Congressional Activities Regarding Misconduct and Integrity in Science
Barry D. Gold
In 1980, four cases of alleged fraud and misconduct in science involving scientists at top research centers, all in the area of biomedical research, captured headlines in the scientific and popular press. The cases included Elias Alsabti, an Iraqi who fabricated his medical and research credentials and who was alleged to have plagiarized almost 60 scientific papers in cancer immunology while working at a half-dozen research centers in the United States;1 Marc Strauss, an oncologist at Boston University, alleged to have falsified patient records to make patients eligible for clinical trials;2 Vijay Soman, an endocrinologist at Yale Medical School, alleged to have fudged and fabricated data and plagiarized a rival's paper;3 and John Long, a cancer researcher at Massachusetts General Hospital, alleged to have fabricated and falsified data and mislabeled the cell line he was studying.4
These cases raised concerns over the capability of existing mechanisms to handle allegations of misconduct in science and maintain integrity in the research environment. Concerns about misconduct in science persisted throughout the 1980s as more allegations were made public and as reports of poor handling of allegations continued to surface. Most of the congressional attention on this subject focused on the role of the National Institutes of Health (NIH) and the National Science Foundation (NSF) to ensure that allegations of misconduct are handled properly and that the integrity of federally funded research is maintained.
Barry D. Gold is a professional staff member for the Subcommittee on Science of the Committee on Science, Space, and Technology of the U.S. House of Representatives. He was the staff officer for the Panel on Scientific Responsibility and the Conduct of Research.
House Science and Technology Committee
In 1981, then Representative Albert Gore, Jr. (D-TN), chairman of the Investigations and Oversight Subcommittee of the House Science and Technology Committee, held oversight hearings entitled "Fraud in Biomedical Research."5 The hearings were motivated by the four widely publicized cases mentioned above, the issuing by NIH of disbarment regulations, and the interest of Representative Gore in ethical issues in science.6 The witnesses were primarily senior spokesmen for science (e.g., Philip Handler, president, National Academy of Sciences [NAS]), but also included one researcher (John Long, Massachusetts General Hospital) who had admitted to having fabricated data on the size of the Hodgkin's immune complex and one scientist (Philip Felig, Yale University) whose career had been affected by a case of alleged misconduct concerning one of his students (Vijay Soman).
These hearings were a subset of ongoing congressional activities designed to examine ethical and institutional issues in science, primarily biomedical research. Other hearings focused on ethical issues surrounding recombinant DNA research.
Most of the witnesses at the Gore hearing held that the problem of fraud in scientific research was not significant and that adequate mechanisms existed within the scientific community to handle these cases. For example, NAS President Handler testified that misconduct in science was "grossly exaggerated" by the press and that it is not a problem when it does happen because "it occurs in a system that operates in an effective, democratic, and self-correcting mode" that makes detection inevitable.7
According to journalists Broad and Wade,8 the scientists who had been called as witnesses and the congressmen presiding at the hearing held strongly divergent views about the nature and seriousness of the problem. They reported that "Gore and his fellow Congressmen were moved to visible amazement and then anger at the attitudes of the senior scientists they had called as witnesses." Representative Gore typified the perspective of his colleagues when he said that "one reason for the persistence of this type of problem is the reluctance of people high in the science field to take these matters very seriously." Jim Jensen, formerly a professional staff member for the committee, stated that committee members expressed concern that the universities and responsible federal agencies were not prepared to handle these cases. In addition, some committee members suggested that scientists convicted
of fraud or misconduct in science should be treated like any other criminal and be put in jail and/or asked to repay public funds.9
In discussing the divergent views between the congressmen and senior spokesmen of science, Broad and Wade identified what they considered to be contributing factors: (1) Congress had been forced into its own self-examination by the Abscam bribery case, (2) the congressmen viewed the scientists as a "fellow group of professionals who apparently preferred to deny a problem existed rather than face up to it," and (3) the congressmen understood that "no matter how small the percentage of scientists who might be fakers of data, it required only one case to surface every few months or so for the public credibility of science to be severely damaged."
The House Science and Technology Committee, while feeling that the scientific community had to do a better job on this issue, decided to adopt a "wait and see" attitude. This decision was due in part to the testimony given by William Raub of NIH regarding actions then being initiated, in part to the fact that the House Science and Technology Committee has no real legislative jurisdiction over NIH, and in part to the concentration of cases in biomedical research.10 It would be eight years before the oversight subcommittee would hold a second set of hearings. In the intervening years the committee and its staff played a role in the development and passage of the Health Research Extension Act and participated in investigations of individual cases and their resolution, Representative Gore was elected to the Senate, and many of the original staff members assumed other positions.11
Senate Labor and Human Resources Committee
Congressional activity in 1981 was not confined to the House of Representatives. The Senate Labor and Human Resources Committee held oversight hearings on the National Cancer Institute (NCI). During these hearings, Chairman Orrin Hatch (R-UT) asked why the NCI had failed to place a hold on federal funds to a researcher alleged to have falsified data. The issue that emerged at the hearings was whether or not to suspend the funding of a researcher accused of fraud or misconduct in science while the investigation of the charge was pending.12 [Note: There are now regulations in place that require suspensions under certain conditions.13]
EARLY LEGISLATIVE AND REGULATORY EFFORTS
The first attempt by Congress to address fraud and misconduct in science was the Health Research Extension Act of 1983. Section 485 of this act included language aimed at ensuring that allegations of fraud and misconduct in science were handled effectively.14 The bill containing these provisions was passed by Congress but was vetoed by President Reagan in 1984.
