Document DDgQnokm23p8JEXxekerepJZN

An Updated Mortality Study of Workers at a Petroleum Refinery in Torrance, California, 1959-97 Otto Wong, Sc.D. 1,2,3 Fran Harris, M.S. !,4 Kim Rosamilia, Ph.D. Gerhard K. Raabe, Dr.P.H. 1) Applied Health Sciences, Inc., San Mateo, CA 2) Department ofEpidemiology, Tulane University, New Orleans, LA 3) Department of Community and Family Medicine, Chinese University ofHong Kong, Hong Kong 4) School of Medicine, University of California, San Francisco, CA 5) Epidemiology and Environmental Health Sciences, Erwinna, PA 6) Health Risk Sciences, Inc,, New Hope, PA Correspondence to: Dr. Wong at Applied Health Sciences, Inc., 181 Second Avenue, Suite 628, San Mateo, CA 94401, USA Short title: Updated Mortality Study of Torrance, California Refinery Workers Key words: refinery workers, petroleum products, benzene, mortality, cancer, lymphohematopoietic cancer, and leukemia. Acknowledgments: The authors are grateful to the National Death Index (NDI) of the National Center for Health Statistics for providing vital status information, to Robert Bilgrad at NDI for his valuable assistance, to state health departments for providing copies of death certificates, to Deirdra Vachal at Industry Dynamics Associates for data preparation, and to Mobil Corporation (now ExxonMobil) for sponsoring the project. January 12, 2001 Summary The present investigation represents an update of a previous cohort mortality study of I employees at the Mobil (nowExxonMobil) .Torrance, California refinery. The updated cohort consisted of 3,328 workers who were employed at the refinery for at least One year between 1959 and ! 1997. There were 2,965 (89.1%) male and 363 (10.9%) female employees. The vital status of the cohort was determined through a variety of sources, including company employment or retirement ! records, the Social Security Administration's Death Master File, and the National Center for Health Statistics' National Death Index. The updated study covered an observation period of 38 years from I 1960 to 1997, with a total of 60,612 person-years of observation. A total of 705 (21.2%) cohort members were identified to have died. Mortality data were analyzed in terms of cause-specific I standardized mortality ratios (SMRs)and 95% confidence intervals (95% CIs), with expected deaths based on US national cause-gender-race-year-age-specific mortality rates. The overall mortality of the cohort was significantly lower than expected when compared with the US general population (SMR=81.9, 95% CI: 76.0-88.2). Overall cancer mortality was also lower than expected (SMR=79.8, 95% CI: 67.9-93.1). For specific cancer sites, significant mortality deficits were observed for cancer of the digestive system (SMR=70.9, 95% CI: 49.498.6) and rung cancer (SMR=76.2, 95% CI: 56.9-100.0). No significant increase was reported for any site-specific cancer. For nonmalignant diseases, no significant increase was observed for any cause. In particular, significant mortality deficits were reported for ischemic heart disease (SMR=87.7, 95% CI: 77.2-99.3), chronic endocardial disease and other myocardial insufficiencies (SMR=8.3, 95% CI: 0.2-46.0), all other heart disease (SMR=64.2, 95% CI: 43.0-92.2), and influenza and pneumonia (SMR=59.2, 95% CI: 33.1-97.6). Detailed analysis by length of employment did not reveal any significant mortality excess or upward trend. Analyses of male workers by job classification (process and maintenance) were conducted. Among maintenance workers, mortality from cirrhosis of the liver (SMR=190.1, 95% CI: 101.2325.1) and suicide (SMR=208.6, 95% CI: 111.1-356.7) was significantly elevated. However, these mortality excesses did not appear to be related to employment at the refinery. No other causes of death showed significant increase among maintenance workers. A similar separate analysis was conducted for process workers, and no significant excess was detected for any cause. The findings from the present study were discussed in conjunction with results from previous investigations of employees at the Torrance refinery as well as with results from other refinery studies. Potential limitations of the study were also discussed. INTRODUCTION Refinery workers are potentially exposed to ~ wide range of petroleum-derived hydrocarbons and chemical substances used in the manufacturing of petroleum fuels and lubricants. Specific substances in the working environment of petroleum refineries have previously been identified by the International Agency for Research on Cancer (IARC), which include: aromatic amines (e.g., anisidines), arsenic compounds, asbestos, benzene, bitumens, butadiene, carbazole, chlorinated hydrocarbons, chromium and chromium compounds, ethylene dibromide, ethylene dichloride, hydrazine, lead and lead compounds, mineral oils, nickel and nickel compounds, phenylenediarnine, polyeyelie aromatic compounds, and silica [/ARC, 1989]. The Mobil (now ExxonMobil) Torrance, California refinery first started operations under General Petroleum Corporation of California in 1929. In 1959 the Torrance refinery became part of Mobil Oil Co. The refinery processes approximately 160,000 barrels ofoil a day and produces a variety of petroleum products including gasoline, diesel fuels, jet fuel, liquefied petroleum gases, petroleum coke and sulfur. In 1985, Enterline and Henderson completed a retrospective cohort mortality study of white male workers at the Torrance refinery who worked for at least one year between January 1, 1959 and December 31, 1978 [Enterline and Henderson, 1985]. The vital status of these workers was ascertained through December 31, 1978. The cohort study was subsequently updated and expanded to include all employees at the refinery who worked for at least a year between January 1, 1959 and December :31, 1987, with vital status follow-up extended to December 31, 1987 [Milcarek, et al., 1994]. However, only analyses for white males were presented. The objective of the present investigation was to update and expand the previous cohort study and to continue monitoring the mortality patterns of the Torrance refinery workers. In the present update, enrollment in the study was extended to cover an additional ten years of employment eligibility (1988-1997). Vital status follow-up was extended from December 31, 1987 to December 31, 1997. MATERIALS AND METHODS The expanded cohort consisted of all Mobil employees who had worked for at least a year at the Torrance refinery between January 1, 1959 and December 31, 1997. The cohort members were identified through a combination of personnel records located on-site and computerized payroll files. Information abstracted fi'om these records included social security number, name, gender, race, date of birth, date of employment, employment status on the closing date of the study, vital status on the closing date of the study, and date of retirement, separation, or death when applicable. Employment histories of individuals in the previous cohort were updated from 1987 to 1997. For each employee, the employment history data consisted of the beginning and ending dates for each job, job titles and departments. The vital status of cohort members as of December 31, 1997 was ascertained through several sources, including company employment or retirement records, records at Pension Benefits Information, Inc., the