NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Committee on Conflict of Interest in Medical Research, Education, and Practice; Lo B, Field MJ, editors. Conflict of Interest in Medical Research, Education, and Practice. Washington (DC): National Academies Press (US); 2009.

Cover of Conflict of Interest in Medical Research, Education, and Practice

Conflict of Interest in Medical Research, Education, and Practice.

Show details

8Institutional Conflicts of Interest

Financial relationships with industry exist at the institutional level as well as the individual level and may create conflicts of interest for academic medical centers, professional societies, and other institutions that carry out medical research, medical education, clinical care, or practice guideline development. Some of these relationships may generate significant benefits to an institution’s primary missions. For example, gifts to endow named professorships or fund the construction of research facilities support the core teaching and research missions of academic medical centers. The committee heard testimony that new kinds of institutional relationships between academia and industry—beyond relationships involving individual faculty members—could promote the translation of basic discoveries into new therapies and thereby benefit society (Benz, 2008; Moses, 2008). The question for institutions as well as individuals is whether a relationship with industry can be maintained in a way that achieves the desired benefits but avoids the risks of undue influence on decision making and the loss of public trust.

Although several cases reported by the news media have called attention to institutional conflicts of interest in medicine, institutional conflicts of interest have generally received less attention than individual conflicts of interest. Institutional conflicts of interest often involve the financial interests of both the institution and its senior officials (Box 8-1).

Box Icon

BOX 8-1

Cases and Controversies Involving Institutional Conflicts of Interest. After the 1999 death of Jesse Gelsinger during a clinical trial involving a gene transfer intervention conducted by a University of Pennsylvania research institute, various investigations (more...)

The risks to core missions posed by institutional conflicts of interest can be as serious as those created by individual conflicts. Moreover, if institutions do not prudently manage relationships with industry and are exposed to public criticism for inadequately or improperly managing conflicts, the work of many individual researchers, educators, and clinicians associated with the institution may unfairly be called into question, even though they were not involved in the conduct that was criticized.

This chapter begins by defining institutional conflicts of interest and describing what has been documented about the extent of such conflicts. The discussion then reviews responses to institutional conflicts of interest and examines some of the challenges in managing such conflicts. The chapter concludes with recommendations, including a recommendation that the National Institutes of Health (NIH) require its grantees to adopt and apply policies on institutional conflicts of interest.


Institutional conflicts of interest arise when an institution’s own financial interests or those of its senior officials pose risks of undue influence on decisions involving the institution’s primary interests. For academic institutions, such risks often involve the conduct of research within the institution that could affect the value of the institution’s patents or its equity positions or options in biotechnology, pharmaceutical, or medical device companies. Conflicts of interest may also arise when institutions seek and receive gifts or grants from companies, for example, a gift of an endowed university chair or a grant for a professional society to develop a clinical practice guideline.

In addition, institutional conflicts of interest exist when senior officials who act on behalf of the institution have personal financial interests that may be affected by their administrative decisions. For instance, a department chair or dean who has a major equity holding in a medical device company could make decisions about faculty appointments and promotions or assignment of office or laboratory space in ways that favor the interests of the company but compromise the overall research, educational, or clinical mission of the institution. Similarly, a hospital official with such a holding would be at risk of undue influence in making decisions about the use of the company’s products for patient care. In situations like these, an individual’s financial relationship also implicates the institution’s interests.

As emphasized in Chapter 2, conflicts of interest are defined in terms of the risk of undue influence and not actual bias or misconduct. Whether they are at the individual or the institutional level, conflict of interest policies seek to prevent compromised decision making rather than to try to remedy its consequences.

Institutional interests can be evaluated for the likelihood of undue influence and the seriousness of potential harms in ways analogous to those applicable to individual conflicts (see Chapter 2). Thus, assessments would consider the nature of the primary interest, the value and scope of the secondary interest, the extent of institutional accountability and discretion involving decisions about the primary interest at stake, and the seriousness of potential harms in relation to potential benefits (see also Emanuel and Steiner [1995]).


