Environmental Health: = A Global Access Science Source | Full text | Mortality among US = employees of a large computer manufacturing company: 1969-2001

Discussion<= /FONT>

Background

Previous=20 studies suggested increased cancer incidence and mortality in = workers=20 exposed to solvents and other chemicals in computer manufacturing = jobs.=20 Most previous studies were of small cohorts and findings were=20 inconsistent. A lawsuit involving a large U.S. company produced a = data=20 file for analysis. This study sought to elucidate patterns of = mortality in=20 workers who were engaged manufacturing computers and related = electronic=20 components in the largest database available to date.

Methods

A = proportional=20 mortality and proportional cancer mortality analysis of deaths in = eligible=20 workers between 1969 and 2001 was carried out, with U.S. = population=20 mortality data as the standard for comparison. Mortality and work = history=20 data was from corporate mortality and work history files produced = during=20 litigation and standard U.S. and state mortality files. The study = base=20 comprised 31,941 decedents who died between 1969 and 2001, who had = worked=20 for at least five years and whose death information was collected = in the=20 corporate mortality file. Proportional mortality ratios (PMRs) and = Proportional Cancer Mortality Ratios (PCMRs) and their 95% = confidence=20 intervals were computed for 66 causes of death in males and = females.

Results

PMRs for all=20 cancers combined were elevated in males (PMR =3D 107; 95% CI =3D = 105=E2=80=93109) and=20 females (PMR =3D 115; 95% CI =3D 110=E2=80=93119); several = specific cancers and other=20 causes of death were also significantly elevated in both males and = females. There were reduced deaths due to non-malignant = respiratory=20 disease in males and females and heart disease in females; several = specific cancers and other causes of death were significantly = reduced in=20 both males and females. Proportional cancer mortality ratios = (PCMRs) for=20 brain and central nervous system cancer were elevated (PCMR =3D = 166; 95% CI=20 =3D 129=E2=80=93213), kidney cancer (PCMR =3D 162; 95% CI =3D = 124=E2=80=93212), melanoma of skin=20 (PCMR =3D 179; 95% CI =3D 131=E2=80=93244) and pancreatic cancer = (PCMR =3D 126; 95% CI =3D=20 101=E2=80=93157) were significantly elevated in male manufacturing = workers. Kidney=20 cancer (PCMR =3D 212; 95% CI =3D 116=E2=80=93387) and cancer of = all lymphatic and=20 hematopoietic tissue (PCMR =3D 162; 95% CI =3D 121=E2=80=93218) = were significantly=20 elevated in female manufacturing workers.

Conclusion

Mortality was=20 elevated due to specific cancers and among workers more likely to = be=20 exposed to solvents and other chemical exposures in manufacturing=20 operations. Due to lack of individual exposure information, no = conclusions=20 are made about associations with any particular = agent.

Outline

Background

Background<= /FONT>

Discussion<= /FONT>

In = 2002,=20 plaintiffs' attorneys in a lawsuit obtained computerized files of = deaths=20 and work histories for US employees of a large company which = manufactured=20 electronic office equipment; mainframe and personal computers; = computer=20 parts, components, and accessories; and software products. The = files=20 contained data on employees who had qualified for pension and = death=20 benefits and died in the US between the years 1969 and 2001. A = unique=20 identifier was included in both files so that death and work = history data=20 for individual decedents could be linked.

Few studies of=20 workers in computer manufacturing and semiconductor fabrication = are=20 available in the published literature. A study of workers at a = Scottish=20 semiconductor plant was conducted by the government occupational = health=20 authority in 2001 [1,2]. This = study=20 showed statistically significant excesses of lung, stomach, and = breast=20 cancers among females and an excess of brain cancer among males; = however,=20 the number of employees in this plant was relatively small and the = number=20 of cases of cancer was limited. Recently, an update of a UK = semiconductor=20 workers study was published which was also small and had few = significant=20 findings with regard to cancer. A previous excess incidence of = melanoma=20 persisted in the update, along with excess rectal and pancreatic = cancer in=20 females [3]. Two = studies of=20 workers who manufactured main boards and integrated circuits in an = electronics company in Taiwan were published [4,5]. In the = cancer=20 incidence analysis there was a significant excess of breast cancer = in=20 female electronics company workers when compared to two other = groups of=20 industrial workers [5]. An = analysis of=20 mortality in three U.S. semiconductor and storage = device-manufacturing=20 plants identified significant excess central nervous system cancer = in one=20 plant and excess prostate cancer in workers in another plant [6].

