Food and Drug Administration Advisory Committees (1992)

This chapter deals with two important issues: How should the FDA protect the deliberations of its advisory committees against potential financial conflicts of interest on the part of individual committee members? And what steps should the agency take to guard against possible intellectual bias of committee members. The committee dealt with the first of these issues extensively, giving it more attention than any other single subject. It was specifically asked to do so by the FDA Commissioner Kessler, because this perplexing issue was affecting the agency's ability to use advisory committees in the evaluation of new products.

We use ''intellectual bias'' to refer to a different concern, namely, the possibility that a committee member may be so convinced about the right answer to a question of science or medicine—or so clearly identified with a particular view—that he or she may not be (or appear to be) able to approach a matter before the committee with an open mind. This concern is sometimes discussed under the label "conflict of interest," but the committee has treated it separately precisely because the federal law contains an elaborate set of restrictions addressed solely to the matter of financial conflict.

Concern about intellectual bias, which is addressed in the final section of this chapter, proves to be equally perplexing and may assume comparable importance at the FDA. But it came into clear focus only near the end of our study, and thus has received less thorough discussion and assessment. The topic is a candidate for further attention in connection with the work of the FDA advisory committees, just as it is now receiving extensive scrutiny and debate among policymakers and academic scientists who confront it in other contexts.

Potential financial conflict of interest and intellectual bias are obviously critical matters for the Food and Drug Administration and for the public. For the FDA's advisory committees to serve their purposes, their judgments must be—and must be seen to be—the product of the members' independent

assessment of the scientific evidence presented to them. At the same time, the criteria by which candidates for committee appointment are screened and the participation of appointed members is regulated must be both realistic and fair. These criteria must protect the agency's processes from real risks of inappropriate influence and yet not disqualify or embarrass all scientists and clinicians who have had any connection to the drug, biologics, or device industries.

The chapter deals first with financial conflict of interest, indicating the origins of the IOM committee's concern for this issue, reviewing the statutory framework that governs the area, examining the system by which the FDA administers the conflict-of-interest laws, analyzing the rapid changes in that system, including a number of controversial cases and some encouraging prospects for improvement, and concluding with a number of recommendations. The chapter addresses the issue of intellectual bias in a concluding section.

The reality facing the FDA is that over the past decade, perhaps longer, sponsors of drugs, biologics, and devices have turned increasingly to academic researchers to help develop and test new products. This pattern is particularly obvious in the biotechnology industry. Consequently, many of the same experts whose advice the FDA wishes to obtain have affiliations with regulated firms, some with many such firms. The recognized expertise of such individuals makes them attractive to both government and industry.

In addition, the agency has sought advisory committee advice on a growing range of scientific and regulatory issues, and it is under pressure to increase the agenda items considered by its committees. One result of these coincident developments has been to generate potential financial conflicts of interest for one or more committee members in connection with every committee meeting.

The tensions that result from this set of relationships cannot be eliminated but must not be ignored. The goals of any system for mediating these tensions must be to protect the integrity of the FDA's decisions and at the same time to allow the agency access to essential expertise. The IOM Committee is concerned that the current system for managing potential financial conflicts of interest, as now administered, may be jeopardizing the latter goal without significantly advancing the first.

When Commissioner Kessler met with the IOM Committee on December 6, 1991, he emphasized his desire for guidance in "solving" the FDA's current problems with conflict of interest and its advisory committees.

The Commissioner stressed that, "If [the committee] does nothing else but solve our conflict of interest problem, then we will have been well served."

However, neither the Commissioner then nor the FDA senior staff later provided the Committee with a detailed picture of this "conflict-of-interest problem." The full dimensions of the problem were not then appreciated, we believe, because they were undergoing significant change even as the study began. It took the better part of this study for the Committee and its staff to gain an understanding, which may still be incomplete, of the "problem."

The IOM committee entertained the hypothesis that numerous members of the FDA advisory committees were participating in decisions in which they had significant personal financial interests—with or without permission to do so. We did not find evidence that this was the case.

The IOM committee also considered the possibility that the "problem" was basically one of perception—a widely held belief that some advisory committee members, even if not in violation of the law, were compromised by their relationships with industry. Although it is not easy to measure public perceptions on such a matter, and we have not attempted to do so, the committee believes this is not a trivial concern. The range of relationships that the current law, as interpreted, treats as presumptively disqualifying financial interests has become so broad that virtually no advisory committee member is untouched. Thus the FDA confronts the need to consider granting conflict-of-interest waivers for one or more members at almost every meeting, creating or bolstering an impression that the system is seriously compromised.

In the committee's judgment, however, the core of the problem, or at least the portion on which thoughtful recommendations might make an immediate contribution, is internal to the FDA and the department. The problem resides in the system for identifying potential financial conflicts, for the agency's determining whether to seek a waiver (which is specifically provided for in the governing law) that would permit the participation of a specific member in advisory committee deliberations, and for evaluating that request for a waiver in a particular case.

A sometimes bewildering number of organizational entities are involved in administering conflict-of-interest laws. They are identified in Table 6-1 on the following page.

A familiarity with the current federal conflict-of-interest law as it applies to the FDA advisory committee members is necessary to understand the "problem" that Dr. Kessler asked us to evaluate. The key statutory provision is 18 U.S.C. §208, which is part of the U.S. Criminal Code, and it applies to all federal government employees. Members of the FDA's technical advisory committees are covered because they are appointed as "special government employees" (SGEs) who serve the government on a part-time or intermittent basis.^* Appointment as an SGE allows an advisory committee member to be paid and compensated for expenses; it also facilitates disclosure to committee members of confidential or proprietary information, which is often the bulk of the material in a drug, biologic, or device application.

Section 208 (summarized in Table 6-2) has two main parts. Subsection (a) prohibits (i.e., makes criminal) a government employee from participating "personally and substantially" in any "particular matter" in which, to his/her knowledge, "he, his spouse, minor child, general partner, organization

in which he is serving as officer, director, trustee, general partner, or employee, or any person or organization with whom he is negotiating or has any arrangement concerning prospective employment, has a financial interest."^* The law does not distinguish among types of financial interests, nor between large and small or significant and insignificant interests. By common consensus it goes well beyond such things as monetary payments or marketable securities.

This broad reach of subsection (a) is qualified by subsection (b), which allows for three exceptions (or waivers) to this general prohibition. Subsection (b)(1) allows the official responsible for appointing an employee to grant an exception to participate in a matter in which the employee's interest "is not so substantial as to be deemed likely to affect the integrity of the services which the Government may expect." Subsection (b)(2) authorizes the promulgation of regulations that categorically except certain types of interests. As the law was amended in 1989, this authority can be exercised only by the Office of Government Ethics (OGE). Finally, subsection (b)(3), which was added to the law the same year, exclusively for advisory committee members, allows the official responsible for appointing a committee member to grant an exception if he/she concludes that the agency's need for the member's service in the particular matter outweighs any risk that this impartiality will be compromised. Each of these three waiver authorities is examined in greater detail below.

Subsection (b)(1) permits the government official responsible for appointing the employee to issue an advance written determination, a (b)(1) waiver, finding that "the employee's interest is not so substantial as to be deemed likely to affect the integrity of the services which the Government may expect." Although this language appears to require an assessment of the magnitude or character of the employee's interest, the section allows consideration of other factors, according to the OGF. These include the magnitude of the employee's other holdings (e.g., as a way of asking "how much would it really matter if his/her stock in Company X doubled in value?"), the likelihood that the interest could be materially affected by a decision made or advised on by the employee, and the type of interest involved). The presence of subsection (b)(1) in the law arguably supports the conclusion that section 208(a) covers any financial interest, no matter how small.

Under a 1990 Executive Order, an agency that contemplates granting a (b)(1) exception must first "consult" with the OGE if it is practical to do so. This does not necessarily mean that the OGE must approve the agency's decision. Any (b)(1) exception granted without consultation with, or even in defiance of advice from, the OGE will nonetheless be valid. However, under DHHS policy, any exception granted under (b)(1) does require approval by the Office of the HHS Special Counsel for Ethics. In other words, the Commissioner is not authorized to approve a waiver without OSCE approval.

Subsection (b)(2) of section 208 authorizes the issuance of regulations that categorically exempt certain classes of financial interests as being "too remote or too inconsequential to affect the integrity of the service of the" employee. Before the 1989 amendment of (b)(2), individual agencies had authority to grant such categorical waivers; if they did so, those regulations remain in effect today. After the legislation, however, the power to issue such regulations was lodged exclusively in the OGE, and agencies without these rules are now precluded from issuing them. Since the FDA had no (b)(2) regulations before the 1989 amendments, and because the OGE has not yet issued such regulations, the FDA has no basis to grant (b)(2) waivers.

The language of subsection (b)(2) allows class exemptions based on either of two criteria. The interest may be "too inconsequential," i.e., too small, which implies an absolute criterion independent of the likelihood that

an interest might be affected or, even if affected, might threaten an employee's integrity. Alternatively, the interest may be "too remote," which seems to speak to the likelihood that the value of the interest would not be affected by the advice given or by any decision based on that advice. Pursuing the latter prong, subsection (b)(2) might allow the OGE to issue a regulation exempting endowment holdings of educational institutions from the class of employer financial interests that would otherwise disqualify an advisory committee member.

The OGE says it is developing regulations to implement (b)(2). As described to the IOM committee, the regulations will speak to both the magnitude and type of interests and to the functions performed. They may initially cover only ownership interests in business enterprises, e.g., shares of stock and perhaps other equity interest such as partnerships.

Thus far the OGE has not developed, and may not even have considered, criteria for evaluating the remoteness of other types of interests, such as research grants. However, the DHHS Special Counsel for Ethics is engaged in discussions with counsel for other science-oriented agencies to explore criteria for waiving research grant "conflicts." Any criteria ultimately developed by this group will, of course, still need the OGE approval and then promulgation as regulations.