Essentially the same language was included in P.L. 99-158, the Health Research Extension Act of 1985.15 Section 493 states:
(a) The Secretary shall by regulation require that each entity … submit … assurances satisfactory to the Secretary that such entity—(1) has established (in accordance with regulations that the Secretary shall prescribe) an administrative process to review reports of scientific fraud in connection with biomedical and or behavioral research at or sponsored by such entity; and (2) will report to the Secretary any investigation of alleged scientific fraud which appears substantial.
This bill was passed but the president vetoed it, stating that the bill "manifest[ed] an effort to exert undue political control over decisions regarding scientific research."16 Congress passed the bill over the president's veto in November 1985. It represents the first time Congress formally directed any federal agency to require that universities and other recipients of NIH research funds have systems in place for handling allegations of misconduct and to report the outcomes of these deliberations to NIH in a timely fashion. This statute required the promulgation of regulations that assign the primary responsibility for responding to allegations of fraud and misconduct in science to the local institution and required a general oversight role for NIH. The institutional assurances requirement draws heavily on the model established in the 1970s with the creation of Institutional Review Boards (IRBs) at local institutions.
Institutional Review Boards were created in response to a series of widely publicized cases of abuses in human subjects research. The Congress adopted legislation in the early 1970s that contained new requirements for grantee institutions sponsoring federally funded human subjects research. Under the terms of this legislation, research institutions must certify to the funding agency that they have established a local IRB that monitors and approves the protocol for human subjects research and that operates in a manner that is consistent with federal guidelines in this area.
The IRB experience appears to have reconciled the needs of local institutions for maintaining flexibility and some autonomy within a
general framework of government standards. Institutional assurances on animal welfare were also implemented in 1986 by the Public Health Service (PHS). These two experiences with research regulation seem to have influenced the early stages of legislative deliberations in establishing guidelines for handling cases of misconduct in science.
The regulations required by the Health Research Extension Act of 1985 were first listed as part of NIH's regulatory agenda in April 1986. Preliminary guidelines were issued in the NIH Guide for Grants and Contracts in July 1986, and the notice of proposed rulemaking was issued in September 1988. The final rule was promulgated on August 8, 1989. Although NIH officials testified in hearings about why it took so long to produce this rule, the length of time contributed to the congressional perception that NIH was not seriously addressing the issue of misconduct in science. Another regulatory step by NIH was the improvement of the ALERT system, a computer-assisted system for tracking individuals who are under investigation or who have been sanctioned for misconduct. Initiated by NIH in 1981, the ALERT system was improved and expanded to include all PHS activities.17
THE NEXT ROUND
House Judiciary Committee
On February 26, 1986, the Subcommittee on Civil and Constitutional Rights of the House Judiciary Committee held hearings to examine the role of libel laws in discouraging publication of reports or allegations of misconduct in science.18 Walter Stewart and Ned Feder of NIH testified before the subcommittee that the threat of libel against journal editors was a major factor in preventing publication of their article on misconduct in science and the quality of the scientific literature. The article resulted from an analysis of papers coauthored by John Darsee, which included 18 scientific papers, 88 abstracts, and three book chapters.19 The Darsee case received wide press coverage in 1983 and involved the fabrication of data while he was a cardiology researcher at both Emory University and Harvard University. Stewart and Feder's study examined the extent to which Darsee's coauthors, journal editors, and referees could have known, or should have known, that his reported research findings were fabricated and inaccurate. Stewart and Feder contended that Darsee was able to publish fabricated material because his coauthors and the peer review system failed to provide appropriate vigilance in examining the data.
The much-disputed conclusions of the Stewart and Feder article are that misconduct among biomedical scientists may be more prevalent than commonly believed and that traditional quality control mechanisms may not work as well as expected. In discussing reviewers' comments on the article, Benjamin Lewin, editor of Cell, stated, "Although they [the reviewers] had criticisms of the article, the overwhelming consensus was that the article says something important about the way science is done, and that it would be in the public interest to have it published."20
House Committee on Science and Technology Task Force on Science Policy
A few months later, on May 14, 1986, Stewart and Feder were among a series of witnesses at hearings on research and publication practices convened by the Task Force on Science Policy of the House Committee on Science and Technology.21 Witnesses at the hearing addressed two issues: (1) "whether current policies allow for the most efficient expenditure of resources for biomedical research" and (2) "the way scientific research is currently conducted and communicated in America."
Stewart and Feder submitted a draft of their still unpublished paper for the hearing record. (Their paper was widely circulated prior to publication, and a revised version of the paper was published in Nature on January 15, 1987.) In describing their study, they said it was an attempt "to measure the frequency of misconduct in a large sample of scientists." They then went on to describe the study and their findings regarding types and prevalence of misconduct among this sample, possible causes, implications for the larger scientific community, the difficulties they encountered in getting their work published, and a recommendation that the science community "start considering solutions to the problem" of the validity of published research and the accuracy of the scientific literature.