Social Security Administration's Death Master File, and the National Center for Health Statistics' National Death Index. The Death Master File is a national database of all deaths reported to the Social Security Administration since 1939. Vital status of an individual is ascertained by matching the last name and social security number. The National Death Index (ND1), established in 1979, is a national death registry designed to facilitate health investigations. Matching is based on the full name, social security number, birth date, gender, race and, in some eases, father's surname. Vital status information as well as causes of death of study subjects are provided by NDI through a service known as "NDI Plus." In the present update, causes of death for known decedents were obtained from either NDI Plus or death certificates. The underlying and contributory causes of death were coded or converted to the 8th Revision of the International Classification of Diseases (ICDS). Statistical analyses were based on cause-specific standardized mortality ratios (SMRs). Person-years of observation were classified by age (5-year groups), gender, race and calendar year (5-year groups). The race of approximately 5% cohort members was unknown, and these workers were assumed to be white in the analyses. Expected deaths were calculated by applying the US ! national age-cause-gender-race-year-specific death rates to the corresponding person-years in the cohort. Cause-specific standardized mortality ratios (SMRs) were computed by expressing the observed deaths as percentages of the expected. The actual calculation was performed through the | University of Pittsburgh's OCMAP program, with standard mortality rates derived from the . Mortality and Population Data System (MPDS) [Marsh et al., 1998]. Analyses were performed for I the entire cohort and for subcohorts stratified by gender, length of employment and time since first employment at the refinery. Trend analyses by length of employment were based on the method I described by Breslow and Day [1987]. To investigate mortality by job category, male hourly employees at the Torrance refinery were classified into two major job categories: (i) process workers, | and (ii) maintenance craft workers. This latter subcohort of maintenance crat~ workers consisted of individuals involved in a broad range of labor and maintenance activities in the mechanical and service I divisions of the refinery. Examples of maintenance jobs included pipefitter, auto mechanic, welder, I insulator, and painter. 1 In addition to analyses conducted using the routine OCMAP program, a special analysis for lymphohematopoietio malignancies was also performed. In the US, because of the way in which t mortality rates by ICD8 categories are tabulated by the National Center for Health Statistics (NCHS), analyses of non-Hodgkin's lymphoma (NHL) and multiple myeloma (MM) in most occupational cohort studies are generally not reported as such. In the OCMAP program, NHL (ICD8 200 and 202) appears in two categories: "lymphosarcoma and reticulosarcoma" (ICD8 200) and "cancer of all other lymphatic and hematopoietie tissue" (ICD8 202, 203, 208, 209). Thus, part ofNHL (ICD8 202, "other lymphomas") is reported together with MM (ICD8 203), polycythemia vera (ICD8 208) and myelofibrosis (ICD8 209). in the special analysis, NHL and MM were analyzed separately. In the OCMAP program, different types of leukemia were analyzed as one single group. Therefore, the following major types of leukemia were also analyzed separately: acute lymphatic (ICD8 204.0), chronic lymphatic (ICD8 204.1), acute myeloid (/CD8 205.0), and chronic myeloid leukemia (ICD8 205.1). US mortality rates for cell-type specific leukemias, NI-IL and MM compiled by the National Cancer Institute (NCI) based on data derived from the "Surveillance, Epidemiology, and End Results" (SEER) program were used in computing the expected deaths from these causes [Young et al., 1981; Reis et al., i994]. In addition, because of the use of asbestos at the refinery in the past, separate analyses for asbestos-related diseases were also performed. The OCMAP program did not provide a separate analysis for asbestosis. The term asbestosis refers to pulmonary fibrosis caused by exposure to asbestos, or pulmonary fibrosis in persons with a documented asbestos exposure history. Since certifying physicians might not be aware of the decedents' asbestos exposure (if any), such deaths could have been coded simply as pulmonary fibrosis or pneumoconiosis. Therefore, mortality from "pulmonary fibrosis" or "pneumoconiosis" (ICD8 515-517) was analyzed. Expected deaths were based on US rates for pulmonary fibrosis obtained from the National Center for Health Statistics, NCHS (unpublished data by special request). For malignant peritoneal or pleural mesothelioma, mortality rates compiled by NCI were used [Connelly et al., 1987; Spirtas et al., 1986]. RESULTS A total of 3,328 workers were eligible for inclusion in the updated study (Table 1). Eighty-five percent (85.9,4) of the cohort members were white (including Hispanics), and 89.1% were male. Selected employment and demographic characteristics of the cohort are provided in Table 1. Approximately 804 were first employed in their twenties or thirties, and only 10.3% were hired at age 40 or older. A great majority (81.3%) of the cohort members were hired in or after 1950. Most cohort members (57.8%) were employed for at least 10 years at the refinery. The maximum length of vital status follow-up was 38 years (January I, 1960 to December 31, 1997). The average duration of follow-up was 18 years. A total of 1,356 (40.7%) cohort members had a follow-up of 20 years or more. The increase in person-years from the previous study (white males only) was 27,281 or 81.8% to a total of 60,612 (all cohort members). By the end of the follow-up period (December 31, 1997), 1,062 (31.9)% cohort members were still employed at the refinery, 1,561 (46.9,4) had separated Or retired, and 705 (21.2%) were known to have died. Information on vital " status was 98.9*,4 complete, with 38 (1.1%) separated employees lost to follow-up. Of the 705 cohort members identified to have died between 1959 and 1997, death certificate information (derived from death certificates themselves or from NDI Plus listings) was available for 695 deaths (98.6%). The causes of death of the remaining 10 deaths (1.3%) were unknown, and they were included in the overall SMR (all causes) calculations but not in cause-specific SMR calculations. The number of observed deaths, SMRs and 95% confidence limits (95% CI) for selected causes for all workers are presented in Table 2. The total number of observed deaths was 705, compared with 860.43 expected (SMR=81.9, 95% CI: 76.0-88.2), indicating a statistically significant deficit in overall mortality. Mortality from all cancers was also significantly lower than expected. The number of observed cancer deaths was 161, versus 201.79 expected (SMR=79.8, 95% CI: 67.9-93.1). For specific cancer sites, significant mortality deficits were observed for both cancer of the digestive system (35 observed vs. 49.38 expected deaths, SMR=70.9, 95% CI: 49.4-98.6) and cancer of the respiratory system (53 observed vs. 71.50 expected deaths, SMR=74.1, 95% CI: 55.5-97.0). No significant increase was reported for