Because institutional conflicts of interest have not received as much attention as individual conflicts of interest, there is less evidence about their characteristics or impacts. The committee found little comprehensive information about the scope and nature of the ties of academic medical centers, professional societies, patient advocacy groups, and other institutions to pharmaceutical, medical device, and biotechnology companies. Such ties may involve various kinds of payments and gifts to an institution, institutional ownership interests in companies, patents, and the relationships of senior officials (for example, service on a company’s board of directors). Most reports focus on prominent and usually egregious cases of misconduct, as illustrated in Box 8-1.

Chapter 4 reviewed the results of a survey of department chairs in medical schools and large independent teaching hospitals that found that 27 percent of preclinical departments and 16 percent of clinical departments received income from intellectual property licensing (Campbell et al., 2007b). (This income may be seen as a benefit of the provisions of the Bayh-Dole Act, which allow institutions to patent discoveries resulting from federally funded research and to grant exclusive licenses for others to develop those discoveries.) The survey also found that ties to industry were common among department chairs, who served as consultants (27 percent), members of a scientific advisory board (27 percent), paid speakers (14 percent), company officers (7 percent), and company board members (11 percent). The committee did not locate institution-level data on company funding of biomedical research, but Chapter 4 reported that the majority of such research in the United States is commercially funded.

For institutions as well as individuals who provide health care, conflicts of interest also arise from provider reimbursement methods, whether these involve fee for service, prospective payment per case, pay for performance, or other arrangements. In addition, conflicts may arise from provider ownership interests, for example, hospital ownership of subsidiary specialty centers to which the hospital’s physicians refer patients. As noted in Chapter 6, however, consideration of payment methods and ownership interests in medical facilities are beyond the scope of this report.

Among universities, a Congressional Research Service report concluded that patents typically account for a small percentage of university research and development funding and that most significant income from patents has tended to come from single “blockbuster” patents (Schacht, 2008). The report did not look specifically at biomedical research institutions. The Association of University Technology Managers, which conducts an annual survey of technology transfer activities (including the licensing of patents and the launching of start-up companies), does not report information by scientific field.1

Most professional societies and disease-focused or patient advocacy groups do not make public the details of funding received from industry, but it appears that many groups depend on medical product companies for a significant share of their overall revenues and for specific activities (e.g., continuing medical education and the development of clinical practice guidelines). In connection with congressional inquiries about its relationships with pharmaceutical companies, the American Psychiatric Association (APA) reported that medical companies supplied about 28 percent of its annual income. An informal APA survey of other medical specialty societies indicated that this figure was about in the middle of the range of the income that companies provide these groups (from 2 to nearly 50 percent) (Stotland, 2008). An Associated Press story on pharmaceutical company spending to promote the awareness of fibromyalgia reported that companies contributed funds that amounted to 40 percent of the annual budget of the National Fibromyalgia Association (Perrone, 2009). Many groups list corporate donors but do not report how much of their income is derived from these donors. Groups that report sources of funding for activities such as clinical practice guideline development usually do not report the amount of company funding for an activity or what percentage of an activity’s cost was accounted for by company funds. These data would assist with assessments of the risk of undue influence.

In a 2006 report for its board of directors, the American Academy of Family Physicians (AAFP) analyzed its resources and activities and concluded that it was not financially possible to forgo industry funding for any of its activities without imposing unacceptable cuts in services to members or increases in member costs. For its fiscal year 2006–2007 budget, AAFP projected that less than 38 percent of its income ($31 million of a total budget of $80 million) would come from dues and sales of products and services to members. Approximately 42 percent ($34 million) would come from the pharmaceutical industry, of which about 60 percent would come from advertising in the academy’s journal and 13 percent would come from payments for exhibits at meetings (AAFP, 2006a). The report noted that the organization had sought to broaden its base of nondues funding beyond pharmaceutical companies by seeking grants from government and foundations for various activities and had also taken other steps to limit the influence of industry. If it stopped accepting all funding from industry, however, including journal advertising, the organization would have had to increase member dues by about $600 (to about $1,000 per year) to maintain the levels of service and the programs (e.g., existing educational activities at the same per program cost to members) that existed at that time (AAFP, 2006a).