In their testimony, Stewart and Feder expressed their belief that these problems had existed for a long time, and they questioned the adequacy of existing institutional mechanisms for preventing and/or responding to the types of alleged misconduct described in their paper. They reported that, on the basis of discussions with other scientists, (1) their estimate of the prevalence of misconduct was not widely off base; (2) pressure to publish may be a contributing factor to the occurrence of the alleged abuses; (3) many research papers are likely never to be read; and (4) they had not investigated what the costs to science might be of researchers trying to replicate or build on falsified results. The other
witnesses testifying at the hearing were Sandra Panem on the efficient use of biomedical research funds and Patricia Woolf, who had been asked to discuss the issues raised by Stewart and Feder.
During her testimony, Woolf stated that Stewart and Feder had raised serious questions concerning misconduct in science. She voiced strong concerns that the normal process of peer review of their paper had been circumvented by threats of libel. In discussing the paper itself, she indicated that it did not answer the question of the prevalence of fraud and misconduct in science, but that the scientific community had in any event "set itself the task of examining and improving its quality assurance programs."22 She suggested that many scientists view publication as a device primarily for professional advancement (i.e., promotion and tenure and successful grantsmanship), as opposed to the communication of results.
Woolf supported the view that the appropriate strategy for maintaining quality in science is to strengthen existing internal review mechanisms in science rather than developing governmental regulations. She indicated that while the peer review system works reasonably well, there is room for improvement and that some efforts are under way to improve it. One suggestion she offered was to provide professional recognition or credit to scientists for the amount of time spent reviewing manuscripts.
INTEREST BROADENS IN THE HOUSE
In 1987, five cases of fraud and misconduct in science, primarily in biomedical research, were reported in the scientific and popular press, and in one instance the supervisor of the individual charged with misconduct committed suicide. With increasing reports that a number of universities and NIH had stumbled in handling allegations of fraud and misconduct, congressional interest reached a new high. Three separate hearings were convened during 1988, two by the House Committee on Government Operations and one by the House Committee on Energy and Commerce, which has legislative and authorizing jurisdiction for NIH.
Witnesses and congressional members participating in the hearings suggested that the science community, in spite of claims to the contrary, had not demonstrated an ability to handle this issue through existing self-regulatory mechanisms, and that government regulation may be necessary. As a result of these hearings, Congress delivered a strong message of displeasure to the scientific community concerning its handling of allegations of fraud and misconduct in science.
House Committee on Government Operations
On April 11, 1988, the Subcommittee on Human Resources and Intergovernmental Relations of the Committee on Government Operations, chaired by Representative Ted Weiss (D-NY), held oversight hearings titled ''Scientific Fraud and Misconduct and the Federal Response."23 This subcommittee has oversight jurisdiction and investigatory authority, but no legislative authority, over the Department of Health and Human Services (DHHS) and other executive branch agencies. The hearings focused on the experience of NIH in handling allegations of fraud and misconduct in science, the treatment of whistle-blowers, and public health and safety concerns that could arise as a result of the publication of fraudulent research data.
During these hearings, Representative John Conyers, Jr. (D-MI), in discussing misconduct in science in the context of white-collar crime, suggested that a federal statute could make fraud or misconduct in science a criminal offense. Witnesses, including two whistle-blowers, dismissed this suggestion as overly harsh and having the potential to "create a damaging climate of fear among scientists."24 Representative Weiss expressed his view that the scientific community, including NIH, the universities, and the professional societies, had shown a general unwillingness to guard against fraud and misconduct in science.
House Committee on Energy and Commerce
The next day, on April 12, 1988, the Committee on Energy and Commerce, chaired by Representative John Dingell, Jr. (D-MI), held oversight hearings titled "Fraud in NIH Grant Programs."25 These hearings called into question NIH performance in handling allegations of fraud and misconduct in science. Representative Dingell contended that NIH had inadequately addressed due process rights and the protection of whistle-blowers in these cases. The hearings focused on the NIH investigation of misconduct charges involving a 1986 Cell paper coauthored by David Weaver, Thereza Imanishi-Kari, David Baltimore, and others.
Stewart and Feder were the first witnesses at the hearing. They presented an analysis of five cases of misconduct in science, including allegations concerning the data in the Cell paper. Stewart and Feder subsequently became advisors to the subcommittee in June 1988 regarding the Cell paper and other cases of alleged misconduct. Margot O'Toole, who raised the initial allegations concerning the Cell paper, testified at the hearings not only about the substance of these allegations
and the process by which they were examined, but also about the subsequent difficulties she faced in securing a research position. One issue raised in the Dingell hearings was whether this case represented misconduct in the reporting of experimental results or a professional disagreement in interpreting data that were breaking new ground.
Also at issue, and of prime importance to Representative Dingell and the subcommittee, was the fate suffered by young scientists who act as whistle-blowers. Dingell's committee had previously drafted the Whistle-blower Protection Act, designed to protect federal employees who disclose waste, fraud, and abuse in government programs. This act was signed into law in 1989.
Although the Whistle-blower Protection Act would not have applied to O'Toole, two witnesses in a June 1989 hearing held by Representative Weiss (see below) suggested that Congress should extend the Whistle-blower Protection Act to university faculty and research staff who are federal grantees. In responding to a letter from Weiss following these hearings, NIH Acting Director Raub wrote that although constrained by the administration's position, "we are very sympathetic toward efforts that would provide appropriate protection to whistle-blowers. The applicability of the provisions of this [Whistle-blower Protection] Act to the scientific/academic community certainly warrants careful consideration."26
The subcommittee members were displeased with the investigations by Massachusetts Institute of Technology, Tufts University, and NIH of the allegations of error in the Cell paper. University officials testified that since O'Toole's initial allegation did not include charges of misconduct, they relied on informal methods of resolving professional disagreements. They indicated that, had a charge of misconduct been made, a more formal investigation would have been provided under a different set of policies and procedures. This distinction drew the interest of the subcommittee. The institutional representatives commented that while the procedures for conducting investigations could use some fine-tuning, they were learning from each case they handled.
