CRS-3

one expert suggests, "the public is really paying for a good-faith effort by highly skilled people...."

Another major issue which arises within the context of discussions about misconduct is fraud. Fraud is a generic term including acts, omissions, and concealments which involve a breach of legal or equitable duty and result in damage to, or the taking advantage of, another. Actual fraud denotes a successful, intentional deception designed to get another to surrender a legal right or to relinquish property. Constructive fraud is a breach of a legally created duty. In other words, the law can declare some conduct to be fraudulent because of its tendency to deceive, to violate confidences, or to injure the public interest. Moral guilt or actual dishonesty and intent to deceive are not essential elements of constructive fraud and the perpetrator may be held to a "reckless" standard, which would denote a wanton or heedless indifference to substantial and unjustifiable risks that may result from misconduct. Some unprofessional behavior is fraudulent in nature. For example, the fabrication of data or results to support a grant proposal might be (actual) fraud, but some acts of misrepresentation may not be fraud. Hence, fraud is a subcategory of misconduct.

The last problematic issue regarding a definition of scientific misconduct is the role that professional judgment plays in analyzing and interpreting data. This issue pertains mostly to accusations of misrepresentation. Is there ever a time when not citing an observation or modifying the data are appropriate based a scientist's professional judgment? Some scientists answer in the affirmative, noting that such judgments are a part of the creative process of the scientific endeavor. Assuming the validity of such a position, then it is possible for professionals to disagree about what constitutes the correct analysis and interpretation of data. It is also possible for a professional to exercise poor judgment. Hence, scientists caution against equating poor or questionable judgment with misconduct.

Scientific Misconduct: Its Extent and
Factors Which Contribute To Its Occurrence

Experts disagree about the seriousness of the problem of misconduct in science. There were 26 cases of alleged scientific misconduct publicly disclosed between 1980-1987. However, the National Institutes of Health (NIH) and

I Woolf, Patricia. Pressure to Publish and Fraud in Science. Annals of Internal Medicine, 1986, v. 104. p. 254-256.

27 C.J.S. Fraud, Secs. 1-3.

Woolf, Patricia. Deception in Scientific Research. Project on Scientific Fraud and Misconduct: Report on Workshop Number One (Washington, D.C., American Association for the Advancement of Science), 1988. p. 52-53.

CRS-4

the National Science Foundation (NSF), the two principal funding sources for research performed by universities and colleges, have received reports of a number of cases which have not been made public. NIH has received 125 reports since 1982 and NSF has investigated 12 allegations since 1980. The question is, do these reported cases of misconduct indicate either of the following:

(1) dishonesty or incompetence in research is significant enough to threaten the soundness (the integrity and quality) of the scientific enterprise; or

(2) the misuse of Federal money for research is widespread.

Is misconduct in science a rare occurrence or do the reported cases indicate the existence of a serious problem?

The science community, in general, tends to believe that misconduct in science is rare. In addition, until recently, it has regarded its traditional formal control mechanisms-peer review and experimental replication-sufficient to prevent and detect an occasional incident of misconduct and to insure the integrity of the scientific enterprise. However, critics point out the limitations of these mechanisms. In particular, they point out that a normal peer review process--for grants or journal publication-cannot detect plausible, internally consistent fabrication,' and that the effectiveness of replication is limited in at least two respects. First, replication can only detect errors; it cannot detect misconduct, or the fabrication of data which elicits correct results. Second, there are few incentives to replicate the work of another. Professional advancement is generally tied to original work and funding is not awarded to projects which are substantially duplicative of what is already known. A counterargument to this position is that if research results are important enough, other scientists will want to followup and build upon the results. Such follow-up work functions as a check on the original work, for if the original work is flawed, the follow-up work will not produce the expected results. A shortcoming of this control mechanism is that it is contingent upon the original work being important. However, this shortcoming may not be significant if "important" original work includes much of the research as well as that research which has implications for consumer use.

Relman, A. S. Lessons from the Darsee Affair. New England Journal of Medicine, 1983, v. 308. p. 1415-17.

'Engler, R. L. et al. Misrepresentation and Responsibility in Medical Research. New England Journal of Medicine, 1987, v. 317, no. 22. p. 13831389.