The data presented in Chapter 6 showed that physician membership organizations obtained 49 percent of their income for accredited continuing medical education from a combination of commercial funding for activities, advertising, and exhibits at meetings. Medical school continuing medical education programs received about 62 percent of their income from these sources; for publishing and education companies, the figure was 73 percent.


Federal regulations and laws have not consistently targeted institutional conflicts of interest. The U.S. Public Health Service (PHS) regulations on conflict of interest, which were issued in 1995 and which are included in Appendix B, cover only individual conflicts of interest and relationships with industry. Institutional conflicts of interest were deliberately not addressed (NIH, 1995). The guidance on financial relationships in research with human participants published by the U.S. Department of Health and Human Services discusses the identification and management of institutional as well as individual financial interests (HHS, 2004). The document suggests questions and procedures for institutional review boards (IRBs), investigators, and institutions to consider in evaluating institutional relationships. Federal antikickback rules apply to illegal payments to institutions as well as individuals. The recommendation by the Medicare Policy Advisory Commission (see Chapter 3) for industry reporting of consulting and other payments covers not only payments to physicians but also payments to medical schools, professional societies, and providers of continuing medical education (MedPAC, 2009). A bill introduced in the U.S. Congress in 2007 (S. 2029) and reintroduced in 2009 (Grassley, 2009) covers payments to individual physicians.

Several academic organizations have issued reports on institutional conflicts of interest, including the Association of American Medical Colleges (AAMC, 2002; AAMC-AAU, 2008), the Association of American Universities (AAMC-AAU, 2008), and the Council on Government Relations (COGR, 2003). The 2002 AAMC and 2008 AAMC-AAU reports dealt with institutional conflicts of interest in research with human participants.

The 2008 AAMC-AAU report was in part a response to evidence that academic medical centers had not implemented the recommendations set forth in the 2002 AAMC report. In an AAMC survey of its members, only 38 percent of the institutions that responded reported that they had a conflict of interest policy that applied to the institution’s financial interest, although another 37 percent reported that they were developing such a policy (Ehringhaus et al., 2008). For institutions that had policies, the documents typically covered equity in nonpublicly held companies (90 percent) or publicly held companies (77 percent), royalties (80 percent), payments for reaching designated milestones in the course of a study (73 percent), and substantial gifts from a research sponsor (73 percent). The majority of institutions that had policies applied them to senior officials (71 percent), governing board members (66 percent), and members of the IRB (81 percent). In addition, the majority of respondents reported creating organizational arrangements to separate institutional responsibility for research from responsibility for investment management (94 percent) or technology transfer (61 percent). Although the most serious problem identified in the survey was the lack of policies at a majority of institutions, another concern was the incomplete coverage by policies of significant institutional interests.

In addition to reiterating the importance of such policies, the 2008 AAMC-AAU report set forth several guiding principles for institutional conflict of interest policies. They were

  • “research and financial decision-making processes and agents must be separated”;
  • “decisions about whether or not to pursue a particular human subjects research project in the presence of an institutional conflict of interest should be governed by a ‘rebuttable presumption’ against doing the research at or under the auspices of the conflicted institution” unless a compelling case can be made to justify an exception; and
  • institutional conflict of interests “will be addressed consistently throughout the institution, such that those subject to institutional financial conflict of interest policies, specifically officials of the institution and the institutions themselves, are subject to substantive reporting, disclosure, and management of their financial interests.” (pp. 14–16)

The report also recommended the creation of a standing institutional conflict of interest committee and discussed procedures for the reporting of institutional financial interests and the managing of relationships that were determined to be conflicts of interest. Strategies could involve divesting the institution of an equity interest in a company, requiring senior officials to remove themselves from involvement with making decisions that might affect their conflicting interest, declining to perform research in which the institution has a financial stake (beyond the funding of the research itself), asking the IRB at another institution to review such research, or disclosing the institutional conflict of interest to research participants.

One university’s policy lists several issues to be considered in evaluations of the circumstances that might justify institutional involvement in a human subjects research project despite a conflict of interest (University of Rochester, 2006). The case for the institution’s participation in the project is stronger to the extent that

  • the work is carried out at multiple sites (e.g., under the auspices of several institutions);
  • the institution takes a relatively passive role in the conduct of the project (e.g., the gathering of data);
  • the number of research subjects under the institution’s supervision is small;
  • an adverse effect on research subjects appears more likely if the institution is not used as a research site; and
  • the investigators conducting the research or the university resources supporting the project are essential and are not readily available elsewhere.