House Committee on Government Operations
On September 29, 1988, Representative Weiss held another set of hearings before the Committee on Government Operations to explore (1) the issue of conflict of interest in U.S. academic research, (2) the effectiveness of local institutions in preventing and investigating instances of fraud and misconduct in science, and (3) whether or not NIH needs to do more to prevent and investigate instances of fraud and
misconduct in science. The hearing focused on conflict of interest issues involving commercial ties by university scientists. In one case it was suggested that commercial incentives may be an important factor motivating scientists to engage in research misconduct.27
On the topic of criminal penalties for scientists found guilty of allegations of fraud and misconduct in science brought up at the earlier Government Operations Committee hearing, Representative Conyers stated that he had "received 2,500 letters from those who supported criminal penalties since our hearings some months ago [April 11, 1988] for withholding information, falsification, [and] misleading research."28
During these hearings, the Office of Inspector General (OIG) of the Department of Health and Human Services presented a draft report on misconduct in scientific research.29 The OIG report included the results of its recent study of (1) the extent to which NIH and its grantee institutions have developed policies and procedures to prevent, detect, and handle scientific misconduct cases and (2) what selected grantee institutions have learned and what changes they have made as a result of their experience in conducting investigations of allegations in misconduct. The OIG report was based on a telephone survey of a random sample of FY 1986 NIH grantee institutions and site visits to nine grantee institutions that had experience with cases of misconduct in science.
DHHS Inspector General Richard Kusserow noted the lack of a central locus of responsibility at NIH for matters of misconduct in science. He stated that 93 percent of the grantee institutions with 100 or more awards have policies and procedures in place to handle allegations of misconduct in science, but that overall only 22 percent of PHS grantee institutions have adopted such policies. [Note: NIH later pointed out that the 22 percent of PHS grantee institutions with policies in place account for 88 per cent of all PHS award dollars.] He identified weaknesses in the misconduct procedures that are in place, often including a failure by research institutions to notify NIH and keep sponsoring officials fully informed of the status or results of a misconduct investigation.
On the question of prevalence, the OIG report contained a strong disclaimer against using their figures as an estimate of the actual prevalence of misconduct in science. The report stated that 36 percent (17 of 47) of the grantee institutions with established procedures reported allegations of misconduct that required their use. Sixteen of the 34 cases (47 percent) investigated by these institutions were substantiated. Based on these figures, the OIG estimated that 95 cases of misconduct in science (47 substantiated and 48 unsubstantiated) had been addressed by PHS grantee institutions. The report was not explicit
regarding the time period over which these cases occurred, but it would appear to be from 1982 to 1988. The OIG report identified an interest among grantee institutions in receiving guidance in developing guidelines for preventing misconduct in science.
Based on these findings, the DHHS inspector general recommended that the secretary of DHHS establish investigatory and oversight functions independent of the research funding agencies and develop a more formal process to deal with cases of alleged misconduct in science. Other recommendations included additional notification requirements for the awardee institutions and the development of alternative methods of detecting possible misconduct, including spot audits of scientific data and specific reviews by editors of scientific journals.
The OIG report also noted that there was no central locus of responsibility or accountability for scientific misconduct within DHHS. While OIG has responsibility to prevent and detect fraud and abuse in DHHS programs and operations, such as Medicare and Medicaid payments, the responsibility for handling reports of misconduct in research programs resided within different components of the Public Health Service. The OIG report recommended that the secretary of DHHS develop a formal centralized process to deal with scientific misconduct.
In his testimony, James Wyngaarden, director, NIH, noted that he had not had an opportunity to review the draft OIG report prior to its release at the hearing. He departed from his prepared testimony to comment critically on the confrontational nature of the OIG report. He indicated that NIH had assumed an active role in addressing issues of misconduct in science. He strongly opposed the recommendation that a body outside of NIH be established to conduct investigations of allegations of misconduct in science.
Rumors of Legislative Activity
Following these hearings, rumors of legislative action abounded. One draft bill was prepared and circulated by Representatives Dingell and Henry Waxman (D-CA) as an amendment to the NIH reauthorization bill, but it was never formally introduced. The amendment would have established a new office of scientific integrity in DHHS. The office would have been responsible for coordinating investigations of allegations of fraud and misconduct involving PHS funds and would have transferred the current monitoring activity from NIH to the OIG or the assistant secretary for health. In addition, the office would have been authorized to conduct random audits, develop
standards of research conduct, and develop retraction guidelines for scientific journals.
In describing this proposed bill, Robert Rosenzweig, president of the Association of American Universities (AAU), said, "To put it bluntly, its passage would have been a calamity for the science community."30 He went on to state, "The powers of this new directorate would substantially reverse the presumption of the 1985 legislation, which recognized an institution's responsibility for the conduct of its members and placed the government in a monitoring role. If enacted as proposed, the new office would have been empowered to receive allegations of misconduct and, if it chose, to investigate them independently of institutional processes."