23-941 0-90-15

CRS-5

For some government officials and a growing number of scientists, the absence of empirical evidence which clearly indicates that misconduct in science is not a problem in the context of the current scientific environment suggests that significant misconduct remains a possibility. This, in turn, implies the consideration of additional or alternative control mechanisms. A few key features of the scientific environment today which have implications for the extensiveness of misconduct are as follows:

(1) Competition for position, tenure, and acclaim all contribute to the great pressure to publish. Such pressure may motivate some individuals to "cut corners" in a variety of ways to increase their list of publications. Cornercutting activities include submitting manuscripts in "least publishable units" (LPU) (which overloads the peer review system) and falsifying data.

(2) An increasing number of financial ties between scientists and
industry means that more scientists may be unwilling to share
their raw data. Such unwillingness prevents others from
obtaining the information necessary to replicate experiments.

(3) The increasing size of research programs may lead to inadequate
`lab supervision. This may be detrimental to the training and
development of new scientists because traditionally lab
supervisors (or senior researchers) took upon themselves the
responsibility of training inexperienced personnel in the methods
and the ethics of research.

(4) Gift authorship is a questionable practice not only because of
the impropriety of accepting credit for work to which one did
not substantially contribute, but also because it may impede
rigorous peer review. Some experts have suggested that
manuscripts "co-authored" by highly respected scientists may
receive less rigorous review.

THE SCIENCE COMMUNITY'S RESPONSE: A DESCRIPTION

Selected Funding Agencies

In 1988, the Federal obligations for civilian and military basic and applied research were approximately $18 billion. Table 2 shows 1988 estimates of Federal obligations for research according to funding agencies, type of research, and the percentage of the research done at universities only.

Some vital statistics are as follows:

CRS-6

53 percent of the basic research as compared with 19 percent of the applied research was funded by NIH or NSF;

Universities/colleges received 73 percent of NSF's funds; and

Universities/colleges received 67 percent of NIH's funds

This paper will focus on NIH and NSF activities concerned with scientific misconduct since universities and colleges receive most of their financial support from these two agencies.

• Not included in the tabulation of the percent of research conducted by universities and colleges are obligations to federally-funded research and development centers associated with universities/colleges.

Source: U.S. National Science Foundation, Surveys of Science Resources Series, Federal Funds for Research and Development: Fiscal Years 1986, 1987, and 1988, v. 36.

The National Institutes of Health (NIH):
Misconduct in Biomedical Science

Much of NIH's activity in the area of misconduct focuses on the issue of investigating allegations. It has investigated a number of cases, endeavored to establish guidelines/regulations for dealing with possible misconduct, and has established an Office of Scientific Integrity, which will have the responsibility of monitoring the investigations performed by universities.

CRS-7

However, investigation is not the extent of its concern. NIH has also expanded its ALERT system which provides the Public Health Service (PHS) with a limited capacity to restrict financial support of researchers who are under investigation for scientific misconduct. In addition, in cooperation with the Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA), it has begun a process for developing guidelines concerning financial conflicts of interest for investigators receiving Federal funds.

NIH's Active Involvement in Investigations. The universities which employ the researchers accused of misconduct have usually had the lead in investigating the cases. However, NIH investigated approximately a third of them. Generally, NIH becomes actively involved in investigations when it perceives major shortcomings in the university's investigation, e.g., the university failed to conduct the investigation in a timely fashion or its investigation left unanswered some significant questions.

Legislation and NIH's Regulatory Involvement in Investigations. In 1985, Congress passed the Health Research Extension Act (P.L.99-158). Part of this legislation (section 493) directs the Secretary of the Department of Health and Human Services (DHHS) and the NIH Director to take specific action in the area of scientific fraud. The Secretary is directed to do the following.

(1) develop regulations regarding how recipients of NIH grants, contracts, and cooperative agreements should respond to allegations of scientific fraud; and

(2) require, by regulation, each institutional applicant for NIH funds to include in the application assurances that it not only has established administrative mechanisms to review reports of fraud in accordance with the regulations specified in (1), but also will report to the Secretary any investigation of suspected fraud which appears substantial.

The Director of NIH is required to establish "a process for the prompt and appropriate response to information provided the Director... respecting scientific fraud." Such process is to include procedures for receiving reports of suspected fraud and for taking appropriate action.

As of May 16, 1989, the Public Health Service of the Department of Health and Human Services (DHHS) had not issued any final rules to

"Newburgh, Janet. Institutional Liaison (Misconduct Policy) Officer, Office of Extramural Programs, NIH. Personal Communication. May 5, 1989. 7 Health Research Extension Act (P.L. 99-158), Sec. 493.