Hearings on the falsification of biomedical research data were conducted early this spring by Congressman Albert Gore, Jr., Chairman of the Subcommittee on Investigations and Oversight of the House Committee on Science and Technology. The hearings lasted two days, March 31 and April 1, during which testimony regarding the seriousness of the problem was taken from a variety of sources, from leaders of the scientific community to persons personally involved in cases of fraud.
The first two witnesses, Dr. Donald S. Frederickson, Director of the National Institutes of Health, and Dr. Philip Handler, President of the National Academy of Sciences, argued that protective measures against fraud are inherent in the very nature of science. Excerpts from their prepared statements follow:

Dr. Frederickson:
"Our subject today is not the history of medicine or science, yet reference to the past is useful for perspective. From the beginning certain patterns inherent in science and scientists werevisible and have persisted: intense competition for priority, a high premium on originality, the insistence that discoveries be greeted skeptically and accepted only after intensive examination, repetition and revalidation of the proofs. The rational, presumably value-free, system of judgment distinguishes the natural sciences from most other ways of deriving knowledge.

The "internal ethic" of science, the systemization of this rationalism, was established early. Occasional violations also did not wait long to begin. By "violations" I do not mean here, Mr. Chairman, the errors in observation or experimental design, or false deductions that occur frequently and are corrected by the very system we describe. These conflicts, which to outsiders may seem often to be arising in science, largely consist of this necessary part of the normal process, the shakedown of findings and conclusions until only the truth is left.

Instead, we are speaking today of violations of the scientific ethic. Among the lesser ones are false claims of priority or plagiarism. One has to read only a little of the sociological studies of science, for example, by scholars like Robert Merton, to realize that conflicts and quarrels over priority-even charges of plagiarism-have occurred over and over in science for hundreds of years. Such quarrels have even involved some of the greatest names in early science, like Galileo, Newton, and Descartes.

Mostly I presume, however, that we are here concerned with that most serious abuse, the fraudulent construction of experimental data. Cases of downright fraud in science have always been rare. Detection of some examples has taken many years, and revelation of old irregularities may be further expected. To give one startling example, modern statistical analyses have made it highly likely that some figures in Gregor Mendel's classic studies, an early window to modern genetics, could not possibly have been generated by the experiments. Whether these were intentional or unconscious errors, and whether they were Mendel's or his gardener's will never be known. Such questions and ambiguities surround most instances of scientific fraud. And some are never answerable.

I do not know. Mr. Chairman, whether scientific fraud, in less spectacular forms, occurs more frequently today than it has in the long history of science. There has been an exponential increase in the number of scientists practicing in the past 30 years over those working previously. The probability of some increase in abuses is therefore high. The likelihood of their being detected is also greater because of an increased density of peers and the development of ever more sophisticated techniques. The strength of the scientific process, however, and the dedication and vigilance of the institutions in which the standards of that process are maintained, do not appear to me to be weakened. Neither, in my opinion, is the public's huge investment in science endangered in any way. Indeed, the current production of useful new knowledge is nothing short of spectacular, and testifies to a vigorous state of health in the life sciences . . .

The NIH cannot guarantee the behavior of scientists or certify the quality of their work through independent analyses, fraud squads, or special statutes. Fortunately, none is necessary, for the natural sciences contain ultimate correctives for any debasement of the knowledge dorived from research. Science is cumulative. It is like a building that is never finished. Any serious flaw in the foundation eventually will be revealed by the weight of the structure above it. If the extension of a wing shows faults in previous construction, the faults are corrected and the design changed. The rational nature of the scientific process makes this feasible and inevitable . . .

I do not believe there is an increase in fraud or other abuses of the scientific method in this work, and I know of no statistical evidence to confirm or deny this opinion. A second opinion is that the system contains safeguards which detect fraudulent data. And a third is that fraud in science carries severe personal penalties for the erring scientist-punishments which are necessarily administered mainly by the scientific community itself-and it is this feature of the system which is the ultimate deterrent."