On March 16, 1989, Assistant Secretary for Health James Mason established two new offices: the Office of Scientific Integrity (OSI) in NIH and the Office of Scientific Integrity Review (OSIR) in his office.31
FEDERAL AGENCIES TAKE ACTION
Public Health Service
In 1987, the PHS drafted a notice of proposed rulemaking (NPRM) on misconduct in science. Following the 1988 hearing held by Representative Dingell, DHHS officials negotiated changes in the NPRM with the Office of Management and Budget (OMB), and a final rule dealing with misconduct in science was published on August 8, 1989.32 The final rule required each institution applying for PHS research funds to certify by January 1, 1990, that it had adopted satisfactory misconduct procedures. The applicant institutions are required to keep these policies current and to provide copies to authorized DHHS officials upon request.
Under the final rule, universities and other research institutions have the primary responsibility for preventing, investigating, reporting, and resolving allegations of scientific misconduct. The new rule does not address issues of the responsible conduct of research and fostering integrity in science (i.e., data sharing and retention, authorship practices, and so on). These issues were addressed in a separate advance notice of proposed rulemaking (ANPRM) issued September 19, 1988, to solicit comments. The ANPRM is a mechanism that a government agency can use to solicit comments and ideas regarding a possible rule without committing to producing final regulations. Some items contained in the ANPRM have subsequently been adopted by DHHS administrative actions.
National Science Foundation
On February 10, 1987, the National Science Foundation published proposed rules for handling allegations of fraud and misconduct in science.33 These rules were subsequently revised and issued as final regulations in July 1987.34 NSF based these regulations on informal policies that had been developed to deal with previous allegations and the 1986 policy guidelines of the NIH. The NSF final rule addresses the definition of fraud and misconduct, roles and responsibilities and procedures for detecting, investigating, and adjudicating allegations of misconduct, remedies and sanctions, and due process rights of the individuals and institutions involved in misconduct cases. The awardee institutions bear the primary responsibility for handling allegations of misconduct.
House Committee on Energy and Commerce
On May 4th and 9th, 1989, Representative Dingell held additional hearings on scientific misconduct before the Committee on Energy and Commerce.35 These hearings reopened questions about the NIH investigation of charges of misconduct stemming from the paper published in Cell by Weaver, Imanishi-Kari, Baltimore, and others. Two panels of witnesses appeared during the first day of hearings. The first panel represented the NIH institutional officers who had conducted the investigation into allegations of wrongdoing, members of the NIH investigatory panel, and Secret Service forensics experts who had examined the laboratory notebooks of Imanishi-Kari. The Secret Service forensics experts presented testimony claiming that pages of Imanishi-Kari's data were prepared out of noted chronological order. Although these data were not cited in the published paper, some of them—including data that the Secret Service experts claimed were produced after publication—were cited in the NIH investigation. Her colleagues largely dismissed the charge as "sloppiness." Noting the potential usefulness of forensic evidence in this case, NIH Director Wyngaarden promised to have a forensics expert included in each NIH investigation.
The second panel was composed of the coauthors of the disputed paper. The hearings were antagonistic, concluding with a series of charges and countercharges between Dingell and Baltimore. O'Toole's
conduct as a whistle-blower became an issue. Baltimore characterized her actions as initially "rational and appropriate." Baltimore was surprised, however, that "having failed to get the desired outcome from two peer reviews, O'Toole would continue to press her case in additional and less conventional forums." O'Toole stated that the "thing is not that the mistake happened, the thing that astonished me was the universal attitude that it didn't have to be corrected." In his summary remarks, Representative Dingell highlighted the whistle-blower's plight, suggesting that it was too "daunting" for a young investigator to allege fraud and reminding the committee that O'Toole found making allegations dangerous, even if she only alleged error.
During the second day of hearings the subcommittee heard from three panels of witnesses. The first panel consisted solely of O'Toole. She asserted that a number of experiments reported in the disputed paper had not been conducted or had been reported incorrectly. The second panel represented the committee from Tufts that had conducted an investigation of the charges by O'Toole and that had found no evidence of misconduct by Imanishi-Kari. The witnesses indicated that they viewed the dispute as involving differing interpretations of data and not misconduct. The third panel represented the committee from MIT that had investigated the allegations and also had concluded that the allegations were the result of normal disagreements over the interpretation of data.
House Committee on Government Operations
On June 13, 1989, Representative Weiss held additional oversight hearings before a subcommittee of the Committee on Government Operations, entitled "Is Science for Sale? Conflicts of Interest vs. the Public Interest."36 Representative Weiss's concerns at this hearing were twofold: first, the possibility that supplementing federal research funding with private (i.e., industry) funding could unduly influence the results of that research; and second, whether or not universities that "sold" the results of federally funded research to both domestic and foreign industries were acting in the public interest. There was general agreement that the conflict of interest issue needs to be looked at more closely and that guidelines may be needed. The issue of university-industry relationships was much more contentious, with some witnesses considering these relationships essential and some questioning whether too much was being given away, especially when taxpayer-supported research was sold to foreign companies.
House Committee on Science, Space, and Technology
On June 28, 1989, Representative Robert Roe (D-NJ) held hearings before the Committee on Science, Space, and Technology, entitled "Maintaining the Integrity of Scientific Research."37 Unlike the adversarial climate in the other hearings described above, these hearings were more cooperative in nature and comprehensive in scope. Witnesses representing the federal agencies (DHHS, NIH, and NSF), working scientists, the institutions and their administrators, journal editors, and individuals who had been involved with cases of alleged misconduct either as whistle-blowers or as investigators testified at the hearings.