Quite obviously, adoption of any (b)(2) regulations is many months, and probably years, away. The process requires consultation, before any proposal is published in the Federal Register, between the OGE and the Office of Personnel Management, the Department of Justice, and the Office of Management and Budget. After publishing the proposal as a Notice of Proposed Rule-Making (NPRM), the OGE must allow for public comment, respond to this comment by changes in the proposal or justification of the proposed action, and proceed once again through the executive branch review process (OPM, Justice, OMB) to develop the final rule.

Subsection (b)(3), the most significant of the waiver authorities, and of most immediate concern to the FDA and the IOM committee, applies specifically and exclusively to members of advisory committees. Under this provision, the appointing official is allowed to grant a waiver for a committee member, who would otherwise be disqualified from discussing a particular matter, i.e., a specific agenda item, to participate in deliberations on that matter without violating the law. "The exercise of this authority calls for a judgment in writing "that the need for the individual's services outweighs the potential for a conflict of interest created by the financial

interest involved." This exemption was added to the law in 1989 to facilitate the participation of members of expert advisory committees.

For an the FDA advisory committee, subsection (b)(3) requires a judgment by the FDA Commissioner that the value of a committee member's participation outweighs the risk of conflict of interest posed by his or her financial interest. It provides the framework within which virtually all "waivers" for the FDA advisory committee members are processed. Subsection (b)(3) clearly calls for a context-specific judgment, one that takes into account not only the interest involved but the contribution that the member can make to the committee's deliberations on the matter before it. Thus, the law presumably allows an assessment of the member's expertise, familiarity with the issues, and uniqueness on the committee in light of the issues to be addressed. And exercise of the authority would seem to call for a personal judgment by the FDA Commissioner—or by the official to whom he delegated his authority.

The Commissioner's authority to grant waivers under (b)(3) is in addition to the authority to grant waivers under (b)(1) and, if the OGE regulations are ever promulgated, under (b)(2). Thus, if an the FDA advisory committee member qualified for a (b)(1) waiver or a categorical (b)(2) waiver, there would be no need to consider his/her eligibility to participate under (b)(3). On the other hand, this also means that a member who could not qualify for a waiver under (b)(1) or (b)(2), e.g., because his or her interest is too large or too likely to be affected, may still be eligible for a (b)(3) waiver based on his or her importance to the committee's deliberations. The FDA Commissioner's authority to grant waivers under subsections (b)(1) and (b)(3) is, under DHHS policies, subject to review by the Office of the Special Counsel for Ethics (OSCE); his authority under the Executive Order requires consultation with the OGE.

Thus, the impact of this legal regime on the FDA advisory committee members, and on the advisory committee system, will be a function of three factors: (1) the types of interests held to fall under the prohibition of section 208(a); (2) the kind and number of relationships that advisory committee members (and their family members and employers ) have with manufacturers of the FDA-regulated products; and (3) the specific issues on which the FDA seeks committee advice. Importantly, for this study, each of these factors has been undergoing change.

It is important to emphasize that 18 U.S.C. §208 is a criminal statute whose violation carries criminal penalties and whose enforcement involves criminal investigation procedures. Accordingly, it is entirely appropriate for the FDA and the DHHS to take seriously their obligations to protect both the agency's decisional process and the members of advisory committees from committing violations.

The granting of a waiver must be understood in this context. What is being waived by the agency is not an individual's conflict (or potential conflict); a waiver is an acknowledgement of a conflict. Instead, what is being waived is the criminal liability of the advisory committee member that would attach to participation with a prohibited interest. Seen in this light, the system has an obvious justification. Some committee members who have complained about the intrusiveness of the agency's questions or delays in approval of their appointments or their participation, may not have fully appreciated the importance of the exercise in protecting them as well as the agency.

Federal conflict-of-interest laws impinge on the FDA advisory committee operations at two stages, each of which has multiple steps. The first stage is when an individual scientist or clinician is being considered for initial appointment and involves screening prospective committee members for potential conflicts of interest. It is this point at which most of the information about and individual's personal, family, and employer or institutional financial interests is sought and provided. The identification of potential conflicts of interest, however, does not result in rejection of many candidates at this stage simply because the specific issues on which their advice will be sought are not generally known.

The conflict-of-interest laws do not forbid the FDA to appoint as advisory committee members individuals who have financial relationships with the FDA-regulated firms. They do forbid the participation of a committee member in a "particular matter" in which he or she has a financial interest. Thus, a judgment of whether the law applies can only be made by considering, in the case of advisory committee members, the specific agenda items on which their advice will be sought. Each committee agenda item, therefore, presents an occasion for evaluating a member's potential conflict-of-interest.

Consequently, agency and departmental administration of conflict-of-interest laws focuses on the second stage, when meeting agendas are known and the "particular matters" to come before the committee have been identified. This stage involves identifying potential conflicts and determining when to seek waivers, preparation of waivers, and processing waivers.

The recruitment, nomination, and appointment of the FDA advisory committee members has been described in Chapter 5. This section focuses on the stage at which the FDA seeks to identify a prospective member's potential financial interests and thereby equip itself to monitor compliance with section 208.

Once a nominee for committee membership has been tentatively approved at the center level, a member of the FDA staff, usually the committee executive secretary or a member of the advisory committee management staff, contacts the individual by telephone to determine his or her availability and to identify any factors that might preclude appointment or diminish the individual's effectiveness. In all three centers, this initial conversation also includes ''prescreening'' questions that solicit information about the nominee's financial interests and relationships. The purpose is to discover future potential conflicts of interest. At the time of this study, the prescreening forms being used were several years old and did not elicit information about spousal or employer financial interests.

Although the FDA has no formal threshold, an advisory committee nominee at this initial stage may be judged to have so many attachments to the drug, biologics, or device industries that the appointment should not be made. The rationale is that numerous potential conflicts will limit the individual's ability to participate in committee discussions. Occasionally a potential member is ruled out because his or her attachments to industry are simply too great to pass an "appearances" test, even though they might not require frequent disqualifications. Any decision not to pursue recruitment of a prospective committee member because of excessive potential conflicts is taken with the concurrence of the division or office director.

Following this preliminary screening of a prospective committee member, he/she is sent an "appointment package" that solicits more specific professional and financial information. Although the appointment packages sent by the three centers differ in small details, they are roughly equivalent. Of particular interest is Form 2637—the Confidential Statement of Employment and Financial Interests. Form 2637 seeks information about financial holdings or business arrangements with any firm, regardless of whether it is known to be regulated by the FDA, as well as employment by and/or consultantships with FDA-regulated firms. (Current instructions for completing the form do not clearly specify that the SGE should include the financial interests of his or her spouse, minor children, partner, and any organization in which he/she serves as officer, director, trustee, general partner or employee, and of any person or organization with whom he/she is negotiating or has any arrangement for prospective employment.) The

prospective advisory committee member returns the completed appointment package, including Form 2637 and a current curriculum vitae.

Upon receiving the appointment package from the potential member, the center staff prepares a nomination package that is sent to the Committee Management Office within the Office of the Commissioner. The Committee Management Office serves as the liaison between the centers and the Commissioner's office and between the FDA and the DHHS. At the same time, relying on Form 2637, an SGE program officer within the center prepares a list of exclusions, i.e., a list of companies, products, members, or topics the discussion of which the member may not participate in because of a potential conflict of interest under section 208 (a). These exclusions, which are listed on a (HHS) Form 410 and hence are referred to as "410 exclusions," are generated by comparing Form 2637 against a now-outdated database of FDA-regulated companies supplied by the FDA's Division of Ethics and Program Integrity (DEPI) and supplemented by further investigations by the SGE program officers. Copies of this "410 exclusions" list and the completed Form 2637 are then returned to the new committee member and are also sent to the division director responsible for the committee and to the DEPI.

The initial decision to appoint committee members is largely the responsibility of the division and or center responsible for the committee. the FDA's Office of Chief Counsel is rarely involved. The Commissioner's office usually plays only a formal role in committee appointments. Beyond signing the appointment package, the Commissioner himself is seldom involved. No DHHS conflict-of-interest review is carried out at this stage. Nor does the OGE have a role in the initial appointment process.

As a result, the FDA officials exercise significant discretion about the magnitude and type of financial interests or relationships that should be considered wholly disqualifying. the FDA Staff Manual Guide, Section 3118.2, part 6, provides general instructions for the screening of individuals for potential financial conflicts, but it offers no concrete guidance regarding acceptable types or magnitudes of financial relationships.

The usual results of the FDA's "financial interest review" at the appointment stage are, first, to confirm the nomination, and second, to create a record of the member's potentially disqualifying financial interests. The latter provides the information on which a determination will be based either to disqualify from participation or to seek a waiver when a specific meeting agenda gives rise to a potential conflict.

The process for review of conflict-of-interest compliance begins with the effort to identify advisory committee members who may have a potential conflict with respect to one or more matters scheduled for discussion at a committee meeting. To identify such individuals, the center's SGE program officer obtains from each executive secretary: (1) a tentative agenda, usually taken from the Federal Register announcement of a meeting, as well as any additional topical information supplied by the executive secretary; (2) the names of company sponsors of products scheduled for review; (3) the products to be discussed and any closely competing products; and (4) a list of other issues to be discussed.

To maintain a current record of the financial/employment relationships of committee members, before each meeting the SGE program officer sends an update form to each member. This and the original Form 2637 completed by a member are the basis on which exclusions are identified. The SGE program officer determines which, if any, committee members are presumptively excluded from a matter scheduled for discussion at the upcoming meeting. An exclusion requires disqualification from that matter unless a waiver is sought and approved.

The SGE program officer, the executive secretary, the division director, and the office director may all be involved in deciding whether to seek a waiver for a committee member who has an exclusion. This decision is ostensibly based on the need for the individual member's expertise and potential contribution to the planned discussion. However, the importance of assuring a quorum at the meeting is also often a factor in assessing the need for a member's involvement. For a committee member who has only a few exclusions, the decision to seek a waiver appears to be almost automatic.

