RESPONDING TO ALLEGATIONS OF MISCONDUCT IN. THE UNIVERSITY

than to get involved in the unrewarding battle to bring someone to justice. A recent editorial that recommends that unproven accusations be denounced as false charges and that whistle-blowers be held to a particularly high standard of competence (Koshland DE, Science 240:585, 1988) appears to offer more indirect support for avoiding involvement than encouragement for bringing questionable practices to the attention of authorities for evaluation.

The requirement that accusers be protected, found in essentially all policy recommendations, is difficult to implement. Nothing in our procedures provides for job security for whistleblowers, nor could we conceive of how to create meaningful guidelines for doing so, and confidentiality cannot usually be maintained after the initial inquiry. Once details of charges become known to the accused, so, usually, is the identity of the accuser. Nevertheless, one of the goals of new policies and guidelines should be to try to protect those who bring accusations, but not to make whistle-blowing a way to ensure tenure. It should be possible to improve the fairness and responsiveness with which accusations of misconduct are treated by improving faculty and administration awareness of ethical issues, without the unwelcome intrusion of self-appointed guardians of scientific morality or of auditors appointed by government agencies. This requires, however, that the burdens of consciousness-raising and pursuing individual issues should be assumed by the institution, and not rest on the

accuser.

E.

Safeguarding rights and reputation of the accused; due process

As detailed by Mishkin (Report of the First Workshop, and JAMA, October 1988), policies should recognize the due process rights of the accused, formalizing the basic fairness of the procedures for dealing with charges. Popular reports of research misconduct and fraud have seemed to equate being accused with being guilty. Perhaps this is because until recently only the most egregious fraud cases have in fact made it to the limelight, and the accused individuals were guilty as charged. That is not always the case, however, and some scientists will be found innocent of intentional wrongdoing. Our procedures follow the usual advice to try to publicize that someone has a clean record if the investigation turns up no misconduct, but we recognize that this cannot compensate for the emotional trauma of an investigation, nor can we make up for the lost research of the accused, interrupted both by having to respond to the investigation, and by the administrative actions of granting agencies, which are entitled to cut off current support and postpone consideration of pending grant applications while an investigation is continuing. Being on the receiving end of fraud charges is inevitably damaging to the morale, if not the career, of the accused.

F. The investigative committee

Who should be on an ad hoc or formal investigative committee? Some schools use a standing committee, we and others try to assemble one of persons with expertise but no conflict of interest. One school seems to rely on a committee largely of Department heads. Our experience suggests that a working committee is required. We

RESPONDING TO ALLEGATIONS OF MISCONDUCT IN THE UNIVERSITY

used senior and distinguished faculty members as far as we could, but when the grinding work of reviewing a large number of papers and correspondence was necessary, the senior faculty were too busy or committed. Critics from both university and government have commented that faculty are not trained as investigators and cannot do a good job of sifting evidence looking for fraud. That is basically an excuse for not spending the time to do the hard work. Scientists are excellent investigators; they understand the rules of research, know the shortcuts and loopholes well, and can add up numbers as well as any auditor that might be sent down from Washington. It is even possible to find statistical expertise in your own University to check on available data. The suggestion that experts from off campus should participate is often raised, and we have invited outsiders to serve on committees dealing with various ethical issues, but have been disappointed in their participation when it was time for the hard work. At times a formal investigation seems to entail a conflict of schedules more than any conflict of interest, but it does require faculty members and leadership that feel a collective responsibility for the quality of the university. The associate dean has organized and chaired ad hoc committees, but the dean's formal investigation should be chaired by such a faculty member.

Bringing an investigation of charges of misconduct to a successful and fair conclusion takes some dedication, not only from the staff and administration, but from the faculty who are essential to the process. We have discussed what sort of faculty should be on an investigative committee, but should note that there are no personal benefits to be gained from serving on such committees, and it requires substantial institutional loyalty to participate. And finally, the actual job of the committee is basically uncomfortable. The suspicions and allegations that the committee hears are unpleasant to consider; fellow-faculty should be above such things. The early or minor transgressions of the accused, typically recalled in investigations of serious misconduct, may stimulate fears that they or their trainees could have been guilty of similar actions. The fear of misjudging someone and ruining a career hangs over the process. The fear of legal missteps or challenge also inhibits active participation by some members. It may be helpful to acknowledge that most institutions have made mistakes in handling their first significant misconduct case.