In a position statement on organizational aspects of physician relationships with industry, the American College of Physicians (ACP) advised that “[m]edical professional societies that accept industry support or other external funding should be aware of potential bias and conflicts of interest” (Coyle et al., 2002b, p. 405). It recommended the adoption of explicit institutional policies on industry relationships, including policies that “avoid reliance on outside sources of support” and that guide the acceptance and disclosure of funding from industry and other outside sources. The ACP position on educational programs is that “it is unethical for academic institutions and educational organizations to accept any support that is explicitly or implicitly conditioned on industry’s opportunity to influence the selection of instructors, speakers, invitees, topics, or content and materials of educational sessions” (Coyle et al., 2002b, p. 405).

In a 2006 statement, the Society for General Internal Medicine (SGIM) reported limits on the share of its annual operating budget that could come from external sources (SGIM, 2006). The limit on external sources of funding was 33 percent overall, with limits of 10 percent from health care-related for-profit entities in combination and 5 percent for any single such entity. (Thus, 67 percent of the operating budget must come from internal sources, such as member dues and fees.) Furthermore, the statement declared that the organization should not accept funds from

for-profit companies (or not-for-profit entities funded largely by for-profit companies) for research or educational projects (including individual precourses, workshops or other presentations at the SGIM national or regional meetings) related to specific diseases, or to pharmaceuticals, medical devices, diagnostics, or other products or services purported to have direct health benefits to patients (regardless of whether the products are sold by that particular external funder). (p. 2)

The statement described such funds as “problematic” because their intent would seem to be “primarily promotional; that is, to directly or indirectly (through greater recognition of the disease in the population) encourage wider use of medical products, to the benefit of the sponsor” (p. 2). The statement stated that general meeting support may be solicited after program planners have determined the content of the meeting.

Chapter 6 discussed the actions that the Accreditation Council for Continuing Medical Education initiated to limit industry influence associated with providers’ solicitation and acceptance of industry funding. Chapter 7 described the steps taken by some professional societies to insulate activities such as clinical practice guideline development from influence associated with industry funding. It also noted that some societies do not accept industry funding for guideline development.


Although the committee found no systematic research on institutional conflicts of interest or the effects of institutional policies, it identified several challenges in managing such conflicts. One challenge is that identifying relevant institutional financial interests and conflicts may be difficult. Particularly in universities or other large institutions, no single individual or office may have knowledge of all such interests. Those responsible for identifying relationships may have to survey various parts of the institution to develop an inventory of relevant interests and relationship. In an academic medical center, for example, this inventory could cover the office responsible for technology transfer and intellectual property, the office or body that manages investments, the offices responsible for purchasing medical equipment, academic departments and other units that may receive gifts, and perhaps other offices or units as well. For senior officials, the usual process for disclosing individual financial interests will apply, although the review of disclosures will be at a higher level, for example, through a committee of the governing board, as recommended below.

Dealing with institutional conflicts of interest may be more difficult in some respects than dealing with individual conflicts of interest. In the case of individual conflicts in large institutions such as universities, medical schools, and major teaching hospitals, opportunities for review usually exist at multiple levels of the institution and involve authorities who are relatively independent and do not stand to gain personally from the secondary interests in question. In contrast, an independent review for institutional conflicts of interest may be difficult because the institutional officers themselves may stand to benefit indirectly from the conflict of interest and may be reluctant to question current or proposed relationships with companies that seem likely to improve the institution’s financial welfare. For example, the reputation and tenure of chief executives and other high-level officials may depend on their success in strengthening the financial health of their institution. If senior officials who oversee technology transfer, intellectual property, and research grants are also charged with managing institutional conflicts of interest, they may find it difficult to resist pursuing a grant or may be reluctant to divest the institution of a property interest even if such actions are necessary to manage the conflict. The leaders of professional societies and patient advocacy groups that depend significantly on member dues or individual contributions may be reluctant to reject grants from industry, even though they create a risk of undue influence over activities such as the development of clinical practice guidelines or educational programs.