Throughout the 1980s the system for identifying potential conflicts of advisory committee members and, in appropriate cases, processing waivers (generally based on subsection (b)(1) of the law) was internal to the FDA. It also appears to have escaped close oversight by the Commissioner's office and there was rarely any legal review of waiver decisions. The system ran smoothly, if in retrospect somewhat unprofessionally.

But later events revealed that the insulation from high-level administrative and legal oversight had resulted in neglect at both the center level—where necessary information about committee member interest was often not solicited—and at headquarters—which continued to adhere to outdated

internal guidelines. Importantly, the responsible the FDA units, for some period of time, ignored, perhaps unknowingly, the important changes Congress made in the conflict-of-interest law in 1989. Absent legal guidance, they continued to operate on procedures based on the now-outdated pre-1989 law. Specifically, they did not modify the procedures regarding the eligibility of advisory committee members for subsection 208(b) waivers.

This system began to come under mounting stress in 1991, shortly after Dr. Kessler was appointed Commissioner of Food and Drugs. Kessler came to the FDA with a commitment to restore the agency's integrity, on the heels of a scandal involving generic drug approvals. He also professed a commitment to seek the advice of the nation's best scientists and supported the FDA's long reliance on expert advisory committees.

In the latter half of 1991, however, the agency brought before several different advisory committees a series of high-profile and deeply controversial issues involving, for example, the approvability of a new drug for Alzheimer's disease, the continued marketing of silicon-gel breast implants, and the alleged suicide-inducing properties of the nation's best-selling antidepressant. These meetings brought the conflict of interest of the FDA advisory committee members, and the agency's system for controlling it, under unprecedented scrutiny.

The scrutiny occurred partly because the FDA officials themselves realized that the controversial nature of the issues required that the agency take precautions to assure committee integrity. Even so, outside parties who were disappointed by the agency's decisions often challenged the objectivity of advisory committee members. the FDA's criteria and procedures for identifying potential financial conflicts and processing waivers caught the attention not only of the Commissioner and his advisers but of other officials inside and outside the agency.

Scrutiny was also heightened because these controversial cases occurred in a very compressed period of time, from late 1991 through mid-1992. In many ways, they were unrelated to each other in substance. But the cumulative effect produced by their rapid, sequential occurrence was substantially greater than it would have been if they had been spaced over a longer period.

The events of 1991 and 1992 exposed the dimensions of the conflict-of-interest "problem" on which Dr. Kessler urgently sought the IOM Committee's guidance. Accordingly, we provide a chronological account of these advisory committee meetings.

This meeting was called to discuss an increased number of reports of adverse reactions linking use of the drug Prozac (fluoxetine), manufactured by Eli Lilly, and other antidepressants to suicidal ideation in clinically depressed persons. The objective was to consider the adverse reaction data to determine the existence of causality between these drugs and suicidal or other violent behavior. Although it was apparently not the FDA's plan to invite discussion of specific product submissions or to solicit advice on remedial actions that would impact manufacturers of Prozac or other antidepressant drugs, the agency realized that the committee could potentially recommend action that might bear on the use of these drugs.

Accordingly, because the advisory committee might recommend actions that could affect manufacturers of antidepressants, CDER's Office of Consultants and Advisors decided to request waivers for all committee members who reported any financial interest with any manufacturer of antidepressant drugs. Therefore, waivers were sought for four members and two consultants who were found to have an exclusion(s).

The exclusions for the members were:

• Member #1: The member was involved in Merck's Phase IV study of the clinical safety and efficacy of a new neuroleptic drug (remoxipride). The study was funded at $95,000 for the period from February 1991 to January 1992. The member received no personal remuneration from the study.

• Member #2: The member was the principal investigator under a grant from Sandoz to study HLA phenotypes and vulnerability to Clozapine-induced agranulocytosis. The grant covered the period from September 1988 to January 1999. The hospital that employed the committee member was named as the grant recipient of $65,790. In addition, the member gives Sandoz-sponsored lectures on an ad hoc basis at various professional societies and medical institutions and is paid for these by Sandoz.

• Member #3: The member had a reported financial interest in Bristol Myers Squibb because his or her spouse is employed by this firm.

• Member #4: The member's employer, an academic institution, had various research grants with antidepressant manufacturers. These included: (1) a grant of approximately $100,000 from A. H. Robins for which the member was a co-investigator on a study of Zacoprid; (2) a grant of approximately $100,000 from Wyeth-Ayerst for which the member was the principal investigator (PI) on the study of Zalosperone;

(3) two pending studies to be funded at $100,000 each from Eli Lilly, the makers of Prozac, one dealing with depressants and the other with sexual dysfunction; (4) a grant of approximately $100,000 from SmithKline Beecham for which the member was the PI on a study of Paroxetine; (5) a grant of approximately $100,000 from Pfizer for which the member was the PI on a study of Tandosperone; and (6) a grant of approximately $100,000 from Ciba-Geigy for which the member was the PI in an ongoing research study of antidepressant drugs.

The FDA process for generating waivers for this meeting differed from the practice that has recently evolved. The identification of exclusions, the need for a waiver, the sufficiency of the documentation justifying a waiver, and the approval of the waiver were handled entirely within the FDA. The following steps were involved:

1. Exclusions were identified for committee members and consultants the week of September 12, 1991.

2. (b)(1) waivers requests were initiated the week of September 12–17.

3. The DEPI signed off on the request for (b)(1) waivers on September 17, 1991.

4. At the time of the meeting, the CDER Director had approved waivers for members and consultants with the concurrence of the Associate Commissioner for Management.

5. The committee meeting was held on September 20, 1991.

Three days after the meeting, on September 23, 1991, the Citizen's Commission on Human Rights wrote to the FDA alleging that certain members of the committee had conflicts of interest. Their specific charge was that a number of individuals on the committee as well as several consultants to the committee had interests in companies that manufacture antidepressant drugs or were conflicted with respect to the "psychiatric industry" because they were psychiatrists.

In reply to these charges, the FDA noted that 18 USC §208(b)(1) permits waivers when the appointing authority certifies that the "interest is not so substantial as to be deemed likely to affect the integrity of the services which the Government may expect from such ... employee." At the time of the September 1991 meeting, the FDA program staffs were operating under the old conflict-of-interest statutes, ignoring subsection (b)(3), which had been added to the law by the Ethics Reform Act of 1989.

The facts that the waiver process operated without legal oversight and was oblivious to the latest changes in the conflict-of-interest law suggested that the process was in severe need of scrutiny. The reliance on subsection

(b)(1), instead of the more liberal (b)(3), seemed difficult to reconcile with the members' and consultants' interests as described above.

The purpose of the meeting of November 12–13, 1991 was to review seven premarket approval applications from manufacturers of silicone-gel breast implants. The agency wished to elicit recommendations regarding the continued marketability of these devices.

Twenty voting members and consultants were to attend the November 12–13, 1991, committee meeting. The program staff identified three voting members and two consultants who had exclusions with respect to matters that would be discussed.

• Members #1–3: Three members were practicing plastic surgeons who were members of the American Society of Plastic and Reconstructive Surgeons, Inc. Since the Society has assessed its membership approximately $4 million to counteract the negative publicity on breast implants generated by the FDA meeting and the media, the center decided that the plastic surgeons had a potential conflict of interest. Consequently, (b)(1) waivers were sought for these individuals, but not for full participation; rather they were to be allowed to participate in the discussion as nonvoting consultants.

• Consultants #1 and 2: Two consultants to CDRH had indicated that they had served as "expert witnesses" in cases involving women allegedly injured by silicone breast implants. Waivers were sought to allow them to participate in the meeting as nonvoting consultants.

The waiver process for the November 1991 meeting involved two "new" layers of review. Given the expected public attention to this meeting, the CDRH took great care in choosing members and solicited legal advice from the FDA's Chief Counsel's Office, which had not previously been involved. In addition, the Office of the Special Counsel for Ethics (OSCE) became involved in these discussions and, for the first time, reviewed the proposed waivers for legal sufficiency. The waiver review and approval process now included the CDRH, the DEPI, the FDA Chief Counsel's Office, the Commissioner, and the OSCE for the department.

The agency was well aware that the subject of silicone breast implants was attracting intense public scruity. In an unusual move, the CDRH scheduled over 100 interested parties to give testimony in the open public session of the meeting. The meeting was covered and reported by

newspapers, television, and the trade press. Committee members sat through the meeting facing television cameras and Kleig lights.

Meeting arrangements provoked sharp negative reactions from the members of the advisory committee. The regular chair of the GPS Devices Panel, Dr. G. Warden, a male, was replaced for this meeting (and for a February 1992 follow-on meeting) by the female chair of the OB-GYN Devices Panel. One apparent reason for this was to assure prominent representation by females. As a direct result of these two meetings, Dr. Warden resigned from the committee. In a highly publicized move before the February meeting, the FDA stripped the vote from another standing panel member because of statements given to the mass media, which prompted the agency to question whether he could render, or would be seen as capable of rendering, objective advice.

The purpose of this meeting was to discuss the Therakos NDA for a combination drug-device treatment of scleroderma using methoxypsoralen in conjunction with photopheresis. The NDA involved the drug, since the device had previously been approved for a related application. This application attracted considerable attention because the drug's sponsor and its clinical investigators charged that the FDA had mishandled the review. The House Subcommittee on Oversight and Investigations of the Energy and Commerce Committee later held a very critical hearing on the issue and has taken a continuing interest in the episode.

The salient conflict-of-interest issues involved CDER's decision not to seek a waiver for a consultant who was (later) alleged to have a conflict of interest. A description of the events of this situation follows.