Reflecting on the legal and due process aspects of a formal investigation leads to the observation that the committee appears to be serving as prosecutor as well as judge or jury. Some rules specify appointment of an investigator to present evidence to the committee, allowing it to function only judicially. This is not the model that we adopted; we viewed the investigation as a fact-finding process rather than an adversarial encounter, with the committee exercising scientific skepticism rather than legalistic judgment, and taking the initiative to uncover and analyze evidence and question witnesses. The meetings of the committee were not designed as formal hearings, though witnesses were interviewed and evidence was presented. We did not specify rules of evidence as might be desired by some, but preferred to let the scientist's judgment of reasonableness prevail. The university's procedures for prosecuting violations of the Faculty Code of Conduct contain such provisions,

RESPONDING TO ALLEGATIONS OF MISCONDUCT IN THE UNIVERSITY

but lack the flexibility to handle the wide variety of potential misconduct issues. Subsequently, the university has adopted additional procedures in compliance with the requirements of NIH/NSF. The details differ from those of the medical school, but they are compatible in general, and go further to encourage development of an ethical research environment.

V.

CONSIDERATIONS IN THE IMPLEMENTATION OF POLICY

A. Concerns of faculty and administrators that must be addressed

When a popular colleague is accused of misconduct, the initial reaction is that someone is out to get him, or he is being set up as the "fall guy" for something improper that has been done in the laboratory. When an unpopular colleague is accused, the usual reaction is to believe whatever the charges may be, perhaps with a glint of satisfaction that the individual has finally been found out. Scientists are as apt to be betrayed in their judgments by personal prejudices as anyone else. The faculty must be assured, therefore, that reviews and investigations will be carried out with the utmost objectivity, and that their personal involvement would be treated with the greatest care and tact. Co-authors of an accused or guilty individual will need advice in making appropriate responses, and may need help in recognizing their responsibility toward the collaborative work which has been questioned. The way the accuser is treated by the institution sends a strong message to others who may be aware of the improper activities but uncomfortable about coming forward.

It is traditional to bring problems first to the laboratory director or the department chair, and it should probably be part of the orientation of trainees and new faculty to reinforce this tradition. The written policies of institutions should contain reference to the steps to be taken by an individual to secure confidential advice for any level of concern, however, since circumstances may preclude the traditional approach.

It

Faculty and administrators who become involved in the evaluation of charges of misconduct may find themselves threatened by lawsuits, or may be almost as uncomfortable in anticipation. is customary for institutions to provide legal counsel and support for all those acting within the framework of their responsibilities, but the unpleasantness of legal challenges cannot be so easily mitigated. Those who become involved must be assured that the advice they receive will minimize their exposure, while allowing them to pursue the real issues with academic disinterestedness. Pursuit of the truth without taking account of the consequences is likely only when concerns about the reputation of the institution and its ability to secure continued research support can be relieved. Threats of sanctions by granting agencies, like threats of lawsuits, could inhibit inquiry, and promote covering up mistakes, as has been evident in the past.

Faculty and administrators must be in sympathy with the regulations or ethical standards that are at issue. Unpopular rules will be hard to enforce by the mechanism of charging misconduct.

