The Division of Investigative Oversight (DIO), ORI, notes problem areas that are interfering with the effective handling of research misconduct allegations:
1. The initial allegations are handled by the laboratory director rather than immediately informing the Research Integrity Officer (RIO).
This can be a serious problem, since it typically results in improper sequestration, and often leads to intemperate action by the lab chief, such as firing an individual without providing due process. Admissions gained at this stage are usually inadequate, as they may be verbal, not witnessed, or lack an explicit admission of falsification or fabrication of data, and may have been coerced under conditions where enormous and inappropriate pressure has been placed on the suspect individual. The best way to deal with this problem is frequent and widespread circulation of information to faculty about proper ways of handling allegations of misconduct and other compliance matters. Having a prominent link on an institution’s web page so that all are constantly reminded of a policy at the institutional level would be very helpful.
2. Sequestration of research records is not conducted in a timely way or is inadequate.
Prompt notification of the RIO of allegations is vital, but in turn, the RIO must act in a careful way to assess the allegations, and, if they warrant action, assemble a team and promptly sequester the relevant records. Taking too long to do this, or talking to witnesses, including the probable respondent, prior to sequestration, can allow guilty parties to alter or destroy critical raw data and other records. Everyone is best served if no one can subsequently claim that so-and-so could have altered or discarded records.
3. Inquiry and investigation committees lack appropriate expertise.
This issue can be due to many factors. When institutions have large standing committees to make findings of misconduct and do not rely on ad hoc experts for critical analyses, their determinations are often based on a lack of understanding of the allegations and the various testimony and analyses provided by the witnesses. Even when ad hoc committees are used, the lack of appropriate expertise can lead to a failure to adequately identify and review the raw data, so that conclusions are based on an incomplete analysis. Such shortcomings translate into poor or no analyses, inadequate documentation of the institution’s findings in its reports, poor judgments, and wrong conclusions.
4. Interviews are poorly conducted and annotated when discussing exhibits. Many are not recorded or transcribed.
The new regulation, 42 C.F.R. part 93, recommends that interviews be recorded during the inquiry and requires that this be done for the investigation. Transcripts should be returned to the witnesses for corrections, since even court reporters rarely know all of the scientific vocabulary used during the interview. Even when recorded, many interviews are virtually worthless because committee members have little or no experience and have been given no guidance on how to conduct an interview. Consequently, committee members constantly talk over one another and interrupt a line of questioning before the witness has a chance to answer crucial questions - such as “Did you falsify this blot?” In addition, unless each object examined and discussed in an interview is explicitly described and tagged during the interview, any discussion will be of no value when read later by others.
5. Committees and institutions are intimidated by aggressive outside counsel.
Officials, committee members, and even the counsels at many institutions are unprepared for the extremely aggressive and obstructive tactics employed by some respondents. As a result, excessive delays are often agreed to and the scope and ultimate resolution of the investigation may become compromised. Although it is obviously important to provide due process to the respondent and other witnesses, a well written institutional policy and procedure, if carefully followed, can help enormously in dealing with a constant stream of letters from the respondent’s attorney. In addition, ORI attorneys in the Office of General Counsel are always available for advice in these matters.
6. Institutions often intentionally do not follow up on new allegations that are uncovered during the inquiry or investigation.
At a number of institutions, the inquiry and/or the investigation committee has been specifically instructed to limit its review to the original allegation(s) and not to pursue any new concerns that arise during the inquiry or investigation. This is contrary to the requirements of the regulation and compromises the ability of the institution to identify the duration and scope of the possible misconduct. It also may fail to identify papers and grant applications containing false or fabricated data.
7. Institutions often provide little or no support to whistleblowers.
Complainants/whistleblowers are often treated harshly by others in their lab, if their identity becomes known, as it nearly always is, and whistleblower protection is very spotty. Admittedly, some whistleblowers have issues that go beyond research misconduct allegations, or their dogged persistence may be hard to endure, but often this behavior is brought on or exacerbated by reluctance on the part of institutional officials to act. Institutions should provide a support mechanism for complainants so they do not feel as isolated as they often do at present, including facilitating their relocation to other laboratories if it is clearly not possible to continue comfortably in their current lab or if their research project has been totally compromised.