After his term expired, the former chair of CDER's Arthritis Drugs Advisory Committee continued as a consultant to the Pilot Drug Evaluation Staff. In that capacity he served as the primary clinical reviewer of the methoxysporalen/photophoresis NDA mentioned above. At the same time, he was involved in preliminary negotiations to participate in a study of D-penicillamine, an alternative therapy for treatment of scleroderma. This study was funded by the FDA's Office of Orphan Drugs. The consultant was to serve as principal investigator at one of several centers participating in a multicenter trial. The grant for the D-penicillamine study would not have been made directly to the consultant's employer, a university, but to another organization. The consultant would have received no compensation from the grant.

CDER decided not to seek a waiver for the consultant because it concluded that, under prevailing the FDA policy, he had no conflict of interest. This decision was not subject to any legal oversight. The consultant had no financial interest in the D-penicillamine study, or in the company that manufactured it, or in any competing firm. All funding for the proposed study would have been supplied by the FDA.

Subsequently, Therakos and its clinical investigators charged that the consultant had been biased against the methoxypsoralen/photophoresis application because he was likely to be involved in research on a rival therapy. Their claim was that if the Therakos application had been approved, the consultant's own research would have been threatened. The allegation precipitated an investigation by the DHHS Office of Inspector General into possible violations of the law.

In this case, the FDA's policies failed to protect its consultant from a criminal investigation. Beyond the personal tribulations of the consultant, many such incidents would surely impair the agency's ability to attract capable clinicians/researchers to serve on its advisory committees.

The purpose of the meeting was to discuss a Product License Agreement (PLA) for two recombinant factor VIII products. The sponsors of the PLAs were Baxter Healthcare/Hyland Division and Genetics Institute, its development partner, and Miles Laboratories, a division of Bayer, A.G.

Three committee members were excluded for specific portions of the meeting:

• Member #1: This member was the principal investigator on a study of Baxter's recombinant factor VIII product. In addition, as a consultant to Baxter he occasionally lectured about the VIII product, for which he received an honorarium and travel expenses. CBER's Division of Transfusion Sciences did not request a waiver for this member.

• Members #2 and 3: These members had an (unspecified) interest in the Genetics Institute. Accordingly, the Division excluded them from discussion of the Baxter product. No waivers were sought for the other parts of the meeting, in which they participated.

This reveals a heightened sensitivity to conflict of interest within the FDA. Largely because of this increased sensitivity, the CBER division elected not to seek waivers for the three committee members. However, due to a miscommunication within the program staff, Member #1 was not

informed of his exclusion until the day before the committee meeting. Believing that his participation in Baxter studies should not have disqualified him from participation in the discussion, the member vigorously protested to the Commissioner.

The purpose of this meeting was to review two PLAs: (1) Proleukin (Interleukin2), made by Cetus/Hoffman LaRoche, and (2) Oncoscint, made by Cytogen. Two committee members, one of them the chair, had exclusions for the Proleukin discussion due to their consulting activities with Hoffman La Roche.

By the time of this meeting, the FDA's system for processing waivers had expanded to included this CBER program area, the DEPI, the FDA Chief Counsel's Office, the DHHS Office of Special Counsel, the Office of Government Ethics, and finally, the Center Director's Office. The process had become both contentious and time consuming. The late discovery of a business relationship between Cetus and Hoffman La Roche triggered an eleventh-hour reevaluation of the members' interests. The chronology in this case follows:

1. Exclusions were identified by CBER committee management staff on January 13, 1992, 4:00 p.m., for an advisory meeting scheduled for the 16th.

2. Waiver requests were initiated by CBER committee management staff for two members on January 14.

3. Because of the scheduled meeting date, the waiver requests were faxed simultaneously on the 14th to the DEPI, the FDA Chief Counsel's Office, and the Office of the Special Counsel for Ethics. Telephone conference calls were held that day to expedite the process. The Chief Counsel's Office sought additional details of the financial interests of the two members.

4. On January 15, uncomfortable with the "appearance of a rush" in getting the waivers approved, the DHHS Special Counsel offered a compromise. The two committee members could be granted waivers to participate and vote, but the chair of the committee would, in exchange, be required to relinquish the chairmanship for this portion of the meeting. Acceding to this "compromise," the CBER committee management staff asked the committee chair to step down for the discussion of the Cetus PLA.

5. The CBER Director signed the waiver requests on January 16, just a few hours before the meeting was scheduled to begin.

This case illustrates several problems. First, the waiver process had become too cumbersome to cope with the last minute ''discovery'' of financial connections. Given the complex relationships between companies, such last minute discoveries may not be an infrequent occurrence. Second, program offices and legal reviewers displayed no inclination to cooperate, leaving both feelings of distrust and frustration. Third, the legal basis for conditioning the chair's full waiver on his acceptance of a diminished role is unclear.

We believe this committee meeting was one of the first instances in which waivers for interests held by employers (university or other institutions) of committee members were required. In an April 6, 1992, memorandum to the CDER committee management office, the DEPI advised: "An additional concern is the requirement under (sub)section 208(a) that the financial interests of an SGE's employer, and other affiliations be considered, and must be addressed in the waiver. The potential impact of these entities has not always been considered in waivers which were requested prior to February, 1992. Therefore, under the new requirements each waiver request must address these concerns, before they are forwarded to the OSCE. We have learned that the Office of Government Ethics plans to draft a class waiver for all university affiliations. Until this class waiver has been approved, each 18 U.S.C. 208 waiver must address the SGE's university affiliations." This was the first written notice that the centers received regarding this change in policy.

Consequently, on April 9, one day before the scheduled Dermatologic Committee meeting, the CDER committee management office was faced with the task of preparing 11 (b)(3) waivers for members who were employed by universities. The waivers were signed by the center Director and DEPI on the 9th, and were delivered to the Commissioner's Office the next morning for his signature.

The requirement of waivers for university affiliations greatly affected the center's workload. Since most advisory committee members are affiliated with universities, whose hospitals dispense products manufactured by various companies with applications before the FDA, from this point on nearly every member required a waiver to participate. By the end of June, such waivers accounted for roughly two-thirds of all waiver activity within CDER.

The purpose of the meeting was to review a PMA for a coronary atherectomy system. This case is a dramatic illustration of how miscommunication among the offices involved in the waiver process delayed consideration of a small manufacturer's application, probably with severe financial consequences for the company. The events leading up to the meeting are:

1. In early March 1992, the CDRH program management office was notified of a panel meeting scheduled for May 11.

2. On March 13, exclusions were identified by the committee management office, and on that basis, 13 (!) waivers requests (for 8 members and 5 consultants) were prepared. One of the requested waivers was for a member who had been designated as the lead reviewer for the atherectomy catheter PMA. The presumptive "conflict" for this member was that his institution was involved in the coordination a large-scale randomized trial evaluating coronary atherectomy versus balloon angioplasty (known as the CAVEAT Trial). This trial, funded by Devices for Vascular Intervention and Eli Lilly in a grant for $2.3 million, is a prospective clinical trial involving 35 clinical sites throughout the United States and Europe.

3. On March 24, the sponsoring companies were notified that their applications would be reviewed on May 11. The notice of the meeting was published in the Federal Register on April 20.

4. By April 21, waivers had not been approved for many panel members. The CDRH Division of Cardiovascular, Respiratory and Neurological Devices informed the committee management office that the meeting might have to be cancelled if the requested waivers were not approved by April 24.

5. On April 24, the center's committee management office conveyed verbal clearance to the program area, which permitted mailing of all meeting material to committee members for review.

6. On Friday, May 8, CDRH's committee management office was informed by the DHHS Office of the Special Counsel for Ethics that no waiver would be allowed for the member who had been assigned as lead reviewer for the atherectomy catheter. Unsuccessful attempts were made to reverse the decision and a final refusal to grant a waiver was received at 3:45 p.m. The center decided to cancel the review of the atherectomy catheter PMA due to the inability to find a substitute reviewer over the weekend before the Monday meeting.

7. Late on Friday, the sponsor of the PMA and the committee members were notified of this change in the agenda. The sponsor was outraged and threatened to take all measures possible to "exact compensation" for the delay.

8. On Saturday, the Commissioner's office contacted the sponsor to allow it to present its application to the committee. An attempt was made to notify committee members of this change in agenda, but most could not be reached.

9. At the committee meeting on Monday, May 11, the members were told that the PMA had been restored to the agenda. The chair polled the members to determine whether they felt comfortable proceeding with the review. All of the members felt that they were inadequately prepared and voted unanimously to postpone review of the application.

The problems evident in this case speak for themselves. The reviewing division apparently assumed that the lead reviewer's presumptive conflict was waiverable. However, the Special Counsel for Ethics Office, exercising independent judgment, determined that his involvement in research on a competing technology precluded a waiver. This judgment was communicated only at the eleventh hour, disrupting the committee meeting and frustrating the sponsor's hopes for product development. The issue of waivers for committee members involved in research on competing technologies still presents problems for the FDA.

The preceding cases illustrate several of the confusing and frustrating events that have occurred within the centers' respective committee management offices between the period from September 1991 and June 1992. The rapidly escalating scrutiny of potential conflicts, the number of parties involved in the waiver process, the expanding criteria for identifying potential conflicts, and the zeal with which these criteria were applied combined to wreak havoc in the FDA's advisory committee operations.

To fully appreciate the impact of these events, it is instructive to look at the waiver process that the centers followed during this period. Before the fall of 1991, the waiver drafting and review process was entirely internal to the agency. The centers (committee management staff) decided when a waiver was necessary and wrote the justification for the waiver. DEPI reviewed the waiver justification and usually recommended its approval without change. Waivers were approved by the authority of the center director with the concurrence of the Associate Commissioner for Management. There was no legal oversight of this process.