In his opening remarks, Representative Roe expressed his belief that the responsibility for maintaining integrity in science belonged primarily to the science community, that it must effectively deal with issues of fraud and misconduct in science, and that Congress had a legitimate concern to see that these issues were resolved because the federal government was funding science at increasing levels. His views were shared by other members of the subcommittee who stated that they hoped the scientific community would effectively deal with issues of fraud and misconduct in science without governmental intervention.
The witnesses were organized into four panels. The first panel reviewed actions taken by the PHS and NSF to establish oversight mechanisms for misconduct in science and the NSF and NIH definitions of scientific misconduct, noting the distinction between fraud and the honest errors that often occur in the scientific research process. The second panel examined the environment within which research is conducted, stating that pressure to publish is cited as one of the leading factors contributing to misconduct in science. The panel also addressed the ways in which ethics and values are taught to students, commenting that this is not a formal process. The panel also reviewed key findings of the Institute of Medicine report The Responsible Conduct of Research in the Health Sciences (National Academy Press, Washington, D.C., 1989), which concluded that "in the long run, the quality of the research environment may be more damaged by sloppy or careless research practices and apathy [lack of attention to traditional monitoring approaches] than by incidents of research fraud or other serious scientific misconduct" (p. 21).38
The third panel addressed institutional responses to incidents of scientific misconduct and commented on the development of institutional policies and procedures for handling allegations of misconduct, experience with implementing these procedures, and the need to improve them. During these hearings one witness suggested that the
effectiveness of these procedures could be improved by providing limited immunity to those who conduct investigations of misconduct and to those who report the findings of misconduct. The fourth panel focused on the role of scientific journals and journal editors in fostering integrity in research. The panel discussed the role of the peer review process in ensuring integrity in research, the limitation of that process in uncovering fraud, and possible changes in authorship practices that might be explored. An experimental audit of papers approved for publication, to develop a better estimate of the prevalence of misconduct, was also discussed.
House Committee on Energy and Commerce
On Monday April 30, 1990, the Subcommittee on Oversight and Investigations, chaired by Representative Dingell, held hearings on allegations of financial conflicts of interest.39 The case in question involved Syed Zaki Salahuddin, a scientist in the laboratory of Robert Gallo, and Pan-Data, a Bethesda-based biomedical company that performed work for Gallo's laboratory. Salahuddin was alleged to have played a role in creating Pan-Data, to have used his position at NIH to bring business to the firm, and to have benefited financially from this arrangement, while not disclosing his relationship with the firm to officials at NIH, including Gallo. According to Dingell, of concern to the subcommittee was "whether monetary interests are undermining the academic integrity and impairing the ability of scientists to carry out health research for which they are supposedly being paid" and to see "that NIH is able to function efficiently, well, honorably and competently in the public interest."
On Monday May 14, 1990, the Subcommittee on Oversight and Investigations, chaired by Representative Dingell, held a fourth hearing on allegations of fraud and misconduct stemming from the paper published in Cell by Weaver, Imanishi-Kari, and Baltimore, among others. Two panels of witnesses presented testimony.40 The first was composed of two forensic experts from the Secret Service, and the second panel consisted solely of Suzanne Hadley, deputy director of the NIH Office of Scientific Integrity. Imanishi-Kari was invited to testify as well but declined. Four days prior to the hearing, Imanishi-Kari held a press conference, charging that the committee had failed to provide her with information about the results of the forensics examination or the purpose for the new hearings.41
The hearings sought to establish the authenticity of the dates for the counter tapes in Imanishi-Kari's notebooks. Secret Service investigators
reported that along with a subcommittee staff member they had traveled to MIT in 1990 to identify the gamma counters used by Imanishi-Kari during 1984 and 1985 and to obtain samples of counter tapes from the laboratory notebooks of other investigators who used these same machines during the period of time in question. They then developed an identity relating the counter number to the date of use of the machine and established a relationship between the fading of type density prior to ribbon replacement. Based on these identities they reviewed the counter tapes in Imanishi-Kari's notebooks and concluded that up to one-third of the counter tapes in the I-1 notebook had not been produced on the dates claimed. As Imanishi-Kari did not testify and as there was no one asking questions of the Secret Service investigators on her behalf, the forensic evidence was not challenged.
The second panel consisted solely of Hadley, who had no prepared statement and simply responded to questions from Representatives Ron Wyden (D-OR) and Dingell. In response to questioning from Wyden, Hadley affirmed that NIH possessed a mounting body of evidence over and above the evidence gathered by the Secret Service and the subcommittee regarding the authenticity of key sets of data associated with the paper in question. She also indicated that one research grant to Imanishi-Kari had been terminated and that NIH lacked the authority to suspend a second because it was in the middle of its funding cycle.
At the conclusion of these hearings, Representative Dingell indicated that the material gathered by the subcommittee would be referred to the OIG at DHHS and to the Office of the U.S. Attorney in Baltimore for possible criminal proceedings.