There are a=20 variety of exposures to chemicals, metals (especially arsenic, = nickel and=20 chromium), and electromagnetic fields (especially ultraviolet = light,=20 radiofrequency, and, in one process, x-ray radiation) in computer = and=20 semiconductor manufacturing [7]. There = have been=20 many studies of workers with these exposures in other industries. = For=20 example, trichloroethylene is used in many industries and there is = a large=20 literature on its health effects [8]. = Similarly, the=20 health effects of methylene chloride, isopropyl alcohol, methyl = ethyl=20 ketone, Freon, as well as metals and electromagnetic radiation = have been=20 studied in other groups of workers and in experimental animals [9,10]. = Furthermore,=20 various US and international agencies have classified several of = these=20 chemicals as to their carcinogenicity [11,12].

Based upon the=20 available literature, workers exposed to the chemicals and = processes in=20 computer and semiconductor manufacturing would be expected to have = elevated mortality from non-Hodgkin's lymphoma [8,11], = brain cancer=20 [9], kidney = cancer [8,11], lung = cancer [2,11], and = breast [5] cancer. = Additionally, exposures in the computer and semiconductor = manufacturing=20 industries are complex and have changed as new processes and = materials=20 were introduced. Thus, the mortality experience of computer and=20 semiconductor manufacturing workers would be expected to vary over = time=20 due to changing exposures.

We = evaluated=20 these expectations and examined the mortality among US employees = of a=20 large computer manufacturing company by calculating overall and = sex-,=20 age-, and time-specific proportional mortality ratios (PMRs) and=20 proportional cancer mortality ratios (PCMRs). We performed = analyses for=20 all eligible decedents and the subgroup of decedents who were ever = employed 30 days or more in manufacturing.

Outline

Methods

Background<= /FONT>

Discussion<= /FONT>

The context of=20 this analysis is unusual because the data were obtained as part of = litigation and the company's attorneys resisted making the files=20 available. A court order forced the production of the files, and = minimal=20 documentation was provided about the content and coding of the = files. Due=20 to the context in which we received these data, we were not able = to obtain=20 additional documentation, access the original or complementary = data, or=20 contact company personnel knowledgeable about the data. Since we = could not=20 resolve problems in the data, we spent considerable time = developing=20 logical rules for handling incomplete and apparently erroneous=20 information. These constraints limited the data available for = subgroup=20 analyses.

Corporate Mortality = File

At = the time it=20 was provided to plaintiffs' attorneys, the IBM Corporate Mortality = File=20 (CMF) maintained by the company had 33,730 records. These records = were for=20 decedents who had been employed for five years or more, who were = actively=20 employed or receiving retirement or disability benefits at time of = death,=20 and whose families had filed a claim for death benefits. CMF data = required=20 for analysis were the identifier, sex, birth date, death date, and = underlying cause of death.

We = reviewed=20 the numbers of deaths by year of death. Based on the apparent = degree of=20 completeness of the CMF by year of death, we defined the study = period as=20 1969=E2=80=932001. We excluded duplicate records and records for = which the sex,=20 birth date, and/or death date remained unknown or discrepant. Of = the=20 33,730 records in the CMF, 1,456 records were excluded for these = reasons.=20 Processing of the work history data (see below) identified 20 = additional=20 CMF records with discrepant birth and/or death dates and these = also were=20 excluded. We then excluded 313 CMF records for which the year of = death was=20 before 1969 or after 2001.

The final=20 number of CMF records used for analysis was 31,941 and comprised = 27,272=20 males and 4,669 females.

Work History Data

The work=20 history data set had a total of 586,258 records representing = 18,928=20 decedents. Work history data required for analysis were the = identifier,=20 effective date, position code, location code, and text notes = indicating a=20 key employment event such as leave of absence, return from leave = of=20 absence, separation, retirement, or death.