any site-specific cancer. For non-malignant diseases, the majority of SMRs were below 100, and no significant increase was observed for any cause. In particular, significant mortality deficits werereported for ischemic heart disease (SMK=87.7, 95% CI: 77.2-99.3), all other heart disease (SMR=64.2, 95% CI: 43.0-92.2), and influenza and pneumonia (SMR=59.2, 95% CI: 33.1-97.6). A separate analysis for male employees is presented in Table 3. Because male employees represented approximately 90% of the entire cohort, their mortality experience was quite similar to that of the entire cohort. In particular, no significant increase in mortality from any cause of death was detected in male employees (Table 3). Similar to the total cohort, significant mortality deficits were observed for the following causes: all causes, all cancers, cancer of the digestive system, cancer of the respiratory system, ischemie heart disease, all other heart disease, and influenza and pneumonia. Among the 36:3 female employees, only 11 deaths were reported. There were two deaths from breast cancer, one from lung cancer and one from cancer of the large intestine. For non-malignant diseases, there were two deaths from cerebrovascular disease, two from ischemic heart disease, one from generalized arteriosclerosis, and one from suicide. The cause of death of the remaining death was unknown. Given the small numbers of death, we will not present cause-specific SMRs among female employees. Table 4 shows cause-specific mortality of male employees by length of employment at the refinery. None of the causes of death showed a statistically significant upward trend. Several causes appeared to show an upward trend, but none was statistically significant: all causes (g:~r=l.86, p>0.05), bronchitis, emphysema and asthma (:g~m~=2.74, p>0.05), motor vehicle accidents (~=0.77, p>0.05), and suicides (~2~=2.25, p>0.05). None of the SMRs presented in Table 4 was significantly devated, regardless of length of employment. Long latent periods (decades) are usually required for chronic diseases to develop. In many instances, it would be more appropriate to examine mortality experience only a.Rer a certain lag period has elapsed. Table 5 shows mortality analysis by interval since hire (latency) for male employees. The category "cancer of other lymphopoietic tissue" showed a significantly elevated SMR in the shortest latency group (SMR=458.6, 95% CI: 148.9-1070.1, 5 observed deaths). This residual category includes non-Hodgldn's lymphoma, multiple myeloma, polycythemia vera and myelofibrosis. As stated in the previous section, a separate analysis of lymphatic and hematopoietic cancers based on a biologically meaningful classification has been conducted and the results will be presented below. Table 5 also shows that male employees with 20-39 years of latency experienced a significantly elevated SMR of 195.5 (95% CI: 115.9-309.0) for cirrhosis of the liver. On the other hand, several causes of death showed statistically significant mortality deficits by time since first employment, including all causes (<20, 20-39 years), all cancers (20-39 years), ischemic heart disease (<20 years), and cirrhosis of the liver (<20 years). - Workers were classified into broad categories of jobs (at least 6 months in the category). Separate mortality analyses were conducted for maintenance and process workers. Some workers were in both categories. Table 6 shows cause-specific mortality of male maintenance employees. Two causes of death showed significant increases: cirrhosis of the liver (SMR=I90.1, 95% CI: 101.2-325.1) and suicide (SMR=208.6, 95%. CI: 111.1-356.7). For cirrhosis of the liver the increase appeared to concentrate in workers with 10-29 years of employment in maintenance jobs (SMR=249.5, 95% CI: 107.7-491.6, 8 deaths); and for suicide, among those with 30 years or longer in maintenance jobs (SMR=ll30.5, 95% CI: 136.8-4084.1, 2 deaths) (not shown in Table 6). No other SMRs were significantly elevated, and no upward trend was reported for any cause of death among male maintenance workers. A similar analysis was carried out for male process employees (Table 7). The overall mortality was significantly less than expected (SMR=81.3, 95% CI: 70.9-92.8). Lung cancer also showed a significant deficit (SMR=46.7, 95% CI: 22.4-85.8). Similar to maintenance workers, process workers also showed an devated mortality fi-om cirrhosis of the liver, but the excess was not statistically significant (SMR=I40.1, 95% CI: 60.5-276.1). The specific categories of lymphatic and hematopoietic cancers used in the OCMAP program are based on the statistical classifications compiled by National Center for Health Statistics (NCHS) and do not permit a specific analysis for major call-type specific leukemias, NHL or MM, which are more appropriate classifications fi'om the biological point of view [Linet, 1985; Heath, 1982; Wintrobe et al., 1981; Wong and Raabe, 1995]. Therefore, analyses for these specific subcategodes of lymphatic " and hematopoietie cancers among male employees were carried out separately. Table 8 shows that there was only one observed AML death among male employees, compared with 2.20 expected (SMR=45.4, 95% CI: 1.1-252.8). For CML, there were 2 observed deaths, approximately one more than the 1.02 expected (SMR=195.6, 95% CI: 23.7-706.6). No death fi'om ALL was reported. For CLL, one death was observed, comparable to the 1.34 expected. There were 4 MM deaths, similar to the expected 3.28 (SMR=122.1, 95% CI: 33,3-312.5). Similarly, there were 9 NI-IL deaths, comparable to the 7.42 expected (SMR=121.4, 95% CI: 55.5-230.4). Table 8 also presents analysis by job category. No significant mortality increase in any disease subcategory was reported by job category. Table 9 shows the mortality fi-om lymphohematopoietic cancers of male employees by length of employment. There was no upward trend for any of the specific categories of lymphohematopoietie cancer. No significant increase was found for any length of employment. Analysis by time since first employment is also presented in Table 9. No significant increase was detected for any categories of lymphohematopoietic cancers for any latency group. Because of the use of asbestos at the refinery in the past, mortality fl'om asbestos-related diseases (asbestosis, lung cancer, and mesothelioma) among cohort members was examined. Although the OCMAP program provided an analysis for lung cancer, it did not provide any separate analysis for asbestosis or malignant mesothelioma of the peritoneum or pleura. Analyses for asbestosis (pulmonary fibrosis) and malignant mesothelioma of the peritoneum or pleura were performed separately. The term asbestosis refers to pulmonary fibrosis caused by exposure to asbestos, or pulmonary fibrosis in persons with a documented asbestos exposure history. The ICD code (Sth revision) for asbestosis is 515.2, and is, therefore, part of the broad category "other non-malignant respiratory disease" (ICD8 460-519) in the OCMAP analysis. In the Torrance cohort, no death was attributed to asbestosis. However, since certifying physicians might not be aware of the decedents' asbestos exposure (if any), such deaths could have been coded simply as pulmonary fibrosis or pneumoconiosis. Therefore, the category "pulmonary fibrosis" was analyzed. For pulmonary fibrosis (ICD8 