IBM says study is flawed
By Kim Howard
Workers in the computer manufacturing industry are at greater risk of death due to cancer, according to a recently published study – a claim disputed by International Business Machines Corp., the source of the data used for the study.
Published Oct. 19 in the journal “Environmental Health: A Global Access Science Source,” the study was based on 31,941 IBM employees who died between 1969 and 2001 and had worked at least five years. Subgroups of workers at plants in Vermont, New York, Minnesota and California also were analyzed. Boston University environmental health professor Richard Clapp authored the study.
The study found higher levels of deaths from cancers of the brain, central nervous system, kidney, skin and pancreas among male manufacturing workers. Among women working in manufacturing, the numbers of deaths from kidney cancer and cancer of lymph and blood cell tissue were elevated, the study says.
Clapp said in an interview that though the written study does not detail plant-specific data, his analysis found that workers at the Essex Junction, Vt., plant showed elevated levels of brain cancer. About 2,000 employees currently work in manufacturing at that plant, according to an IBM spokesman.
Clapp said the study is not “conclusive proof” that there is a connection between the chemicals used in chip processing and increased rates of death from cancer among computer manufacturing workers.
“We can’t say exactly what it is that might be causing excess deaths,” Clapp said. “We don’t know about individual chemicals. We can’t isolate out solvents… But it’s more evidence.”
Clapp’s work indicates that the few studies that do exist about workers in computer and semiconductor manufacturing showed excesses of certain cancers, but that the study samples were small. Clapp’s analysis, however, is based on the largest database of its kind available to date, according to the study.
Earl Mongeon, who works in manufacturing at IBM’s Essex Junction plant, said he is not surprised by the study’s assertion that manufacturing workers exposed to solvents and other chemicals are at higher risk for cancer.
“The stuff we work with at IBM is pretty nasty,” said Mongeon, who is vice president of IBM’s workers’ union. “I used to hand-dip the stuff in the acid tanks.”
Mongeon said substances used include sulfuric acid, hydrochloric acid, hydrofluoric acid, nitric acid, hydrogen peroxide, ammonium hydroxide, and gases from arsenic. Mongeon has worked at IBM for 28 years.
IBM spokesman Ari Fishkind said he could not respond to questions about Clapp’s study beyond a written statement produced by the company. The statement says the study is “based on flawed methodology and woefully incomplete data.”
The basis of the study is what is known as IBM’s “corporate mortality file,” data on IBM employees who died between 1969 and 2001 and who qualified for pension and death benefits.
IBM was required to produce the data for analysis as part of a lawsuit. Workers in California sued IBM, alleging manufacturing conditions led to cancer. The data on which Clapp’s study was based ultimately was not allowed as evidence in the trial. IBM won the case.
Clapp acknowledges in the study that the plaintiffs’ attorneys obtained the raw data from IBM, and that he was paid to review it. He says it does not affect how the data was analyzed, however.
“The law firm did not design or conduct the study, nor review or approve the manuscript,” Clapp wrote in the study. “The author received no remuneration for the preparation of this manuscript.”
IBM initiated its own lawsuit to prevent publication of Clapp’s study, but lost.
Clapp disagrees with IBM’s assertions that the study’s data is incomplete and the methodology flawed. Clapp said the study was published in a peer-reviewed journal, meaning professionals in his field reviewed and required changes to the article prior to its publication.
One of the reviewers was Robert Herrick, a co-author of a previous IBM commissioned study on cancer rates. Herrick’s pre-publication comments say “in general I think this is a very interesting manuscript that is a well-written presentation of an important set of findings.” Herrick required a number of changes to the manuscript prior to its publication.
IBM also said in its statement that a study of IBM plants by researchers at the University of Alabama at Birmingham and Harvard University (the latter conducted by Herrick) found among its workers overall lower mortality rates and a lower risk of cancer than the general population. A review of those studies published by “Occupational Health and Safety,” a national magazine, indicates employees in certain work groups were at elevated risk of specific cancers, however.
Clapp, who worked for nine years as the director of the Massachusetts Cancer Registry, said the reason for his research is simple.
“My goal in all of this as a public health research person is to figure out a way to prevent unnecessary deaths,” he said.