We = added the=20 birth and death dates from the CMF to the work history data set. = We then=20 identified irrelevant, missing, invalid, and inconsistent work = history=20 data to the extent feasible.

We = deleted all=20 work history records for decedents that were excluded from the CMF = (see=20 above). Likewise, we deleted all work history records with = irrelevant,=20 missing, invalid or inconsistent data. Upon completion of these = steps, the=20 work history data set included a total of 375,563 records = representing=20 10,219 decedents. There did not appear to be any systematic = pattern in the=20 missing data that would produce bias in the analysis of this = subset of=20 decedents.

Cause of Death = Coding

The underlying=20 causes of death were coded to the Ninth Revision of the = International=20 Classification of Diseases (ICD 9).

Comparison = Population

We = selected=20 the US as the most appropriate comparison population for the = overall CMF=20 analyses and we used it as the comparison population for the = exposure=20 subgroup analyses for consistency. We obtained cause-, sex-, age-, = and=20 time-specific mortality data, adjusted to ICD 9, for the US = comparison=20 population from the Mortality and Population Data System (MPDS) at = the=20 University of Pittsburgh [13]. We = used=20 default categories for sex, age, and time, and default and = "custom"=20 categories for cause of death. The mortality data were not = categorized by=20 race because race was not known for the decedents included in the = CMF.

Analysis

We = analyzed=20 the data with the Occupational Cohort Mortality Analysis Program = (i.e.,=20 OCMAP-PLUS version 3.10) software package from the University of=20 Pittsburgh [13]. In = addition to=20 the 63 OCMAP-PLUS default cause of death categories, we defined = the=20 following six "custom" cause of death categories: non-Hodgkin's = lymphoma,=20 multiple myeloma, diseases of the nervous system, multiple = sclerosis,=20 Parkinson's disease, and amyotrophic lateral sclerosis.

Standardized=20 proportional mortality ratios (PMRs) were calculated from the = ratio of the=20 observed number of deaths due to a particular cause among the = workers to=20 the expected number of deaths based on the age-, sex-, time-, and=20 cause-specific proportional mortalities of the US population. The = expected=20 number of deaths was computed by multiplying the proportional = mortalities=20 of the US population by the total number of deaths for the = workers. The=20 PMRs were standardized on age, sex, and time. Standardized = proportional=20 cancer mortality ratios (PCMRs) also were calculated. For the = PCMRs,=20 site-specific cancer proportional mortality was computed as a = proportion=20 of all cancer mortality. The underlying cause of death was used = for=20 calculation of the PMRs and PCMRs. Workers with an unknown cause = of death=20 were included in the "all causes of death" and "unknown causes"=20 categories. The associated 95% Poisson confidence intervals (95% = CIs) for=20 the PMRs and PCMRs were calculated as well.

Exposure Subgroups

We = defined the=20 following exposure subgroups for both men and women:

=E2=80=A2 ever worked=20 in manufacturing for 30 days or more;

=E2=80=A2 ever worked=20 at a selected California, Minnesota, New York and Vermont location = for 30=20 days or more.

The latter=20 four state sub-groups were defined for analyses of specific plants = of=20 interest in the litigation (data not shown).

Outline

Results

Background<= /FONT>

Discussion<= /FONT>

Figures

Figure=20 1
Comparison of = SMRs and=20 PCMRs

Tables

Table=20 1
Number and = Percent of=20 Decedents with Work History by Year of = Death

Table=20 2
Cause-Specific=20 Proportional Mortality Ratios (PMRs) for = Males

Table=20 3
Cause-Specific=20 Proportional Cancer Mortality Ratios (PCMRs) for=20 Males

Table=20 4
Cause-Specific=20 Proportional Mortality Ratios (PMRs) for = Females

Table=20 5
Cause-Specific=20 Proportional Cancer Mortality Ratios (PCMRs) for=20 Females

Table=20 6
Cause-Specific=20 Proportional Cancer Mortality Ratios (PCMRs) for Male = Manufacturing=20 Workers

