WASHINGTON-A still-unpublished paper by two NIH scientists on professional misconduct has spawned sharp debate within the scientific community on the responsibilities of co-authors and the role of lawyers in the publications process.
The authors of the 1983 report, Walter Stewart and Ned Feder, have appeared in recent months before two congressional committees and a steadily growing number of university gatherings to discuss their findings and the larger issues it has raised. But the possibility of lawsuits from a few of the principals discussed in the study so far has prevented the details of the Stewart-Feder paper from becoming known to the audience the authors most want to reach-their colleagues.
The authors based their study on public records in the case of John Darsee, a cardiology re searcher at Emory University and Harvard University medical schools. One incident of fraud in 1981 that he admitted led to investigations by both universities and the National Institutes of Health that eventually uncovered dozens of other cases in which Darsee had falsified or fabricated data. Darsee lost his Harvard fellowship and in 1983 was debarred for 10 years from receiving NIH funds or serving on any type of NIH advisory group.
Stewart and Feder examined Darsee's publications, finding in them examples of obvious errors and implausible data, cases of"honorary" authorship for scientists who had done little work on the project, instances where old data were misrepresented as new results, and attempts to reach firm conclusions from ambiguous results; The authors speculate on whether current practices foster such misconduct, and discuss possible solutions to the problem. They focus not on Darsee himself but on 47 of his co-authors.
In an unrelated case that raises many of the same troubling issues, a committee at the University of California at San Diego medical school has found that 68 papers have been tainted by the misconduct of researcher Robert Slutsky. The careers of dozens of co-authors have been thrown into question by the October 9 report of an investigative panel, which required scientists to demonstrate their contributions to the research effort to escape suspicion of misconduct. In
its recommendations, the committee said departments should develop a way to measure "the type and degree of participation" of faculty in every published work.
The Stewart-Feder paper, innocuously entitled "Professional Practices Among Biomedical Scientists: A Study of a Sample Generated by an Unusual Event," was first submitted in September 1983 to the British journal Nature. It was withdrawn in February 1985 after an extended round of revisions and repeated delays in publication dates. It was resubmitted last March, and has undergone a similar process since that time. John Maddox, the editor, said last month from London that "we shall be publishing it, I hope before the end of the year."
A year-long round of negotiations with Cell editor and publisher Benjamin Lewin ground to a halt last fall, although Lewin said recently that the study "has important things to say and should enter the public discussion." The authors of more than a dozen other publications indicated in letters to the two men that they were not interested in wrestling with the subject.
Stewart and Feder, who have acted independently of their positions at the National Institute of Diabetes and Digestive and Kidney Diseases, have sought other outlets for their work. Their testimony during a congressional hearing, as well as their 49-page study, is expected to appear as part of the re port of the Task Force on Science Policy conducted by the House Science and Technology Committee. The committee hopes to complete its massive two-year effort by the end of the year.
"We've been advised that the Task Force is immune from legal action," responded one staff member when asked about the possibility of suits. At the same time, the task force plans to include an unsolicited, 20-page analysis of the Stewart-Feder testimony prepared by the law firm acting on behalf of one of the scientists whose research is mentioned in the study.
This fall Stewart and Feder have taken to the road, speaking to both scientists and the public at several universities and research laboratories.
A series of talks at the University of Virginia Medical School, for example, has strengthened interest there in setting an upper limit on the number of publications that will be considered for tenure, explained Jeremy Tuttle of the department of physiology.
"What people here found most interesting were the factors that lead to overabundant authorship," Tuttle said. "They don't find the results [the errors and improprieties described by Stewart and Feder] surprising at all." The committee in San Diego also recommended an emphasis on quality over quantity in tenure decisions.
The Stewart-Feder study is not without its critics. The fact that all the subjects were connected to Darsee may make them an unrepresentative group, several scientists have observed. They also wonder whether the situation at Harvard is
atypical, and whether two re searchers with backgrounds in organic chemistry are qualified to speculate on the psychological and sociological factors that affect the publications process.
The paper's odyssey has fanned resentment by researchers against the growing presence of lawyers in the publications process. "It's wildly inappropriate," said Lewin. "It makes things much more complicated, increases the costs enormously, and exerts an inhibitory effect. But it can be quite effective, too."
The intervention may even alter the quality of the work. Judy Willlis, an entomologist at the University of Illinois at Urbana-Champaign who invited the pair to the campus, noted that "the threat of libel takes the edge off using all the data that is available." Remarked Tuttle, "Any time lawyers get involved in writing something, it becomes less than useful for public consumption."
For related article, see page 12.