Med School Dean Issues Letter on Research Fraud

Following recent revelations about research fraud in a Harvard-affiliated laboratory, the dean of the Medical School has called on all members of the medical faculty to take personal responsibility for the validity of the work they supervise.

Dean Daniel C. Tosteson said in a December letter to Harvard doctors that he will appoint a small committee of senior faculty members to help him develop procedures to prevent fraud.

Tosteson nevertheless said he doubted that the medical community could find an effective way to avoid recurrences of the problem.

"Like you, I feel frustrated that there seems to be no straightforward actions that we can take to stop dishonesty in science," Tosteson wrote.

The letter marked the third time Tosteson has addressed the faculty on professional misconduct. He was reacting to the second major case of falsified research at the medical area in five years.

A five-member panel at the Harvard-affiliated Dana Farber Cancer Institute is currently investigating an apparent falsification of data in a laboratory run by Associate Professor of Medicine Ellis L. Reinherz.

Reinherz and his associates last February reported that they had discovered a new hormone, Interleukin 4-a, which they claimed could be effective in combating cancer.

The team of scientists publicly retracted their reported findings in the November 28 issue of Science magazine, saying that they could not reproduce their data.

The director of research at Dana Farber later said that one member of the team, Claudio Milanese, since departed from Harvard, had apparently manipulated experimental results in the laboratory.

Tosteson said that lack of coordination and monitoring within large research teams have been cited as factors that contribute to fraud.

"I urge each of you to accept personal responsibility for the validity of all data gathered in and published from your laboratories," Tosteson said.

"Each offense of this kind does grave damage," Tosteson said. "It corrodes the reputations of the institutions where it takes place" and "erodes the trust on which science and medicine are based," Tosteson said.

Tosteson said that he intends to bring the matter of fraud prevention before the steering group of the Medical SchoolFaculty and the department chairmen. He askedHarvard doctors to offer suggestions for copingwith the problem.

Although the pressures of competition and thelack of supervision of "young investigators" havebeen cited as causes of fraud, Tosteson said, "itseems unlikely that a single remediable cause willbe found."

Tosteson was out of town and could not bereached to discuss his letter.

The medical community here was last shaken byrevelations of fraud in 1981, when John R. Darsee,then an associate professor, admitted tofabricating research data from a study on heartattacks.

Harvard dismissed Darsee and subsequentlyestablished a standing committee on facultyconduct to investigate and punish research fraudcases.

In his letter, Tosteson said that thosemeasures are adequate for looking into andpunishing fraud but "contribute little toprevention."

Despite the two widely publicized instances offraud in recent years, Harvard medical professorssaid that the Medical Area is not subject to morefraud than other schools.

"Harvard is a preeminent institution. Cases atother institutions don't make the same splash,"said White Professor of Biological ChemistryManfred L. Karnovsky.

Several Medical School faculty memberssuggested that closer supervision of researchwould cut down but not eliminate instances offraud.

"I agree with the idea that the [heads oflaboratories] are responsible for the accuracy ofall data that comes out of their labs," said DavidM. Knipe, associate professor of Microbiology."But [fraud] is very hard to legislate away."

Research supervisors "should where possiblelook at raw data, especially when the finding isclaimed to be important," said Krayer Professor ofPharmacology Irving H. Goldberg. "If you just seethe data turned into a table of results, you'reasking for trouble."

Large laboratories which rush to publish theirresults make themselves vulnerable to fraud,several professors said.

"My prejudice is that there are feweropportunities for fraud in smaller laboratorieswith closer observation," Goldberg said.

"In a very hot field, people want to publish asfast as possible, so a lot of the checking [ofdata] is not done," Karnovsky.

However, too much supervision could be harmful,said Kuhn Professor of Biological Chemistry EugeneP. Kennedy. "There's a need for balance. We needsome super-vision but we can't have too much. Wecan't have an Orwellian atmosphere," he said