Document omeMeKvy1pLooRBKG3m98bm0E

A FOLLOW-UP STUDY OF PETROLEUM REFINERY WORKERS Philip Cole, M.D. Elizabeth Delzell, S.D. Chansuda Wongsrichanalai, M.D. July 8, ]987 SH ELL-LEBLANC-01103,! EXECUTIVE SU~RY This study evaluated the mortality experience of ].0,7].8 employees of the Shell Oil Company Wood River Refinery. The subjects were all employees who had worked at the refinery. for at least six months before December 31, 1983, and who were employed in 1940 or later if staff or in 1942 or later if hourly. We determined the vital status of 94% of the cohort. Of the 2,925 employees who were known to be deceased, death certificates were available for 2,774 (95%). We compared the overall and cause- specific observed numbers of deaths among th~ refinery workers with the numbers expected on the basis of US general population mortality rates. ~ong the 9,484 white men included in the study, mortality rates for all causes and for most major dause of d~ath categories were substantially lower than those of white men in the US general population. Overall, there were 2,874 observed deaths, compared with 3,568 expected (SMR:0.8). The SMR for all cancers combined was 0.9 (95% confidence interval: 0.8-1.0). The generally favorable mortality experience of the white men occurred among long-.term employees, and it was still evident. 40 years after first hire. These results are similar to those of most previous studies of petroleum refinery workers, including an earlier investigation of Wood River Refinery employees. The only statistically significant results for white men for specific cancers were: (I) a 60% excess of leukemfa, based on 44 observed and 28.2 expected deaths; (2) a 70% deficit of deaths SH ELL-LEBLANC-O 1103! from cancer of the buccal cavity and pharynx, based on 6 observed "and 22.5 ekpected d~aths; and (3) a 60% deficit of deaths from cancer of the liver, based on six observed and 15.5 expected deaths. There were small excesses of skin cancer (SMR=I.2), Hodgkin's disease (SMR=I.2) and cancer of other lymphatic and hematopoietic tissues (SMR=I.I) , but none of these was sta%isti- ally signfficant. -There were no excesses of certain c~ncers found to be increased in other groups of refinery wor~ers, including cancers of thestomach, pancreas, lung, bladder, ~idney and brain. The overall excess of leukemia began in'[he 1950s (18/3.7; SMR=2.7) and persisted until 1980. In 1980-1984, there was no excess (6/5.7; SMR=I.I). Analyses of leukemia mortality by cell type, period bf death and ~eriod of hire revealed several interesting patterns. For lymphocytic leukemia,, there was an excess among deaths occurring before 1970 (]1/3.1; SMR=3.5] but no excess in 1970-i984 (5/4.5; SMR=I.I). .Lymphocytic leukemia was increased both among men hired before 1940 and amon~ men hired in 1940 or later. In contrast, for myelocytic leukemia, there was a deficit ~efore 1970 (2/4.8; SMR=0.4) and an excess, due entirely to acute myelocytic leukemia (12/5.0; SMR=2~4) in 1970-]984. Furthermore, the acute myelocytic leukemia excess was limited to men hired in 1940 or later. Monocytic leukemia was increased from 1950 through 1979 {6/]..05; SMR=S.7) ~and did not appear to be related to period of hire. If these patterns are real and not due to misclassification of leukemias by cell type, they have several implications. It is SHELL-LEBLANC-01104( possible -that the restriction of the acute myelocytic leukemia excess to men hired in 1940 or later is related to process changes which occurred at the refinery after 1940. These changes resulted in an increased percentage of benzene in certain refinery streams. However, it is not known whether there was a concomitant increase in workers' e~posure to benzene. The fact that there was an apparent excess of lymphocytic, but not mye!ocytic, leukemia among men hired before 1940 suggests that some exposure other than benzene was responsible. This study did not assess the detailed job histories or exposures of workers at the refinery. Therefore, it does not p~ovide any direct information which would allow us to determine if the overall or cell type-specific leukemia excesses are linked to specific jobs or to exposure ~ benzene or othe~ chemicals. Future observation of cell type-specific leukemia mortality will be necessary to confirm the apparent decrease in the lymphocytic and monocytic leukemia excesses and to determine if the acute myel0cytic leukemia excess persists among Wood River Refinery employees. Additional work will also be required to evaluate the possible association of lymphocytic, myelocytic and monocytic leukemias with benzene and other chemicals at the refinery. SH ELL-LEBLANC-01104' TABLE OF CONTENTS INTRODUCTION ..................................................1. METHODS ........................................................3. Definition of study groups ..................................3. Sources of identifying and employment information ...........4. Ascertainment of vital status ................................ Death certificate acquisition.: ............................. Data analysis ..............................................i.i RESULTS ..................~ ..............." .....................14 White men ..................................................1.5 White women ................................................2.1 Black men ..................................: .... ' ............21 DISCUSSION ' 22 REFERENCES FIGURE .. TABLES APPENDICES ............. A. Classification of prodt~ction-related and nonproduction-related jobs B. Recozd abstract form C. Documentation of merging procedures used to combine files D. Vital status ascertainment E. Nosologic coding for leukemia F. Mortality patterns among always staff employees in PR and NPR jobs G. Wood River Refinery employees who worked for fewer than six months H. Characteristics of Wood River Refinery employees with leukemia SHELL-LEBLANC-O 1104: INTRODUCTION This report .describes the mortality experience of employees of the Shell Oil Company Wood River Refinery in Roxana, Illinois. The refinery, which has been in operation since 1918, is a manufacturing complex whose operations include crude oil refining and production of gasoline, aviation fuel, diesel fuel, lubricating oils, asphalt and various petrochemicals, including benzene. Both a retrospective follow-up study (RFS) and a case-control study' of leukemia among Wood River Refinery workers have been conducted (1,2). The ~JFS examined the cause-specific mortality experience of 3,976 white men who had worked at the refinery for at least one day from January i, 1973, through December 31, 1982, or who had retired from the refinery before 1973 and were alive on January I, ]973 {i). The employees included in the study had mortality rates which were lower than those of United States (US) white men for all causes, all cancers and all specific major organ system diseases. leukemia deaths: However, there was a two-fold increase in the standardized mortality ratio (SMR) for leukemia was 2.13, with a 95% confidence interval (CI) of 1.17 to: 3.58, based on 14 observed and 6.6 expected deaths. Eight of the 14 men who died of leukemia had acute myelo- cytic leukemia, and the SMR for this cell type was 3.94 (95% CI, 1.72-7.88). SH ELL-LEBLANC-01104: The findings pertaining to leukemia at the Wood River Refinery are unusual in two respects. First, eight earlier studies of refinery workers have reported inconsistent results for leukemia (3-10). Three of these studies found an excess, with estimated rate ratios of 1.14 (3), 1.]8 (4) and 1.73 (5), and only the latter rate ratio was statisti- cally significant at the 0.05 probability level. Thus, the leukemia excess among Wood River Refinery employees is the largest reported in the published literature to date. Second, the study of Wood River Refinery employees is the only published report of an excess of" acute myelocytic leukemia among oil refinery workers. Because benzene is present in peroleum refineries and is generally recognized as a cause o~ acute myelocytic leukemia (ii), the observa- tion of an excess of this ceil type was of particular interest and was investigated further in a nested case- control study (2). This s~ldy evaluated the association of leukemia with jobs, departments and estimated benzene levels among workers at the refinery. No specific job or department could be identified as responsible for the leukemia excess. Comparison of the benzene exposure levels of cases and controls did not indicate any association with benzene. The purpose of the present investigation was to examine further the overall and cause-specific mortality patterns of employees at the Wood River Refinery. The study is an expansion .