8. Inquiry and investigation reports are poorly documented, failing to provide either a meaningful summary of the evidence or a rationale for the findings.
It is pretty clear that, particularly for institutions with no prior experience with ORI, committees and officials expect DIO’s oversight to be perfunctory, and their reports frequently do not meet the guidelines in the regulation or those routinely transmitted to the RIO when acknowledging the institution’s notification of ORI of its progressing to an investigation. When such reports are received, it is generally not possible to define PHS issues or to evaluate the evidence and basis for the institution’s findings even when the issues are adequately defined. This generally leads to substantial delay because DIO will request that the report be redone or will request additional records and analyses from the institution so that DIO can better determine the scope and significance of the case.
9. Institutions do not handle admissions well.
ORI has learned from many cases to treat so-called admissions with caution. Many are legally insufficient, lack specificity and fail to admit to falsification or fabrication of data. Others are insufficient because they were either not written, or were not signed in the presence of witnesses, and thus are easily retracted. Lastly, many are made by respondents anxious to limit the scope of the findings made against them, and institutions often wish to minimize their effort and wish to accept the admission - generally accepting statements that nothing else was falsified or fabricated - and do not independently review the rest of the respondent’s research for evidence of a broader set of misconduct issues.
10. Many institutions are lax in following NIH guidelines on retaining research records.
Historically, one of ORI’s most enduring problems is the lack of primary data. This can result from several causes, including the respondent becoming aware of concerns before being confronted and having the data sequestered, allowing for the destruction of incriminating notebook pages, raw data, and computer files. However, it is also clear that in many laboratories the lab chief does not routinely review the notebooks of his students and fellows, has not promulgated guidelines for recording and retaining data, and is unaware or does not care about the fact that the university owns the data, or at least the data paid for by PHS funding. In addition, many lab chiefs do not object to their students or fellows taking their notebooks and data with them when they leave the lab, even though NIH guidelines require that these records be maintained in the institution receiving the grant(s).
11. Institutions could do a better job of providing guidelines to senior staff on their mentoring skills and responsibilities.
DIO believes that the single most effective tool in reducing the level of research misconduct by junior scientists is through better mentoring and monitoring of students and fellows. Mentoring - leading by example - is crucial for instilling in nascent scientists a code of behavior and good research practices. Monitoring is important to ensure that students’ and fellows’ notebooks are properly maintained and that the raw data match the textual claims, tables, and figures presented to the laboratory and outside scientific community in the form of talks, posters, grant applications and papers. Monitoring may seem onerous, but senior scientists who have been victimized by dishonest junior lab members swear by it.
12. Many institutions have inadequate and out-of-date policies and procedures; student disciplinary procedures are often inappropriately applied when students are accused of research misconduct in PHS funded research.
Institutional policies and procedures vary widely in terms of their adequacy in how to properly deal with allegations of research misconduct and in their adherence to the appropriate Federal regulations. Most have not yet been updated to take into account the significant changes made in the June 2005 regulation (42 C.F.R. § 93). To the extent that an institution’s policies and procedures are inadequate or fail to adequately address the requirements of the new regulation, the institution becomes vulnerable to being manipulated and frightened by aggressive attorneys working on behalf of the respondent. DIO also notes that even when an institution’s policies and procedures are adequate, it remains necessary for institution officials and committee members to refer to those policies and act according to their guidelines.
13. IRB committees are often given inadequate guidance and training in recognizing research misconduct issues, in contrast to protocol violations and other matters they are expected to handle.
In the arena of clinical research, the role of the IRB in approving protocols and ensuring that research subjects are treated appropriately is widely appreciated. However, members of an IRB are generally ill equipped to recognize or handle allegations of research misconduct. Thus, it is important that IRB members be both trained to recognize the difference between research misconduct and the issues they normally deal with, and to refer such matters immediately to appropriate institutional officials, rather than trying to handle the misconduct matters directly.