By the end of June 1992, drafting and review of waivers had become a much more arduous process. Each waiver must now state the precise exclusions for which the member is being waived as well as providing a clear explanation of the "need" for the individual's participation. In current waivers the exclusions are listed in extreme detail; e.g. what percentage of the individual's income is represented by a particular financial interest. Additionally, the "need for the member" portion of the waiver is being reviewed more critically; in the past, statements to the effect that a member was a preeminent scientist and a member of the committee generally sufficed as a rationale.

A very serious flaw in the current waiver process is the lack of relevant written standards, at any level, for granting waivers, i.e., for ranking potential conflicts, for deciding whether a member's importance outweighs any risk of conflict, or for explaining decisions to grant (or deny) waivers. This has resulted in a customized, time-and resource-consuming process of case-by-case discovery and rationalization. Not infrequently, the process leads to burdensome iterations; one new discovery raises several more questions about other possible financial connections, requiring that the committee member be contacted again and asked to provide more information.

A likely, but hard to quantify, cost of this system is the disillusionment, and perhaps ultimate withdrawal, of advisory committee members who resent the disclosure of personal financial information, the repetitious requests for more information and clarification, the eleventh-hour decision about their eligibility (or disqualification) for participation, and the residual innuendo that they cannot be trusted.

Another consequence of the lack of written standards for approving waivers is that the process often becomes a hurried, sometimes frantic, rush to complete the waiver request just before an advisory committee meeting is scheduled to begin. This frustrates all participants. The FDA staff responsible for initiating waiver requests are most frustrated because they view the process as obstructing program goals and because decisions get made late, often long after their involvement, and without clear explanation. Those near the end of the process display less frustration, save with those who initiate waiver requests, because they have very different program objectives, such as preservation of department- or government-wide uniformity, avoidance of embarrassment for the administration, and maintenance of decisional integrity.

Even if one examines a number of prior decisions, as we have done, the current operational criteria for approving a waiver are elusive. The statute requires a judgment, ostensibly by the Commissioner, that a committee member's participation in a particular agenda item is important enough to justify the potential conflict. Recently, this decision has been made at the

level of the OSCE or the OGE and the OSCE, several levels removed from responsible authority.

On the other hand, the participation of these offices in the waiver process has introduced a level of legal rigor that was lacking in the pre-1991 process. The desire to conform existing practices to the 1989 law is commendable given the stakes involved, although the zeal with which this task has been approached may have obscured the need to fashion an orderly and predictable system for determining whether waivers are needed and justified.

The OSCE and the OGE, based on guidance from the Department of Justice, have interpreted Section 208 as reaching a very wide range of interests. The statute has been interpreted as embracing nearly all personal, spousal, minor child, and employer financial relationships, regardless of size, as presumptively disqualifying. For example, a member's employer may have a financial relationship with the company whose submission will be the subject of the committee meeting. Because there is no threshold limit on the size or remoteness of such employer connections, the practical outcome is that every member who is employed by a university whose hospital dispenses drugs made by the manufacturer whose submission is under review requires a waiver. Similarly, waivers are required for members whose institution may own stock in, or received an endowed chair from, the manufacturer.

The reach of the law has also spread with imaginative analysis into the matter of competing products and technologies. The OSCE has taken the position that a member with a connection, e.g., a research grant, with a company that is developing a technology that could be substituted for the product before his or her committee will require a waiver to participate. Rigorous implementation of this theory means, for example, that if a device pending approval will be offered to treat the same condition as an existing drug, members of the device panel must be screened for their, their spouse's, and their employer's financial relationship with the maker(s) of those drugs.

The OSCE and the OGE have also displayed concern about "appearances" of conflict of interest. In the proposed regulations (56 FR 33778) that followed Executive Orders 12674 and 12731 and the Ethics Reform Act of 1989, the OGE announced that even employees (including SGEs) who would not violate the law if they participated may nonetheless be disqualified from participating because of an apparent conflict. Usually, concerns about appearances of conflict surface in connection with high-profile advisory committee meetings in which potential press coverage may cause the OSCE to be more cautious than usual. In some such cases, the OSCE has insisted on "restricted" waivers: i.e., waivers that limit the member's participation in some way, typically by excluding the member from voting on a particular matter. The OSCE has argued that appearances of conflict

demand an "appearance solution" and, further, that restricting the participation of members with an appearance of conflict will reduce the chances for an actual conflict of interest.

The foregoing picture of the system managing potential financial conflicts of interest by members of the FDA advisory committees vividly reveals why Commissioner Kessler reported to the IOM committee that the agency faced a serious problem. It would be easy to assign blame for the emergence of the problem—to the divisions and centers for failing to appreciate the sensitivity of potential conflicts and viewing waivers as a matter of routine; to the DEPI for adhering to outdated policies and failing to appreciate the requirements (and perhaps even the enactment) of subsection (b)(3); to lawyers in the Office of the Chief Counsel for failing to provide either the centers or DEPI systematic legal guidance; to the Office of the Special Counsel for Ethics for excessive conservatism and failing to develop and convey general standards for approving waivers; to the Office of Government Ethics for last-minute and often unexplained second-guessing of waivers on which it was consulted; and to every level of the process for indifference to any reasonable set of deadlines for the development and approval of waivers.

There is, however, an alternative, less critical account. It is a story of offices and agencies caught suddenly in a confluence of forces that were moving too rapidly for any one to step back from the cascade of individual waiver cases to explain what was occurring and decide how the system should be righted. These forces included the heightened concern, within the administration and in Congress, over conflicts of interest involving federal employees; President Bush's decision to centralize in the OGE oversight of the waiver process for the entire government; Secretary Sullivan's decision to create a Special Counsel for Ethics responsible for reviewing all waivers granted within the department; the ripple effects of Dr. Kessler's own demands that the FDA officials and procedures should be, and should be seen to be, free from any hint of conflict of interest; a revived appreciation that the carelessness in identifying conflicts and granting waivers might not only jeopardize agency decisions but leave committee members exposed to criminal prosecution; and the reintroduction of lawyers, who were themselves confronting issues for the first time, into a system that had previously displayed an amateur understanding of the law. Under the circumstances, it is hardly surprising that confusion, acrimony, and frustration resulted.

As the IOM committee completed its work, there were signs that the participants in this process had themselves come to appreciate the nature of the "problem" and had resolved—within the practical limits of current law and executive orders—to improve the system's operation. The most promising of these steps was a meeting held between Mary Pendergast, Senior Advisor to Commissioner Kessler, and Jack Kress, DHHS Special Counsel for Ethics. Representatives of all five the FDA centers, including the three whose committees are the subject of this report, were also in attendance. Based on individual accounts of the meeting, it is possible to sketch the broad outlines of the reforms tentatively agreed on.

OSCE representatives agreed to consider one-time waivers for advisory committee members whose disqualifying interest is that of their university or hospital in sales of medical products of a company with a product under review by the FDA. Such one-time waivers would allow the committee member to participate in all future committee meetings. Since over half of all (b)(3) waivers now sought by the centers are for employer interests of this sort, approval of one-time waivers should dramatically reduce the waiver workload.

The FDA's centers will remain responsible for the initial screening of advisory committee members, for determining whether a member confronts a potential conflict relative to an agenda item, for deciding whether to recommend a waiver to the Commissioner, for preparing waiver requests, and for obtaining information from committee members. The OSCE will have final authority for the agency and the department to determine whether a member's participation—absent a waiver—would violate the law, i.e., whether a potential conflict of interest exists. Authority to decide whether to grant a waiver will rest with the Commissioner or, at his choice, the Deputy Commissioner for Operations. This understanding represents a constructive clarification and allocation of responsibility between the Commissioner, who under the law is empowered to grant waivers, and the OSCE, which Secretary Sullivan has made responsible for assuring that the conflict-of-interest laws are complied with.

Within the FDA, waiver review and approval is to be expedited. Centers will forward recommended waivers to the OSCE at the same time they are

submitted to the DEPI. The DEPI will be given 48 hours to respond, and silence will be construed as acquiescence. Meanwhile, presumably, the OSCE lawyers may carry out their own review. This arrangement would appear to subordinate the role of the DEPI, as well as to expedite review, and we consider both to be sensible. With the OSCE able to provide authoritative legal review of the need for, and form of, any waiver, the FDA's concern that waivers should be granted judiciously and only when necessary to assure effective committee functioning can be fully protected by the Office of the Commissioner.

Centers are to submit recommended waivers to the OSCE (and to the DEPI) at least two and preferably three weeks before the committee meeting. The OSCE has committed itself to review such recommendations promptly and to use its best efforts to complete all corrections at least three days before the meeting. Under this arrangement, the Deputy Commissioner should have all waiver recommendations, approved as to need and form by the OSCE, three days before the meeting, which should provide adequate time to exercise independent judgment and, if necessary, confer with the OSCE. Understandably, exceptions to this schedule may be needed for consultants invited to assist the committee at the eleventh hour.

Waiver forms will be divided into two parts. Part I will be a straight forward recitation that the Commissioner or Deputy Commissioner has granted a waiver for a committee member to participate in the discussion of an agenda item with respect to which he/she would otherwise have a conflict. Part II will consist of a more detailed explanation of the circumstances that give rise to the potential conflict and the reasons why the center (and agency) concludes that the member's participation ought nonetheless to be approved.

The IOM Committee views this meeting, and the agreements reached, as a significant first step. Given Secretary Sullivan's decision to lodge conflict of interest oversight authority for the entire department in the OSCE, this office will inescapably have a central role in the FDA's administration of conflict of interest restrictions applicable to advisory committees. Accordingly, for any system to work effectively, it must have the support and active cooperation of both OSCE and the FDA. The agreement to grant one-time waivers for certain attenuated employer interests should cut the FDA's waiver caseload significantly, but there will remain a number

of individual cases—intersections between committee agendas and member exclusions—that will require sensitive judgment and active, timely cooperation between the FDA, including the centers, and the OSCE. Our recommendation below that the two organizations make an effort to agree on, and in some fashion codify, the criteria for evaluating waivers in these more difficult cases will demand a much greater commitment to cooperative policymaking than has been evident so far.

