IBM responds in study dispute

In Science, the company says data for worker mortality study were inadequate

By | July 16, 2004

This week's issue of Science carries a letter from IBM presenting its side of an ongoing controversy triggered when an occupational medicine journal refused to publish a study reporting high cancer mortality in workers at the computer firm's semiconductor plants.

Last month, other researchers who were slated to contribute to a November issue of Clinics in Occupational and Environmental Medicine withdrew their submissions in protest over the Elsevier publication's refusal to include the article by Richard Clapp.

IBM spokesperson Chris Andrews told The Scientist that Clapp gained access to company data as part of a recently resolved court case in California between IBM and former employees diagnosed with cancer. Clapp, an epidemiologist at the Boston School of Public Health, was asked to review the data and testify on behalf of the plaintiffs. However, the judge eventually forbade Clapp from testifying, and the plaintiffs lost the case earlier this year.

Andrews added that Clapp signed a confidentiality agreement as part of the court case, and publishing this paper would violate the terms of that agreement.

In IBM's letter to Science, Scott R. Brooks responds to the journal's May 14 article about the controversy, saying that Clapp's data are “incomplete and inadequate for reliable study,” and that Clinics is not a peer-reviewed journal. He adds that that Clinics' guest editor, Joe LaDou, of the University of California School of Medicine at San Francisco, was a plaintiffs' expert in litigation against IBM. The fact that Clapp's study was driven by litigation raises “serious questions about its objectivity,” he says.

Brooks writes that IBM has been funding its own mortality and cancer incidence study since 1999, headed by researchers at the University of Alabama and Harvard. Alabama's Elizabeth Delzell confirmed to The Scientist that she is participating in the study, but declined to comment on ongoing research.

IBM also declined The Scientist's request for comment on the letter.

LaDou told The Scientist that he was asked to serve as a historical expert to present background information during the deposition for the California case because he was one of the few people in practice in Silicon Valley during the late 1960s, when many of the allegedly detrimental exposures occurred. However, he never testified in the actual trial.

“I have never been an expert witness in my entire career,” he said.

He added that the journal itself is not peer-reviewed, but as guest editor, he was told to “review” articles, so he asked all contributors to review each others' papers before submitting them to the journal. “It was a less formal process,” he said.

Indeed, in a response to IBM's letter, the author of the May 14 Science article, Dan Ferber, writes that he asked three epidemiologists with no relationship to the litigation to look at the study, and all said the results were preliminary but “scientifically valid and deserving of further study.”

Robert Harrison, of the University of California, San Francisco, who was slated to contribute an article on medical monitoring to the November issue of Clinics, told The Scientist that he spent 21 days answering questions during the recent IBM deposition, and this letter was the first he'd heard of any mortality study that IBM was conducting itself.

He said that he did not believe that data obtained in the context of litigation represented a “fatal conflict,” and argued that IBM should let people see Clapp's study so they can decide what is fair for themselves.

“What is IBM afraid of with the publication of Dr. Clapp's article?” he asked.

Donald Kennedy, Science's editor-in-chief, asks the same question in an accompanying editorial, also noting that this debate fits into the ongoing “war” about science's role in law. On one side are some environmentalists and plaintiffs' representatives who believe that judges tend to bar testimony from expert witnesses. These people are often opposed by advocates for tort reform, who claim that plaintiffs' bad studies are fueling oversized jury awards.

“The decisions to disallow the testimony and to shield the work behind a gag order may have served the cause of justice, but the scientific community and the public have been prevented from reaching their own conclusions about an important matter,” he writes.

As for Clapp, he said that, as far as he knows, the data he received from IBM were close to 100% complete, and that Alabama's Delzell published a previous article using the same dataset.

He added that his lawyers argue that he can release the data without violating the terms of the confidentiality agreement, and two on-line, peer-reviewed journals have requested his article, along with at least one other print journal.

However, he told The Scientist that he doesn't plan to publish the article until everyone involved gives him the okay. “I don't want to take that step without having my attorneys and all the other attorneys agree,” he said.

Elsevier has not yet made a decision regarding what they will include in the November issue of Clinics, said spokesperson Eric Merkel-Sobotta.

Popular Now

  1. GM Mosquitoes Closer to Release in U.S.
  2. Judge Recommends Ruling to Block Internet Access to Sci-Hub
  3. Opinion: Microbiology Needs More Math
  4. Do Pathogens Gain Virulence as Hosts Become More Resistant?