OTHER SIGNS OF LEGISLATIVE INTEREST
In addition to the hearings described above, the Congress has from time to time provided formal guidance to research agencies through quasi-legislative language included in committee reports. For example, in 1981 Senator Lawton Chiles asked NIH Director Wyngaarden to prepare a report describing NIH's handling of allegations of fraud and misconduct in science.42
The Senate Appropriations Committee also included language in the 1989 report accompanying the FY 1990 appropriations bill for DHHS. Under the heading "Misconduct in Research," the committee wrote that it "is concerned about the effects on research that may result from recent investigations into allegations of misconduct in the biosciences."43 The report went on to say, "The Committee deplores those instances where scientists have been guilty of fabrication or falsification of data, or of
plagiarism, but believes the vast majority of America's biomedical researchers are honest, dedicated, and hardworking individuals of the highest integrity.'' After mentioning the possibility for honest error, the self-correcting nature of science, and the tension between free inquiry and regulation, the report concluded: "Thus, the Committee will closely monitor the work of the new NIH Office of Scientific Integrity and the DHHS Office of Scientific Integrity Review to make certain that these agencies carry out their proper roles, but do not take actions that thwart or interfere with the continued creativity and excellence that are the hallmarks of biomedical research in this country."
In commenting on the funds provided to establish the NIH Office of Scientific Integrity (OSI), the language in the 1989 Senate Appropriations Committee report states that the mission of OSI is "to promulgate policies and procedures designed to safeguard the integrity of research conducted and supported by the NIH."44 However, the federal notice announcing the establishment of OSI and OSIR states that OSIR has responsibility for "propos[ing] policies and procedures for preventing, detecting, reporting, and handling instances of alleged or suspected misconduct in science and present[ing] such policies to the Assistant Secretary of Health for approval."45
The majority of the hearings to date have been held in the House of Representatives. The committees most interested in this issue are:
House Committee on Science, Space, and Technology, chaired by Representative Roe until 1991 and currently chaired by Representative George Brown (D-CA); Subcommittee on Investigations and Oversight, also chaired by Roe until 1991 and currently chaired by Representative Howard Wolpe (D-MI). The Science, Space, and Technology Committee has legislative and authorizing jurisdiction over a number of federal research agencies, including NSF and NASA.
House Committee on Energy and Commerce, chaired by Representative Dingell; Subcommittee on Oversight and Investigations, also chaired by Dingell. The Energy and Commerce Committee has legislative and authorizing jurisdiction over DHHS, as well as aspects of federal research in the Department of Defense and Department of Energy.
House Committee on Government Operations, chaired by Representative Conyers; Subcommittee on Human Resources and Intergovernmental Relations, chaired by Representative Weiss. The
Government Operations Committee has oversight jurisdiction and investigatory authority over all aspects of government. The Weiss subcommittee has jurisdiction over DHHS, among other executive branch agencies. The Government Operations Committee has no legislative authority.
Since the 1981 Hatch hearings, no Senate committee has undertaken formal activities in this area. The Senate Labor and Human Resources Committee, chaired by Senator Edward Kennedy (D-MA), has expressed concern over the regulatory reforms discussed in the House and the way institutions have been handling allegations of fraud and misconduct in science.
In December 1989, Representative Dingell inquired about the status of the NIH's investigation of allegations concerning Robert Gallo's discovery of the AIDS virus.46 This resulted from an extensive 16-page investigative report by John Crewdson of the Chicago Tribune.47 In that article, Crewdson suggests that NIH did not do an adequate job in investigating concerns about Gallo's discovery of the AIDS virus and that there may be substance to some of the allegations. As of February 1990, NIH had requested that the NAS and the IOM nominate a panel of qualified independent experts "to verify the independence and thoroughness of the NIH's own investigation."48 It is expected to be a number of months before this review is complete.
The Subcommittee on Human Resources and Intergovernmental Affairs of the House Committee on Government Operations, chaired by Representative Weiss, released a committee report on misconduct in science based on hearings convened by the subcommittee over the past few years.49 The report addresses conflict of interest issues, questions concerning substandard research practices and their relevance to misconduct, and the role and plight of whistle-blowers, among other issues. The findings were critical of the existing systems within government and universities for handling these issues, and the committee made the following recommendations: (1) to restrict by statute or regulation "honoraria, consulting fees, stocks/equity, and other conflicts of interest for Federal health research grantees"; (2) that NIH should more carefully enforce its guidelines and regulations on disclosure of nonfederal funding sources; (3) that "NIH should implement and enforce data-sharing requirements"; and (4) that "Congress should initiate
legislation to protect whistle-blowers and introduce penalties for grantees who cover up misconduct."
In May 1990, Representative Roe solicited comments on a draft bill that would provide limited immunity to institutions and individuals that conduct investigations of alleged misconduct and to the journals and editors that report the findings of these investigations. The bill was not formally introduced.
In the 102nd Congress, the National Institutes of Health Revitalization Amendments of 1991 (HR 2507) contain a subtitle dealing with scientific integrity. The bill would provide legislative authority for the Office of Scientific Integrity (OSI) within NIH; direct the secretary of DHHS to establish guidelines for whistle-blower protection; define scientific misconduct as seriously deviating from "standards of conduct that are recognized within the scientific community … includ[ing] fabrication, and plagiarism"; direct the General Accounting Office (GAO) to study the effectiveness of OSI; and direct the secretary of DHHS to establish guidelines for journals listed in the National Library of Medicine referencing systems for retractions and corrections. As of fall 1991, the bill had been approved by the House and was awaiting action in the Senate. Its prospects of becoming law, however, may be tied to two controversial subtitles on the bill not directly related to scientific integrity, the moratorium on fetal tissue research and federal reimbursement for university indirect costs.