Dr. Handler:
"The matter of falsification of data, I contend, need not be a matter of general societal concern. It's rather a relatively small matter which is generated in and is normally effectively managed by that smaller segment of the larger society which is the scientific community. This occurs in a system that operates in a highly effective democratic, self-correcting made-the very "peer review system" .

The well publicized instances of apparent falsification of research data [are] an aberration that is difficult for the rest of us to understand. For those scientists for whom the very doing of science is not a sufficiency, of itself, as its own reward, all else derives from the esteem of his peers. Such esteem is enormously gratifying and, moreover, it is from that very esteem that all other forms of reward-promotion, income, position, status-necessarily flow. If the subject of the research and the doliberately contrived data are trivial, the act may go overlooked for some time. But in that case the rewards will also be negligible; hence there must be little temptation for so doing. How frequently such may occur is beyond my knowing. On the other hand, if motivation is to be found in an intense desire for recognition and esteem, only falsification of data relevant to some question regarded as of major significance by the relevant scientific community will serve the purpose. But in that case, glory must be short-lived, indeed. Then, the matter will very soon be found out in other laboratories which will either repeat the contrived experiment or find that the reported results are incompatible with subsequent developments.

It is surely true that the circumstances of our times have generated pressures which, one might think, might occasion an overall increase in the frequency of such undesirable occurrences. The reward system, the "publish or perish syndrome," the sensed, indeed quite real need to present evidence of scientific progress in seeking continuing financial support of one's research, the knowledge of a decline in the availability of research funds,-all of these could be imagined to so weigh upon the mind of a troubled individual as to lead him or her to succumb-and then engage in frank falsification of data. But as we have seen, in the hot cauldron of the operation of peer review in the scientific world, such data cannot long survive undetected. And it is the culprit, not society, that is injured on the rare occasion in which this occurs."

Dr. Felig:
Dr. Phillip Felig of the Yale School of Medicine was among the others who offered testimony during the first day of hearings. In 1980 Dr. Felig had co-signed a research paper with his junior colleague, Dr. Vijay Soman, in which the latter, without the knowledge of Dr. Felig, had invented some of the data.

In answer to the Chairman's question as to whether competition for shrinking research funds is altering the research environment, and whether that competition is influencing data interpretation, Dr. Felig had the following to say: ". . . it is my belief that it is not the system of academic advancement which inherently influences the use or interpretation of research data. Where such misuse or misinterpretation does occur, it is the unfortunate reaction on the part of some individual to that system."

During the course of Congressman Gore's interrogation Dr. Felig revealed that after he had discovered the fraud, he sent letters of retraction to two journals. These journals, Nature and the Journal of Clinical Investigation, refused to publish the retractions. The latter did not publish correspondence and the former needed definitive proof of fraud. In response to this situation, Dr. Felig expressed the opinion that "a journal which previously published an article has an obligation to publish a subsequent communication if there is evidence or reason to doubt the reliability of the original work."

Dr. Fetig then considered a proposal for detecting and preventing fraud. "To prevent such situations might require all scientists to submit their data notebooks together with grant requests or manuscripts for publication. Such a system, I believe, would be counterproductive because it would be extremely cumbersome, and foster an environment of mistrust . . .

But more importantly, Mr. Chairman, what we have to realize is that if we require such submission it is possible that there would be those individuals who would falsify such sources of data.

In other wards, how could we then be sure ultimately? It has to be based on trust, and the question that I don't have the answer far and which obviously you are addressing is, to what extent do we build in safeguards so as to make sure that that trust is not being violated to any substantial or significant or counterproductive way. I don't know the answer.

Mr. Capron:
On the second day of hearings, Mr. Alexander M. Capron, Executive Director of the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, reparted on the results of his investigation of scientific fraud. Among other things he said, Mr. Capron criticized the new Department of Health and Human Services deparment regulations for failing to provide for cases involving falsification of data.

