IBM case highlights health fears in computer industry

Silicon Valley owes its might to front-line clean room workers such as James Moore and Alida Hernandez who worked for IBM Corp. in the 1970s and 1980s. Next week, a jury trial begins to determine whether that work caused health problems that almost killed them.

The trial will also set the stage for more open discussion about whether clean-room work has sickened thousands of other semiconductor manufacturing employees worldwide whose stories have yet to be heard.

Ample evidence points to the historical dangers of working with chemicals used in the manufacturing of chips and other components. Due to automation and increased awareness, current conditions are thought to have improved. But no comprehensive study of the long-term effects of working in a clean room has been undertaken, partly because the industry has been unwilling to supply the necessary data, critics charge. The Semiconductor Industry Association (SIA) taken preliminary steps to fund a study, but not in time to stave off a first wave of litigation.

“Generally speaking, with these kinds of chemical exposures, there is a long latency period, a long period of exposure before the harm comes to the surface. These cases would just be arising,” said Robert L. Rabin, a professor of law at Stanford Law School.

Hernandez and Moore allege that IBM failed to acknowledge higher-than-usual instances of cancer and birth defects among workers in its semiconductor and hard drive manufacturing facilities employed during the 1970s and 1980s. Hernandez, 73, endured a mastectomy after developing breast cancer, and Moore, 62, suffers from non-Hodgkins lymphoma.

The two employees worked in one of IBM’s clean rooms. Clean rooms are manufacturing facilities that are kept free of any possible contaminants that might affect the purity of computer components.

Workers in such environments were routinely exposed to chemicals that have been identified as carcinogens by several different federal and California agencies, according to court documents. Some of those chemicals were known to cause cancer during the 1980s, while others were removed from manufacturing facilities as their effects became known.

The trial starting Monday, in Superior Court for the State of California for the County of Santa Clara, does not focus on whether IBM was aware of such conditions, but on whether it withheld data from employees that IBM workers were suffering from higher rates of cancer and birth defects than normal segments of the population, said Richard Alexander, an attorney with Alexander, Hawes and Audet LLP and lead counsel for the plaintiffs.

IBM knew, or should have known, that these workers were suffering from illnesses as a result of their work at IBM, and had a legal and moral duty to inform those workers, Alexander said. His firm also represents hundreds of former IBM workers in California, Minnesota, and New York that are in the process of suing IBM.

IBM denies that it misled employees about their working conditions, and also denies that the work environment caused its former employees’ illnesses.

A potentially explosive court document dubbed the “corporate mortality file” by the plaintiffs was revealed during preparations for the trial. According to the plaintiffs, this file containing death records of IBM employees over the last 25 years or so proves that IBM was aware of elevated cancer rates, or should have been aware.

Dr. Richard Clapp, an epidemiologist at Boston University hired by the plaintiffs, analyzed the file and found “significantly” elevated instances of cancer among IBM manufacturing employees as compared to the general population, according to court documents.

IBM disputes that a sound medical diagnosis can be made from that file, said Bill O’Leary, an IBM spokesman. As in many companies, families had to submit a death certificate to IBM to collect death benefits, he said.

“It has been suggested that some people have drawn conclusions from the data that there were higher death rates. There is no way from those records that you can legitimately medically draw those kinds of conclusions,” O’Leary said.

Though the file will not be an issue at trial, as Judge Robert Baines granted an IBM motion to exclude any mention of the file to jurors, there is other evidence that might support Moore and Hernandez’s case.

In documents filed with the court, Beth Diesner-Gee reported that IBM conducted extensive employee surveys about, among other things, workplace conditions while she was a manager at IBM facilities in Tuscon, Arizona and Sunnyvale, California, in the 1980s.

Many workers complained about exposure to fumes and chemicals, Diesner-Gee reported. IBM revealed the results of one anonymous survey conducted at an unidentified location in court documents requested by the plaintiffs.

Verbatim comments from the survey raise the following health-related questions:

— “Employees at curebake and clean line are exposed daily to hexavalent and not enough research is being done to determine the extent of the exposure. Cure operators are forced to handle disks with no gloves on first cure, where they constantly come into contact with alodyne residuals that are embedded into their skins.”

— “There seems to be a few holes in the ‘right to know’ policy. I have asked some engineers some direct questions about chemicals and got fairy tales for an answer.”

IBM instructed managers to provide a stock response to employee questions about clusters of cancer or birth defects, such as “there are no data available that IBM employees suffer death from chemical exposure nor has there ever been a proven case of chemical exposure to any IBM employee,” Diesner-Gee said in the documents.