Table=20 7
Cause-Specific=20 Proportional Cancer Mortality Ratios (PCMRs) for Female=20 Manufacturing = Workers

The number of=20 decedents and the percent of decedents with available and usable = work=20 history information are shown in Table 1 by = year of=20 death and sex. Of the 27,272 male decedents, 9,219 (33.8%) had = work=20 histories. For the 4,669 female decedents, 1,000 (21.4%) had work=20 histories. In general, work history information was most complete = for the=20 years 1980 =E2=80=93 1999 and was more complete for men than = women. There were=20 3,652 male and 723 female decedents whose work history information = indicated that they worked 30 days or more in manufacturing.

The all cause=20 average age at death was 64.5 years for male decedents and 63.4 = for female=20 decedents. The all cancer average age at death was 63.9 years for = male=20 decedents and 59.5 for female decedents.

Cause-specific=20 PMRs and 95% CIs for male decedents in the CMF were calculated and = the=20 results for major causes of death are summarized in Table 2. = There were=20 27,272 deaths from all causes. The PMR for all cancers was = significantly=20 greater than 100. There were 7,697 cancer deaths and 7,206 were = expected=20 (PMR =3D 107; 95% CI =3D 105, 109). This excess was due to = statistically=20 significant excesses for cancer of the digestive organs and = peritoneum,=20 testicular and other male genital cancer, kidney cancer, malignant = melanoma of the skin, brain and central nervous system cancer, = thyroid and=20 other endocrine cancer, and lymphatic and hematopoietic = cancer.

In = addition,=20 statistically significant excesses were observed for benign = neoplasms,=20 nervous system diseases, ischemic heart disease, and all other = causes of=20 death, although the excess for ischemic heart disease was quite = small.=20 Statistically increased PMRs for multiple sclerosis, Parkinson's = disease,=20 and amyotrophic lateral sclerosis contributed to the excess for = diseases=20 of the nervous system. There also were deficits of several causes = of death=20 as estimated by the PMR method.

In = Table 3, = we present=20 the cause-specific PCMRs and 95% CIs for males in the CMF. There = were=20 7,697 deaths from all cancers in this group. Except for testicular = and=20 other male genital organ cancer, all of the significantly = increased PMRs=20 for specific cancers also had significantly increased PCMRs, but = the=20 magnitudes of the PCMRs were less than those for the PMRs. The = testicular=20 cancer PMR was statistically significantly increased, whereas the = PCMR was=20 not (PCMR =3D 91; 95% CI =3D 63, 130).

The=20 cause-specific PMRs and 95% CIs for the 4,669 female decedents in = the CMF=20 also were calculated and the results for major causes of death are = summarized in Table 4. = The PMR for=20 all cancers was significantly greater than 100 for women as well. = There=20 were 1,667 cancer deaths and 1,454 were expected (PMR =3D 115; 95% = CI =3D 110,=20 119). This excess was due to statistically significant excesses = for=20 respiratory system cancer, breast cancer, cancer of other female = genital=20 organs, brain and central nervous system cancer, and lymphatic and = hematopoietic cancer. A statistically significant excess also was = observed=20 for all other causes of death. As in males, there were = statistically=20 significant deficits for several causes as estimated by the PMR=20 method.

In = Table 5, = the=20 cause-specific PCMRs and 95% CIs for females in the CMF are shown. = There=20 were 1,667 deaths from all cancers in this group. The PCMR for = breast=20 cancer (PCMR =3D 115; 95% CI =3D 106, 125) remained significantly = elevated,=20 while the PCMRs for several other cancers were no longer elevated = compared=20 to their respective PMRs. These cancers included the following:=20 respiratory system cancer; cancer of the trachea, bronchus, and = lung;=20 cancer of other female genital organs; brain and central nervous = system=20 cancer; lymphatic and hematopoietic cancer; leukemia and = aleukemia;=20 non-Hodgkin's lymphoma; and all other lymphatic and hematopoietic = cancer.=20 Furthermore, the PCMRs for cancer of the buccal cavity and = pharynx, and=20 cancer of the digestive organs and peritoneum were significantly = decreased=20 compared to the corresponding PMRs.