Commenting on a memorandum regarding this problem written by Dr. Frederickson to the Secretary of Health, Education and Welfare, Mr. Capron said these new "regulations were designed to preclude persons guilty of fiscal mismanagement or fraud from receiving further HHS grants or contracts. As written, the regulations do not apply well to the kind of scientific fraud or abuse we are discussing here.

Moreover, although the HHS department rules specifically provide debarment (from NIH grants and contracts) for individuals debarred by other HHS components far irregularities in conducting an activity supported by departmental funds, debarment by the FDA would not appear to trigger further debarment under this provision.

Speaking for myself, it seems dubious to erect on paper a regulatory structure that has so little effective substance. If the protection against fraud and abuse of subjects promised by the regulations is important, the absence of actual mechanisms is worrisome; if the enforcement of the rules as written would be too burdensome, then one would be wary of the false assurances provided by the apparent reality of the written rules."

Dr. Raub:
The next person to make a statement, Dr. William H. Raub of the National Institutes of Health, also discussed the issues raised by Dr. Frederickson's memo. Like Drs. Frederickson and Handler on the previous day, he refused to see fraud in science as presenting a serious problem.

Today's hearing on fraud in science is timely, in view of the recent press reports about possible deceptive research practices. We do not know with certainty if this represents an increased incidence or simply a higher rate of public interest, although it is my strong belief that the latter is the case. I find little reason to be concerned that we are faced with a burgeoning epidemic of fraudulent research practices by scientists.

In fiscal and business matters it has long been recognized that awarding agencies have a responsibility to protect the financial interests of the public, and that actual or potential recipients of Federal funds are entitled to procedural safeguards and appeals. Present procedures are designed primarily to handle violations of business manage ment policies (or to adjudicate varying interpretations of administrative requirements), and they serve those purposes well.

Fraudulent research practices present a different set of problems. In the first place, the process by which such violations are discovered is quite different. It has been NIH's experience that derogatory information is typically brought to light not by an agency administrator or auditor, but by an institutional colleague, a member of the public, or through the validation processes built into scientific publications. And while the scientific method is subject to a general body of ethics (as I believe was described by several panelists yesterday), accepted medical practice, consensus as to the real or apparent existence and seriousness of specific violations may not be easily attained.

One may legitimately ask what, if any, role should allegations of misconduct play in consideration of grant applications and contract proposals, and how should that information be introduced into the review process? Traditionally, the peer review process has included an evaluation of the qualifications and past performance of the applicant investigator as a key factor in the assessment of scientific merit. Implicit in (and essential to) this review, however, is a measure of objectivity and consensus about the facts concerning alleged misconduct, as well as the presumption of innocence until guilt is established. In cases of suspected scientific fraud, it is very difficult for reviewers to reconcile allegations about inappropriate scientific practices with a principal investigator's assertion of innocence and traditional measures of his other performance as an investigator, especially when these measures indicate high quality performance and excellent productivity. And the advisory councils and boards, while technically responsible for advising on the policy aspects of particular awards, are in no better position than initial review groups to look behind the allegations.

We believe that the responsibility for dealing with such issues rests squarely on the agency officials. We also recognize that this causes considerable discomfort to all concerned. In general, scientistsincluding those who manage our research programs-are not at home in the legal/regulatory milieu. On the business management side, complaints and violations are rooted, for the most part, in the chronically imperfect fit between legal and accounting requirements on the one hand and the academic research environment on the other. Despite continuing and often vociferous debate on specifics, the research community never has contested the principle that the agencies dispensing the funds have the right and responsiblity to protect the public interest by exacting certain requirements affecting their expenditures and accounting. Where the substance of science is involved, however, a more strongly held bias against administrative intervention comes into play.

Commenting on recent procedural changes in the awarding of grants, Dr. Raub went on to say: "Debarment procedures are appropriate for cases in which proof of wrongdoing exists. For cases in which such charges are yet unproved. NIH has long recognized that some sort of tracking system is desirable. In general, NIH science administrators have felt that decisions that might adversely affect the application should not be made too early in the review process, and the derogatory information should be handled on a "need to know" basis."