515-517), there were 3 observed deaths among all male cohort members, compared with 4.24 expected. The corresponding SMR for pulmonary fibrosis among male employees was 70.8 (95% CI: 14.6-207.0). One of the 3 male employees who died fi'om pulmonary fibrosis had worked in both maintenance and process jobs, and the remaining two in administrative or technical jobs. As stated above, lung cancer results were provided by the routine OCMAP program. The results for the overall cohort are presented in Table 2 (SMR=76.2, 95% CIi 56.9-100.0). The lung cancer SMR for all males was 76.0 (95% CI: 56.6-99.9.). Exposure to asbestos has been linked to malignant peritoneal and pleural mesothelioma in epidemioiogic studies. Malignant peritoneal mesothdioma is part of "malignant neoplasm of the peritoneum and retroperitoneal tissue" (ICD8 158), and malignant pleural mesothelioma is part of"malignant neoplasm of the pleura" (ICD8 163). The most comprehensive mortality rates for malignant peritoneal and pleural mesothelioma for the US are those compiled by the National Cancer Institute in the Surveillance, Epidemiology and End Results (SEER) program [Connelly et ai., 1987; Spirtas et al., 1986]. In the SEER program, cases were selected having the International Classification of Diseases for Ontology 0CT)O) morphology code 905 (mesothdial neoplasms) and topography codes 158 (peritoneum) or 163 (pleura). In addition, only malignant cases were included [Spirtas et al., 1986]. In the present study, diagnostic information was based entirely on death certificates, and no information on morphology was available. In the Torrance cohort, there was one death coded as ICD8 163 (malignant neoplasm of the pleura). Based on the SEER rates, approximately 1.08 deaths fi'om malignant mesothdioma of the pleura were expected. The corresponding SMR was 92.6 (95% CI: 2.4-516.0). i ! DISCUSSION I In the discussion below, we will first compare some major results in the updated study to those previously reported for the Torrance cohort by Miicarek et al. [1994]. We will then discuss several specific findings in the updated study, which are of particular interest in epidemiologic studies of refinery workers. Extended enrollment and follow-up in this update on Torrance refinery workers added 1,337 workers (+67.2%), 27,281 person-years (+81.8%) and 297 additional deaths (+72.8%). One of the reasons for the sizable increases was that only white males were included in the previous study, whereas no restriction on gender or race was imposed on cohort eligibility in the update. In the previous analysis 20.5% of cohort members were deceased, compared with 21.2% in this update. Updated results for the total cohort were consistent with earlier reports, which demonstrated a favorable overall mortality experience of the Torrance refinery workers. Based on national rates, the updated SMR of 81.9 for all causes showed a significant deficit of 18.1%, which was similar to the deficit previously observed (SMR=81). As in the previous study, the favorable overall mortality experience of these refinery workers might be attributed to the "healthy worker effect," whereby working populations exhibit decreased overall mortality due to initial selection into the workplace and maintenance of this healthier status through benefits derived fi'om employment. In terms of specific causes, for the cohort as a whole or the male subcohort, no significant mortality increase was observed for any cause of death. The earlier reports showed non-significant deficits for cancer of the digestive system and cancer of the respiratory system. In the updated study, similar deficits were observed, and, because of the larger numbers of deaths, the deficits for both cancer of the digestive system and cancer of the respiratory system were statistically significant. In the previous reports, a significant deficit of ischemic heart disease was reported. In the current update, a similar significant deficit was observed for ischemic heart disease. In addition, a significant deficit was also reported for the category "all other heart disease." 1 1 Analysis by length of employment also indicated that there was no work-related mortality 1 pattern for male employees. No significant increase of mortality was found for any cause for any length of employment group, and no upward trend was detected for any cause of death. Analysis by latency ! (time since hire) showed that there was a significant mortality excess from the miscellaneous category "cancer of other lymphopoietic tissue" for the group with less than 20 years of latency (5 observed v. .! 1.09 expected, SMR=458.6). Of the 5 deaths in this category, one was a multiple myeloma (ICD8 203), and the other 4 were "other lymphomas" (ICD8 202). As discussed earlier, a proper analysis of NHL should include "lymphosarcoma and reticulosarcoma" (ICD8 200) and ".other lymphomas" (ICD8 202). In the group with less than 20 years of latency, no death was observed for I "lymphosarcoma and reticulosarcoma" (ICD8 200). Thus, according to the disease grouping provided in the OCMAP analysis, NI-IL was reported in two places: no death was attributed to "lymphosarcoma I and reticulosarcoma" (ICD8 200) and 4 to "other lymphomas" (ICD8 202). When NHL (ICD8 200 and 202) was analyzed as a group (Table 9), the SMR was 298.3 (95% CI: 81.3-763.9) and not significant. The more important observation was that there was no upward trend by length of employment for I'q'HL. The finding of no association between NI-IL and employment at the Torrance refinery is consistent with the conclusion reached in a recent review of NHL in petroleum workers [Wong and Raabe, 2000a]. The review was based on more than 308,000 petroleum workers from the US, the UK, Canada, Australia, Italy and Finland. None of the individual studies showed a significant increase of NHL. For all studies combined, there were 408 observed NHL deaths, comparable to the 412.20 expected (SMR--99, 95% CI: 90-109). Specifically for US refinery studies, there were 317 observed deaths, slightly fewer than the 330.79 expected (SMR---96, 95% CI: 86-107). There is no evidence that employment at petroleum refineries is associated with an increased risk of NHL. For male employees with 20-39 years of latency, mortality from cirrhosis of the liver was significantly elevated, but that for male employees with less than 20 years of latency was significantly low (Table 5). There was no upward trend of mortality from cirrhosis of the liver by length of employment (Table 4). Cirrhosis of the liver is associated with alcohol consumption. Unfortunately, no information on alcohol consumption was available in the study. Cirrhosis of the liver has not been linked to any petrochemicals present at refineries. No other epidemiologic studies of petroleum refinery workers have reported an increased mortality from cirrhosis of the liver. Because of potential differences in exposures and exposure patterns, separate .analyses were performed for maintenance and process workers at the refinery. Among maintenance workers, mortality from cirrhosis of the liver was significantly elevated (13 observed v. 6.84 expected, SMR=190.1, 95% CI: 101.2-325.1). As discussed above, the excess of cirrhosis of the liver was not likely to be related to employment at the refinery. There was a significant increase of