--2-- SH ELL-LEBLANC-01104, of the previous RFS, in terms of both the observation period and the-number of workers included. It is more than twice as large as the earlier investigation, and it covers a longer follow-up period, 1940 through 1984. It also examines in more detail the effects on overall mortality and leukemia mortality of certain employment features such as duration of work, calendar period of work and potential exposure to refinery chemicals, as indicated by hourly versus staff (salaried) payroll classification and type .of staff position. METHODS Definition of study groups The study included all employees who met the following criteria: (a] employed for at least six months before December 31, 1983; (b) hourly and employed at any time during 1942 through 1983; and (c) staff and employed at any time during 1940 through 1983. Employment records of hourly workers who terminated before i942 and of staff workers who terminated before 1940 were not routinely available; therefore, we did not include these employees in the study. As mentioned previously, the present investigation included most of the workers who were part of the previous RFS. "'3-- SH ELL-LEBLANC-O 1104{ We identified the following groups of employees and conducted all analyses separately for these groups: (a) white men and men of unknown race ( these groups are referred to hereafter as "white men"); (b) white women; and (c) black men. For white men, we also defined two mutually exclusive subgroups based on payroll classification: (a) "ever hourly" employees comprised men who had ever been hourly, including men who had started hourly and then transferred to staff positions; as (h) "always staff" employees included men who staff positions and who had never b~e~ hourly. held We further classified the always staff employees as: (a) ever employed in production~related (PR) jobs and (b) empl6~ed only in nonproduction-related (~PR) jobs. We did not consider strike d[]ty in designating staff workers as PR or NPR. Appendix A provides further information on the classification of PR and NPR jobs. Sources of identifying and employment information We sought the following information for all workers included 'in the study: name, social security number, company identification number, gender, race, birth date, dates of first employment and of termination at the refinery, employment status (active, retired or terminated)., payroll classification (hourly or staff), ~ype of staff job (P!~ or NPR) , date of death and state of death. Sources used to SHELL-LEBLANC-01104( obtain hese data are described below. The figure sun~arizes the cohort identification and data sources. We used a computer file developed for the previous RFS to obtain most of the data for employees who terminated after ].972 or who were retired and alive as of January I, 1973. The file was modified to include women and nonwhites and was [~dated through January, 1985. The updated version contained information on 4,585 employees. Both the original and the updated .files of the previous RFS lacked certain job history data needed for the present study, including the following: precise d~es of beginning and terminating work at the Wcod River Refinery; last payroll classification; information on whether or not staff employees had "~ver been on the hourly payroll at theWood River Refinery; and, if ever hourly, the last date hourly. Employment dates on the computer files of the previous study were uncertain for two reasons. First, the files contained the "equated date" of employment rather than the actual hire date at the Wood River Refinery for many employees. The equated date is the hire date, adjusted for periods of unemployment. For staff employees, it refers to the date of starting work at the Shell Oil Company, not at a particular facility. Therefore, the difference between the date of hire at Wood River and the equated date ranges from zero to many years. This difference usually would be larger for staff employees, among whom transfers from location to location are common, than for hourly workers -5- SHELL-LEBLANC-01104: Second, the termination dates were the dates of termina- tion with the Shell Oil Company and did not necessarily specify the date on which employees left the Wood River Refinery. As with equated dates, differences between the Shell Company termination date and the termination date at Wood River tend to occur among staff but not among hourly workers. To verify employment dates and to obtain additional job history information on staff workers included in the previous study, we. obtained additional work records from the refinery and from the Shell Corporate Office. From these records, we abstracted information on hire date; termination date; employment status; last payroll classification; whether or not the employee ever had been on the hourly payroll; if so, the last date on the hourly payroll; and type of staff job (PR or NPR). We did not attempt to verify the employment dates of all hourly workers included in the original RFS. It is likely that errors in their employment dates are few and small. We identified employees who were not included in the previous RFS and obtained information on them from personnel records stored at the refinery. These records are described below. Employee Service Records (ESRs) : ESRs were available for hourly workers who terminated in 1942--1965. The ESRs usual].y contained all of the information sought except race. SHELL-LEBLANC-O 1104~ Personnel Files: Personnel files were available for hourly workers who terminated from 1955 through..1965 and for hourly workers who retired from 1942 through 1965. These files contained most of the information sought but did not provide job history information in as much detail as ESRs. Consolidated Files: Since 1966, hourly workers' ESRs and personnel files have been combined in a singlei "consolidated" record. These were available for hourly workers who terminated or retired from ].966 to 1976. Staff workers have always had consolidated files, and the files were available for those who terminated from 1940 to 1976. Pensioners' Records: Pensioners' records were available f6r some workers who were active pensioners in 1946-1972. These records provide~ essentially the same information as consolidated files. We prepared record abstracts for all available ESRs, personnel files, consolidated files and pensioners" records for workers who terminated or retired before 1973 and for some who terminated after 1973. Appendix B displays the record abstract form. We completed 10,348 record abstracts (employees who worked fewer than. six months included) and developed these into a computer file, referredto hereafter as the "abstract file." One hundred seventy-seven individuals had more than one record on the abstract file. We merged these duplicate records, using procedures which maximized data completeness and accuracy. For example, personnel file data were useful SHELL-LEBLANC-01104~ in supplying information on race, which was not usually availab!6 from ESRs. Pensioners ' records were useful i.n supplyin.g data on race and date of death for a small number of cohort members. The abstract file contained 10,066 eligible subjects after the 177 pairs of record abstracts were merged and after exclusion of workers with unknown SSN (N=39), staff ?~orkers who terminated before 1940, hourly workers who terminated before 1942 (N--19] and employees with missing birth or employment dates (N~4"7). We then combined the records of these i0,066 individunls with records in the updated file from the previous RFS and merged the duplicate records of 1,497 individuals whb appeared on both files. Appendix C providos additional documentation of the merging process. After combining these files, the total number of subjects was 13,154. Of these, 2,400 who had worked for fewer than months were removed from the cohort. This resulted in a final study size of ]0,754, including 9,484 white men, 1,060 women and 210 black men. Thirty-six of the women were known to be black and were excluded, restllting in a total of 1,024 "white" women. Appendix G describes the results of data analysis for the 2,042 white men and the 286 white women who worked.at the refinery for fewer than six months. we nsed several additional sources of information to check the completeness of cohort identification. These records included the Wood River Refinery green books" (1 ........of names of employees who terminated from 1940-1983 with hire date, and reason for termination); lists of hourly and staff workers active -8- SHELL-LEBLANC-01105( as of January, 1985; and note books containing data oh Wood River research laboratory workers durin~ 1950-1975. By crosschecking the names compiled from these three additional sources with the abstract and previous RFS files, we found that records were not available for 261 employees. Ascertainment of vital status We used company, Social Security Administration (SSA) and National Death Index (NDI) records to determine the vital statns of subjects. Iniia].ly, active workers and pensioners were followed using records available at the Wood"River Refinery and the Head Office. We used these records to identify workers who were active either at the Wood River Refinery or at other Shell locations as of January I, 1985; to identify Shell pensioners who were alive as of the same date; and to identify deaths among active workers and pensioners. We submitted to the SSA a list consisting of the 10,272 cohort members who were not active as of January ], 1985, and who did not have a death certificate on file. The SSA provided information on deaths that occurred through 1984 and identified individuals assumed to be alive at the end of 1983. The SSA search identified 2,479 individuals as deceased, 6,904 .as alive and 889 with