The IBM case is fueling debate that has been simmering for years.

Most companies won’t admit they maintain detailed employee health records, but the disclosure that IBM had such a system that tracked employee health information suggests that most companies do in fact keep such records, said Joe LaDou, director of the International Center for Occupational Medicine at the University of California at San Francisco.

An Intel Corp. spokesman denied that Intel maintained such a file. Representatives from Motorola Inc. and Texas Instruments Inc. also denied their companies kept the kinds of records shown in the IBM file.

Concerns date back to the 1980s

Researchers such as LaDou have been trying for years to get companies such as IBM to provide data such as what was listed in the so-called “corporate mortality file,” but the companies had steadfastly refused, LaDou said.

In 2001, after its own study of clean room workers, the SIA concluded that “there is no affirmative evidence of increased risk of cancer among U.S. semiconductor factory workers,” according to its Web site. But in the next sentence, the SIA said “insufficient data exists at the present time to conclude whether exposure to chemicals or other hazardous materials has or has not increased such a risk of cancer.”

As a result of those findings, the SIA commissioned the Worker Health Project, said Molly Tuttle, an SIA spokeswoman. Johns Hopkins University researchers are currently compiling industry data to determine if there is enough information to begin an epidemiological study next year, she said.

In later portions of the project, the SIA plans to identify opportunities to limit exposure to existing chemicals as much as possible and find ways to better manage the introduction of new chemicals into the manufacturing process, Tuttle said.

The SIA believes the Johns Hopkins study will either lay to rest any concerns about widespread health problems in clean rooms, or identify areas where improvements can be made, Tuttle said.

But few other studies have been conducted by parties unaffiliated with the semiconductor industry that examine these issues. A study conducted by the U.K.’s Health and Safety Executive in 2001 was unable to definitively state that workers at National Semiconductor (UK) Ltd. developed higher rates of cancer than the general population.

The study found evidence of increased rates of female lung cancer and breast cancer, but said that further research was needed to determine whether the illnesses were inherently work-related.

Completing that research has proved a frustrating exercise for LaDou and his colleagues. “We’ve been deprived of any research opportunities for the past eight years. There have been no cancer studies related to the entire electronics industry, because the companies were unwilling or unable to provide the data,” he said.

An awareness of dangers caused by overexposure to chemicals is nothing new. In 1988, the U.S. Department of Labor (DOL) published the “Safety and Health Guide for the Microelectronics Industry,” which reported some of the potential hazards involved in semiconductor and microelectronic manufacturing.

“The inhalation or absorption of solvents may affect the central nervous system, acting as depressants and anesthetics causing headaches, nausea, drowsiness, dizziness, complaints of irritation, abnormal behavior, general ill-feeling, or even unconsciousness. These symptoms should be viewed as visible signs of potential disease. Excessive and continued exposure to certain solvents may result in liver, lung, kidney, and reproductive damage, as well as cancer,” the report said.

Glycol ethers were voluntarily removed from the manufacturing process in 1989 after a study conducted with the University of California at Davis concluded the chemical was responsible for a slightly elevated risk of miscarriage among women working in clean rooms, the SIA’s Tuttle said.

Nevertheless, based on data compiled by epidemiology researchers at the Massachusetts Institute of Technology (MIT), the plaintiffs might have a hard time demonstrating that clean room workers as a whole suffered higher-than-usual instances of common cancers such as breast cancer and non-Hodgkins lymphoma.

The death rate from non-Hodgkins lymphoma has risen dramatically for all U.S. residents since the mid-19th century, according to data assembled and maintained by MIT at http://epidemiology.mit.edu/, said Bill Thilly, a geneticist and epidemiologist at MIT.

A study conducted in 1997 showed that an average white male born in the U.S. between 1930 and 1939 had a 0.4 percent chance of dying from non-Hodgkins lymphoma by age 65, Thilly said in an interview.

If a group of 10,000 males reached the age of 65, epidemiologists would expect 40 males from that group to die from the disease, he said. But since it is a fairly common form of cancer, the number of non-Hodgkins lymphoma cases in that group would have to exceed 85 to exclude the possibility that the disease was caused by chance, he said.

With smaller sample sizes, it’s even harder to demonstrate that disease wasn’t caused by pure chance, Thilly said.

The findings of the Johns Hopkins study commissioned by the SIA will hopefully shed more light onto the risks faced by clean room workers over the last three decades. Any damage caused by chemical exposure has already been done, while concerns going back a quarter of a century continue to linger and to lack an answer. With the IBM case going to trial, however, the debate about workplace health risks is set to get a thorough airing.