The potential for conflicts of interest among senior institutional officials is one reason for the committee’s recommendation below that the key responsibility for oversight of institutional conflicts of interest be lodged with an institution’s governing body. It is also a reason for the recommendation that independent members—individuals not affiliated with the institution—be included on board committees that review and manage institutional conflicts of interest.

Because the potential financial gain from a secondary institution-level interest may not be personal for institutional officials, their decisions may be more easily rationalized as serving the institution rather than themselves—even when officials also stand to gain in personal reputation. In fact, the gains often do serve the institution’s primary mission, for example, when returns on investments or licenses are distributed to worthy research, educational, or patient care activities. Nonetheless, it is precisely because this argument for benefit is so plausible (and often valid) that serious institution-level conflicts of interest may be ignored or may not be reviewed carefully to assess whether they might, on balance, undermine rather than promote the primary missions of the institution.

For similar reasons, the public may—at least initially—be more tolerant of institutional conflicts of interest than individual conflicts of interest and may expect that institutions will pursue relationships to advance research, expand educational activities, or increase clinical resources. This tolerance may, in turn, reinforce the inclination of institutional leaders to downplay or ignore the resulting conflicts of interest. Because it is clear that universities and other health care institutions require resources to fulfill their missions and because society has encouraged institutions to pursue such resources, “[s]ociety may not view this as self-interested behavior and consequently may erroneously be more tolerant of circumstances in which an institution’s financial interests may compromise the integrity of its missions than of similar situations involving individual conflict of interest” (Emanuel and Steiner, 1995, p. 263).


Because no decision maker in an institution is fully free of conflict in the case of institutional conflicts of interest, it is not possible to establish a fully independent process for assessing such conflicts. Although no perfect solution exists, the committee concluded that, on balance, the most suitable authority for making judgments about institutional conflicts is the board of trustees or an equivalent governing body.

In their fiduciary role, members of the board are responsible for giving priority to the longer-term interests of the institution. Because they stand at a greater distance from the daily pressures of decision making than an institution’s senior officials, they should be able to assess more judiciously the positive or negative effects of financial interests on the institution’s core mission. Board members also have access to comprehensive information about the finances of the institution, some of which may be confidential and not revealed to senior institutional officials. They may also be better positioned to help an institution’s chief executive resolve disputes about conflicts of interest that involve different units within the institution. For example, in a university, faculty in the school of public health may be more concerned than faculty in the school of business about the potential for investments in certain products to create a risk to the missions of the whole institution.

In addition, the decisions made by a governing board are more salient within and beyond the institution than decisions made by staff. When the board takes up an issue, the concerned public is more likely to take notice.

RECOMMENDATION 8.1 The boards of trustees or the equivalent governing bodies of institutions engaged in medical research, medical education, patient care, or practice guideline development should establish their own standing committees on institutional conflicts of interest. These standing committees should

  • have no members who themselves have conflicts of interest relevant to the activities of the institution;
  • include at least one member who is not a member of the board or an employee or officer of the institution and who has some relevant expertise;
  • create, as needed, administrative arrangements for the day-to-day oversight and management of institutional conflicts of interest, including those involving senior officials; and
  • submit an annual report to the full board, which should be made public but in which the necessary modifications have been made to protect confidential information.

The standing board committee (or subcommittee) would regularly review the financial relationships of the institution itself to identify conflicts of interest with its primary mission or missions and would likewise review the financial relationships of senior officials. The board committee would also evaluate the adequacy of the policies and procedures established to deal with these relationships. This board committee would be different from the committee established to address individual conflicts of interests, as suggested in Recommendation 3.1.

