Chicago Tribune

Science on the hot seat
U.S. crackdown on research fraud raises concerns

By John Crewdson

March 19, 1989

After his conviction late last year for defrauding the United States government, Stephen E. Breuning was ordered to repay $11,000 of the money he stole, serve 60 days in a halfway house and perform 250 hours of community service.
As crimes go, Breuning’s might not have been noticed – except that he was a former University of Pittsburgh psychologist who had been charged with obtaining federal research grants on the strength of research he had never conducted.
A few days after Breuning was sentenced, Shervert Frazier, a psychiatrist and former director of the National Institute of Mental Health, resigned abruptly as head of McLean Hospital, Harvard University’s psychiatric teaching facility.
Frazier’s resignation followed the discovery, made by a graduate student and confirmed by a faculty committee, that he had plagiarized parts of four articles published between 1966 and 1975.
Then, last month, the National Institutes of Health (NIH) announced that an internal investigation had identified “several serious inaccuracies” in an important paper co-authored by David Baltimore, a Nobel Prize-winning biologist at the Massachusetts Institute of Technology.
The NIH director, James B. Wyngaarden, wrote Baltimore a letter criticizing him for not having taken seriously a junior researcher’s concerns about the paper, and directing him to send a detailed correction to the journal where it appeared.
Never before had a scientist been convicted of criminal charges for falsifying his research. Never had a scientist so eminent in his field been forced to resign because of unethical behavior. Never, certainly, had the head of NIH ordered one of the world’s most distinguished scientists to correct a scholarly article.
Though unrelated to each other, these events and others suggest that the question of how to ensure the integrity of scientific research, until now an issue debated mostly among members of the elite scientific fraternity, is becoming the subject of broader public concern.
Among scientists, the decisions to prosecute Stephen Breuning, to ask for Shervert Frazier’s resignation and to publicly chastise David Baltimore did not have many advocates.
It is not that science does not admit to shortcomings. It does, but it has long preferred that they be investigated by discreet committees of scientists who render their verdicts behind closed doors.
Many scientists say they fear that prosecutors, reporters and other lay investigators who do not understand the complexities of laboratory research or the subtleties of the scientific method cannot distinguish between willful fraud and unavoidable error.
Outsiders, they say, do not comprehend the degree to which science is “self-correcting,” nor can they appreciate the need for scientists to make mistakes along the twisting path to truth.
“The halls of Congress,” said Baltimore in a letter to colleagues after a congressional subcommittee invited him to testify in public about the disputed article, “are not the place to determine scientific truth or falsity.”
For outsiders to investigate such matters, Baltimore warned, “could cripple American science.”
But some other scientists privately acknowledge a greater fear: the erosion of public confidence that might result from a better understanding of how science in this country is funded and carried out.
“Scientists are people,” says a researcher at the National Cancer Institute. “There are scientists who aren’t terribly moral. The really good scientists are locked away in their labs and dedicated to their research. The people who get the money and the power are the ones who are the wheelers and dealers.”
Few scientists are willing to speak out against their own establishment, but their numbers are growing. When they do speak, they describe a closed system in which it often seems that no one has a vested interest in publicizing fraud, sloppiness, or even honest error.
Not many scientific journals, for example, will eagerly admit to having published a fraudulent article, and many journals are reluctant to retract such ar-ticles or to correct inadvertent mistakes once they have been pointed out.
The result is that most scientists can point to dozens of published articles that are acknowledged by everyone, including their authors, to be wrong from top to bottom, but that never have been formally withdrawn from the literature.
“Being wrong in print doesn’t have anything to do with success or failure in science,” one government scientist says. “This isn’t journalism.”
Nor are there many universities that actively will investigate, without a strong impetus from outside, suspected misconduct on the part of their own faculty members, especially those with tenure or who are skilled at obtaining research funds.
A report issued last month by the highly respected Institute of Medicine, an arm of the National Academy of Sciences, raises serious questions about the ability of academic institutions “to conduct objective investigations of misconduct by their own faculty members or research staff.”
Even the NIH, which dispenses the bulk of the money that pays for this country’s biomedical research, has been reluctant to act when cases of suspected misconduct are brought to its attention. The Baltimore investigation, for example, was spurred by pressure from a congressional subcommittee, and the National Institute of Mental Health required more than three years to take final action in the Breuning case.
When the federal agency eventually concluded that Breuning had committed some of his research fraud while at the University of Pittsburgh, it reserved some criticism for the university, which had assured the institute that it could find “no serious fault with Dr. Breuning’s activities here.”
Those who know the most about scientific misconduct are the scientists themselves, often those who work in the same laboratories or on the same research projects with the miscreants.
But the few researchers who have turned suspicions about their colleagues into official complaints have not been treated as heroes, and their experiences are likely to make others think twice before sounding the alarm.
The man who uncovered Breuning’s misdeeds and who, for three years, pushed the National Institute of Mental Health to act on his case is Robert Sprague, a professor of psychology at the University of Illinois at Urbana-Champaign who had given Breuning’s career its initial boost by hiring him as a research assistant.
“The science mechanism either refuses to adjudicate fraud or is very, very sluggish,” Sprague says of his experience, “in spite of the myth that science is beautifully self-correcting. I’ve been at Illinois for 25 years. I’m a full professor, a research professor at a major university, and I almost didn’t get the job done.”
At the time he blew the whistle on Breuning, Sprague had received 16 years of continuous funding from the institute. The following year, his grant was slashed by 90 percent.
“They said it had nothing to do with this,” Sprague says, “but I won’t buy the Brooklyn Bridge from you, either.”
There are significant distinctions between misconduct in science and other intellectual wrongdoings, such as the fabricating of important historical or literary documents like Adolf Hitler’s diaries or Shakespearean sonnets.