unknown vital status. To complete vital status ascertainment, we submitted a total of 937 names to the NDI, inc].uding the names of workers whose status cou].d not be determined by Company or SSA records and of workers Whom the SSA identJ [~ed as having died after 1978 but -9- SHELL-LEBLANC-01105' whose death certificates were not located initially. Appendix D describes the vital status aacertainment procedures in more detail. Deathcertificate acquisition The total number of deaths identified was 3,451. This includes 526 deaths among subjects employed for fewer than six months. We obtained 71] death certificates from the Shell Corporate Of Jice. In addition, for 233 employees identified" as deceased during record abstraction, we located death certificates at the refinery or obtained them from Illinoi~or Missouri vital records offices. For the additional 2,479 deaths identified by the SSA, we requested cer{~~icates from the vital records offices of 42 - states. Of the 2,479 decedents identified by the SSA, we found death certificates for 2,253. Of the 226 [9%) not located, 48 died after ].978. We submitted these 48 names to the ,NDI, together with those of workers who were lost to follow-up. The NDI was ah!e to confirm 44 of the deaths, and we obtained death certificates for 40 of them. The NDI identified 28 additional deaths,- and we located death certificates for 27 of these. The totalnumber of death certificates located was 3,267 (95%) of the 3,451 reported deaths. Among workers employed for at least six months, the total number of death certificates located was 2,771 (95%) of the 2,925 reported deaths. -i0- SHELL-LEBLANC-01105: A nosologist coded the underlyihg cause of death according to the Eighth Revision of the International Classification of Diseases, using coding rules in effect at the time of death. A contributing cause of death was also coded if it was a blood disease or a cancer. Appendix E describes the nosologic coding of leukemia. Data analvsis We compared the mortality experience of each of the study groups with that of'the appropriate segment, of the US population, using the SMR as the measure of association, lh this.report, the SMR is expressed as the ratio of the observed to the expected number of deaths. To obtain"{he expected numbers of deaths for the SMR, we multiplied the age-, gender-, race-, calendar time-, and cause-specific mortality rates of the US general population by the corresponding number of person-years (PY) of observation of the refinery workers. We commuted PY of observation by summing the total years of follow-up of workers as they passed through each five-year age group, and calendar period. For staff workers, follow-up began 6 months after the first date of employment or on Januaryl, 1940, for employees hired more than six months before For hourly workers, follow-up began six months after the first date of employment or on January i, 1942; for employees hired more than six months before 1942. Observation continued until death or nntil January ], 1985. For t|iose ].ost to follow-up, PY were accumulated up to the last date of contact, usually the SH ELL-LEBLANC-O 1105: termination daSe at the Wood River Refinery, or for transferred employees, the termination date at another Shell location We calculated approximate 95% confidence intervals (Cl) of the SMRs assuming a Poisson distribution for the observed nttmbers of deaths. When the number of deaths was small, such as in the analysis of leukemia by cell type, we calculated exact 95% CI (12). Throughout this report, results are designated as statistically significant when the two-sided 95% CI of the SMR excludes the null value of 1.0. We used a computer progr=~m developed by Monson' to calculate the SMRs (13). The most recent version of this program contains US mortality" rates updated to 1980. The leukemia rates for 1975-1979 and 1980, however, were erroneously low because of the inadvertent exclusion of unspecified lei~mias from the overall leukemia.rates, starting in 1979 (R. Monson, personal communication). Therefore, we corrected the 1975-1979 leukemia rates as suggested by Monson (Memorandum to .USDR users, 1-5.-87) and rdplaced the 1980 fates with rates calculated from published US vital statistics for 1980 (14). We used rates based on the Surveillance, Epidemiology and End Results (SEER) program to derive expected numbers of cell type-specific leukemia deaths occurring in or after 1970 [15). For leukemia deaths occurring before 1970, we used published ].968 US mortality data to calculate cell type-specific leukemia rates and to generate expected numbers (16). General population morta- lity rates for specific leukemia cell types were not available before 1968. It is .likely that cell type-specific rates would -]2- SHELL-LEBLANC-01105, have increased duri~{g 1940 to 1970 as a result of more accurate d~.agnosis and reporting and, possibly,, a true rise in incidence and mortality. If this were the case, the nse of 1968 rates as referent rates for the period of 1940 through 1969 would overestimate the true expected numbers. The SSA provided information on deaths occurring through 1984 and identified other employees as alive or lost to follow-up as of December 3], 1983. We included all dea[hs occurring in 1984 in the analyses; and, for the pur~6se of computing PY of observation, we assumed that men, identified as alive in 1983 and not reported as deceased in ].984, remained alf~e until December 3], 1984. This assumption could produce an underestimate of SMRs if there were underascertainment of deaths in 1984. However, the SMR for all ca~~s for the overall study period of 1940 (or' 1942) through 1984 was 0o81, whereas the SMR obtained when follow-mp stopped at the end of 1983 was 0.77. This suggests that under- ascertainment did not occur. We were not able to locate the death certificates of 151 (5%) of the reported decedents with at least six months of employment. In the anaiyses, we included these deaths in the SMRs for all causes. However, we did not adjust cause-specific observed numbers of deaths or the corresponding SMR to include a percentage of the deaths from unknown causes. Therefore, cause-specific SMRs are underestimated by an average of 5%. -13- SHELL-LEBLANC-01105( RESULTS Table 1 summarizes the vital status of the cohort as of January I, 1985. Of the 9,484 white men, 6,266 (66%) were alive at the close of follow-up, 2,874 (30%) had died and 344 (4%} had unknown vital status. We located the death certificates of 2,733 (95%) of the white men who had died. Of the 1,024 white women, 728 (71%) were alive, and only 45 deaths had occurred. Followiup was less complete for white women than for men: vital status was unknown for 25% of the women, and death certificates could not be located for i0 (22%) of the reported deaths. Of the 210 black men, ]91 (91%) were alive, 6 [3%) had died and 13 (6%) were lost to follow-up. We located death certificates -for the six decedents. Table 2 p~vides information on the employment status and payroll classification of the cohort. Of the white men, 7,524 (79%) were ever hourly; 1,871 (20%) were always staff; 47 were staff, unknown if ever hourly; and 42 had unknown payroll class- ifications. Of the white men who were always staff, 1,109 (59%) had held PR jobs. Of the white women, only 167 (16%) were ever hourly, 852 were always staff and 781 (92%) of the women who were always staff had held NPR jobs. Of the black men, 89% were ever houmly. Table 3 provides further information on employment characteristics and age distribution of the cohort. About 60% of the white men and 45% of the white women were hired before 1950, whereas all black men were hired after 1949o Sixty-eight per cent of the white men had worked for five or more years, and 27% had worked -14- SHELL-LEBLANC-01105( for at least 25 years. In contrast, white women and black men tended to have worked for shorter durations, with 66% of the white Women and 49% of. the black men having fewer than 5 yeazs of employment. White men and women had been followed for an average of about 28 and 21 years, respectively. Black men had a considerably shorter average period of observation (9.5years). Among white men, 43% of the total 261,737 PY were contributed at ages greater than 49 years. White women and black men were younger, with only 23% and 2%, respectively, of their PY contributed at ages above' 49 years. Results of mortality analyses are presented below, separately for white men, whi'te women and black men. White men Table 4 displays observed .and expected numbers of deaths among white men. The total number of observed deaths was 2,874, whereas 3,568 deaths were expected. The corresponding SMR for all causes was 0.8, reflecting approximately a 20% deficit in overall mortality among these workers compared to US white men. Observed numbers of deaths from most specific disease categories were also 9ppreciably fewer than expected, and most of these deficits" were statistically significant. For all cancers combined, there were 661 observed and 728~6 expected deaths (SMR=0.9, 95% CI=0.8-1.0) . If the 141 men for whom death certificates were not found were allocated to disease categories according to the cause distribution of the deaths from known causes, the can~er