The IOM committee believes that it is essential that the members of technical advisory committees be impartial and objective and not compromised by financial conflicts of interest. To achieve these ends, the IOM committee has addressed the standards and procedures for controlling conflict of interest.

Any attempt to address the problem must deal with issues of law, of bureaucratic procedure, and of administration. The IOM committee considered reforms that would require new legislation and those that could be implemented within existing statutory authority.

The IOM committee considered several options that would require new legislation. For example, a recent report to the Administrative Conference of the United States advocated a system under which there would be no disqualification of any advisory committee member for financial conflict of interest, but each member would be required to make full public disclosure of all financial dealings, holdings, and relationships.¹ This proposal differs from the current system in two ways. First, full public disclosure of all of each member's financial interests goes well beyond the present extensive disclosure to the FDA and public disclosure only of agenda-specific conflicts that may disqualify or constitute the basis for a waiver. Second, it holds that no interest would preclude a member from participating in committee deliberations.

The IOM committee concluded that the latter feature of the proposal was unacceptable. It would permit an advisory committee member to serve in instances in which his or her financial interests would constitute a clear conflict of interest and in which the remedy should be disqualification from participation. Such a system would undermine the appearance of objectivity of the committee's advice.

A second option would be a system that coupled full disclosure of all interests with a general rule barring participation by members with significant financial interests.

Although this proposal may contain the core of a promising reform of the system for regulating conflict of interest, the IOM committee did not explore fully its ramifications. The committee's judgment—confirmed by many we spoke to, including officials of the FDA, OSCE, the OGE—was that such a major legislative overhaul of this magnitude was simply not a possibility in the near term. Thus, given the FDA's pressing needs, our charge, and our timetable, it seemed imperative for the committee to turn to reforms that could be implemented within the existing statutory framework. However, this possibility is clearly a candidate for further study.

What can be done under existing authority? Potentially a good deal, as the following recommendations suggest. Although the first and second recommendations below could be implemented by the FDA itself, the successful implementation of the other recommendations would require the active involvement of the Commissioner and his office, the collaboration of the OSCE, and at least the tolerance of the OGE.

A theoretical option for the FDA would be to avoid appointing advisory committee members as special government employees, thus circumventing the restrictions of the federal conflict-of-interest law. This solution has the notable disadvantage of attempting to define the problem away, hardly a way to instill confidence in the system. Moreover, new legislation might be needed to allow payment of members and sharing with them of trade secret information.

Second, the FDA itself could exercise greater care in the initial appointment of advisory committee members. It could demand even more information that would enable it to identify in advance potential members whose financial interests would clearly disqualify them for some committee meetings. On the other hand, because the interpretation of a prohibited interest is already extremely broad, and because potential conflicts cannot be identified before meeting agendas are set, vigorous pursuit of this approach might disqualify valuable members and produce no gain in integrity.

Third, the FDA, working with the OSCE, could formulate and codify criteria for granting 208(b)(3) waivers. The IOM committee believes that this is essential. Codification would be a lengthy process, but some mutual understanding of the grounds for justifying a waiver is badly needed. A checklist of variables should be formulated that includes: the size of the

interest; the character of the interest; the likelihood than an interest will be affected by agency action based on the committee's advice; and the actual importance of the member to the committee deliberations. Regarding the latter point, we believe that committee membership alone should not be taken automatically as a decision measure of a member's importance.

Of immediate importance is the need to clarify the criteria for dealing with potential conflicts arising from institutional or employer financial interests, research grants and contracts to committee members, and member involvement with competing products and technologies.^* Most advisory committee members are university employees; most of their employers operate medical schools, hospitals, and hospital pharmacies. The OSCE, with the FDA, should develop clear criteria for dealing with waiver requests that arise because a committee member is affiliated with an institution that operates such subordinate entities, which in turn derive income from the dispensing of use of the FDA-regulated products. Most universities also own diversified endowment funds and it is common for some portion of these to be invested in pharmaceutical, biotechnology, or medical device securities. The OCSE, with the FDA, should clarify the criteria for dealing with these ''employer interests'' as well.

Fourth, the FDA has the authority to streamline its own internal policies and procedures for deciding when to seek waivers and how to prepare their justifications. The IOM committee believes that this also is essential. Responsibility for preparing the initial waiver request should reside with the division. The decision to request a waiver should be made by the center director. The IOM committee sees no need for independent review of this decision by the DEPI or by the FDA's Chief Counsel, so long as the OSCE continues to exercise an oversight role. Central agency review of waiver requests should be by a high-level policy official in the Office of the Commissioner.

Fifth, the FDA should develop and adhere to strict schedules for processing waivers. It should present waiver requests to the OSCE no later than three weeks in advance of a meeting.^* The Commissioner should seek agreement from the OSCE that it will review any proposed waiver within three days. The Commissioner, who has the ultimate responsibility for approving waivers, may even wish to establish default rules that penalize centers for failure to complete their part of the process in a timely way (e.g., the member is disqualified or the agenda item is postponed). This or other default rules should be designed to ensure an expeditious process and also to guarantee that at no time would an advisory committee member of questionable impartiality be allowed to participate on the committee.

Sixth, the FDA must update its training of officials who have responsibility for implementing conflict-of-interest policies with respect to advisory committee members. Training programs should build around the substantive and procedural changes suggested above. Participation should be required of all the FDA professional staff who deal with advisory committee members.

Seventh, the FDA must also initiate and maintain a formal orientation program for advisory committee members. Individual members should clearly understand the laws that govern financial conflict-of-interest and the justifications for granting waivers. However, the IOM committee believes that guidance on conflict of interest should be part of a broader orientation

program (discussed below and at length in Chapter 8). This linkage is important because an exclusive focus on conflict of interest will necessarily emphasize the risk of criminal prosecution and the need for intensive inquiry into personal financial matters—an emphasis that would surely obscure the public service dimension of advisory committee membership.

Eighth, the FDA and OSCE, on behalf of the department, should continue to press the OGE to issue government-wide general (b)(2) waiver regulations as soon as possible. This yet-to-be-exercised statutory authority is intended to remove certain classes of potential conflicts from a case-by-case determination. Employer financial interests and some research grants and contracts could be dealt with by such a rule.

Finally, the FDA, and the department, should seek the revision of Executive Order 12674, which requires case-by-case consultation with the OGE on all waiver requests.

This chapter thus far has focused on the methods for protecting committee deliberations against just one threat to impartiality—the possibility that committee members will modify their advice because of the prospect of personal or employer financial gain or loss. This focus is justified because

the administration of the financial conflict-of-interest restrictions applicable even to short-term federal employees is the main source of the problem on which our advice was sought.

But the FDA needs to guard against another potential threat to advisory committee objectivity. As a recent series of articles in Science magazine recounts,²,³,⁴,⁵ there is a growing concern about—and considerable publicity surrounding—the subtly and perhaps even overtly biasing effects on objectivity of a scientist's prior research and public positions, particularly positions taken in formal administrative or judicial proceedings.

For convenience we have termed this potential effect "intellectual bias," which is meant to distinguish the problem at hand from financial conflict of interest. Too frequently, we think, both members and observers of the scientific community apply the term "conflict of interest" to the problem of intellectual bias, which is more complex and elusive than the sorts of financial conflicts we have heretofore been discussing.

Making this distinction is particularly important in the present context, because the FDA (and its advisory committee members) are subject to a set of formal criminal restrictions that apply only to financial conflicts of interest. Section 208(a) says nothing about possible intellectual bias or prejudgment that is untainted by financial interest. A committee member may be incapable of entertaining a particular hypothesis, however convincing the evidence, but he or she does not violate section 208(a) by participating in committee deliberations on an issue to which the hypothesis' plausibility is crucial. By participating, however, he or she not only threatens the committee's capacity to render impartial and thus useful advice to the agency, he or she may in so doing thereby jeopardize the validity of any decision that the FDA may reach based on the advice it has heard.

What legal restrictions apply in this context is a matter of some uncertainty. It could be argued that a committee whose members include someone with a closed mind on an important issue is not "balanced" as the Federal Advisory Committee Act requires. Even if this were plausible, recent cases have cast doubt on the enforceability of the FACA's "balance" requirement. Furthermore, the FACA requirement of balance cannot be translated easily into operational safeguards against possible committee member intellectual bias. When a new committee is formed, or a new member is appointed to an existing committee, it is impossible to anticipate all of the issues or applications on which the FDA will seek the committee's advice. The general jurisdiction of the committee will of course be known, but its future agenda cannot be. And it is possible bias with respect to a particular agenda item that the agency should be concerned about. Any practical approach to this problem must operate at the point that the

committee's agenda is set and the issues to be addressed at a specific meeting are determined.

The vulnerability of any decision reached by the Commissioner, the official decision-maker, is likely to be a function of his own rather than any committee's impartiality. The key factor seems to be whether the Commissioner's decision is required to be based on a formal adjudicatory record. Most of the decisions on which the FDA seeks advice from its committees do not require the agency to hold a formal adjudicatory hearing. Even so, it cannot be said with assurance that the participation of committee members whose views on crucial matters were already invariably formed might not provide a legal basis for setting aside the FDA's ultimate decision. Proving intellectual bias of such character would, of course, be difficult, and perhaps in all but clear cases, impossible.

Even assuming the risk of judicial reversal is small, however, there are very good reasons why the FDA should be concerned to assure that committee members are capable of maintaining an open mind in evaluating the theories and evidence brought before them. One is that a committee whose advice is not impartial defeats the very purpose of seeking independent expert advice. A second is that the widespread belief that some committee members are, if only rarely, immune to persuasion by evidence would surely erode public confidence in a mechanism the FDA has devised precisely to enhance confidence in its own decisions.

