More Errors Found In Work of Med School Prof

The reports of a former Medical School professor who confessed to falsifying research in 1981 were riddled with errors beyond his admitted fraud, according to a study published yesterday of the nearly 100 documents.

Two doctors at the National Institutes of Health (NIH) examined the papers, book chapters and abstracts that former Medical School professor John R. Darsee co-authored with 47 scientists. The study also criticized the work of many of Darsee's co-authors.

After Ned Feder '48 and Walter W. Stewart '67 conducted their study in 1983, they circulated it privately to about 100 top researchers and submitted it to several scientific journals. The study was published in yesterday's edition of Nature, a weekly British journal.

Feder and Stewart said that Darsee's co-authors failed to make their methodology clear, included large numbers of errors in their data and published results without making it clear that some of the data had been published before.

One of Darsee's co-authors, Blumgart Professor of Medicine Eugene Braunwald attacked the study in a commentary which accompanies it in Nature and in an interview yesterday.

"The study is grossly flawed and inaccurate. The republication charge is absurd. The relationship between my two studies is indicated in the opening paragraph of the second article," he said yesterday. "Many of the errors they cite are not errors at all and others were trivial or typographical mistakes."

But Stewart said in an interview yesterday, "We're not saying the errors themselves were dangerous to the scientific community. The presence of a large number of visible errors in the studies suggests a lack of care on the part of the co-authors."

Braunwald also said that the study does not present an accurate picture of the scientific world because it deals solely with papers Darsee worked on.

Darsee, who was subsequently dismissed from Harvard and stripped of his medical license, confessed to falsifying data heart attack experiments in 1981. Most of the papers Stewart and Feder studied have already been retracted.