SM~ would have been 1.0 (0.9-1.1). -15- SHELL-LEBLANC-01105: There were similar mortality defici{s among the 4,845 ~*hite men who had worked at the refinery for 10 or more years (table 5) and among the 6,999 men who had been followed for at leas[ 20 years since starting work (table 6). In each of these two groups, there was a 30% excess of deaths from diseases of the blood and blood-forming organs, based on 8 observed and 6.2 expected deaths (men employed for I0 or more years) and on 9 observed and 6.9 expected deaths (men followed for at least 20 years since starting work). The following specific diagnoses were reported on the death certificates of the nine decedents with this underlying cause of death: aplaitic anemia secondary anemia (unspecified priggery cause) (N=2), thrombocyto- penic purpura (N:I) , pernicious anemia (N=I), acute splenitis (N=I) and Felty'.s syndrome (N=]). Table 7 contains a further description of mortality patterns for all causes and all cancer by duration of employment and time since starting work. The observed number of deaths from all causes was lower than the expected number by 14% to 25% i~ all duration of employment categories. The deficit was largest (44%) within I0 years of star9ing-work but was still substantial {17%) among men who had been followed for 30-39 years since starting work. Table 8 displays observed and expected numbers of deaths and SM~s for site-specific cancers. The observed number was approximately equal to or was less than the expected number for all sites except skin (observed/expected, ]5/].2.1), Hodgkin's disease (9/7.~), leukemia (44/.28.2) and other lympheric and hematopoietic -16- SHELL-LEBLANC-011 tissues (20/17.8) . The SMR for leukemia was 1.6 (95% CI=I.I-2.1), reflecting a statistically significant 60% excess of deaths from this disease. For skin cancer, Hodgkin's disease and cancer of other lymphatic and hematopoietic tissues, the SMRs were only slightly elevated (SMRs: ].2, ].2 and I.i, respec- tively) , and none was statistically significant. There were statistically significant, deficits of deaths from buccal cavity and pharynx cancer (6/22.5; SMR:0.3, 0.I-0.6) and from liver cancer (~/15.5; SMR--0.4, 0.i-0.8). There were no marked differences between the sitespecific cancer SMRs for men who had worked at the refinery for at least 10 years (table 9] and the SMRs of the overall cohort. Among these men there were 33 observed and 18.9 expected leukemia deaths (SMR=I.7, 1.2-2.5). As shown i.n t.~h]e ]O, men who were followed for at ]east 2(] years since starting work had a leukemia SMR of 1.5 {I.0-2.]) , based on 34 observed and 22.7 expected deahs. This was not materially different from the leukemia SMRs for the overall cohort and for men employed for at least i0 years at the refinery. These men also had slight excesses of cancers of the large .intestine and rectum (85/73.4) and central nervous system (CNS) (17/15.0). The excesses of skin cancer (12/9.3), Hodgkin's disease (7/4.3] and cancer of other lymphatic and hematopoietic tissue (19/16.1) were slightly higher in this group of workers than in the overall cohort of white men. The mortality patterns of ever hourly employees were generally similar to those of staff employees (tables ii and 12). -17- SHELL-LEBLANC-O 1105~ However, the re].ative excess of leukemia was higher among staff (9/3.5; SMR=2.6, 1.2-4.9) than among hourly workers (35/24.5; SMR=I. 4, i. 0-]. 9) (table 12) . A direct comparison of the leukemia mortality rates of staff employees with the rates of hourly employees yielded a rate ratio of 2.8, with a 95-% CI of 0.9-4.1. All deaths from diseases of the blood and blood forming organs occnrred among honrly employees, for whom there was a slight overall excess (9/8.2) (table 11). Within the subcohort of always staff employees, there were no striking differehces in mortality patterns between men in PR and NPR jobs (Appendix F, tables F1 and F2) ....The leukemia SMR was 2.]., based on-4 observed and 1.9 expected deaths, among men in PR jobs and 3.2, based on 5 observed and 1.6 e~<pected deaths, .among men in NPR jobs. Table ]3 further describes the lemkemia mortalty patterns of the overall cohort. The leukemia excess appeared to be largest in the 1950s (10/3.8, SMR:2.6), although excesses also occurred in the 1960s (11/6.8, SMR=I.6) and the 1970s (16/10.5, SMR=I.5]. In the 1980s there was virt~ally no excess of leukemia (6/5.7, SMR=I.II) . There was only one leukemia death . in the 1940s, with an SMR of 0.8. The overall leukemia excess did not vary ma~erially by period of hire, although the largest number of observed deaths and the only statistically significant excess occurred among men hired before 1940 (24/14.3; SMR:f~7, 1.1.-2.5). This was due primarily to a 2.3-fold increase in le<tkemj_a among men hired ~n 1935-1939 (Ii/4.7). Among men hired before 1935, the leukemia SMR was 1.4 (13/9.6} . -18- SHELL-LEBLANC-01106( Although there was no clear pattern of increasing leukemia SMRs with increasing duration of employment, the SMR was somewhat larger among men who had worked at the refinery for at least I0 years (SMR=I.7, 1.2-2.5) than among men employed for fewer than I0 years (SMR=I.2, 0.6-2.1). There was no excess of leukemia within I0 years since first hire (1/2.0; SMR=0.5, 0.0-2.8). The largest excess occurred during the period of ].0-39 years since starting work (SMR=I.9, 1.3-2.6) and persisted until at least 40 years since first hire had elapsed (SMR=I.I, 0o5-2.1). Table ]4 di>plays leukemia SMRs jointly by period of hire and period of death. Among men hired before 1940, there was a 2.5-fold increase in leukemia (]7/6.8) during ]940-1969 and no excess thereafter (7/7.5). In contrast, among men hired in 1940 or later, no e~cess of leukemia was present before 1970 (5/5.2), whereas a 70% excess, based on ]5. observed and ~.7 expected deaths, occurred after ].969. Table 15 displays cell type-specific observed and expected numbers of leukemia deaths by period of death. Overall, the workers experienced a slight excess of myelocytic leukemia (15/12.4, SMR=I.2), a 2.i-fold increase in lymphocytic leukemia (16/7.6) and. a 4.8-fold increase in monocytic leukemia (6/1.3). Theexcess of lymphocytic leukemia was highest in 1950-1959 (5/1.0; SMR=5.1), declined in 1960-1969 (6/2.1; SMR=2.8) and disappeared in 1970-1984 (5/4.5; SMR=I.I). There was no excess of myelocytic leukemia in 1950-1959 (0/]..8) or in 1960-1969 (2/3.0). In contrast, in ]970-1984, most of :he overall leukemia excess was due to increased mortality from acute myelocytic -19SHELL-LEBLANC-01106' leukemia (12/5.0; SMR:2.4). There was an excess of monocytic leukemia front 1950 through 1979 (6/i.0; SMR=5.8), but no deaths from t~is cell type of leukemia occurred after 1979o Analysis of cell type-specific leukemia by period of hire also revealed different patterns for. lymphocytic and myelocytic leukemia. Men hired in or after 1940 had a 1.7-fold increase in deaths from myelocytic leukemia {]1/6.4), whereas there was no excess of deaths from this cell type amon~ men hired before 1940 (4/5.9). In contrast, the excesses of lymphocytic and mono- cytic leukemias ,were not related .to period of hire (lymphocytic leukemia: hired before 1940, 9/%.3; hired in or after 1940, 7/3.2) (monocytic leukemia: hired before 1940, 4/0.6; hired in or after 1940, 2/0.6). There wer6 slight differences in the leukemia ceil type distribution between ever hourly and always staff employees Among the eve~ hourly, . 13 (37%) of the leukemias were lymphocytic, 13 (37%) were myelocytic, 4 (1].%) were monocyti~ and 5 (14%] were other or unspecified. Among the always siaff, 3 (3S%) were lymphocytic, 2 (22%) were myelocytic, 2 (22%) were monocytic and 2 (22%) were other or unspecified. Six men with leukemia reported on the death certificate as a contribntlng, rather than underlying, cause of death were not included in any of the above analyses. Of these men, two had lymphocytic leukemia and were hired before 1940; one had acute myelocytic leukemia and was hired after 1940; one had unspecified leukemia and was hired before 1940; and two had unspecified leukemia and were hired after 1940. Appendix H provides further -20- SH ELL-LEBLANC-O 1106: details on employment and other characteristics of all st~bjects with leukemia mentioned on their death certificates. Table 16 displays observed numbers of deaths and SMRs for diseases of the blood and blood forming organs among hourly workers. An excess occurre among men who were hired before 1940. (7/4.8; SMR=I.5, 0.6-3.0); among those who were employed for 20 or more years (7/4.0, SMR=I.7, 0.7-3.6); and among those who were followed for at least 30 years since starting work (8/4.4~ SMR=].8, 0.8-3.6). White women White women employed at the Wood River Refinery had lower mortality rates than US white women (table 17). The SMR for all causes was 0.6 (0.5-0.9), based on 45 observed and 71.2 expected deaths. There was a 20% excess of breast cancer (7/5.8; SMR=I.2, 0.5-2.5) and a two-fold increase in hematopoietic malignancies (4/2.0; SMR=2.0, 0.5-5.1) (table 18). The leukemia SMR was 1.2 (0.0-6.9), based on one observed and 0.8 expected deaths. Black men -Blac~ men employed at the refinery had lower mortality rates than US black men (table 19). The SMR for