We do not consider intellectual bias to be a common problem among members of the FDA advisory committees. Scientists are trained to be skeptical, to insist on evidence to support hypotheses, and to be rigorous in their assessment of evidence. We are convinced the overwhelming majority succeed. Furthermore, the FDA advisory committee context probably presents fewer occasions for challenge to long-held views than many other contexts in which individual scientists are called on to offer their opinions. One reason is that advice-giving by a committee is a collective process and not an individual exercise.

However, it cannot be said that the FDA has no basis for concern about intellectual bias or no reason to take precautions to guard against it. The IOM committee believes that the agency should be sensitive to the possibility that, on particular issues, an advisory committee member might be so deeply committed to a point of view, or so publicly identified with that view, that his or her objectivity cannot be assumed or will not be credited by those who are interested in the committee's deliberations. What steps the FDA should take when such a case arises probably cannot be prescribed in advance; the appropriate remedy is likely to depend on the circumstances. Perhaps even more difficult is designing a formal system to screen for potential intellectual bias.

Because no clear set of legal restrictions is operative here, the agency has a wider range of remedies from which to choose than it has under section 208. If the determination of bias rests on publicly stated positions, exclusion is probably warranted. There may be other cases in which this may also be appropriate, e.g., when a member is an inventor of the technology under review (even if he or she has no financial interest in its approval). And denial of participation surely would reduce the risk of recrimination and embarrassment.

There may, however, be instances in which exclusion of a member for possible bias would deprive the committee members who do participate of information helpful in their independent evaluation. If exclusion stems from the member's prior research, especially as a principal investigator, the FDA should not have to forego that individual's expertise. This can be solved by inviting the person to address the committee as a witness (or as a "guest"). In such an instance, it would be desirable to situate the individual so that he or she does not appear to occupy his usual role as a voting member of the committee.

On the other hand, there is some disadvantage to creating too large a set of roles at committee meetings to accommodate various perceived levels of partiality. A sensible rule of thumb might recognize just three roles for committee members in the case of intellectual bias: (a) full voting participation; (b) full exclusion from a meeting or an agenda item; or (c) appearance as "witness" or "guest" of the agency.

Short of complete exclusion, the success of any more limited remedy will depend on full public disclosure of the facts that give rise to the concern that the objectivity of an erstwhile committee member may be, or may be thought to be, in doubt. It may be possible to say that "Dr. Jones has agreed to recuse himself from the discussion of Product Y because of concerns that, based on prior work in the field, his objectivity may be challenged. He therefore will not participate in the committee's formulation of advice, nor will he vote. He has been asked by the FDA to be available as a witness to answer questions from voting committee members."

The development by the FDA of a system to screen for potential intellectual bias will require considerable thought. A member's prior research will probably be revealed in the screening for financial conflict of interest. All relevant publications presumably will be included in a potential member's curriculum vitae. However, it may be necessary for the agency to

inquire specifically about public positions, especially any taken before regulatory agencies or judicial proceedings.

This brief discussion barely penetrates the surface of a complex and very sensitive issue. It is sensitive, in part, because committee members whose objectivity might be challenged on other than financial grounds are often likely to resent the suggestion that they cannot be trusted. This is particularly true if the suggestion comes from the agency, which has appealed to their spirit of public service. Perhaps the heightened attention given to the subject within the scientific community will generate greater sensitivity on the part of individual committee members.

One further observation is in order. We consider the matter of intellectual bias to be a problem for the FDA to address and resolve. It is not covered by the federal conflict-of-interest laws and its possible occurrence is therefore not properly addressed through the formal waiver process. It obviously has legal ramifications to the extent that agency decisions might be subject to attack because of the participation of a committee member who lacked, or was accused of lacking, the requisite objectivity. But these are ramifications that the FDA's Chief Counsel is very capable of assessing and providing guidance on to the agency.

The IOM committee recommends, appropriately in my view, that the FDA and the DHHS Office of the Special Counsel for Ethics (OSCE) "immediately begin the process of codifying the criteria for granting 208(b)(3) waivers, especially with respect to employer interests, research grants and contracts, and competing products and technologies." This statutory provision, specifically enacted for members of federal advisory committee, holds that an agency head may waive the potential financial conflicts of an advisory committee member if he determines that "the need for the individual's services outweighs the potential for a conflict of interest created by the financial interest involved." It is this authority on which the FDA now exclusively relies in deciding whether to allow members with potential financial conflicts to participate in committee deliberations on a particular matter.

The IOM committee report itself does not offer concrete guidance on how this might be done. This apparent deficiency of the report becomes understandable when one grasps the difficulty of the exercise and recalls that the committee's schedule allowed for only four face-to-face meetings. Framing a discussion of which kinds of potential conflicts should be considered serious and which not, and of how to assess the importance of a single member to a committee's deliberations is a complex undertaking. Reaching judgments on these issues requires extended discussion and debate. There was scarcely time to attempt the first of these challenges, and no opportunity at all for the full committee to engage in the extended discussion needed to reach agreement on the second.

What follows is one member's attempt to outline the analysis he would follow in attempting to carry out the IOM committee's recommendation. It does not necessarily reflect, either in its approach or in the normative judgments implied, the views of any other committee member. It is offered to provoke further analysis within the FDA and OSCE rather than to prescribe a solution to the problem that they jointly confront.

The problem of codifying the criteria for approving waivers under 208(b)(3) is complicated by two main facts. First, the range of matters on which the FDA seeks advice from its advisory committees, when coupled

with the prominence and diverse activities of the members of these committees, generates a substantial number of presumptive "exclusions" that dictate nonparticipation or require consideration of waivers. An FDA committee meeting seldom occurs today without one or more members facing exclusion from one or more items on the agenda. In short, the agency faces a large "caseload" of potential waivers.

Second, the caseload is large in major part because the conflict-of-interest law, section 208(a), sweeps extremely broadly, embracing as potentially disqualifying of an individual committee member not only small personal (and family) financial interests but interests or relationships of the member's employer. As most committee members work for universities or other research and health-care providing institutions, most have, through their employers, traceable if indirect ties to multiple research grants, clinical research arrangements, and a vast array of paid-for health care services. As the law is now interpreted, any of these interconnections can give rise to a potential conflict—and thus require either exclusion or a waiver—for an advisory committee member.

Relatively few such interconnections, in my judgment, ought realistically to be viewed as jeopardizing the impartiality of a committee member's advice. And this, as I understand it, is the central concern of the conflict-of-interest laws, i.e., a concern to prevent governmental decision making, or in this context advice obtaining, from being compromised by the self-interest of the advice giver. The implication of this judgment is that a large, but indeterminate, percentage of the presumptive exclusions revealed by the FDA's system of comparing committee member interests (including employer interests) with committee agendas are plausible, if not compelling, candidates for the exercise of the Commissioner's waiver authority.

The goal sought by the IOM committee's recommendation is the development of general criteria or guidelines that facilitate decision making about the appropriateness of waivers in individual cases. We believe that it should be possible to identify and articulate categories of interests that ordinarily ought not be considered disqualifying, i.e., should be considered waiverable. It also may be possible to identify other categories that ordinarily should be considered disqualifying. And it may even be possible to enunciate criteria for deciding whether a committee member should be allowed to participate even with a significant potential conflict in a particular matter because of his or her importance to the committee's deliberations.

The recent history of the development, review, and approval of (b)(3) waiver requests for members of the FDA advisory committees, recounted in the IOM committee report, illustrates the consequences of the failure to develop general criteria for approving such requests. For some months now, each new waiver request appears to have been treated as a novel case,

requiring extended negotiations over the appropriateness and terms of granting a waiver and the content and format of the document explaining the agency's decision. Such a case-by-case approach virtually assures that the process will prove both burdensome and deeply frustrating.

The IOM committee found signs that officials in the FDA and OSCE recognize the need to regularize the waiver review process and reach agreement on the treatment of certain categories of potential conflicts. The committee's recommendation is essentially that this effort be extended and given priority.

The product we visualize would be a series of written guidelines, or even a grid, for decision making. For example, one ''guideline'' now under consideration by the FDA and OSCE would say something like: "The fact that a committee member's institutional employer operates a hospital or clinic that dispenses and charges for the FDA-regulated products, including products of the manufacturer whose application is to be reviewed, will not ordinarily be deemed disqualifying. Accordingly, a waiver to allow him or her to participate in committee deliberations is appropriate." Without necessarily endorsing this illustration, the IOM committee's hope is that other classes of interests that under the law would be presumptively disqualifying can be generically categorized as waiverable or as not waiverable.

There are many obstacles to the achievement of this goal. Some are empirical. It requires comprehensive knowledge that may not be easily assembled about the types and magnitudes of interests that the FDA committee members report that now trigger exclusions. We were given many examples, but no information that would allow a judgment about which potential conflicts were representative or how often any one occurred.

Another set of obstacles is institutional. Since many decisions about whether a type or size of interest should be viewed as disqualifying are, ultimately, matters of judgment, it is to be expected that individuals will disagree about the proper disposition of paradigm cases. The present arrangements arguably require the concurrence, or at least the acquiescence, of three offices—the FDA Commissioner, the OSCE, and the Office of Government Ethics—before any waiver can be approved. Achieving agreement at this level on any set of generic guidelines is likely to be a long-term task.

This Appendix addresses a third set of obstacles to the achievement of what the IOM committee has termed "codification." For lack of a better word, I will label these "analytical." In order to decide whether a particular kind of interest should disqualify a committee member, or, since the statute treats most interests as disqualifying, whether a waiver is appropriate for a given committee member, one needs to have some understanding of the underlying goals of the conflict-of-interest law. I suggest that the primary

goal should be to prevent the participation of committee members whose advice might, because of self-interest, be distorted. One could add to this formulation, as the present law does, a qualification: If a member's value to the committee's deliberations is great enough, even some risk of distortion may be accepted.