RESPONDING TO ALLEGATIONS OF MISCONDUCT IN THE UNIVERSITY

The recent controversy at MIT over published work that contained subsequently acknowledged errors illustrates that when there is no consensus about ethical requirements for action (i.e., retraction of minor components of a paper), then academics will disagree on whether failure to do so constitutes misconduct. This argues for the utility of open debate and discussion about what constitutes good scientific ethics, to enhance consensus and good practice. B. Interests of granting agencies, government, and the public

Accountability closely follows support. Science has received immense support, financial and moral, from private and governmental agencies, with the enthusiastic interest of the taxpayer. This is most marked in the area that has been highlighted in discussions of fraud and misconduct, biomedical research. Those who have provided this support have a legitimate interest in the quality of research and the occurrence of misconduct. The media, doing their traditional job of trying to uncover whatever has been covered up, function as a nagging conscience, though their occasional sensationalism and inaccuracies damage those who are involved. While institutions responding to charges of misconduct would prefer to lock the doors and work everything out fairly but privately, the reality of accountability precludes this approach, so that guidelines and procedures must take into account the relationship of the institution to its sponsors, and provide for responsible reporting at appropriate times.

VI. THE EVOLUTION OF POLICIES AND PROCEDURES

A. Institutional reactions to the procedural issues

"Policies for responding to allegations of fraud in research" is a compendium and commentary by P.J.Greene and colleagues, published in Minerva (Vol. 23, No. 2) in 1985, based on a survey of 747 academic institutions and hospitals. One hundred sixteen of the 493 respondents had written policies which were summarized in the paper. Greene identified eight issues about which there was considerable difference of opinion among the institutions. These are worth comment in the light of current ideas. .cp4

1. Definition of fraud in research: the issue was whether the distinction between fraud and error or bad technique is really straightforward or not and can be spelled out in a set of rules. My opinion is that intent to deceive others can be hard to prove.

2. Desirability of having a policy: this may have been debated in 1985, but is moot now, in view of federal requirements. About half the institutions with written procedures used the AAMC report for a starting point.

3. Distribution of statement of policy: it is curious that some were satisfied that if policies existed there was no need to disseminate them, and others thought that unwritten consensual rules were sufficient. We published our procedures in the medical school newspaper when they were completed, but have failed to disseminate them again at intervals. It would be useful to include them with

RESPONDING TO ALLEGATIONS OF MISCONDUCT IN THE UNIVERSITY

the other materials given new faculty, such as the faculty code of conduct, recognizing that it is very difficult to get faculty to read regulations. This communication problem deserves further consideration, especially since promulgation of research conduct guidelines is increasingly likely.

4. Participants: whether individuals close to the laboratory of the accused or the department chair should participate as members of review committees was still controversial in 1985. It is not appropriate for any close colleagues or former collaborators to belong to the investigative committees; they should all be given adequate opportunity to present their opinions to the faculty committee as witnesses. We have no written requirement for outside members, and our difficulty in getting their full participation has since justified this position.

5. Notification: this appears much less controversial now. The standard is that notification of appropriate agencies and journals should occur when a formal investigation is initiated, and again when it is completed (with the results). NIH and NSF require even earlier notification if circumstances warrant it.

6.

Publicity: it is not surprising that some felt an obligation to notify the public while others were less inclined to do so, with a view toward legal liability. Our procedures do not address this issue; our policy has been to acknowledge that an investigation was in progress if asked by the press, but not to volunteer additional information until the investigation was completed. We would feel obliged to announce the results of a formal investigation in any case. We cannot, of course, prevent committee members or other faculty from speaking unofficially, but most have been quite happy to accept our recommendation to refer all questions from the press to the associate dean or public information officer.

7. Legal issues: these included matters such as whether an accusation should be accepted unless put in writing and signed; the most diligent approach today, as urged in the AAU report, is to take notice of and investigate any accusation of serious misconduct, even if it is not formally delivered.

The right to counsel was not uniformly granted, though the great majority allowed at least the accused to be accompanied by counsel at all hearings. The extent of due process provided for in written procedures varied widely; in our attempt to keep the UCSD document short and readable we specified less than would be desirable by Mishkin's standards. We failed to define rules for proper conduct of research, which she points out should be the first requirement of due process, to serve notice to those who might become liable to charges of misconduct. We have also not provided for cross-examination of witnesses in committee hearings nor provided for the creation of a written transcript which could be provided to the accused. The hearing and appeal procedures under the Faculty Code of Conduct do provide these quasi-legal procedural safeguards, however, and would be available to an accused faculty member who wished to appeal the decision of a medical school investigative committee.