all causes was 0.7 (0.2-1.4), based on six observed and 9.3 expected deaths. -21- SHELL-LEBLANC-01106'. DISCUSSION This @tudy evaluated the mortality experience of ]0,718 oil refinery workers, white men, who accounted for 88% of the overall cohort, had an average of nearly 28 years of follow-up during the 45-year study period of 1940-1985. Their mortality rates for all causes combined and for most major cause of death categories were much lower than those of white men in the general US population. Th& other noteworthy results of this study pertain to t'h~ statistically significant excess of leukemia among white men, present since 1950, and to the patterns of this excess by cell type, period of death and period of hi~e. These patterns include: (a) a marked shift from an excess of l~phocytic leukemia among deaths occurring before 1970 to an excess of acute myelocytic leukemia in 1970-]984, (b) the presence of lymphocytic and monocytic leukemia excesses among, men hired both before 1940 and in or after 1940 and (c) restriction of the myelocytic leukemia excess to men hired, in 1940 or later and dying in or after ]970. The overall favorable mortality experience of the white men included in this study persisted among long-term employees, and it was st~ll present after 40 years of follow-up. These results are: consistent with most previous, studies of oil refinery workers (3-10,17) and with the earlier investigation of Wood River Refinery employees (i) . The strength and consistency of the findings suggest that the favorable mortality e.~.p~rlcnce" ~ ~ of oil refinery workers is real, ra6her than an artifact of incomplete fo] low-up (only 4% were lost to follow-up in the -22- SHELL-LEBLANC-01106, present study). General factors which may be responsible for the apparent health advantage of oil refinery workers include selection of relatively healthy individuals for employment in the industry and maintenance of their healthier status through benefits deriving from the economic advantages of employment (18). Moreover, .these findings point strongly to the absence of pervasive occupational hazards within the industry which might have a negative impact on employees' health. We found a sta,tistically significant 60% excess of leukemia deaths among white men at the refinery. This was smaller than the approximately two-fold increase reported in an earlier study of men at the same refinery (I). However, the present study was about 100% larger and included a longer period of observation than the previous study. The previous study also included a relatively, large proportion of retirees. Although we did not examine the mortality experience of retirees per se, we did evaluate leukemia mortality by duration of employment. Among men employed for at least 20 years, a group 'which includes most retirees, the leukemia excess was also about 60%. Several features of the association 6f leukemia with employmeni at {~e Wood River Refinery are important to consider further. First, there was no consistent increase in the leukemia SMR with duration of employment. The fact that there is little evidence of "dose-re~=ponse" does not detract greatly from a causal interpretation of the obs(:rved association, however, because overall duration of employmen may not be highly correlated with leukemogenic exposures, if any,-at the refinery. The SHELL-LEBLANC-01106{ pattern of leukemia mortality with time since first hire was more consistent with an excess attributable to occupational factors. There was no leukemia excess within the first ]0 years of hire; variable excesses, ranging from 60 to 150%, but no consistent trend, occurred 10-39 years after hire; and there was, again, essentially no increased leukemia-mortality 40 or more years after hire. The leukemia excess was higher among staff than among hourly workers, who probably experienced the higher exposures to chemicals at the re{inery, and there was little difference in the magnitude of the leukemia excess between men i~ PR and NPR jobs. These observations appear to be inconsistent with the hypothesis that the leukemia excess is the result of work-related exposures. However, class~'{ication of workers by payroll status and job categories may not have been adequate to distinguish their potential for leukemogenic exposures at the refinery. It is also possible that unidentified nonoccupational exposures or exposures at work sites other than the Wood River Refinery were responsible for-the observed leukemia excess. The relatively large number of leukemia deaths identified by this study. (N:44 among white men) made it possible to evaluate the leukemia excess by cell type and by period of death. Of particular interest were the findings of a shift from an excess of lymphocytic leukemia among deaths occurring before 1970 to an excess of myelocytic leukemia in 1970 through 1984 and the presence of an excess of monocytic leu]~emia until the 1980s. The excess of myelocytic leukemia in 1970-1984 was due primarily to a -24- SHELL-LEBLANC-O 1106( 2.4-fold increase in acute myelocytic leukemia, as previously reported (]). The difference in the celi type distribution between the two time periods could be accounted for in part, but not entirely, by the higher proportion of unspecified leukemias in 1950-1969 than in later years. In 1950-1969, there were three unspecified leukemias. If all of these were, and had been classified as mye!ocytic, the per cent myelocytic would have been 23%, and the myelocytic leukemia SMR. would have been 1.0 (5 observed/4.8 expected). In cont}ast, 13 (59%) of the leukemias in 1970-1984 were myelocytic, and the SMR was 1.7. There w6~e two unspecified leukemias during the latter time period. Diagnostic techniques for cell type-specific leukemia have improved during the past few decades, and mi~classification between myelocytic and lymphocytic leukemias in 1950-1969 is another possible explanation for the change over time in the cell type distribution. Benzene, which is present along with other aromatic hydro- carbons in oil refineries, is considered a cause of acute myelocytic leukemia (II) but not of leukemia of other cell types. Studies of rubber workers exposed to benzene, toluene and other 6rganic solvents have suggested that these agents are associated with chronic lymphocytic leukemia (19,20). However, there is no firm evidence relating specific chemical agents to leukemia cell types other than that relating benzene to acute myelocytic leukemia. Manufacturing processes at the Wood River Refinery have changed over time, leading to changes in the number of [~nits -25- SHELL-LEBLANC-01106~ that produce or use benzene and in the percentage of benzene in refinery streams. The concentration of benzene in refinery streams was not more than 5% until 1941, when tolt~ene units were built with one stream of 5% to 10% benzene. Benzene was also supplied by tank cars for the cumene plant (operating from 1943 to 1945) and .the dodecyl benzene unit (1950-1954), and a coke oven product containing 65-88% benzene was delivered to the refinery from 1966 to ea}ly 1983. Benzene extraction units and catalytic reforming units,.in operation since 1960, have benzene streams of 5% to 100%. The increased amounts and concentrations bf benzene present at the refinery were accompanied by enhanced efforts to control exposure. Thus, the percentage of benzene in refinery streams may not have .....been correlated directly with actual exposure. However, it is of interest that an excess of myelocytic leukemia was found only among workers hired in 1940 or later and that this excess was largely attributable to acute myelocytic leukemia. The present study did not include an assessment of work-related exposures at the refinery. Therefore, it is not clear which specific exposure or exposures at the refinery were responsible, for the overall leukemia excess or for the marked shift ih the cell type distribution. The fact that an excess of lymphocytic leukemia was found among men hired before 1940 suggests that benzene was not entirely responsible for the overall excess o.ff this cell type, since there was no co~res- ponding increase in myelocytic leukemia among men hired before 1940.. -26- SH ELL-LEBLANC-O 11061 Only three (3-5) of eight (3-10) studies of other groups of refinery workers, in addition to the earlier study of Wood River employees (1), have reported leukemia excesses. Morgan and Wong (5) found an SMR of 1.73 (1.09-2.60) in a study of about 6,00.0 refinery workers followed from 1945 to 1978, a result which is quite similar to the SMR of 1.6 in the present study. However, in the other two studies, the leukemia excesses were less than 20% and were not statistically significant. Divine (4) reported an SMR of 1.18 (0.87-1.56) in a cohort of almost 20,000 refinery and petrochemical workers followed from 1947 to 19'77. Wen (3) found a leukemi~ SMR of 1.14 (0.81-1.57) a~ong about 15,000 refinery workers followed from 1937 to 1978. In the present study, there was no excess of certain cancers found to be increased in othe~ groups of refinery workers, including cancer of the brain, lung, stomach, pancreas, bladder and kidney (21). However, there were excesses of skin cancer, Hodgkin's disease and other lymphatic and hematopoletic tissue cancer. For