This formulation of the statutory goal is not quite congruent with the language of 208 (b)(3), which arguably calls for an individualized judgment about a member's value to committee deliberations in every case. However, because the determination of a member's value appears to be a more complex inquiry, I believe it would be more fruitful to concentrate first on the dimension that requires consideration of a waiver in the first place—the presumptively disqualifying financial interest—and see if it is not possible to categorize and then rank such interests in terms of their potential to undermine impartial advice-giving. I also believe that this is not only possible but compatible with the statute.

It seems to me quite plausible to argue, for instance, that for some sorts of interests—though perhaps not many—the threat to impartiality is so negligible that the fact of selection for committee membership should be taken as sufficient evidence of a member's value to the committee's deliberations. The willingness of the FDA and OSCE to consider agreeing that employer health care delivery activities, e.g., university hospitals, should never (or rarely) be viewed as disqualifying—i.e., should be automatically waiverable—is evidence that this legal interpretation is not preposterous.

I should add that even if it proves difficult to reach agreement on many other automatically waiverable classes of financial interests, the exercise of categorization and ranking should help improve relations between the agency and the OSCE. One frequent complaint that we heard from the FDA officials was that they never knew what the rules were. This can be translated as "we never know what sorts of interests would really raise eyebrows at OSCE." The waiver review process would be greatly improved if it were possible for the OSCE to say, and the FDA officials to know (even if they do not agree), what sorts of interests will be most difficult to grant waivers for.

On the other hand, OSCE staff members might develop a greater understanding for the agency's position if the FDA officials were able to articulate the factors that they consider important in assessing a committee member's value to committee deliberations. The attitude that committee membership—another live body eligible to vote—is all that is necessary to convince center personnel that an individual member, despite a significant potential conflict, is absolutely crucial to deliberations cannot inspire confidence that the FDA is exercising the sort of discriminating judgment that the law seems to contemplate.

The effort suggested could have value beyond assisting the FDA and the OSCE in preparing and reviewing proposed (b)(3) waivers. The part of the task discussed in this Appendix—classification of disqualifying financial interests in terms of their likely effect on committee member impartiality—would be directly relevant to the Office of Government Ethics' consideration of possible (b)(2) waiver regulations. The reader will recall that section 208(b)(2) allows the OGE—and only the OGE—to promulgate regulations categorically exempting certain types of magnitudes of financial interest as "too remote or too inconsequential' to affect a government employee's honest performance of his or her functions. This sort of waiver does not require an assessment of the employee's, e.g., the committee member's, importance to committee deliberations. Thus the effort to identify types of financial interest whose potential influence is so improbable that mere membership can be considered outweighing is a logical prelude to the exercise that the OGE must eventually undertake to implement (b)(2).

The OGE should welcome the FDA/OSCE effort, even if it does not agree with every part of their classification. The FDA/OSCE analysis should advance thinking in this most difficult area and provide examples of (b)(2)-waiverable interests that are common among medical and scientific researchers but perhaps not frequently encountered among other federal personnel.

What can be said, if anything, about the sorts of interests that ought to be considered as jeopardizing a committee member's impartiality? Although the following discussion reflects personal judgments, it may offer the beginning of a framework for thinking about that question.

For me, certain generalizations seem plausible, though not incontestable. The magnitude of a financial interest surely is likely to make a difference; we would worry more about a committee member's objectivity, in assessing a company's product, if he or she owned shares of stock in the company than if he or she owned one share. The law may not see a difference here, but most people do. And the law would appear to allow this difference to be accorded weight in a decision whether to grant a waiver.

Differences of magnitude—at least in ownership interests or direct payments—are relatively easy to discern and deal with. At least they are easy to array on a chart or grid. It may not be easy to agree on what threat to impartiality is presented by interests of different sizes. And searching for agreement at several different levels may not be worthwhile. Perhaps it should be enough to reach agreement on "de minimis" levels that would, if not exceeded, ordinarily allow a waiver. (We are not speaking about establishing a de minimus standard for applying the presumptive disqualification of 208(a), but rather are seeking one measure of the presumptively disqualifying interest that define its eligibility for waiver. In short, we are not

quarreling here with the prevailing interpretation of 208(a), which holds that even a single share of stock or a $100 speaking honorarium is a prohibited interest, although in another context many of us surely would do so.)

Matters get more complicated when one tries to categorize financial interests by type. But the exercise is not futile. It is possible to frame generalizations about what kinds of interests are more worrisome than others. The central question that seems likely to help clarify thinking about which interests really threaten impartiality is "whose interest is it?" This question can be examined with reference to employer interests and personal interests. The statute forbids a committee member's participation in any matter in which he or she, a family member, or an employer holds any financial interest. I suggest that personal (including family) interests are more likely to threaten an individual's objectivity than the financial interests of his or her employer. To be sure, one can think of examples of both sorts that would defy this generalization, and perhaps—on close examination—those examples would swallow the principle. Even so, as one starting place for analysis, it is likely to prove helpful to make the personal-employer dichotomy one of the dimensions of a grid of financial interests, all of which under the current law are deemed presumptively disqualifying.

We have been provided examples of three types of employers' interests that are believed to trigger section 208(a): (1) sales by the employer, or by a subordinate unit of the employer, of the FDA-regulated products made by companies that have new product applications pending before the FDA; (2) research or other grants from such companies; and (3) gifts from such companies to support institutional programs, e.g., an endowed chair. No doubt there are other many others, among which it ought to be possible to draw distinctions based on the likelihood that a committee member might modify his or her advice in order to protect an employer's relationship with a company. It would not be imprudent, in my view, for the FDA to take the position that interests falling in the first category ought always to be waiverable absent clear evidence that the employer receives a substantial amount of its income from such sales or that the committee member him-or herself benefitted personally from decisions affecting the usage or sale of the company's product. A similar judgment might be supportable for research grants other than those made to a committee member personally.

My goal here, however, is not to offer conclusions about which kinds of employer interests ought to be considered routine, possible, or unlikely candidates for waiver, but simply to suggest an approach to thinking about this question. The approach involves, first, the categorization of the various types of employer interests the FDA committee members have displayed and, second, thoughtful assessment of the likelihood that interests within a

particular category will undermine an individual committee member's impartiality.

The same sort of exercise is appropriate in assessing the personal interests of committee members. To simplify the task, it might well be prudent to conflate member interests and family interests, i.e., to assume that a spouse's or minor child's financial interest is as likely (or unlikely) to affect a member's impartiality as his or her own. It would be a crude generalization, to be sure, but crude generalizations will be necessary to develop a framework that can guide—and, which is the ultimate goal, simplify and thus expedite—review of individual waiver requests.

Within the category of personal interests, individual research grants are apparently a common source of presumptive disqualification. The dollar value of a grant probably ought to be a consideration in assessing the likelihood that it may affect a committee member's impartiality. But equally important, it seems to me, is the extent to which a grant contributes to a researcher's personal income, as distinct from institutional income. I would be inclined, as well, to differentiate between research grants provided in the past and grants that currently support a member's research. The influence, if any, of the former must be in the member's hope for future research support from the same source, and I do not find it implausible that an individual's judgment is less likely to be influenced by a hope that support might someday be renewed than by the fear that current support may be terminated.

I would, at least tentatively, draw a similar distinction between other sorts of company payments to a committee member. A concluded consulting arrangement that once paid a $2,500 yearly honorarium strikes me as less worrisome that an on-going relationship that provides rewards, even of smaller magnitude, in future years.

Indeed, it may be appropriate to draw a broader distinction between interests that a member already owns, and whose value will not significantly change, and interests whose enjoyment, or whose value, may depend on the success of the company that is the source of the interest. A consulting fee paid in the past may be the source of hope for future beneficial relationships, but its value will not be diminished if the company never provides support again. By contrast, the value of stock owned by a committee member in a company whose products he or she is asked to evaluate is clearly affected by the future success of the firm, and very possibly by the profitability of the new product.

The range of personal financial interests that one can assemble from examples provided by the FDA is large, and their variety may appear to defy any systematic effort at description, much less a categorical assessment of their likely affect on impartiality. But one cannot know this without making

the effort. And there is a value to the intellectual exercise even if a formal decision making grid or set of decisional guidelines remains incomplete. It will force those involved to articulate and explain their judgments about the appropriateness of granting waivers in specific cases. It may also yield—and this would be no small achievement—a common vocabulary for describing and analyzing individual cases. And, if engaged in jointly by officials from both the FDA and the OSCE, it may help to reveal common ground and to clarify differences.

The discussion thus far has focused on only the first element of the statutory formula for granting (b)(3) waivers—the potential of different types and magnitudes of financial interests to undermine a committee member's impartiality. Section 208(b)(3) also requires consideration of a member's importance to committee deliberations. The suggestion made here is that some interests can be classified as so unlikely to threaten impartiality that selection for committee membership can be taken as sufficient evidence of importance to offset the remote risk. But there may be few such interests, and they are likely to be an employer's rather than personal or family interests. Thus, in evaluating many waivers attention must be given to a member's importance to committee deliberations. I believe that this second element should also be susceptible to categorical analysis, i.e., it should be possible to formulate guidelines for evaluating individual cases. And the committee's recommends that this should be done.

There is one additional point to be made. Many readers may ask how the exercise sketched in the foregoing paragraphs fits with the statutory regime for regulating financial conflict of interest. The answer has already been suggested but warrants reiteration. Section 208(a) of the conflict-of-interest statute sweeps very broadly and, I acknowledge, as construed makes the sorts of distinctions discussed above irrelevant to a determination of whether an interest presumptively disqualifies a committee member from participating in advice on a particular matter. But section 208(b)(3) calls for an assessment, by the Commissioner of Food and Drugs, of the likelihood that a disqualifying interest will in fact affect the member's objectivity, as when as of his/her importance to the committee's deliberations. Such an assessment logically invites, and surely permits, consideration of the sorts of distinctions I have suggested.