mortality from suicide among maintenance workers (13 observed v.6.23 expected, SMR=208.6, 95% CI: 111.1356.7). However, there is no reason to believe that the increase was related to exposures at the refinery. Among process workers, no significant increase of mortality from any cause of death was detected. As stated above, the OCMAP program does not provide analyses specific to subgroups of lymphohematopoietic cancers. When analyses were performed separately for ALL, CLL, AML, CML, MM and NI-IL, no significant increase was found for any category for the cohort as a whole or for the subcohorts of maintenance or process workers. However, it must be noted that the number of deaths for each category was small, with the possible exception of NFIL (9 observed deaths v. 7.42 expected). Nevertheless, the data indicated that there was no increased mortality fromany of the subcategories of lymphohematopoietic cancers. Similar to other industrial facilities, asbestos was used at the Torrance refinery in the past, and detailed analyses of asbestos-related diseases were performed for male employees. Although specific exposure measurements were not available, data from the industry as a whole have indicated that asbestos concentrations at refineries were generally extremely low [IARC, 1989]. Among Torrance refinery employees, no death was attributed to "asbestosis" as the underlying cause of death. The term asbestosis refers to pulmonary fibrosis in persons with documented asbestos exposure history. Since certifying physicians might not be aware of the decedents' asbestos exposure, such deaths could have been coded simply as pulmonary fibrosis or pneumoconiosis. In terms of pulmonary fibrosis, 3 deaths were reported, compared with 4.24 expected (SMR=70.8, 95% CI: 14.6-207.0). One of the 3 fibrosis deaths worked in both process and maintenance jobs, and the other two in administrative or technical jobs. One death in the cohort was coded as ICD8 163 (malignant neoplasm of the pleura), whereas 1.08 deaths were expected. This individual worked for approximately 2 years in the late 1950s at the Torrance refinery, and died 4 years after termination. Epidemiologic studies have reported the average latency of malignant pleura/mesothelioma to be around 30 to 40 years even among heavily exposed workers. The extremely_ short interval of 6 years between employment at the Torrance refinery in 1958 and death in 196_.,.0 does not support an association between the two. In addition to the malignant pleural cancer case, the words "malignant mesothelioma" were mentioned on another death certificate. The cause of death listed on the death certificate was "malignant mesothelioma, metastatic" (with no specific site), and coded as ICD8 199.1. According to the SEER criteria, the information provided on the death certificate was not sufficient to classifij this death specifically as either malignant mesothelioma of the peritoneum or malignant mesothelioma of the pleura. For lung cancer, an asbestos-related malignancy, no increase was found for the overall cohort or the subcohorts of maintenance or process workers. In fact, for all three groups, lung cancer mortality was slightly less than expected. Equally important is the observation that no upward trend by length of employment was detected. Thus, the data indicated that there was no relationship between lung cancer mortality and employment at the Torrance refinery. This finding is consistent with the conclusion reached in a recent review of petroleum workers [Wong and Raabe, 2000b]. The review was based on more than 350,000 petroleum workers from the US, the UK, Canada, Australia, Italy and Finland. Lung cancer SMRs from individual studies ranged from 44 to 107. For all studies combined, there were 5,695 observed lung cancer deaths, significantly fewer than the 7,056.25 expected (SMR=81, 95% CI: 79-90). For US refinery workers, there were 2,696 observed lung cancer deaths, significantly fewer than the 3,573.27 expected (SMR=75, 95% CI: 73-78). Thus, there was no indication that petroleum workers were at an increased risk of lung cancer. It should be pointed out that there are some potential limitations in this study. Most limitations are typical of a historical mortality study of industrial populations. First, although both the percentage ! ! I I !1, |] I Ii I of individuals with unknown vita~ status (1.1%) and the proportion of deaths with no death certificate information (1.3%) were low, it was possible, but unlikely, that some deaths from causes of interest might have been missed. We do not believe that such low percentages would have any significant impact on our results. Second, the health endpoint in our study was mortality, which is a reasonable surrogate for incidence of many cancers. Analyses were based on the underlying cause of death listed on death certificates. As such, the investigation inherited the problems associated with death certificates (diagnostic accuracy, for example). No detailed clinical information was available on the deaths in our study. However, it must be pointed out that although detailed information derived from medical records or pathology reports may be more accurate than that based on death certificates, it would be inappropriate to use such information for comparisons with national mortality rates. In our study we compared diagnoses based on death certificates to national statistics which were derived from death certificates as well. Furthermore, our analysis was adjusted for calendar time, thus at least partially controlling for changes in survival and diagnostic practices. As in most historical cohort mortality studies, little quantitative exposure data were available during the early part of the study, and analyses by quantitative exposure indices were not possible. In the present study, we relied on duration of employment and the broad classification of jobs (process and maintenance). Although such exposure surrogates could be informative, they might lack both sensitivity and specificity. I, For some analyses in the overall cohort as well as in some subcohort analyses, particularly those stratified by job category or by length of employment, the numbers of deaths from some rare causes were small. As a result, the corresponding 95% CIs were wide and the findings must be interpreted with caution. However, for common disease categories such as all cancers, digestive cancer, lung cancer, prostate cancer, heart disease and non-malignant respiratory disease, the statistical power of the study was quite adequate. Finally, information on lifestyle or exposures from employments elsewhere was not available, although such information would have been extremely helpfi~l in interpreting some of the findings in the present study, such as alcohol consumption in relation to cirrhosis of the liver. From the methodological point of view, the most appropriate approach of incorporating lifestyle risk factors and exposures from employments elsewhere is the nested case-control study design. CONCLUSION This updated study has confirmed earlier findings of an overall favorable mortality experience of employees at the Torrance, California refinery, when compared with the US general population, In addition, significant mortality deficits were found for several causes of death. For the entire cohort, no significant