The Office of Scientific Integrity

For most of the 1980s, the specter of scientific fraud popped into public view every few years, usually only to submerge again. Faced with several well-publicized cases of scientists who blatantly faked their data—among the best-known being Harvard cardiologist John Darsee (whose colleagues watched him forge data) (Broad and Wade 1982, p. 14) and Sloan-Kettering Institute immunologist William Summerlin (who painted black spots on white mice to simulate tissue grafts) (Broad and Wade 1982, p. 155)—the scientific community as a whole tended to dismiss the phenomenon as the work of a few deviants. Called in 1981 to testify before then-Representative Albert Gore Jr.'s subcommittee on investigations and oversight, Philip Handler, then-president of the National Academy of Sciences, told the panel that science was perfectly capable of regulating itself, and that the problem of scientific fraud had been "grossly exaggerated" by the press (Broad and Wade 1982). The message was clear: Congress should stick to its knitting, and science would take care of itself.

Now, however, scientists are faced with a very different situation: a permanent federal bureaucracy, known as the Office of Scientific Integrity (OSI), devoted to the investigation of alleged scientific wrongdoing. The transition from the days of laissez-faire, when universities were entrusted with all investigations unless they were handled so badly that the federal government was forced to take notice, has been an uncomfortable one for many researchers. Lately, this discomfort has taken shape as a vocal protest against OSI. Largely as a result of this protest, federal officials recently promoted a plan to restructure OSI and change some of its operating procedures, a move with unpredictable consequences for universities and researchers alike.

OSI was born largely because of investigations by two influential congressmen: Representatives John Dingell of Michigan and Ted Weiss of New York. While both claimed they acted mainly to make sure taxpayer dollars were not spent on fraudulent research, both also maintained a strong interest in protecting "whistleblowers," the often vulnerable, and usually junior, scientists who first raise questions about bad research. And Dingell seemed to take special pleasure in pointing up the deficiencies of investigations run by universities and by the National Institutes of Health (NIH) itself. (For reasons that are not well understood, biomedical research seems far more prone to fraud than any other field of science. As a result, NIH, which funds most biomedical research in the United States, has held ultimate responsibility for most prominent cases of fraud.) Plagued as they often were by conflicts of interest, shoddy methodologies, and, in some cases, a willingness to believe any excuse by the accused, no matter how far-fetched, these investigations made easy targets. And the notion that science could "govern itself" grew harder to sustain.

In late 1988, Dingell and other congressmen began circulating draft bills that would have created a new office for investigating scientific fraud. Dingell would have placed such an office under the aegis of the Inspector General of the Department of Health and Human Services, raising the possibility that scientific fraud might be probed by the same criminal investigators who unravel Medicare fraud. Unsettled by this vision of the "science police," NIH officials moved quickly to pre-empt Congress. On March 16, 1989, they published a notice in the Federal Register announcing the creation of OSI, an office with responsibility for ensuring thorough university investigations of misconduct and the authority to conduct its own investigations when necessary.

From the beginning, OSI took an unusual approach to its work. For one thing, the office had a broad mandate: Although located within NIH, it was responsible for research funded throughout the entire Public Health Service, including the Food and Drug Administration, the Alcohol, Drug, and Mental Health Administration, and the Centers for Disease Control. Furthermore, it was not restricted to scientific fraud. Departmental lawyers, wishing to avoid the legal connotations of the term "fraud"—in particular, a requirement that fraudulent actions not only mislead others but also cause them noticeable harm—decided instead to adopt the term "scientific misconduct," defined as "fabrication, falsification, plagiarism, or other practices that deviate...