The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 29.0°F | A Few Clouds

Panel finds no evidence of fraud in biology paper

by David P. Hamilton

Although a National Institutes of Health investigative panel has concluded that a research paper co-authored by Whitehead Institute Director David Baltimore '61 contains "significant errors," it found "no evidence of fraud, conscious misrepresentations, or manipulation of data" by the authors.

The panel characterized the flaws of the research effort as "significant errors of misstatement and omission, as well as lapses in scientific judgment and interlaboratory communication." In particular, it noted factual and clerical inaccuracies in two tables of data presented with the paper, and it raised questions regarding the sensitivity and specificity of certain reagents and assays used in the experiments.

These inaccuracies were "inadvertent errors of understanding and communication" between the three principal authors, the panel's report stated. It went on to recommend that the authors submit a letter of correction containing substitute data to the journal Cell, which originally published the paper in 1986.

In addition, a memorandum filed by Katherine Bick, the NIH deputy director for extramural research, criticized the paper's authors for their failure to "address and resolve questions of error and interpretation in a collegial manner."

Bick also voiced concern that MIT policy, which requires an individual to make an allegation of fraud in order to invoke any formal investigative process, could unnecessarily constrain junior scientists with legitimate questions about the validity of research conducted by senior scientists.

"While every disagreement about error clearly should not be escalated into an investigation of fraud, research institutions have the responsibility of providing an evironment in which scientists, especially junior ones, have the freedom to air their concerns without feeling they may be branded as troublemakers," Bick wrote.

Despite the critical remarks, the authors felt vindicated by the NIH report, according to Alfred Kildow, a spokesman for the Whitehead Institute. "The report demonstrates the accuracy of the authors' judgment that their research was worth publishing," he said.

The authors have drafted a suggested letter of correction and passed it on to the director of NIH for comment, Kildow said. Some scientific points still need to be discussed and clarified, he added.

According to Bick's memorandum, the authors initially disputed the significance of some of the errors pointed out by the panel and disagreed with the need to send a corrective letter with substitute data. However, they did agree to provide Cell with a report outlining the problems related to the reagents and assays used in the experiments.

Kildow disagreed with Bick's criticism that the authors had failed to suitably address questions of error and judgment, saying that the authors felt they had "seriously considered" the objections raised by a postdoctoral fellow which led to the investigation of the research.

History of the controversy

The paper detailed the results of NIH-sponsored research into changes in the production of antibodies by mice after certain genetic alterations. Some statements in the paper and its conclusions were disputed by a member of the research team, who claimed she was unable to duplicate the results of the experiment.

When Margot O'Toole, then a postdoctoral researcher for one of the paper's authors, attempted to bring her differences to her superior, she was rebuffed and sent to tend the breeding of laboratory mice, she claimed in testimony before a congressional subcommittee, which held hearings on the matter last April.

O'Toole accused Baltimore of obstructing her attempts to correct errors in the paper. She said Baltimore told her he would write a rebuttal if she attempted to submit a letter of correction to Cell.

Last November, the authors did submit a letter of corrections to Cell which was printed on Nov. 18, 1988. The researchers acknowledged three "incidences of misstatement," but concluded that they "are not material alterations" and "do not affect the conclusions of the paper." The report of the NIH scientific panel states that this letter did not fully explain the paper's deficiencies.

Investigative committees at MIT and Tufts found no evidence of fraud or misconduct on the part of the paper's authors, although the MIT investigation turned up a minor error deemed too insignificant to warrant a letter of correction.

The dispute eventually attracted the attention of the House Subcommittee on Oversight and Investigation, which was investigating the adequacy of institutional response to charges of scientific fraud. NIH also began an investigation of the incident shortly thereafter, although the first panel it assembled had to be dismissed because two of its three members had previously published papers with Baltimore.