increase was detected in any cause-specific mortality, which could be attributed to employment at the refinery. Furthermore, analyses by job category and length of employment also indicated that there was no mortality increase from any cause of death associated with employment at the refinery. | | Breslow NE and Day NE [1987]: Statistical Methods in Cancer Research, Volume 11- The Design and Analysis of Cohort Studies. /ARC Scientific Publications No. 82. International Agency for | Research on Cancer, Lyon. Connelly RR, Spirtas R, Myers MH, Percy CL, and Fraumeni JF [1987]: Demographic patterns for | mesothelioma in the United States. J Natl Cancer Inst 78:10531060. Enterline PE and Henderson V [1985]: A report of the mortality experience of workers from a petroleum refinery at Torrance, California. Technical report submitted to Mobil Oil Corporation. Heath CW [1982]: The leukemias. In: Cancer Epidemiology and Prevention (Schottenfeld D, | Fraumeni IF eds.). WB Saunders. International Agency for Research on Cancer [1989]: 1ARC monographs on the evaluation of ! carcinogenic risks to humans. Occupational exposures in petroleum refining," crude oil and major petroleum fuels, Volume 45. World Health Organization. I Linet MS [1985]: The Leukemias: Epidemiologic Aspects. Oxford University Press. Marsh GM, Youk AO, Stone P,A, Sefcik S, and Alcon C [1998]: OCMAP-Plus: A program for the I comprehensive analysis of occupational cohort data. d Occup Environ Meal40:351-362. Milcarek BI, Collingwood KW, and Raabe GK [1994]: An updated cohort mortality study of white male workers at the Torrance, California refinery, 1959-1987. Technical report, Mobil Oil Corporation Medical Department. Reis LAG, Miller BA, Hankey BF, Kosary CL, Harras A, and Edwards BK [1994]: SEER Cancer Statistics Review, 1973-1991. NIH Publication No. 94-2789, National Cancer Institute, Washington. Spirtas R, Beebe GW, Connelly RR, Wright WE, Peters .iM, Sherwin RP, Henderson BE, Stark A, Kovasznay BM, Davies JNP, Viarma NJ, Keehn R J, Ortega LG, Hochholzer L, and Wagner JC [1986]: Recent trends in mesothelioma incidence in the United States. Am dlndMed 9:397-407. Wintrobe MM, Lee GR, Boggs DR, Bithell TC, Forester J, Athens JW, and Lukens JN [1981J: Clinical Hematology (Sth ed.), Lea & Frbiger. Wong O and Raabe GK [1995]: Cell-type specific leukemia analyses in a combined cohort of more than 208,000 petroleum workers in the United States and the United Kingdom, 1937-1989. Regulatory Toxicology and Pharmacology 21: 307-321. Won80 and Raabe GK [2000a]: Non-Hodgkin's lymphoma and exposure to benzene in a multinational cohort of' more than 308,000 petroleum workers, 1937 to 1996. J Occup Environ Med 42: 554-568. | Wong O and Raabe GK [2000bJ: A critical review of cancer epidemiology in the petroleum industry, with a meat-analysis of a combined database of more than 350,000 workers. Regulatory Toxicology and Pharmacology 32: 78-98. Young JL, Perry CL, and Asire AJ [1981]: Surveillance, Epidemiology, and End Results.. Incidence and2~lortaliO~ Data. NIH Publication No. 81-2300. Table 1 Characte~istica of the Torrance, Caligornia, cohort Total employees Total person-years White Nonwhite Male Female Year of hire < 1950 ~ 1950 Duration of employment (yr) <10 10-29 30 + Age at hire <20 20-39 40+ Age at death <50 50-59 60-69 70 + Vital status (December 31, 1997) Alive Dead Unknown Number 3,328 ~/ 60,612 2,859 ~'"'~ 469 ~,." 2,965 "~ 363 ~,..-" 623 ~ 2,?05 1,403 1,472 453 317 2,669 342 64 91 186 364 2,585 ~/ ..... 705 ~ 38 ~- %total 100.0 100.0 85.9 14.1 89.1 10.9 18.7 81.3 42.2 44.2 13.6 9.5 80.2 10.3 9.1 12.9 26.4 51.6 77.7 21.2 1.1 | ! Table 2 Cause-speciflc mortality of the entire aohort of employees at the Torranae refiner~ Cause of death OBS EXP ~ Lows= Upps= 95% Confidenoe l~m/te All Causes of Death All Malignant Neoplam Canoe= of Buooal Cavity & Pharynx C~oer of Emopha~m Canomw of B~=h C~ao~ of ~go Intestine C~ae: of ~o~ C~oo: of Bili~ Pagmagem & Livo~ C~oer of ~10~hew Digestive Organm C~oeE of ~mp~Eato~ ~st~ C~n~ of ~ C~naor of B~onohum, T~ah~a, Lung ~1 UtoEino Canoewm (F~Iom only) Censer of Co,ix U~e~i (Fmlom only) CanGow of O~hm~ Fmlo ~n~tal Organs ~noe= of PEom~ate (~lmm only) Canoe~ of Tom~om and Othe~ ~io ~nital O~ganm C~omz of BZa~ and O~he~ Urina~ Organs ~li~t ~lan~ of Skin Canoe~ of E~ C~oeE of CentEal Ne~oum Symt~ ~noe= of ~oia & Othe: Endoo=ine Glan~ CanoeE of Bone ~noew of ~I L~hatio, Ha~topoietio Timmue L~hom~o~ & ~ioulomaro~ H~n~ D~moame ~a & ~o~a ~I O~he~ ~l~ant Neoplam Beni~ Neopl~ Di~etem ~11i~u~ Ce:obrovasoul~ Disease ~1 Hoa:~ Dimoamo ~e~tia He~t Disease Imoh~o Hoa=t Dimosme ~:onio Endooa~. Dim. ;Otho: ~ooa:~. Inmuff. ~o=~nmlon with Hoa:t Dimeame ~I 0thee He~t D~eaee ~o:~eneion w/o Heart Disease Non~li~an~ ~mp~wa~o~ Disease Influenza & Pne~nia B=onohiti:, ~m~, ~t~ a:onohiti~ ~s~ Otho~ Non~li~ant b:pi=ato~ Disease Ul=e~ ~f Stash & Du=den~ CiE~o~is of Live~ ~1 External ~umom of Death ~oiden~ ~toE Vehlole ~odents ~I OthsE ~oidents H~ioidem & OthoE External Causes 705 860.43 81.9 ** 161 201.79 79.0 ** 3 4.97 60.4 70.9 * 3 5.05 59.4 5 7.24 12 18.20 2 4.23 69.0 65.9 47.3 1 4.28 23.4 120.1 0 1.39 53 71.50 0.0 74.1 * 0 2.56 0.0 52 68.20 76.2 * 137.4 0 0.34 192.4 0,0 0 0.22 0.0 0 0.35 18 18.98 0.0 94.8 0 0.54 0.0 62.7 2 5.73 34.9 4 2.73 146.3 1 0.13 799.6 6 4.91 122.1 0 0.55 0.0 0 0.54 0.0 19 18.65 101.9 1 2.11 47.4 1 1.13 5 7.37 88.4 67,8 14 15.28 149.2 91 ~ 6 0 2. l0 10 14.30 0.0 69,9 40 50.70 289 326.64 70.8 88,5 * 5 4,64 107.7 '249 1 5 283.06 12.12 6.63 87.7 * 8,3 ** 75.4 29 45.16 64.2 * 3 2.30 130.4 55 69.56 79.1 15 25.35 59.2 * 12 14.23 04.3 3 2.28 9 lo. 9S 131.5 82.2 0.0 28 32.37 0 3,35 89.3 0.0 19 17.25 110.2 59 65.15 115.5 90.6 37 38.97 94.9 22 17.65 124.6 15 21.47 69.9 108.1 5 9.65 51.8 * slgnlfoant at 5% level ** signifioant at It level No. of employees m 3,328 Pereon-yoa=e m 60,612 76.0 67.9 12.5 49.4 12,3 22.4 34.1 5.7 0.6 62.1 0,0 55,5 0,0 56.9 3.4 39.? 0.0 0.0 0.0 56.2 0.0 12.9 4.2 39.9 20.0 44.8 0.0 0.0 61.3 1.2 2.2 22 0 77.1 50,1 0.0 33.5 56.3 70.6 35.0 77.2 0.2 24.5 43,0 26.9 59.6 33.1 43.6 27.1 37.6 0.0 59.3 0.0 66.3 37.5 68.9 66.8 78,1 39.1 63.0 16.8 88.2 93.1 176.4 98.6 173. E 161.1 115.2 170.8 130.3 209.8 265.2 97.0 144.4 100.0 765.7 562.1 1073.6 1693.6 1050.0 149.9 678.7 183.2 126.0 374.7 4455.3 265.8 665.9 689.2 159.1 264.3 492.8 158.3 260.6 153.7 175,3 128, 6 107.3 99.3 251,4 99.3 46.0 176.1 92.2 381.0 102.9 97.6 147.3 384.3 156.0 361.7 129.0 110.2 172.0 269.5 116.8 130.9 188.7 115.2 173.1 120.9 | | T&131m 3 Cause-specifio ~orta~it~ of ~e e~plo~eea at the Tar=anne refiner~ ! Cause o death OBS EXP S)~ Lower U~per All Cauaes of Death | All Malignant Neoplam Canner of Buooal Cavity & Pharynx 694 843.36 157 196.22 3 4.90 82.3 ** 80.0 ** 81.2 Canner of Digestive Organs & Peritoneum 34 48.34 70.3 * Canner of Emophagum 3 4.99 60.1 Canner of St~maoh 5 7.13 70.I Canoe= of La~go Intoetine II 17.76 61.9 Canner of Reotum Canner of Biliary Passages & Liver 2 4.15 1 4.16 48.2 24.0 Canner of Panoroam 12 9.76 122.9 Canner el All Other Digemtive Organm 0 1.36 0.0 Canner efReapiratorySymtem 52 70.36 73.9 * Caneew of L~rlrnx 0 2.53 0.0 ! Canner efBronehum, Traohea, Lung Canner of All Other Respiratory Canner of Breast 51 87.10 76.0 * I 0.71 139.9 1 0.24 418.7 Canner of Prostate 18 18.98 94.0 ! Canner of Tomtom and Other Male Genital Organs Canner of Kidney Canner of Bladder and Other Urinary Organs .0 0.54 3 4.70 2 5.69 0.0 63.8 35.2 Malignant Melanoma of Skin Canner of Eye 4 2.66 1 0.12 150.4 BiB. 6 Caneer of Central Ne~voum Symtm 6 4.76 126.0 Canner of Thyroid & Other Endnorine Gland~ Canner of Bone 0 0.53 0 0.52 0.0 0.0 Canner of All Lymphatio, Haematopoietio Tissue 19 18.18 104.5 I Lymphomaro~ma & Retioulomarooma Hodgkinm Dieeame 1 2.07 1 1.10 48.2 90.8 Leu~em/a &A leukemia 5 7.19 69.5 Canner of All Other LymI)hopoietio Tissue 12 7.82 153.5 ! All Other Malignant Neoplam Benign Neoplam Diabetes Mallitue Cerabrmrasoular Disease 14 14.87 0 2.04 10 13.82 38 49.72 94.1 0.0 72.4 76.4 All Heart Disease 287 322.42 89.0 * Rhet~atlo Heart Diaeame 5 4.47 111.7 Imohemlo Heart Dimeame Chrenlo Endooard. Dia.