it has promised to reissue revised guidelines in the Federal Register.

Early in 1990 both the Association of American Medical Colleges (AAMC, 1990) and the Association of Academic Health Centers (AAHC, 1990) published reports based on their ad hoc committees' considerations of conflicts of interest and commitment. Because PORTs are based in academic medical centers, the subject of the two publications overlaps with that of this report but is not coextensive with the issues addressed by this committee. Both the AAMC and the AAHC guidelines focus on financial conflicts of interest (though their definitions of conflict include nonfinancial concerns), and both stress the need for full disclosure of financial ties.

There are two mechanisms of disclosure. The first and more traditional is faculty initiated, although sometimes only if the faculty member is taking a management position or assuming an equity interest. The second is university initiated in which periodic reports are required of each faculty member. In some cases the university must approve consulting or sponsored research before it can be undertaken.

The guidelines recommend that institutions revise conflict-of-interest policies to enhance disclosure and argue that any prohibitions should be at the discretion of the institutions.

The AAMC guidelines call for disclosure of financial and professional interests. Such reviews are to be conducted annually and at the time any new relationships are anticipated; they should include both personal holdings and those of the researcher's immediate family. Questionable cases should be reviewed by the appropriate individual and, if necessary, referred to a standing committee that would have three roles: evaluation, adjudication and arbitration, and policy development. In its evaluative role the committee would review all information. In its adjudicative and arbitrative role it would determine whether a given situation was (1) unacceptable and thus prohibited, (2) permitted with the implementation of one or more committee recommendations to preclude

unreasonable levels of bias or inappropriate activities, or (3) permitted as disclosed. New policies or changes to existing policy would be formulated or reviewed by the standing committee, which would advise administrative officials on policy issues.

In part, the guidelines call for institutions to have appropriate mechanisms in place and, in particular, to develop and disseminate policies that clearly articulate the institution's position on "1) sponsored research, 2) acceptable types and levels of outside financial and professional interests, 3) the need to recognize and deal openly with real or apparent conflicts, and 4) the relationship of faculty and staff to outside institutions and third parties" (AAMC, 1990:9). Under the guidelines, institutions must "develop procedures for full disclosure to the institution, and to the interested public, of financial and professional interests that may influence, or may be perceived to influence, research activity or other scholarly responsibilities" (AAMC, 1990:9); implement enforcement procedures, including appropriate sanctions; assure management or resolution of conflicts of interest; respond expeditiously to questions raised; and avoid institutional conflicts of interest.

The AAHC guidelines call for the development of policies identifying activities that require prior approval and activities for which disclosure is sufficient. Thereafter, the guidelines recommend periodic disclosure of "[s]ignificant financial, personal, or professional relationships that raise a potential conflict of interest" not only to persons within the institution but also "in all speeches, writings, advertising, public communications, or collegial discussions . . ." (AAHC, 1990).

Many universities have published or begun revisions of their conflict of interest and conflict of commitment guidelines to strengthen their oversight mechanisms and provide guidance to faculty members; to date, however, these guidelines are reportedly not widely known among researchers. Most research institutions expect to review contractual arrangements involving the institution to be certain there is no conflict of interest or conflict of commitment that would affect the university. A few have gone further and instituted required review of consulting arrangements. The emerging standards point to maximum scrutiny of a researcher's financial relationships with those organizations that produce

the drugs, devices, or other interventions the investigator is studying (McNeil and Roberts, 1990).

Johns Hopkins University and the Harvard University Faculty of Medicine issued revised guidelines late in 1990. The Hopkins guidelines ban ownership by a faculty member, a faculty member's spouse, and minor children of stocks or stock options in companies supporting a researcher's work, including the work of supervised faculty, nonfaculty employees, and students (Johns Hopkins University School of Medicine, 1990). The Harvard guidelines elaborate on categories of sensitive situations that are subject to particular scrutiny, including equity interests, consulting income, executive positions tied to the investigator's research, and certain forms of sponsored research (Harvard University Faculty of Medicine, 1990). These guidelines are described further in Chapter 5, "Managing Conflicts of Interest: General Models and Approaches."

Hugh Clegg, former editor of the British Medical Journal, has stated "A medical editor has got to be a keeper of the conscience of the profession" (Lundberg, 1989:33). In keeping with this assertion, at least two journals publish explicit rules regarding disclosure of conflicts of interest in their "Instructions for Authors" (Lundberg and Flanagin, 1989; Relman, 1990). For example, the Journal of the American Medical Association (1990) requires a cover letter with submitted manuscripts that includes the following statement: "I certify that affiliations (if any) with or involvement in any organization or entity with a direct financial interest in the subject matter or materials discussed in the manuscript (e.g., employment, consultancies, stock ownership, honoraria, expert testimony) are disclosed below.'' The instructions further state that "[r]esearch or project support should be listed in an acknowledgement."

In April 1989, the New England Journal of Medicine adopted a similar disclosure policy for authors of articles. The following July, editor-in-chief Arnold Relman announced an even stricter policy for authors of review articles and editorials, who must have "no financial association with a company whose product figures prominently in the article or with a company making a competitive product" (Relman, 1990:56). In explaining the editorial decision, he wrote, "When authors have a financial as well as a scientific interest in their subjects, questions inevitably arise that cast doubt on this presumption of objectivity."

In the field of law an attorney may decide that a conflict of interest is serious enough to disqualify him or her from representing a client, despite full disclosure and informed consent. Individual law firms have guidelines for their staff attorneys and complex systems for tracking representation of the various interests of their clients to ensure that conflicts do not arise. Similarly, judges, in certain cases, recuse themselves because of a possible conflict of interest. During the IOM workshop, Arnold Relman noted that both the judicial system and clinical research bear responsibilities as public institutions:

This chapter reviews financial conflicts of interest, particularly those that affect university-industry relationships, and nonfinancial conflicts of interest. It has described the ways that academic and professional organizations and other groups with fiduciary responsibilities have attempted to manage conflicts. Chapter 4 examines how these general categories of conflicts may apply to PORT research.