each of these, the excess was small and was not statistically significant. Moreover, there were no marked trends with period of hire, duration of employment or time since first hire. Studies of other groups of refinery workers have reported inconsistiht results for each of these disease categories. For skin cancer, there have beenthree positive studies (3,5,6), with SMRs ranging from 1.22 to 2.16, and three negative studies (4,8,9), with SMRs ranging from 0.84 to 0.99. For Hodgkin's disease, two studies found slight excesses (3,4) (SM]{s; I~08 and 1.45, respectively), and four studies reported deficits SH ELL-LEBLANC-01106! (3,6,9,10). For other lymphatic tissue cancers, two studies reported negative findings (7,8), and there were five positive studies with SMRs ranging from 1.15 to 1.58 (3-5,9,10). None of these SMRs was statistically significant. This study found a slight excess of nonmalignant diseases of the blood among certain subgroups o~ white men. Only two studies of other groups of oil refinery workers have found an excess of these diseases. In the first study, the SMR was 2.2, based on 6 observed and 2.7 expected deaths (22). In the second, the SMR was 1.2, based on 9 observed and 7.5 expected deaths (8). The present study found no increased mortality from diseases of the blood in the overall cohort. However, there was a 30% excess among men employed for at least I0 years and. among men followed for at least 2~"-~ears after hire. Of the nine men with diseases of the blood, three had aplastic anemia, which may be caused by exposur~ to benzene or by certain drugs. We have no information on the specific jobs or work-place exposures of these three men; therefore, can not determine whether their disease was occupationally related. Like the white men included in this study, white women and black men..imployed at the refinery had favorable overall mortal- ity experiences. However, because the numbers of women and black men were small, the study contained little information on their cause-specific mortality patterns. White women had no exc~ss of leukemia (i observed, 0.8 e..p..cL<.d) , and no leukemia deaths were 9eported among black men. However, black men tended to be -28- SH ELL-LEBLANC-01107( relatively young and rece.nt].~ hired, and their expected number for leukemia was only 0.06. In summary, this study reinforces other available evidence that p(]troleum refinery workers have an overall mortality experience which is favorable relative to that of the general population. It also supports previous work which suggests that some, but not all, groups of oil refinery workers have an exce~s of leukemia. Assuming that the leukemia mortality patterns by cell type, period of hire and period of death seen in the present study are real. and. not due to misclassification by cell. type, they have several implications: (a) The restriction of the myelocytic leukemia excess to men hired in or after 1940 may be related to changes in exposures at the refiner~-in the 1940s or later. One possible explanation is an increased exposure to benzene resulting from the process changes which occurred at the refinery after 1940. However, this explanation is speculative because the changes may not ..have resulted in a concomitant increase in workers' benzene exposure levels and because other exposures, possibly relevant to leukemia, may have been introduced during this period. Also, the present study did not include a direct evaluation of the relation between l~ukemia and benzene or other chemical exposures, and an earlier case-control study of leukemia at the Wood River ~efinery found no association with benzene (2). (b) The excess of lymphocytic, but not myelocytic, leukemia among men hired before ]940 suggest-.~ that some exposure other than benzene was responsible. -29- SH ELL-LEBLANC-01107' (c) Finally, the lymphocytic leukemia excess among white men appeared to diminish in 1970-1984. It is possible that this trend is real and reflects a reduction in some unknown leukemogenie exposure related to hhis cell type. The acute myelocytic leukemia excess was also slightly lower in 1980-.1984 (3/]..7, SMR=I.7) than in 1970-1979 (9/3.3; SMR=2.7), and there have been no deaths from monocytic leukemia in the 1980s. However, the expected numbers of cell type-specific leukemia deaths in 1980-1984 were rather small, and future observation of leukemia mortality will be ne~cessary to evaluate the apparent reduction in the lymphocytic ind monocytic leukemia excess4s and to determine if the acute myelocytic leukemia excess will persist among Wood River Refinery employees. -30- SHELL-LEBLANC-01107: REF~RENC S McCraw DS, Joyner RE, and Cole P. Excess ].eukemia in a refinery population. J Occup Me 27:220-222, 1985. Austin H, Cole P, and McCraw DS. A case-control study of leukemia at an oil refinery. J Occup Med 28:1169-].173, 1986. Wen CP, Tsai SP, McClellan WA, et al. Long-term mortality study of oil refinery workers: I. Mortality of hourly and salaried workers. Am J Epidemiol 118:526-542, 1983. Divine BJ, Barron V, and Kaplan SD. Texago mortality study I. Mortality among refinery, petrochemical, and research workers. J Occup Med 27:445-447, 1985. o Morgan RW., and Wong O. An epidemiologic analysis of the mortality experience of Mobil Oil. Corporation employees at the Beaumont, Texas refinery. Oakland, CA: Environmental Health Associates, Inc. 1984. Q Rushton L, and Alderson MR. An epidemiological survey of eight oil refineries in Britain. Br J lad Med 38:225-234, 1981. Schottenfeld D, Warshauer ME, Zauber AG, et al. prospective study of morbidity and mortality in A petroleum industry employees in the United States - a preliminary report " In Peto R., Schneidermann M (eds) : Quantification of Occupational Cancer. Cold Spring Harber, N. Y : Cold Spring Harbor Laboratory, 1981. Hanis NM, Shallenbezger LG, Donaleski DL and Sales EA. A retrospective mortality. study of workers in three major US refineries and chemical plants. Part i: comparisons with US population. J Occup Med 27:283-292, 1985. Wong O, Morgan RW, Bailey WJ, et al. An epidemiological study of petroleum refinery employees. Br J Ind Med 43:6-17, 1986. i0. Kaplan SD. Update of a mortality study of workers in petroleum refineries. J Occup Med 28: 514-516, 1986. ii. International Agency for Research on Cancer. IARC Monographs on the evaluation of th,~" carcinogenic risk of chemicals to humans: Seine ~ndustrial chemicals and dyestuffs. Vo].. 29. Lyons, France : International Agency for Research on Cancer, pp 93-148, 1982. -31- SHELL-LEBLANC-01107', 12. Liddell FDK. Simple exact analysis of the standardized mortality ratio. J Epidemiol Community Health 38:85--88, 1984. 13. Monson RR. Analysis of relative survival and proportional mortality. Comput Biomed Res 7:325-332, 1974. 14. National Center for Health Statistics. Vital Statistics of the United States, 1980, Vol. If, Mortality, Part A. DHHS Pub. No. (PHS) 85-]101. Washington, D.C.: Department of Health and Human Services, 1985. 15. National Cancer Institute. Surveillance, Epidemiology and End Results: Incidence and Mortality Data 1973-1977, NIH Publication 81-2330. Bethesda, Md.: National Institute of Health, 1981. 16. National Center for Health Statistics. Vital Statistics of the United State~ 1968, Vol.ume II-Mortality, Part A, Publication (HSM) 72.-]101. Washington, D.C.: Dept. of Health, Education and Welfare, 1972. 17. Thefiault G, and Provencher S. Mortality study of oil refi~ery Workers: five-year follow-up. J Occup Med 29:357-360, 1987. 18. McMichael AJ. Standardized mortality ratios and the "healthy worker effect": scratching beneath the surface. J Occup Med 18: 165-168, 1976. 19. McMichael AJ, Spirtas R, Kupper LL. An epidemiologic study of mortality within a cohort of rubber workers, 1964-72. J Occup Med 16:458-459, 1974. 20. Checkoway H, Wilcosky T, Wolf P, and Tyroler H. An evaluation of the associations of leukemia and rubber industry solvent exposures. Am J Ind Med 5:239-249, 1984. 21. Savitz DA, and Moure R. Cancer risk among oil. refinery workers: a review of epidemiologic studies. J Occup Med 26:662-670, 1984. 22. Hanis NM, Holmes TM, Shallenberger MPA, and Jones Epidemiologic study of refinery and chemical plant workers. J Occ[~p Med 24:203-212, 1982. 23. W.H.O. Manual of the Ihkernational Statis<ical Classification of Diseases, Injuries and Causes of Death, Sixth Revision, Vols. 1,2. Geneva, Switzerland: W.H.O., 1948. -32- SHELL-LEBLANC-01107, 24. National Center for Health Statistics. International Classification of Diseas-es, Adapted for Indexing Hospital Records by Diseases and .Operations. Vols. ].,2. DHEW Publication No. (PHS) 719. Washington, D.C.: Departnlent of Health, Education and Welfare, 1959. 25. National Center of Health Statistics. Eighth Revision, International Classification of Diseases, Adapted for use in .the United States, Vols. ],2. DHEW Publication No. (PHS) ].693. Washington, D.C.: Department of Health, Education and Welfare, 1967. 