~ Other Myooard. Insuff. Hyper~onmion with Heart Disease 247 280.81 1 11.87 5 6.42 88.0 * 8.4 ** 77.9 All O~her Heart D~seame Hypertension w/o Heart Disease 29 44.19 3 2.24 65.8 * 133.9 Non-malignant Respiratory Disease 55 68.45 80.3 InfXuenza & Pnem~onia Bronchitis, Emphysema, ~thma 15 24.96 12 14.01 60.1 * 85.8 Bronehite 3 2.25 133.1 Emph~aema A~t}sna Other Non-malignant ReepiratoryDieeaee 9 10.84 0 0.96 28 30.85 83.0 0.0 90.8 Ulnar of Stomach & Duodenmn Cirrhosis of Liver Nephritis & Nephrosie 0 3.31 0.0 19 16.86 - 112.7 5 4.24 117.9 All External Cauaee of Death 58 63.69 91.1 Anoidente Motor VehAeleA~oidentm 37 38.13 22 17.18 97.0 128.1 All Other Aooidentm 15 21.10 71.1 Suioidee 16 15.41 103.8 Hmnioidee & Other External Causes 5 9.35 53.5 * signifioant at 54 level ** signifioant at 14 level No. of employees - 2,965 Person-years -- 55,984 954 Condenoe limits 76.3 68.0 12.6 48.7 12.4 22.8 30.9 5.B 0.5 63.5 0.0 55.2 0.0 56.6 3.5 10.5 56.2 0.0 13.2 4.3 41.0 20.5 46.2 0.0 0.0 62.9 1.2 2.3 22.6 79.3 51.5 0.0 34.7 54.1 79.0 36,3 77.3 0.2 25.3 44.0 27.6 60.5 33.6 44.2 27.5 38.0 0.0 60.3 0.0 67.8 38.3 69.1 68.3 80.3 39.8 59.3 17.4 88.8 93.8 178,9 98.3 175.5 163.7 110.8 174.0 133.8 214.8 271.2 96.9 145.6 99.9 779.5 2332.9 149.9 678.7 186.5 127.1 385.2 4561.3 274,2 693.1 706.1 163.2 268.7 505.9 162.2 268.2 157.9 180.8 133.1 104.9 99.9 260.7 99.6 46.9 181.8 94.3 390.9 104.6 99.1 149.6 389.0 157.6 386.1 131.2 111.5 176.0 275.1 117.7 133.8 193.9 117.2 168.6 124.8 Table 4 Cause-specific mortality of male employees at the Torrance refinery by length of employment Cause of death All malignant neoplasm~ Cancer of bu=cal cavit~r & pharynx Cancer of digestive organs & peritone~n Cancer of elophagus Can=mr of stomach Can=mr of large intestinm Cancer of rectum Cancer of biliarypassagss &iiver Cancer of pancreas Cancer of all other digestive organs Cancer of respiratory myst~n Cancer of larynx Cancer of bronchus,.trachea & lung Cancer of breast Cancer of prostate Cancer of kidney Cancer of bladder & other urinary organs Malignant melanoma of skin Cancer of central nervous myst~n Cancer of lymphatic & h~matopoieti= tissues Lymphosarcoma & reticulosarcoma Hodgkin's disease Leukemia Cancer of o~her l~phopoietic tissue Benign neoplasms Diabetes mellitus CeE~rovas=ular disease All heart disease Isch~mio heart disease Nonmalignant respiratory disease Influenza & pneumonia Bronchitis, emphysem~ & asthma Other non-malignant respiratory disease Cirrhosis of liver All external causes of death Accidents Motor vehicle accidents All other accidents Suicides Homicides & other external causes Number of employees Person-years <I0 years OBS SMR 128 72.7 ** 38 96.5 0 0.0 8 87.4 0 0.0 1 76.8 3 93.6 0 0.0 1 111.2 3 160.2 0 0.0 13 90.2 0 0.0 12 87.3 0 0.0 ' 2 83.5 1 95.9 1 122.8 1 115 . 6 2 142.1 6 141.9 " 1 244.4 1 258.1 1 62.3 3 164.1 0 0.0 3 103.0 3 42.9 38 70.1 * 25 54.9 ** 8 76.8 0 0,0 * 1 51.3 7 141,9 1 19.6 25 82.9 15 87.9 9 I01.6 6 72.6 6 86.8 4 66.7 2,452 28,430 10-29 years OBS ~ 425 84.2 ** 87 75,6 ** 3 102.7 19 65.5 2 70.1 4 90.5 5 47.0 0 0.0 . 0 0.0 8 137.2 0 0.0 29 71.4 0 0.0 29 75.0 1 694.2 i0 83.8 2 74.2 1 27.9 3 224 . 1 3 117.5 Ii 106.3 0 0.0 0 0.0 4 96.6 7 163.5 0 0.0 5 61.7 22 67.5 192 94.3 171 95.6 34 79.4 12 74.5 6 65.6 16 86.8 16 169.6 26 93 .O 19 10B. 6 II 155.9 B 76.2 6 86.3 1 33.3 1,393 23,073 * significant at the 5% level; ** significant at the 1% level 30+ years OBS SMR 141 86.8 32 76.6 0 0.0 7 68.9 I 97.4 0 0.0 3 76.8 2 241.6 0 0.0 1 48.6 0 0.0 i0 65.2 0 0.0 I0 68.1 0 0.0 6 129.2 0 o.0 0 0.0 0 0.0 1 124.7 2 55.6 0 0.0 0 0,0 0 0.0 2 117.2 0 0.0 2 71.3 13 128.4 57 88 .i 51 90.4 13 85.5 3 58.8 5 171.6 5 66.8 2 86.4 7 127.0 3 R3.8 2 157.8 1 42.8 4 258.2 0 0.0 283 4,482 Table 5 Cause-specific mortality of male employees at the Torrance refinery by time since first employment Cause of death All malignant neoplasm~ Cancer of buccal cavity & pharynx Cancer of digestive organs & peritoneum Cancer of esophagus Cancer of stomach Cancer of large intestine Cancer of rectum Cancer of biliarypassages & liver Cancer of pancreas Cancer of all other ~igeetive organs Cancer of respiratory. Syst~n Cancer of larynx Cancer of bronchus, trachea & lung Cancer of breast Cancer of prostate Cancer of kich%ey Canoer of bladder & other urinary organs Malignant melanom~ of skin Cancer of central ne~"~ous system Cancer of lymphatia & hematopoietic tissues Lymphosarcoma & reti=ulomaracma Hodgkin's disease Leukemia Cancer of other lymphopoisti= tissue Benign neoplasms Diabetes mellitus All heart disease Ischemic heart disease Nonmalignant respiratory aiseass Influenza & pneumonia Bronchitis, emphysema & asthma Other non-malignant respiratory disease Cirrhosis of liver All external causes of death Accidents Motor vehicle accidents All other accidents Suicides Homicides & other external causes <20 years OBS S~ 84 59.5 ** 20 78,4 0 0.0 2 32.3 0 0.0 1 93.6 1 53.8 0 0.0 0 0.0 0 0.0 0 0.0 6 68.9 0 0.0 5 61.0 0 0.0 1 140.4 2 286.9 0 0.0 0 0.0 0 0.0 8 246.4 * 0 0.0 1 216.1 2 166.3 5 458,6 * 0 0.0 1 46.5 2 40 .I 26 65.0 22 65.5 2 35.2 0 0,0 0 0,0 2 109,1 0 0,0" 25 73.9 17 89.2 12 119.9 5 54.9 5 67.5 3 43.7 Number of en~ployees Person-years 2,795 31,161 * significant at the 5% level; ** significant at the 1% level 20-39 years OBS SMR 322 83.3 77 79.1 * 3 108.2 19 77.7 2 76.1 3 81.0 5 58.1 1 44,7 1 52.2 7 137.5 0 0.0 27 71.8 0 0.0 27 75.4 1 831.1 4 57.7 1 41.1 1 38.2 4 327.5 6 245.4 4 48.1 1 82.0 0 0.0 0 0.0 3 88.2 0 0.0 3 48.4 18 81.2 139 87.8 119 83.8 21 76.9 5 59.3 5 67.2 Ii 93.5 18 195.5 * 22 103.0 14 106.4 6 110.7 8 102.8 6 103.6 2 93.0 1,599 19,640 40+ years OBS SMR 288 91,2 60 81,7 0 13 73.5 1 62.8 1 42.5 5 68.6 1 74.2 0 0.0 5 145.7 0 0.0 19 79.0 0 0.0 19 82.3 0 0.O 13 114.7 0 0.0 1 37.5 0 0.0 0 0.0 7 105.5 0 0.0 0 0.0 3 I08.0 4 120.3 0 0.0 6 109.6 18 79.7 122 98.3 106 100.7 32 90.2 i0 71,4 7 136,5 15 86,9 1 40,0 Ii 129,8 6 101.6 4 228.4. 