By Therese Poletti
SAN JOSE MERCURY NEWS
Posted on Sun, Feb. 01, 2004
Sometime in the next month or two, the jury in the IBM toxics trial will render a verdict. But despite exhaustive testimony from a battalion of scientists, a fundamental question will go unanswered: Do the hundreds of thousands of people who have worked in computer industry clean rooms since the 1960s face a higher risk of developing cancer?
The reason: Neither plaintiffs Alida Hernandez and James Moore -- former IBM employees who contend they contracted cancer from working in the computer giant's San Jose plant -- nor IBM can cite a single definitive study on cancer rates among semiconductor industry workers.
After two decades of controversy over clean-room safety, the state of the science about the health risks of exposure to toxic chemicals used in computer manufacturing remains sketchy.
A few small-scale studies have been done over the years. But efforts to conduct a comprehensive review of death and disease rates for clean-room workers have been stymied or are proceeding in fits and starts.
Part of the problem lies with the industry's well-founded fear of litigation and a reluctance to participate in research that could prove costly in the courtroom, scientists said. Then there are the not-inconsiderable logistical hurdles of tracking down tens of thousands of former clean room employees, their medical records and work histories.
According to the U.S. Department of Labor's Bureau of Labor Statistics, about 66,600 people worked in semiconductor production in the United States in 1972. That number jumped to 103,200 by 1987 and 121,700 in 2002. Because a decade or more may pass between a worker's exposure to carcinogenic chemicals and the onset of cancer, scientists -- and lawyers -- have focused on individuals who worked in the industry between the 1960s and 1990s.
"We have been promised a cancer study for I don't know how long," said Joe LaDou, director of the International Center for Occupational Medicine at UC San Francisco's School of Medicine.
LaDou, who gave a deposition to the plaintiffs in the IBM toxics case, is one of the researchers who has been pushing the industry to do more health studies.
Semiconductor company executives argue that their industry is one of the safest. In fact, the Bureau of Labor Statistics released data last month showing that in 2002 semiconductor employees had one of the lowest injury rates among manufacturing workers.
Chip manufacturing employees work in particle-free "clean" rooms, where they wear full-body "bunny suits" and other gear designed so nothing contaminates the silicon wafers.
These workers are surrounded by a brew of toxic chemicals, such as benzene, trichloroethylene, arsenic, methylene chloride and acetone.
In the 1970s, '80s and early '90s, workers had more direct contact with the many chemicals used in manufacturing chips, disk drives and other electronic parts. IBM trial plaintiffs Hernandez, 73, and Moore, 62, both testified about daily interaction with acetone -- used to clean disk drives -- and trichloroethylene, used to clean printed circuit boards.
Others who worked in the industry during that era tell similar stories. "We used dipping tanks that had vapors, and everyone in manufacturing was exposed to that," said Sylvia Anderson, 66, a retired San Jose electronics worker, describing the process for cleaning off printed circuit boards in the 1970s and '80s. "The average worker didn't get too much information."
Since then, the manufacturing process has become increasingly automated and workers these days do not directly handle as many chemicals.
One of the first significant industry health studies was published in 1995. Commissioned by the Semiconductor Industry Association, the study's authors spent four years investigating miscarriage rates by screening 7,269 chip manufacturing workers from 14 companies. For 904 women selected for further investigation, the study found a higher rate of miscarriage than in women working elsewhere in the semiconductor industry. The study recommended the industry stop using certain glycol ethers, which were linked to miscarriage.
A 1996 study by Johns Hopkins University for IBM reviewed miscarriage rates between 1980 and 1989 for 561 employees at two of the company's chip plants. Researchers found workers' potential exposure to ethylene glycol ethers was associated with a higher risk of miscarriage.
Two years later, in 1998, a United Kingdom health agency began studying clean-room workers' cancer rates at a National Semiconductor plant in Greenock, Scotland, after pressure from a workers' support group attracted the attention of the British Parliament.
That study analyzed employment and mortality records of 4,388 National Semiconductor employees.
Researchers found more cases than expected of female lung cancer, female stomach cancer and breast cancer. They also discovered more incidence of brain cancer in men than anticipated.
The Semiconductor Industry Association subsequently formed a panel to "assess whether more extensive evaluation" of chip industry health risks was warranted. That group took a year and a half to determine that more study was needed.
In March 2002, the Semiconductor Industry Association said it would follow the panel's recommendations and "conduct a preliminary review to determine if it is possible to conduct and go forward with a meaningful retrospective epidemiological study."
Almost two years later -- the time it takes chip giants like Intel to double the number of transistors on a single computer chip -- the industry still has not revealed whether it will initiate the study.
When asked to talk about the feasibility study, the lead researcher at Johns Hopkins University for the Semiconductor Industry Association referred all questions back to the group.
"We have not yet assembled any of the information in a format that would allow us to answer your question," wrote Genevieve Matanoski, Johns Hopkins' program director of Occupational and Environmental Epidemiology, in an e-mail.
Semiconductor Industry Association spokeswoman Molly Tuttle said the researchers should issue a report to her group in February or March.
"People ask why this is taking so long, critics accuse the industry of foot-dragging. It's not the case," said IBM spokesman Bill O'Leary. "They have a more ambitious task; they are looking at environments across companies."
IBM is in the process of conducting its own study on cancer and death rates among 10,000 employees at plants in Burlington, Vt.; East Fishkill, N.Y.; and San Jose. Researchers at the University of Alabama in Birmingham are managing the study for IBM. University of Alabama researchers declined to comment on their work for the company.
"This is a retrospective study that goes back 30 years," O'Leary noted.
But some occupational health researchers are suspicious of any research overseen by industry.
"The companies have long figured out that they can control the research, they can pick the scientists ... they control the research protocol," said LaDou of UC San Francisco.
But to do any kind of major study, the scientific community needs to work with the electronics industry.
IBM, for example, has kept corporate mortality records for its pension plans, and for a few years it maintained a database that tracked worker health history and chemical exposures.
There are also other difficulties in doing retrospective medical studies, such as intermittent workers and inconsistent data from different chip companies.
A large-scale study proposed in 1998 by the U.S. Environmental Protection Agency and the California Department of Health Services failed to get off the ground because of differences with the industry. The idea was to cross-check the California cancer and birth defects registry with employee records provided by semiconductor companies.
O'Leary of IBM said the industry did not think the EPA should be doing such a study and that such research was the purview of the Occupational Safety and Health Administration. By then, IBM was starting its own cancer study with the University of Alabama. The computer giant had been sued in 1996 by plaintiffs who worked at its East Fishkill, N.Y., plant and developed cancer or gave birth to children with birth defects.
Now the question is whether the Semiconductor Industry Association will fund a comprehensive study of its own.
"This is an industry that is fully equipped to gather, store and analyze information. That is one of their specialties," said Amanda Hawes, one of the plaintiffs lawyers involved in the San Jose and East Fishkill litigation. "We are now in 2004 and waiting for the first study on cancer (and) semiconductor workers in the U.S."
copyright- San Jose Mercury News - 2004