In = Table 6, = PCMRs are=20 presented for males in the CMF who worked in manufacturing for 30 = days or=20 more. There were 1,180 deaths from all cancers in this group. In=20 comparison to the PMRs, only the PCMRs for pancreatic cancer (PCMR = =3D 126;=20 95% CI =3D 101, 157), kidney cancer (PCMR =3D 162; 95% CI =3D 124, = 212),=20 malignant melanoma of the skin (PCMR =3D 179; 95% CI =3D 131, = 244), and brain=20 and central nervous system cancer (PCMR =3D 166; 95% CI =3D 129, = 213) remained=20 statistically significant.

In = Table 7, = PCMRs are=20 presented for female manufacturing workers. All cancers accounted = for 302=20 deaths in this group. The PCMRs for kidney cancer and the other = lymphatic=20 and hematopoietic cancers (i.e., leukemia and aleukemia and all = other=20 lymphatic and hematopoietic cancer) were significantly increased = in this=20 subgroup.

In = Fig. 1, = we compare=20 the PCMRs for males and females combined who worked 30 days or = more in=20 manufacturing in the CMF analyses with the SMRs from the combined = CA, NY=20 and VT plants analyzed by other authors [[6], Table = 2]. = Although=20 there is considerable overlap in the two studies, the time = periods,=20 numbers of deaths, and sources of information are different, so = only a=20 graphical comparison of results for specific cancers is=20 presented.

Outline

Discussion

Background<= /FONT>

Discussion<= /FONT>

We = found an=20 overall excess mortality for all cancer in males and females in = the entire=20 corporate mortality data. The PMRs for all sites were = significantly=20 elevated when comparing the computer company workers to the US = population=20 experience during the time period 1969=E2=80=932001. In addition, = there was excess=20 mortality in both males and females due to several types of = cancer.

Because of the=20 potential for lowered mortality due to major causes such as = ischemic heart=20 disease or respiratory disease, we calculated PCMRs, which focused = specifically on cancer mortality. The PCMR is less susceptible to=20 distortion by the "healthy worker effect," and focuses on the = health=20 outcome of most interest in these analyses. The results of the = PCMR=20 analyses in males were for the most part quite similar to the PMR = results=20 for the same cancers. For cancer of the testis, however, the PCMR = was 91=20 (95% CI 63,130) as compared to a PMR of 146. In females, the PCMR = for=20 breast cancer was significantly elevated, while the PCMRs for = several=20 other cancers were not, compared to their respective PMRs. A = recent SMR=20 analysis of workers at three of the plants over an earlier time = period in=20 this large company [6] = indicated that=20 there was a substantial "healthy worker effect" although the data = to do an=20 evaluation of this in our dataset was not available to us. From a = graphic=20 comparison of the SMRs from one table in this analysis with our = nearest=20 equivalent PCMR results, provided in Figure 1, = it appears=20 that the overall pattern is very similar in the two mortality=20 analyses.

Cancer=20 mortality patterns in the general population over the time period=20 1969=E2=80=932001 changed somewhat. For example, lung cancer = mortality increased=20 dramatically in US females during this time period while cervical = cancer=20 mortality decreased. In males, lung cancer mortality increased in = the=20 early part of the time period, but began to decline in the 1990s.=20 Accordingly, we examined the workers' mortality experience in = five-year=20 time intervals, and we looked at patterns by five-year age groups = during=20 these intervals. The excesses of some cancers were greater in the = younger=20 age groups and in the 1970s and 1980s rather than the latter years = of the=20 study period. These patterns are consistent with a work-related = cancer=20 mortality effect that has diminished in recent years.

Additional=20 analyses of specific subgroups of decedents were carried out, = primarily=20 because of concerns that toxic exposures used in manufacturing, = such as=20 solvents, metals, electromagnetic radiation, and the components of = photoresist might be associated with increased cancer mortality. = For=20 example, numerous halogenated solvents are used in computer = manufacturing=20 [1,2,10] and = some of=20 these are associated with increased risk of kidney cancer and = cancer of=20 the lymphatic and hematopoietic tissue in workers [8]. No = information=20 on individual exposure to specific chemicals or processes was = available=20 for this analysis. Accordingly, we examined the mortality patterns = among=20 decedents for whom work histories were available and grouped them = into=20 manufacturing jobs and presented results for this sub-group.