2 47.5 5 225.7 0 0.0 674 5,182 Table 6 Cause-speclfic mortality of male ma/ntenance employees at the Torrance refinery Cause o death OBS EXP SMR Lower Upper 95% Confidenoe limits All Causes of Death All Malignant Nooplasm~ CanDor of Buses1 Cavity & Pharynx Canoe= of Digestlvo Organs & Peritoneum Canoe=of Esophagus Canoe= of Stom~oh Canoe= of Large Intestine CanDor of Raot~ Canoe= of Billary Passages & Liver Canoe= of Pane=see CanDor of All Other Digestive Organs Canoe= of Respiratory Symte~n Canoe= of Larynx Canoe= of Bronohus, Trashes, Lung Canoe= of All Other Respiratory Canoe= of Breast Canes= o Prostate Canoe= of Testes and Other Male Genital Organs CanDlE of Kich~ey Censer of Bladde: and Other Urinary Organs Malignant Malan~ of Skin CanDor of Eye Canoe= of Central Nmrvoum System Canoe= of Thyroid & Other EndeD=ins Glands CanDor of Bone Canoe= of All Lymphatio, Haematopoieti= Tissue LymphomaEooma & Rstioulomarooma Hodgkinm Disease Leuken~La & Aleukemia Canoe= of All Other Lymphopoietio Tissue All Othe=Mallgnant Neopl~m Benign Neoplasms Di~otem Mallitus Cerebrovasoula= Disease All Heart D~saase ~heu~atio Heart Disease Isohemlo Heart Disease Chronlo Endooard. Dis.; 0the= Myocara. Insuff. /po~tznelon with Heart Disease All O~he= Hea=~ Disease Hyps~tenslon w/o Heart Disease Non-mallgnant Respiratory Disease Influenza & Pneumonia Bronohitim, Emphysema, Asthma Bronohitis Emphysema Asthma Other Non-malignant Respiratory Disease Uloer of Sto~aoh & Duoden~n Cirrhosis of Liver Nephritis & NaphEosis All External Causes of Death Ao~idsn~m Moto= Vehlole Aaoidents All Othe=Aooidenta Suloldes Homloi~em & Other External Causes .at 5% levsl at 1% level 301 313.04 67 73.62 1 1.89 18 18.06 3 1.93 2 2,68 6 6.55 1 1.54 0 1.57 6 3.67 0 0.51 22 26.74 0 0.96 22 25.50 0 0.27 0 0.09 7 6.59 0 0.23 1 1.79 1 2.03 3 1.06 1 0.05 2 1.89 0 0.21 0 0.20 5 6.87 0 0.79 0 0.45 2 2.70 3 2.93 6 5.62 0 0.76 3 5.16 15 17.76 125 117.84 O 1.70 105 102.54 1 4.08 3 2.41 16 16.30 2 0.02 18 24.22 5 8.66 4 5.06 0 0.80 4 3.90 0 0.36 9 10.97 0 1.21 13 6.84 0 1.54 34 26.34 17 15.67 12 7.24 5 9.49 13 6.23 4 4.10 96.2 91.0 53.0 99.6 155.5 74.7 91.6 64.9 0.0 163.7 0.0 ~2.3 0.0 86.3 0.0 0.0 106.3 0.0 55.7 49.3 282.9 2172.8 106.1 0.0 0.0 72.8 0.0 0.0 74.2 102.2 106.7 0.0 58.2 84.5~ 106.1 0.0 102.4 24.5 124.5 96.2 243.4 74.3 57.7 79.1 0.0 102.7 0.0 82.1 0.0 190.1 * 0.0 129.1 108.5 165.7 58.9 208.6 * 97.6 No. of employees - 1,075 Person-years . 23,419 85.6 70.5 1.3 59.1 32.1 9,0 33.6 1.6 0.0 60.1 0.0 51.6 0.0 54.1 0.0 0.0 42.7 0.0 1.4 1.2 58.4 54.3 12.8 0.0 0.0 23.6 0.0 0.0 9.0 21.1 39.2 0.0 12.0 47.3 88.3 0.0 83.8 0.6 25.7 56.1 29.5 44.0 18.7 21.6 0.0 28.0 0.0 37.5 0.0 101.2 0.0 89.4 63.2 85.6 19.i 111.1 26.6 107.7 115.6 295.2 157.5 454.4 270.0 199.5 361.6 235.7 356.3 725.5 124.6 382,7 130.6 1347.7 4112.6 219.0 1616.9 310.6 274.5 826.8 12107.0 38R,i 1793.7 1844.4 169.8 465.8 819.9 268.0 298.7 232.3 482.7 170.0 139.3 126.4 216.5 124.0 136.5 363.8 159,4 879.3 117.5 134.8 202.5 458.6 262.8 1012.7 155.8 303.9 325.1 239.2 1B0.4 173.7 289.4 137.5 356.7 249.9 Cause of death OBS EXP S)~t All COUlee of Death All Malignant Neoplamns. Censor of Buooal Cavity & Pharynx Censor of Digestive Organs & Peritoneum Censer of Esophagus Canoe: of Stomaoh Censer of Largo Intestine Canoe~ of Reotum Censer of Biliary Passages & Liver Censer of Panoreae Canoew of All Other Digestive Organs CanoeE of RespiratorySystem Censer of Larynx Censer of Bronohus, Trashed, Lung Censer of All Other Respiratory Censer of Breast . Censer o Prostate Censer of Testes and Other Male Genital Organs Canoe= of Kidney , CanosE of Bla~der and Other Urinary Organs Malignant Melanom~ of Skin Censer of Eye CanoeE of Central Nervous System Censer of Thyroid & Other Endoorine Glands Canoe~ of Bone . Canaew of All Lymphatio, Haematopoietio Tissue Lymphoearooma & Retiouloearooma Hodgkine Disease Leukemia & Aleukemia Canoe= of All Other Lymphopoietio Tissue All O~herMalignant Neoplasms Benign Neoplam Diabetes Mellitus Cerobrovaeoulmr Disease All Heart Diaeaeo Rhemnatlo Heart Disease Isohentlo Heart Disease Chronlo Endooard. Dis.; Other Myooard. Insuff. Hype=tenslon with Heart Disease All Other Heart Disease Hypertension w/o Heart Disease Non-mallgnant ReepiratozyDieease Influenza & Pneumonia Bronohitie, Emphysema, A~hma Bronohltim En~hyeoma A~thma Other Non-malignant Resplrator~ Disease Uloer of Stomaoh & Duodenum Cirrhosis of Liver Nephritis & Nophroeis All External Causes of Death Aooldents Motor Vohiale Aooidente All Other Aooidonte Suioidem Homioidem & Other External Causes 219 42 0 13 1 3 3 1 0 5 0 10 0 10 0 0 3 0 2 0 1 0 2 0 0 7 0 1 2 4 4 0 0 14 92 1 78 0 0 13 0 18 4 7 2 5 0 7 0 8 3 25 17 12 5 6 2 269.45 62.44 1.59 15.24 1.61 2.22 5.60 1.29 1.35 3.08 0.42 22.48 0.81 21.43 0.23 0.08 5.82 0.19 1.51 1.75 0.89 0.04 1.58 0.17 0.17 5.88 0.65 0.37 2.32 2.54 4.78 0.66 4.41 15.03 100.22 1.42 06.86 3.73 2.07 14.26 0.69 21.34 7.67 4.29 0.68 3.30 0.31 9.80 1.01 5.71 1.33 23.70 13.89 6.48 7.47 5.60 3.98 81.3 67.3 ** 0.0 05.3 62.1 135.3 53.6 77.5 0.0 162.3 0.0 44.5 ** 0.0 46.7 ** 0.0 0.0 51.5 0.0 " 132.7 0.0 112.5 0.0 126.4 " 0.0 0.0 119.0 0.0 267.8 86.4 157.3 83.6 0.0 0.0 * 93.2 91.8 70.6 89.8 0.0 0.0 91.2 0.0 84.4 82.1 163.2 292.7 181.6 0.0 71.4 0.0 140.1 226.0 105.5 122.4 185.3 67.0 107.2 50.3 * signifioant at 5% level ** signifioant at lq love1 No. of employees = 1,069 Person-y, are= 20,815 Lo~r 48.5, 0.0 45.4 1.6 27.9 11.1 1.9 0.0 52.7 0.0 21.3 0.0 22.4 0.0 0.0 10.6 0.0 16.1 0.0 2.8 0.0 15.3 0.0 0.0 47.8 0.0 6.7 10.5 42.9 22.8 0.0 0.0 50.9 74.0 1.8 71.0 0.0 0.0 48.5 0.0 50.0 14.2 65.6 35.4 49.2 0,0 28.7 0.0 60.5 46.6 68.3 71.3 95.8 21.7 39.4 6.1 " '232.3" 145.| 346.3 395.4 156.6 433.1 273.6 378.8 872.0 81.8 455.3 85.8 1592.5 4883.0 150.6 1933.0 479.5 211.1 627.0 9620.5 456.5 2142.1 2166.8 245.1 564.4 1492.1 312.1 402.7 214.2 562.7 83.6 156.3 112.6 393.5 112.1 98.9 178.3 155.9 532.1 133.3 133.5 336.2 1057.5 353.7 1178.8 147.1 364.6 276.1 660.4 155.7 195.9 323.7 156.2 233.4 181.6 Table 8 Mortality from llm~hohematop~ietic can~ers of male employees at the Tozrance refiner~ Total cohort Cause of death Acute lymphoid leukemia Chronic lymphoid leukemia A~ute myeloid leukemia Chronic myeloid leukemia Multiple myeloma Non-Hodgkin's iymphoma OBS EXP SMR %5% CI 0 0.36 0,0 (0.0-I035.2) I 1.34 74.4 (1.9-414.3) I 2.20 45.4 (i. 1-252.0} 2. 1.02 195.6 (23,7-706.6} 4 3.28 122.1 (33.3-312.5) 9 7.42 121.4 (55.5-230 . 4) Maintenance Cause of death Acute lymphoid leukemia Chronic lymphoid leukemia Acute myeloid leukemia Chronic myeloid leukemia Multiple myeloma Non-Hodgkin's lymphoma Process Cause of death Acute l~m~hoid leukemia Chronic lymphoid leukemia Acute myeloid leukemia Chronic myeloid leukemia Multiple myeloma Non-Hodgkin's lymphoma OBS 0 0 0 1 0 2 OBS EXP 0 14 O. 48 0,63 0,39 1.22 2 . 81 EXP SMR O. 0 0.0 0.0 253.6 0.0 71.2 SMR 95% CI (0 . 0-2"/21.0) (0.0-769.7) (0.0-443. O) ' (6.3-1412.8) (0.0-302.4) (8.6-25./. 2) 95% Cl (0.0-3042.5) (6.0-1340.6) (3.5-"/'/6.5) (0.0-1083.0) (2.4-539.2) (45.3-425.4) L~ nc~ch of r'ausm of dea~ Table 9 Morta.l-~t'1r f','om 11rmp'hohematopo-iet:Lc cancers of a~l.e employ~a a.t the Torrance refinery by length of mmplol~ent and ~ s~nce h~cm <i0 ~reare ~ 0 0.0 0 0.0 0 0.0 1 344.2 1 166.6 4 225.9 95q CI (0.0-3332.3) (0.0-1869.5) (0.0-706.9) (8.6-1910.0) (4.2-928.1) (61.1-578.5) 10-29 ~/ea=s OBS ~ 0 0.0 1 120.5 1 80. G 1 179.3 1 51.1 5 119.4 95~t C1 (0.0-1989.3) (3.0-671. G) (2.0-449.4) (4.5-999.3) (1.3-204.7) (36.8-270.7) 30+ ~r-_=,-s OBS ~ 0 0.0 0 0.0 0 0.0 0 0.0 2 277.9 0 0.0 95q CI (0.0-6125 ~ 3) (0.0-1160.3) (o.o-o34.1) (0.0-2115.9) (33.6-1003.9) (0.0-253.0) T,_'~,e s'ince hi:e ~:uts lymphoid Chronic lymphoid ~e ~elo~d le~a ~oni: ~eloid le~a Non-Hogan' s <20 ea:s OaS ~ 0 0.0 0 0.0 0 0.0 2 729.5 1 271.3 4 298.3 95q CZ (0.0-3613.3) (0.0-4546.3) (0.0-958.5) (93.1-2780.0) (6.8-1511.6) (81.3-763.9) 20-'~.o yea:s OILS ~ 0 0.0 0 0.0 0 0.0 0 0.0 2 125.3 2 57.9 95q CI (0.0-2892.9) (0.0-620.1) (0.0-357.'/) (0.0-061.3) (15.2-452.7) (7.0-209.1) 40+ yea:s OBS ~ 0 0.0 1 147.8 1 126.9 0 0.0 1 76.2 3 114.5 95t CZ (0.0-2910.3) (3.7-823.8) (3.2-707.2) (0 . 0-1103.6) (1.9-424.6) (23 6-334 . 6)