Most scientists who design their research on the basis of experiments reported by others do not first try to replicate those experiments but simply accept the results as given. The same is true of physicians who base their treatments and prescriptions on the experiences of colleagues with similar patients.
It was because of Breuning’s fraudulent articles, for example, that doctors in some state institutions ceased their historical practice of prescribing tranquilizers for mentally retarded children. Though there is no evidence that any children were harmed, such harm is difficult to measure.
“There are tens of thousands of mentally retarded people who are on these medications,” says Sprague. “They can’t defend themselves. They can’t even speak, most of them.”
The Breuning case and others like it lead to an inevitable question: Is there more misconduct among scientists now than before? Rep. John Dingell (D., Mich.), whose subcommittee is responsible for overseeing NIH’s $7 billion annual budget, believes that there is.
“We are only seeing the tip of a very unfortunate, dangerous and important iceberg,” Dingell says.
Daniel E. Koshland Jr., the editor of Science magazine, the prestigious weekly journal of the American Association for the Advancement of Science that serves as the voice of the scientific establishment, does not agree.
“We must recognize,” Koshland has written, “that 99.9999 percent of reports are accurate and truthful.”
But not many of Koshland’s fellow scientists would make the same remark.
“All of us make mistakes, and we don’t necessarily write to the journal and say ‘Fix this,’ ” says a National Cancer Institute laboratory chief.
Even when misconduct does come to light, the scientific press may be reluctant to make the information known. As Sprague points out, Science magazine held an article on the Breuning case for 10 months before it was finally published in December, 1986.
The mounting pressure to publish research has left scientists more vulnerable to temptation than ever before, to cut corners if not to make things up. It is now nearly impossible for any academic researcher to renew a government research grant without having published at least one major paper in an important journal during the previous year.
“Science is so competitive,” says a despairing scientist in California. “You get grants in small increments, you have to submit incredible amounts of paperwork and progress reports and publish a continual stream of stuff, or the money stops.”
For junior researchers who have five years to win university tenure or clean out their labs, job performance is measured mostly by the number of publications rather than their content.
“It really doesn’t matter so much what is being published, which leads to a lot of crap getting into print,” the California scientist says. “You really feel the pressure continually to produce, and that makes problems.”
Even government researchers are not spared such pressures. A few years ago, the director of one of the institutes that make up NIH sent a memo to two of his tenured scientists warning that, while they had better begin publishing some papers, “journeyman works for publication in second-, third- or fourth-line archival publications will be quite satisfactory.”
A second reason may be a quantum change in the potential rewards for scientific fraud. Until recently, the most that a successful scientist had to look forward to were his own laboratory and postdoctoral research staff, membership in the National Academy of Sciences and a mantel filled with prizes – perhaps a Nobel Prize among them.
But as advances in genetics and immunology create new products and processes, there are signs that some scientists are willing to manipulate their research in return for the riches that modern biotechnology can confer. Indeed, it is difficult to underestimate the impact in recent years of money on the biological sciences.
“The biologists have been literally hounded,” says Harvard biologist Temple Smith. “You cannot publish a paper that sounds like it has anything to do with solving AIDS, cancer or anything but you will have four millionaires in your office before noon trying to give you money.”
The arcane world of bioscience, in which new developments are often known to, and understood by, only a handful of researchers long before they become public, has also opened new venues for insider trading that have not yet been pursued by federal regulators.
“What you find today is a tremendous blurring between the marketplace and the academic ivory tower,” says Harvard scientist Norman Letvin. “One doesn’t know what to believe when one goes to scientific meetings anymore. In the good old days, the scientist was making $20,000 a year and he had nothing to lose by being honest. Now he’s got everything to lose.”
The answer, according to some in Congress, is simply to level the playing field by giving the potentially dishonest scientist something to lose. One of Dingell’s colleagues, Rep. John Conyers (D., Mich.), has suggested enacting a federal law that would make it a crime to falsify any scientific research, no matter where it is done or by whom.
At the moment, the only criminal laws that cover falsification of scientific data are the general fraud statutes that were used to prosecute Stephen Breuning. According to NIH officials, those laws can be applied only to nongovernment researchers who use bogus research to obtain federal funds.
NIH staff scientists are currently exempt from such prosecutions, one official explained, “because publication’s not part of their job. They do that for their own personal recognition. There’s nothing that says you must publish, although it is the criterion by which they’re measured.”
For the moment, the attention of Congress has been captured by cases of outright fraud. But to focus on the Stephen Breunings is to overlook a more consequential problem: the accumulation of uncorrected errors in the scientific literature that result from sloppiness, haste or downright incompetence.
“There’s almost a tradition of sloppiness now because of the high flux of personnel,” says Harvard’s Temple Smith. “The turnover in postdocs, visitors from other labs and graduate students is just immense. It’s unfortunate, but it’s part of the game. These people are under a lot of pressure to find answers, because they know if they don’t get it done in three months somebody else will.”
The gulf between principle and practice is currently so wide that scientists cannot bring themselves to utter the word “correction,” calling their corrections “errata” to make them as obscure as possible.
Most scientific journals make an effort to keep out bad science by submitting prospective articles to peer review, an anonymous evaluation of their strengths and weaknesses by at least two other researchers working in the same field. But Illinois’ Sprague questions the value of such reviews.
“I think the evidence is pretty clear on a case history basis that you don’t find out fraud by peer review,” he said, noting that several of Breuning’s fraudulent papers had been passed by peer reviewers before they were published.
“Somebody on the inside has got to blow the whistle and it’s usually a young person, and the history is that those young people are crushed.”