Microscope Focuses on Researchers : Science: A new agency has unprecedented power to police scientific behavior. Some feel this will discourage bold and innovative projects.
A decade ago, biologist Jonathan Singer had an experience that every honest scientist dreads. A colleague had falsified research data, and the information ended up in one of Singer’s published papers.
Such first-hand knowledge of scientific misconduct, by most accounts, is not unheard of. And neither was Singer’s response.
He did nothing.
Today, the veteran UC San Diego researcher still believes he did the right and “humane” thing. “It was minor,” Singer said of his colleague’s transgression. “It just didn’t make any difference in the end. Nobody was harmed.”
Right or wrong, such attitudes are coming back to haunt the nation’s biomedical research community.
Responding to congressional complaints that scientists can no longer be trusted to police themselves, the National Institutes of Health set up an office with unprecedented powers to supervise scientific behavior and conduct investigations into allegations of misconduct.
Denounced as a “science FBI” by its detractors, the Office of Scientific Integrity opened its doors in August and soon will be staffed by 10 investigators with advanced degrees in fields such as medicine, immunology, pharmacology, cell biology and physiology--all areas with high incidences of alleged fraud.
“The credibility of the biomedical community is at stake,” said Suzanne Hadley, acting director of the agency.
“Scientists in this country have been used to an extraordinary amount of freedom, and the public has been well served by that,” said Patricia Woolf, a Princeton sociologist who has studied fraud in science. “But I don’t think there’s any doubt that there’s a mood that is emphasizing oversight.
“This is probably the best solution to a situation of developing mistrust of science by the public.”
Still, Woolf and others fear the office may end up deterring not fraud but innovative, breakthrough science because honest errors are an ineluctable element of progress.
“The line between error and fraud isn’t always very sharp or apparent,” said Bernard D. Davis of the Harvard School of Public Health. “The danger is, if we try too hard to police fraud, we will discourage bold research, and scientists will be prone to doing more pedestrian things. And I fear we may be crossing the critical line between fraud and matters of scientific judgment.”
Guesses about the extent of scientific misconduct are just that. But many scientists say privately that, at some point in their careers, they have either suspected or become intimately aware of misconduct.
“You hear that a lot,” said Washington attorney Barbara Mishkin, a former NIH researcher who now specializes in biomedical ethics. “You go to meetings or even parties and there always will be one or two tales of incidents. . . . But you really don’t know.” Hadley added: “The relative incidence of misconduct is small. But that is not the point. Even one or two is too many.”
Whatever its dimensions, fraud in science is not always simply an issue of academic concern.
Last year, for instance, Dr. Stephen E. Breuning, one of the country’s leading experts on mental retardation, was criminally prosecuted for falsifying data on drug treatments for mentally retarded children.
Because of Breuning’s reputation, many states had based their treatment policies on his reported findings.
Breuning pleaded guilty to two counts of submitting false statements to the government in exchange for the dropping of one count of obstructing a federal investigation. He was placed on five years’ probation and ordered to serve a 60-day work-release sentence and to pay $11,300 restitution to the University of Pittsburgh. The two-year investigation of Breuning had been led by Hadley, who was the misconduct officer at the National Institute of Mental Health.
Also last year, a one-time Harvard ophthalmology fellow was found to have distorted his findings in experimental treatments using an eye medication while he owned shares in the company that made the drug.
The most controversial case of alleged fraud involved David Baltimore, a Nobel laureate in medicine and president of the Whitehead Institute for Biomedical Research, which is affiliated with the Massachusetts Institute of Technology.
In 1986, Baltimore was among six authors of a study on genetically altered mice and immunity that appeared in Cell magazine. The article came under fire because of alleged misconduct and erroneous data.
Numerous panels, including ones at MIT and NIH, eventually cleared the authors of fraud or misconduct but found technical errors and omissions.
The authors, including Baltimore, whose role in that paper was tangential, later sent correction letters to the journal, conceding that the notes gathered by a co-signer were sloppy and disorganized.
The case attracted much attention not only because of Baltimore’s involvement but also because of a subsequent investigation by the House Energy and Commerce subcommittee on oversight and investigations. The probe itself became highly controversial as the subcommittee brought in the U.S. Secret Service to examine the lab notebooks of one of Baltimore’s colleagues.
The investigation, Davis said, is a typical example of the blurring of the line between fraud and error. “The dispute was far too technical for a congressional forum,” he said.
Largely because of the controversy, Baltimore faced strong opposition from the faculty of New York’s Rockefeller University when he became a candidate to be its president. Ultimately, he got the job, and will start in July.
Baltimore declined to comment publicly about the controversy or the newly created Office of Scientific Integrity.
If OSI has a forerunner, it is the NIH’s Institutional Liaison Office, whose role has been much misunderstood, according to Brian Kimes, OSI’s first acting director.
The liaison office had--and still retains--broad responsibilities, including dealing with research institutions on issues such as conflict-of-interest regulations and grant management, said Kimes, now a senior official at the National Cancer Institute.
Almost as an afterthought, the liaison office also was given the responsibility to investigate misconduct. But as the number of allegations increased over the years, its small staff became overwhelmed, Kimes recalled. “People just got the wrong impression,” he said.
What the NIH also lacked in the past, according to attorney Mishkin, were investigators who were properly trained to deal with charges of misconduct. As a result, she said, many investigations were bungled. In one case, for instance, the lab notebooks of an accused researcher ended up in the hands of the accuser.
At the same time, Mishkin and others said, few specific federal guidelines existed to assist academic institutions, which often were left largely to their own devices when confronted with possible misconduct by a faculty member.
The upshot was wildly differing procedures and results that aroused skepticism and suspicion.
The Office of Scientific Integrity, acting director Hadley said in a recent interview, is a necessary first step toward restoring Big Science’s once pristine public image.
The office may not investigate every allegation of misconduct at the 4,000-plus institutions throughout the country that receive approximately $7 billion in federal biomedical research funds doled out annually. But it is empowered to supervise such investigations, which can be conducted by the institution involved, she said.
“The primary duty is up to the institutions,” Hadley said. She noted, however, that her office will not hesitate to get directly involved.
These academic institutions now also must develop, for OSI approval, specific administrative guidelines to govern the handling of misconduct charges. The first of these annual “assurances” are due by Jan. 1.
And starting about a year from now, such universities must submit annual reports on all allegations, inquiries and investigations involving possible misconduct that occur during a calendar year. These reports will provide the first real measure of the extent of fraud in science.
In addition, the Office of Scientific Integrity, which reports to the NIH director, will actively promote guidelines for “responsible science” on issues such as how best to retain one’s research data, Hadley said.
“But 95% of our efforts will be on doing investigations,” she said. Such investigations, Hadley said pointedly, could involve applications for grants as well as actual research and the reporting of data.
The office already has more than 70 cases to handle.
At the end of an investigation, the office will send its results--and recommendations, if any--to James O. Mason, assistant secretary of Health and Human Services, who will have the final say on questions of discipline.
“There’s no doubt that some oversight is appropriate. But it has to be done judiciously,” said Princeton sociologist Woolf.
“They must beware of allegations that may be the result of professional jealousies or personal enmity,” said Mishkin, of the Washington law firm of Hogan & Hartson. “What is perceived as misconduct may be merely a disagreement about methodology or statistical analysis. It’s possible that both the accusers and the accused are guilty of exaggeration, stubbornness and poor judgment--but nothing more.”
Hadley agreed: “Look, one man’s honest error is another’s willful mistake. A lot of what we must sort out is just that. But somebody’s got to make that cut.”