26. Health Care Financing Administration: The International Classification of Diseases, 9th Revision, Clinical Modification (I<D-9-CM). Vols. i,2. DHHS Publication No. (~HS) 80--1260. Washington, D.C.: Department of Health and Human Services, 1979. -33- SHELL-LEBLANC-01107( Missing records N ~ 26[ Record abstracts N ~ i0,3&8 lJnknown SStt / N = 39 T~rmlnatlon < [940 (staf) or < 19A2 (hourly) N = 19 Missing birth or~ employment dates N ~ ~7 Potentlaily eli g Ib i e suhject, s N = 10,066 Dupil~utes N= Duplicates N~3 ev ious RF~ 3,976 llired lO83 N = 22 Updated RFS N = 4,5~5 Dup [ic.a tes N = I ,497 Employed < 6 monthu N = 2,400 CombJ ned file N = 14,651 Prel Iminary cohort N = 13,154 Ylnal cohor t~ N = 10,754 Whl t e men N = 9,484 Women I,I = 1,060 Black men N = 210 Figure: Data sources and cohort identification ..... -34- SHELL-LEBLANC-01107( Table. 1 Distribution of Wood River Refinery workers according to vital status Total: Vital Status: Alive: , Active Not active Death certificate obtained Death certificate not 6btained Unknown: White Men No. % 9,484 I00 White Women No. % 1,024 I00 6,266 1,383 4,883 2,874 2,733 "141 344 66 (22) (78) 30 (95) (5) 4 728 142 586 45 35 I0 251 71 (20) (81) 4 (78) (22) 25 Employed at the Wood River Refinery as of 1-]-85. Black Men No. % 21.0 !OO 191 124 67 6 6 0 91. (65) (35) 3 (I00) ( 0] ].3 6 -35- SHELL-LEBLANC-01107~ Table 2 Distribution of Wood River Refinery workers according to emplo.yment status and payroll classification Total: Employment Status: Active Pensioned Terminated Deceased* Transferred Unknown Payroll Classification: Ever hourly: Always hourly Hourly and staff Unknown if ever staff Always staff: PR%., NPR'~ Unknown if PR Staff, unknown if ever hourly Unknown if hourly or staff White Men No. % 9,484 I00 1,383 2,695 4,389 323 683 11 15 28 46 3 7 0 7,524 6,416 I,I01 7 1,871 1,].09 754 8 47 79 (85) (14) (0) 20 (59) (40) (0) 1 42 0 White Women No. % 1,024 I00 Black Men No. % 210 i00 134 29 844 4 12 1 13 3 82 0 123 1 84 0 0 59 0 40 0 0 ].67 16 154 (92 lO ( 6 3 (2 852 83 71 ( 8 78] (92 0 (0 20 30 186 89 180 (97) 6 (~) o (o_). 22 (i0) 9 (41) 13 (59) 0 (0) 00 21 Died while actively employed. %PR, ever employ6~ in production-related jobs. #NPR, employed only in nonproduction-related jobs. SHELL-LEBLANC-01107~ Table 3 Distribution of Wood River Refinery workers according to selected characteristics Total no. of workers: White men No % 9,484 i00 White women No % 1,024 i00 Black men No % 210 ]00 Year of hire: <!940 ]940-1949 >1950 Duration of employment (yrs) : 2,515 3,219 3,750 0.5-4 5-14 15-24 >25 3,043 2,483 1,425 2,533 of follow-up: Total: Average: Age (yrs) : 26],737 27.6 <4O 40-49 50-59 60-69 >70 83,986 66,666 58,919 36,968 15,198 27 34 40 3"2 26 15 27 i00 - 32 25 23 14 6 28 435 "561 3 42 55 0 0 210 0 0 I00 680 270 40 34 66 26 4 3 102 97 II 0 49 46 5 2].,231 20.7 i00 - 1,996 9.5 I00 11,632 4,653 3,396 1,344 206 55 1,632 22 325 16 39 60 10 82 16 2 0 0 PY, person-years -37- SH ELL-LEBLANC-01107! Table ~ Observed (obs) and expected (exp) numbers Df deaths and SMRs for selected causes among white men emp]_oyed at the Wood River Refinery Cause of death All causes All cancers Infective and parasitic diseases Benign neoplasms Diabetes mellitus ]Diseases ofthe blood and blood forming organs Diseases of the nervous system Diseases of the circulatory system Diseases of the respiratory system Diseases of the digestive system Diseases of the genitourinary system Senility and ill defined conditions All external causes Other kno~n causes Obs 874* 661 14 8 48 9 14 1,441 136 108 28 19 228 17 Exp 3,568.0 728.6 59.9 SMR 0.8 0.9 0.2 9.7 50.6 9.3 0.8 0..9 -~ 1.0 31.7 0.4 1,850.3 0.8 224.5 0.6 169.1 0.6 53.5 0.5 35.7 0.5 30] .5 32.8 0.8 0.5 (95% Cl%) (0.8-0.8) (0.8-i.0) (0.i-0.4) (0.4-1.6) (0.7-1.3) (0.4-].8) (0.2-0.7) (0.7-0.8) (0.5-0.7) (0.5-0.8) (0.4-0.8) (0.3-0.8) (0 . 7-0 . 9) (0.3-0.8) Includes 141 deaths for whom the death certificate and cause of death were not available. %CI, confidence interval of the SMR. -38- SHELL-LEBLANC-01108( Table 5 Observed and expected nnmbers of deaths and SMRs for selected causes among white men who worked for ]0 or more years at the Wood River Refinery [N : 4,845) Cause of death All causes All cancer Infective and parasitic diseases Benign neoplasms Diabetes mellitus Diseases of the blood and blood forming organs Diseases of the nervous system Diseases of the circulatory system Diseases of [he respiratory system Diseases of the digestive system Diseases of the genitourinary system Senility and ill-defined conditions All external causes Other known causes Obs ].,910" 436 Ii 6 34 8 Ii 1,021 90 67 2] 13 112 8 Exp 2,459.2 .505. i 36.2 6.3 35.5 6.2 20.0 1,335.2 162.8 110.2 37.5 23.8 153.1 21.2 SMR 0.8 0.9 0.3 1.0 1.0 1.3 0.5 0.8 0.6 0.6 0.6 0.6 0.7 0.4 (95% CI ) (0o7-0.8) (0.~-1.0) {0.2-0.5) (0.4-2.1) (0.7-1.3) (0.6-2.5) (o.3-1.o) (0 7-.0.8) (0 4~0.7) (o 5-o. 8) (0 4-0.9) (0 3-0.9) (0 6-0.9) (0.2-0.7) Includes 72 deaths for whom the death certificate and cause of death were not available. -39- SHELL-LEBLANC-01108' Table 6 Observed and expected number of deaths and SMRs for selected causes among white men employed at the Wood River Refinery, 20 years after first hire (N---6,999) Cause of death All causes All cancer Infective and parasitic diseases Benign neoplasms Diabetes mellitus Diseases of the blood and blood forming organs Diseases of Lhe nervous system Diseases of the circulatory system Diseases of the respiratory system Diseases of the digestive system Diseases of the genitourinary system Senility and ill defined conditions. All external causes Other knowh-causes Obs 2,490* 587 9 7 45 9 14 1,311 127 96 27 17 125 18 Exp 2,930.0 632.5 30.5 7.2 42.5 6.9 23.1 1,601.7 ]98.9 129.0 39.2 29.6 i55.7 27.5 SMR 0.8 0.9 0.3 1.0 I.I -~ 1.3 0.6 0.8 0.6 0.7 0.7 0.6 0.8 0.7 (95% C~ ) (0.8-0.9) (0.9-I.0) (0.4-2.0) (0:8-1.4) (0.6--2.5) (0o3-i.0) (0. -0.9i " (o._,-o.8) (0.5-i.0) (0.3-0.9) Includes 102 deaths for whom the death certificate -~nd cause of death were not available. SH ELL-LEBLANC-01108~ Table 7 -Observe<] and expected numbers of deaths an{] SMRs for all causes and all cancer among white men employed at the Wood River Refinery by duration employment and time since starting work All causes Obs Exm SMR Duration of employment (yrs) : 0.5-4 5-14 15~24 >25 612 580 744 938 711.9 687.8 931.6 1,245.9 0.9 0.8 0.8 0.8 Years since starting work: <i0 10-19 20-29 30.-39 >4O ]15 269 618 937 935 208.5 434.3 863.7 1,123.8 946.9 0.6 0.6 0.7 0.8 1.0 All cancer Obs Exm SMR 146 134 151 230 1.43.9 ],36.9 178.4 271.1 1.0 1.0 0.8 0.8 12 62 138 243 206 24.6 72.3 ].76.7 2549 201.8 0.5 0.9 08 0.9 ~..0 SHELL-LEBLANC-O1108~ Table 8 Observed and expected numbers of deaths and SMRs for se.lected cancers among white men employed at the Wood River Refinery Cancer site All sites Buccal cavity and pharynx Esophagus Stomach Large intestine and rectum Liver Pancreas Larynx Lung Skin Bladder Kidney CNS Lymphosarcoma Hodgkin's disease Leukemia Other lymphatic and . hematopoie .t.i c tissue Other and unknown Obs 661 6 13 27 66 6 31 5 209 15 17 18 19 14 9 44 20 142 ExD 728.6 22.5 17.2 38.6 67.1 15.5 39.4 i0.8 235.4 12.1 22.5 17.9 20.5 13.7 7.8 28.2 17.8 141.8 SMR 0.9 0.3 0.8 0.7 1.0 0.4 0o8 0.5 0.9 1.2 0.8 i. 0 0.9 i. 0 1.2 1o6 I.i 1.0 (95% CI ) (0.8-].0) (0ol-0.6) (0.4-1.3) (0.5-1.0) (0.8-1.3) (0. i-0.8) (0.5-1. i) (0.2-1. I) (0.8-1.0) (0.7-2.0) (0.4-1.2) (0.6-1.6) (0.6-1.5) (0.6-1.7) (0.5-2.2) (1.1-2. I) (0.7-1.7) (0.8-I~2) Includes other lymphoma (ICD=202) , multiple myeloma (ICD:203) and polycythemia vera (ICD=208) . SHELL-LEBLANC-O1108, Table 9 Observed and expected numbers of deaths and SMRs for selected cancers among white men who worked for i0 or more years at the Wood River Refinery Cancer site All sites Buccal cavity and pharynx Esophagus Stomach Large intestine and rectum Liver Pancreas Larynx Lung Skin Prostate .Bladder Kidney CNS Lymphosaxcoma Hodgkin's disease Leukemia Other lymphatic and hematopoietic tissue Other and unknown Obs 436 4 '7 15 60 4 25 2 130 9 38 12 ii 12 9 5 33 13 47 ExD- 505.1 15.4 12.0 27.9 64.0 ii.I 27.7 7.6 160.9 7.6 40.0 16.8 12.2 12.5 9.1 4.3 18..9 12.2 44.1 SMR 0.9 0.3 0.6 0.5 0.9 0.4 0.9 0.3 0.8 i~2 1.0 0 ..7 0.9 1.0 1.2 1.7 I.i I.i (95% CI ) (0.8-1.0) (0.I-0.7) (0.3-0.9) (0.7-] .2) (0.1-0.9) (0.6-1.3) (0o0-1.0) .- (0.7-i.0) (0.5-2.3) (0.7-1.3) (0.4-1.3) (0.5-1.6) (0.5-io 7) (0.5-1.9) (0.4-:2.7) (1.2-2.4) (0.6-1.8) (0.8-1.4) -43- SH ELL-LEBLANC-01108( Table i0 Observed and expected number of deaths and SMRs for selected cancers among white men. employed at the Wood River Refinery, 20 years after first hire Cancer site All sites Buccal cavity and .pharynx Esophagus Stomach Large intestine and rectum Liver Pancreas Larynx Lung .......... Skin Prostate Bladder Kidney CNS Lymphosarcoma Hodgkin's disease Leukemia Other lymphatic and hematopoietic tissue Other and unknown Obs 587 5 13 22 85 6 29 4 193 ].2 43 16 13 17 I0 7 34 19 59 ExD 632.5 19.2 15.2 31.5 75.4 13. i 34.7 9o 5 211.6 9.3 49.5 20.5 15.4 15.0 ii.0 4.3 22.7 16.1 55.8 SMR 0.9 0.3 0.9 0.7 I.I " 0.5 0.8 0.4 0.9 1.3 0.-9 0.8 0.9 1.1 0.9 1.6 1.5 1.2 ].I (95% CI ) (0.9-1o0) (0.].-0.6) (0.5-1.5) (0.4-I.I) (0.9-1.3) (0.2-1.0) (0.6-1.2) (0. I-1. I) (0.8-].1) (0.7-2.3) (0.6-1.2) (0.5-]. 