The overall=20 PMR and PCMR analyses for manufacturing workers revealed a pattern = that=20 was similar to the pattern in all workers combined. In most = instances, the=20 PMRs and PCMRs that were elevated in the total CMF were somewhat = more=20 elevated in the manufacturing workers. The decreased mortality due = to=20 respiratory system cancer in males in the CMF was similar in the=20 manufacturing subgroup. However, all the manufacturing subgroup = analyses=20 were based on fewer deaths and were therefore statistically less = stable.=20 In males employed in manufacturing, there was consistent increased = mortality from cancer of digestive organs, kidney cancer, melanoma = of skin=20 and cancer of the brain and central nervous system. The PCMRs for = several=20 of these cancers were significantly elevated and there was also = excess=20 mortality due to non-Hodgkin's lymphoma within the broader = category of=20 lymphatic and hematopoietic malignancies. Among female = manufacturing=20 workers, we found the breast cancer PMR to be slightly elevated, = while the=20 PMRs for kidney cancer and lymphatic and hematopoietic tissue = malignancies=20 were significantly elevated but based on few deaths. The female=20 manufacturing workers PCMRs were based on very few deaths and = followed a=20 similar pattern.

Mortality=20 analyses such as the one we conducted are limited by the = availability of=20 information on death certificates, and typically this does not = permit=20 controlling for confounding factors such as prior occupational = exposures,=20 non-occupational exposures, or cigarette smoking [14]. = Similarly, the=20 cause of death information is often limited and may be incomplete = for=20 certain types of cancer. These limitations are well-documented and = apply=20 to these CMF analyses, as well. Work history information available = for=20 grouping manufacturing workers or locations was also incomplete, = but we=20 did not detect any systematic pattern in the missing data that = would=20 produce biased estimates of effect. We examined the overall = patterns of=20 mortality among those for whom work history information was = available and=20 they appeared to be similar to those for whom this information was = not=20 available. Consequently, we do not consider the manufacturing = analyses to=20 be distorted in any important way; the lack of information reduces = the=20 number of deaths available for inclusion in these analyses and = therefore=20 reduces the power to detect some associations.

As = noted=20 above, this study was conducted in the context of litigation, so = we were=20 unable to communicate with the record providers to clarify = inconsistencies=20 or fill gaps in the information provided. Here again, we do not = consider=20 this to be a significant source of bias, but it limits the power = of the=20 sub-group analyses. If additional work is carried out on this = database, it=20 would be desirable to do this in an atmosphere of open=20 communication.

List of Abbreviations

PMR,=20 Proportional mortality ratio

PCMR,=20 Proportional cancer mortality ratio

CI, Confidence=20 Interval

IBM,=20 International Business Machines

CMF, Corporate=20 Mortality File

ICD 9,=20 International Classification of Diseases, 9^{th =}revision

OCMAP,=20 Occupational Cohort Mortality Analysis Program

SMR,=20 Standardized Mortality Ratio

Competing interests

The author was=20 previously paid a consultancy by the plaintiffs' law firm; the law = firm=20 did not design or conduct the study, nor review or approve the = manuscript.=20 The author received no remuneration for the preparation of this=20 manuscript.

Authors' contributions

RC = was=20 responsible for the overall conception and presentation of the=20 work.

Acknowledgements

The author=20 would like to acknowledge the major contribution to these analyses = by Dr.=20 Rebecca Johnson. The graph in Figure 1 = was=20 prepared by Ms. Courtney Ledger. This work was supported in part = by paid=20 consultancy to Drs. Clapp and Johnson by the law firm of = Alexander, Hawes=20 & Audet, LLP in San Jose, CA. The funders obtained the raw = data from=20 the company but played no role in the design or conduct of the = study nor=20 in the management, analysis or interpretation of the results. The = funders=20 did not prepare, review or approve the = manuscript.

Outline

References

Background<= /FONT>

Discussion<= /FONT>