Although the board should be accountable for institutional conflicts of interest, the committee recognizes that board members may not be well suited to carry out day-to-day oversight or conduct special investigations, especially in academic medical centers and other large institutions. The board may therefore decide to establish a mechanism for the day-to-day oversight of institutional conflicts of interest. This mechanism could take different forms at different institutions. For example, as AAMC and AAU have recommended, an academic institution might establish a faculty-staff committee that would oversee institutional conflicts of interest and that would be separate from any committee responsible for individual conflicts of interest. Such a committee (and any other support staff) could report to the board committee or to an officer of the institution who is not directly responsible for institutional investments, technology transfer, or research. Various options are reasonable; and the choices made may depend in part on the size, organization, and scope of an institution. In any case, the option selected should be consistent with the objectives of establishing and supporting governing board oversight of institutional conflicts of interest.

The recommended annual report from the board committee will provide an incentive for that committee to report on both what it has decided with respect to newly identified conflicts of interest and how its previous decisions (e.g., plans for eliminating or managing an institutional conflict of interest) have been implemented. Such reporting will also provide an incentive for rigorous review and accountability. The board committee is more likely to be diligent in its reviews if its members know that if they miss potential problems, their failure may be publicized for all to see, should these problems become the subject of official investigations or media reports. In certain cases, a tension may exist between the countervailing goals of public disclosure and keeping confidential certain personnel information and certain facts about current or pending intellectual property. Thus, the board’s public reports may exclude some details because the information is confidential, but such exclusions should be rare.

To speed the adoption of institutional conflict of interest policies, NIH should extend the 1995 PHS regulations on conflict of interest to cover institutional as well as individual financial interests for institutions that receive PHS research grants. Such rules would also call attention to the issue and encourage institutions that do not receive research funds but that are engaged in medical education, clinical care, or the development of practice guidelines to voluntarily take action to avoid and oversee potential conflicts of interest. Ideally, the development of new PHS rules would be harmonized with corresponding revisions in the regulations of the National Science Foundation.

RECOMMENDATION 8.2 The National Institutes of Health should develop rules governing institutional conflicts of interest for research institutions covered by current U.S. Public Health Service regulations. The rules should require the reporting of identified institutional conflicts of interest and the steps that have been taken to eliminate or manage such conflicts.

Although the new PHS rules should be consistent with the recommendation in Recommendation 8.1 and other recommendations in this report, they need not be highly prescriptive or rigid, particularly given that experience with institutional conflict of interest policies appears to be more limited and is less well documented than policies governing individual conflicts. Provisions for monitoring and enforcement are, however, important both at the level of the NIH extramural program and within research institutions. Consistent with current PHS rules on individual conflicts of interest, Recommendation 8.2 calls for grantee reporting to NIH of identified institutional conflicts of interest.

NIH can encourage the appropriate and reasonably consistent implementation of the regulations by providing supplementary explanations and guidance, as it has recently done for its policies and regulations on individual conflicts of interest (see Chapter 3). It can also bring grantee representatives together to discuss their experiences and identify good practices in policy development and implementation. In addition, NIH can develop or commission case studies on common situations that raise concerns over conflicts of interest, such as institutional stakes in start-up companies that seek to sponsor research at the institution.

Although the 2008 AAMC-AAU report did not explicitly recommend governing board responsibility for policies on institutional conflicts of interest, their report can still provide useful guidance to NIH and to grantee institutions and a model for developing case studies to provide education on the evaluation of conflicts of interest. Because experience with and evaluations of institutional conflict of interest policies are limited, the investigation of such policies should be one focus of the research agenda recommended in Chapter 9. In addition, continued attention to this area—for example, further surveys of policy adoption—by AAMC would also be constructive.

The intent of the recommendations in this report is to promote a culture in which conflicts of interest are taken seriously by institutions and individuals engaged in medical research, education, and practice and practice guideline development. For this to happen, institutions must effectively manage their own conflicts and be seen to be doing so. The board and the senior officials set the tone for the institution. They should be accountable for making sure that their own institutional interests are in order.



On the basis of its 2006 survey, the Association of University Technology Managers reported 12,672 actively managed licenses from patents as well as the introduction of 697 new products and 553 start-up companies (AUTM, 2007). It did not report the extent to which the institutions had financial stakes in the new products and companies. The survey covered 190 institutions, including 161 universities and 28 teaching hospitals and research institutions.

Copyright © 2009, National Academy of Sciences.
Bookshelf ID: NBK22934


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (2.2M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...