ACTIVITIES OF CONGRESSIONAL SUPPORT AGENCIES
Congressional Budget Office
To the best of our knowledge, the Congressional Budget Office has conducted no studies in this area.
Congressional Research Service
In response to a request from the House Committee on Science, Space, and Technology, the Congressional Research Service (CRS) prepared a report entitled Scientific Misconduct in Academia: Efforts to Address the Issue.50 In the report, the authors review responses by the PHS, NIH, NSF, professional societies, universities and colleges, and journals to the issue of fraud and misconduct in science. They evaluate the adequacy of these efforts, discuss additional options, and describe the policy issues facing Congress as "(1) assessing the effectiveness and
efficiency of recommendations and actions and (2) considering incentives to motivate the appropriate parties to adopt and implement the most effective recommendations."
CRS has also completed a study addressing the current status of activities and efforts to address the issues of misconduct in science across all relevant federal agencies. This report, which includes a survey of federal activity in this area, has not yet been released.
General Accounting Office
The General Accounting Office (GAO) has not conducted any studies on fraud or misconduct in science. From time to time, in response to congressional requests, GAO has conducted studies of the peer review system and the distribution of federal research funds.51 As noted above, legislation pending in the House would direct GAO to study the Office of Scientific Integrity to determine its effectiveness in investigating and preventing scientific misconduct and report by the end of fiscal year 1992 to the House Energy and Commerce and the Senate Labor and Human Resources Committees.
Office of Technology Assessment
In response to a request from the Task Force on Science Policy, of the House Committee on Science and Technology, the Office of Technology Assessment (OTA) prepared a report in 1986 entitled The Regulatory Environment for Science.52 The report examines the social and legal forces acting to restrict or regulate scientific research from the postwar period to the present. The report discusses regulatory actions implemented in the 1960s and 1970s "that began to constrain not just what topics scientists should pursue, but also how they should be pursued and the results disseminated." The roles of institutions, professional societies, citizen groups, and the government, and the mechanisms they use in controlling research, are reviewed and discussed.
OTA has recently completed a study entitled Federally Funded Research: Decisions for a Decade, examining the U.S. system for supporting basic research.53 The report "analyzes what OTA identifies as four pressing challenges for the research system in the 1990s: setting priorities in funding, understanding trends in research expenditures, preparing human resources for the future research work force, and supplying appropriate data for ongoing research decisions" (p. iii). The
report signifies an interest and capacity of the Congress for addressing the contextual issues of the research system and not focusing merely on smaller segments of research policy such as misconduct.
This review of congressional activities concerning misconduct in science is characterized by the divergence in perspective between the Congress and the science community regarding the prevalence of misconduct, the competency of the science community versus Congress to judge instances of misconduct, the effectiveness of the institutions of science to implement oversight and investigative mechanisms (i.e., the ability to be self-policing), and the legitimacy of questions concerning the use and misuse of public funds.
Representatives from the research community have indicated that cases of research misconduct are extremely rare and that they usually involve isolated acts by psychologically disturbed individuals. Some witnesses have suggested that these actions may be encouraged by career advancement incentives within the research enterprise and that systemic pressures such as "pressure to publish" may require attention. Several members of Congress have expressed views that challenge these findings. It has been suggested in congressional testimony that the incidence of research misconduct is higher than has been publicly reported. It has also been suggested that whistle-blowers who expose wrongdoing in the research environment experience reprisals that subsequently inhibit the disclosure of cases of misconduct.
Representatives of the research community have indicated that since many misconduct cases involve disputes about the significance of reported research findings, individuals with scientific credentials are needed to distinguish error from misconduct and to determine appropriate research practices. While recognizing mistakes and inexperience in the early stages of institutional investigations of these cases, many witnesses have stated repeatedly that local institutions are capable of handling misconduct investigations in an appropriate manner. Based on the procedural delays, and the informality that has often accompanied local investigations, some congressional members and staff have concluded that the research community does not currently have an adequate system of safeguards to ensure against the waste of federal research funds. Congressional observers have commented that misconduct investigations conducted by research institutions do not follow basic standards of other investigations into waste, fraud, and abuse involving government funds. They have suggested that more
formal policies and procedures are needed and that such investigations require certain standards of evidence to support claims by individuals accused of misconduct.
Over the past decade exchanges between the scientific community and the Congress over questions of misconduct in science have been marked by tension over the legitimate roles and responsibilities of each group to address these issues. Similar tension has characterized other discussions of congressional oversight or regulation of scientific research.
Throughout these discussions Congress and its committees have affirmed their belief in the veracity of scientific research and the importance of the contributions of the research community to the nation. In fact, more often than not congressional committees have argued that they are supporters of science and that they have an obligation and an interest to see that the enterprise remains healthy.
Many authors have noted the role of the media in disclosing accounts of research misconduct as a key factor in bringing the misconduct issue to the attention of Congress. Another factor in the legislative review process is the growth in staff and oversight responsibilities that occurred in the Congress in the mid to late 1970s. A third factor is reports by congressional staff that they receive a constant, if somewhat low-level, stream of complaints from scientists who believe there is something wrong with the system. As long as these allegations have substance and are not seriously addressed by the institutions of science, it is likely that Congress will remain interested in the issue of misconduct. Furthermore, misconduct in science is one of several issues on the congressional agenda with regard to universities and the academic research enterprise that include indirect cost payments, facilities maintenance, conflicts of interest, and education, all of which will continue to receive increased attention.