3) (0.5-1.5) (0.7-1.8) (0.4"1.7) (0.7-3.3) (I.0-2.]) (0..7-.i.9) (0.8-1.3) -44- SHELL-LEBLANC-01108( Table ii Observed and expected numbers" of deaths and SMRs for selected causes amonq white men employed at the Wood River Refinery according to payroll classification All causes All cancer Infective and parasitic Diseases Obs Ever hourly Ex,p $~,iR 2,589 3,126.8 0.8 ~85 636.3 0.9 12 53.1 0,2 (95% CI ) (0.8-0.9) (0.8-I,01 (0.1-0.4) Obs 263 68 2 Always Staff ED SMR !95% CI) 420.2 0.6 (0.6-0.7) 87.6 0.8 (0.6-1.0) 6.5 0.3 (0.0-I.I) Benign neoplasms 7 8,4 0.8 (0.3-1.71 1 1.2 0.8 (0.0-4.5} Diabetes me!!itus 44 44.5 1.0 (0.7-1.3 4 5~8 0.7 (0.2-1.8) Diseases of the blood and blood formng organs 9 8.2 (o.5-2.1 0 1.0 0.0 (0.0-3.5 Diseases of nervous system 1'3 27.3 0.5 (0.3-0.8 4.2 0.2 (0.0-1.3 Diseases of the circulatory system i~316 1,638.8 0.8 I0,8-0.9) 117 201,8 0.6 I0,5-0.7] Diseases of the respiratory system 129 199.1 0.6 (0.5-0.8) 6 -24.2 0,2 (0.1-0.5} Diseases of the digestive system 93 146.1 0.6 .,(0.5-0.8] !5 21.9 0.7 (0.4-!.i) Diseases of the genito-urinary system Seni]~, and ill defined condltions I m F- All eterna" causes ro Other known causes 25 ~7.7 19 30.8 199 249.0 14 27.9 0.5 0.6 0.8 0.5 (0.3-0.8) (0.4-1.0 3 5.5 0.5 [0.i-1.6) 0 4.7 0..0 (0.0-0.8) 26 49,8 0.5 3 4.7 0.6 (0.I-1.9) Table 12 Observed and expected numbers of de@ths and SMRs for selected cancers amonq white men employed at the Wood River Refinery according to payroll classification Cancer site Ever hourly Always Staff Obs . Exp SMR (95% CI ) Obs ExD SMR (95% CI l All sites ' 585 636,5 0.9 (0.8-i.0) 68 87.6 0.8 (0.6-i.0} Buccal cavity and pharynx Esophagus 5 19.6 0.3 (0.i-0.6) 13 ,-5.0 0.~ (0.5-1.5) 1 2.7 0.{ (0.0-2.0) 0 2.1 0.0 (0.0-1.8) Stomach 22 34.4 0~6 (0.4-I.0) 4 4.0 1.0 (0.3-2.5) Large intestine and rectum 84 79.1 i.i (0.9-1.3 8 I0.i 0.8 (0.3-1.6) Liver Pancreas 6 13.7 0.4 (0.2-1.0 26 34.6 0 8 (0.5-1.1 0 1.7 0.0 (0.0-2,2) 4 4.6 0.9 (0.2-2.2) Larynx 5 9.5 O5 (0.2-1.2 0 1.3 0.0 (0.0-2.91 Lung 188 204.1 0 9 (0.8-1.1) 19 29.7 0.6 (0.4-1,01 Skin Prostate Bladder 13 10.3 1 3 (0.7-2.2] 41 46.4 0 9 (0.6-1.2) 17 20.1 0,8 (0.5-1.4) 2 1.8 i.i (0.i-4.1) 3 4.9 0.6 (0.1-1.8) 0 2.3 0.0 (0.0-!.6] Kidney 16 15.6 !.0 (0.6-!.7) 2 2.2 0.9 (0.1-3.2) CNS 14 17.4 0.8 {0.4-!.4) 4 3.0 1.4 (0.4-3.5) Lymphosarcoma 13 11.9 i.i (0,6-1.9) i 1.8 0.6 (0.0-3.2) T m Hodgkin's disease 8 6,5 1.2 (0.5-2.4) 1 1.2 0.9 (0.0-4,8) P P Leukemia 35 24.5 1.4 (1.0-1.9) 9 3.5 2.6 (1.2-4.9) m Other lymphatic and 19 15.4 1,2 (0o7-1.9) 1 2.3 0.4 (0.0-2.4] ,o hematopoietlc tissue Other and unknown 60 1.0 (0.8-1.3) 7 8.6 0.8 (0,3-1.7) ~ CNS. centz-~! nervo'as Table 13 Observed numbers of leukemia deaths and SMRs among white m6n employed at the Wood River Refinery, according to selected characteristics Period of death: 194~-1949 1950-1959 1960-1969 1970-1979 >1980 Period of hire: <1940 1940-1949 ->- 1950 Duration of.employment (yrs) : 0.5-9 10-19 >20 Years since starting work <I0 10-19 20-29 30-39 >4O Obs l I0 I] 16 6 24 15 5 SMR 0,8 2.6 1.6 1.5 I.I 1.7 14 1.6 I] 1.2 i0 1.8 23 1.7 1 0,5 9 2.5 I0 1.6 ]6 1.8 8 I.i 95% CI 0.0-4.3 1.3-4.9 0.8-2.9 0.9-2.5 0.4-2.3 1.1-2.5 08-23 0.5-3.8 0.9-3.4 ]ol.-2.5 0.0-2.8 1.2-4.8 0.7-2.9 1.1-3.0 0.5-2.1 -47- SHELL-LEBLANC-01108~ Table 14 Observed numbers of leukemia deaths and SMRs among white men employed at the Wood River Refinery, according to period of hire and period of death Period of hire <1940 >1940 Total : Period of death 1940-1969 1970-1984 Obs SMR Obs SMR 17 2.5 7 0.9 5 1.0 15 1.7 22 1.8 22 1.4 Total Obs SMR 24 1.7 20 ] .4 44 1.6 -48- SH ELL-LEBLANC-O 1109( Table 15 Observed and expected numbers of leukemia deaths by ceil type and decade of death among white male employees of the Wood River Refinery Decade of death 1950-1959 1960-1969 ]970-1994 Cell type All lymphocytic All myelocytic Monocytic Other and unspecified All lymphocytic All myelocytic Monocytic Other and unspecified Lymphocytic Acute Chronic Other Myelocytic Acute Chronic 0thez Monocytic Other and unspecified Obs Ex~ 5 1.0 0 1.8 1 0~23 4 0.75 6 ' 2.1 2 3.0 3 0.4] 0 1.4 5 4.5 1 0.70 4 3.1 0 0.69 13 7.6 12 5.0 1 2.1 0 0.44 2 0.63 2 3.3 SM R 5.1 0.0 4.4 5.3 2.8 0.7 7.3 0.0 i.i 1.4 1.3 O.0 1.7. 2.4 0.5 0.0 3.2 0.6 One death which occurred in 1947 is not included. The expected numbers are based on US Vital Statistics, 1968 (16), for deaths that occurred in the 1950's and 1960's and on SEER data (15) for deaths that occurred in 1970-1984. SHELL-LEBLANC-01109' Table Observed numbers of deaths and SMRs for diseases of the blood and blood forming organs among white male hourly employees at the Wood River Refinery, according to selected Characteristics Period of hire: <1940 1940-1949 >1950 Duration of employment: 0.5-9 10-19 >20 Years since starting work: <20 20-29 >30 Obs 7 2 0 1 1 7 0 1 8 SMR (95% Cl ) 1.5 (0.6-3.0) 0.7 (0.8-2.4) 0.0 (0.0-10) 0.4 0.6 (00-3.3) 1.7 (0.7--3.6) 0.0 (0.0-2.3) 0.6 (0.0--3.1) 1.8 (0.8-3.6) -50- SH ELL-LEBLANC-01109: Table I 7 Observed and expected number of deaths and SMRs for selected causes among 1,024 white women employed at the Wood River Refinery All causes All cancers Diseases of nervous systems Diseases of circulatory system hespiratory diseases Diseases of digestive system Diseases of genito ............ urinary system Senility and ill-defined conditions All external causes Other known causes Obs 45 19 1 5 1 1 ~ 1 3 2 E~p 71.2 23.7 1.2 23.8 3.3 4.] 1.3 0.8 6.7 1.2 SMR (95% Cl) 0.6 (0.5-0.9) 0.8 (0.5-1.3) 0.9 (0.0-4.8) 0.2 (0.]-0.5) 0.3 (0.0-1.7) 0.3 (0.0-1.4) o.8 (o.o-4.~) ] .3 (0.0-7.1) 0.5 (0.1-]..3) 1.7 (0.2-6.1) Includes 10 deaths for whom the death certificate and cause of death were not available. -51- SH ELL-LEBLANC-01109~ Table 18 Observed and expected number of deaths and SMRs for selected cancers among 1,024 white women employed at the Wood River Refinery Cancer site All cancer Digestive system Lung Breast Lymphosarcoma Hodgkin's disease Leukemia Other lymphatic and hematopoietic tissue Other and unknown sites Obs 19 0 2 7 0 1 1 2 6 ExD 23.7 4.8 3.1 5.8 0.4 0.3 0.8 0.5 8.0 SMR (95-% CI ) 0.8 (0.5-]..3) 0.0 (0.0-0.8) 0.6 (0.I-2.3) 1.2 0.0 (0.0-8.9) 3.6 (0.]-20 ] 1.2 (0.0-6.9) 3.8 (0.4-].4 ) 0.8 0.3-1.6) -52- SHELL-LEBLANC-01109, Table 19 Observed and expected number o~ deaths and SMRs for selected causes among 2]0 black men employed at the Wood River Refinery Cause o,_ death All causes All cancers Infective and parasitic diseases Diseases of circulatory system All external causes Obs 6 0 i 1 Exp 9.3 0.8 0.i 1.8 4.3 SMR 0.7 0.0 7.2 0.6 0.9 (95% Cl ) (0.2-1.4) (O.O-47) (0. ~-40 ) (0.0-3.2] SHELL.LEBLANC-01109( APPENDIX A Classification of Production-related (PR) and Nonproduction-related (NPR) Jobs -54- SHELL-LEBLANC-01109( Classification and nonproduction-rela~ed (NPR) jobs Job Title Product ion-rel.a l:c,d* Accountant Accounting Assistant Adminis trativ e Sup t. Adv, Products Mgr. Alkylat ion Mgr. Analyst Assistant Asst. Mgr. Buyer Carpenter and Rigging Cat Cracking Ass't. Mgr. Chemist Chemist Co-O? Student Clerk C].erk Summers Clerk-Sliippln~ Console Operator Craft ~oreman Data Proc. Assistant Data Proc. Chief Designer Draftsman Draftsman ..Duplleating Operator Editor Employee Relations Represen. Employee Supervisor Eng. Office Mgr. Engo Office Ass't. I.Igr. En~inee~ Engineer Co-Op Student N N N N Y N N Y N Y Y Y Y N N N N Y N .N N N N N N N? N N N N -.55- SH ELL-LEBLANC-O 1109: Job Title Engineer ~leld Ass't. Mgr. Engineer ~oreman Engineer Se~Jice Mgr. Engineer Student Trainee E~ginee r Supe~isor Envir. Conservation Mgr. Estimator File and Mail Clerk P r oduc t io~-rela ted*Y Y N N N N N N Bile Clerk Foreman and Safety Inspector Foreman Supe~isor Fuels Mg~. N Y Y N Gas Asst. Mgr. Gate Captain Gatema~ Glass B,lower Y N N Y Group Leader Ind. Rel. Asst. Mgr~ Info. Se~ices Assistant. N N Inspector Y Inspector Inst. Y Keypunch Operntor Lab A-Co-Op Student N Y Lab %echnician Lab Assistant Lab Technician Y Lab Technician - Ref. Lab Y Laboratory Assistant .. Laborer Laborer Operator Y Y Y Mall Carrier N Manager Compounding Y Manager Construction N ~4anager ~.mploy~ Relations N Manazer I r~as. N -56- SH ELL-LEBLANC-01109~ Job Title Manager Utili[ies Material Analyst Medical Directo-r Mgr. Aromatics Mgr. Cat. Crkg~ Mgro Compounding Mgr. Construction Mgr. Econo & Schedule Mgr. Emp. Relations Mgr. Eng. O~fice Mgr. Fuels Mgr. Ind. Rel. Ass't. Mgr. Indl. Prod. Mgr. Lube Oils Ass't. Mgr. Pipefitter & I R Mgr. Product Planning Mgr. Purchasing Ass't. Mgr. Refinery Lab Mzr. Safety Nurse Office Assista~ Office Emp. Head Office Secretary Office Supervisor Oil Accountlng Supervisor Oil Mvmt. Supervisor Operations Foreman Operations Supervisor Operator Assistant Operator Foreman Operator Machine Operator' Manager Operator Superintendent ()~):~.r;~ [or Supe r~ l.so r Ope[ntor Telephone Production- re la ted* N N N Y Y Y N N N N N N N N N N N N N N N N N N Y Y Y Y Y Y N -57- SHELL-LEBLANC-01109~ Job Title Production-related* Opt. Duplicating O~s. Supervisor N Y Patrolman Personnel Asst. Mgr. N N Pipefltter Y Pip.e fitter Foreman Y Planning Coordinator N Process Superintendent N Process Manage[ Y Process Superintendent N Programmer N Purchasing Assistant Purchasing Mgr. Ref. Auditor N Refinery Lab Mgr. N Refinery Mgr. N Refinery S~p erint end,-~nt N Refinery Supe[visor Re f in~~y Su>erintendent N Res. ~echnician Resenrch Cl~emis t Y Y Research Director N Research Engineer Research Lab Research Physicist, Asst. Chief Y Research Services Mgr. Research Technician Safety Inspector ..Safety Mgr. N Y Y N Safe~y Supervisor Sched. Assistant Sec.re~ary Security Super'~Isor N N N Shift Foreman Y Steno-Clerk N SHELL-LEBLANC-011 lot Job Title Steno-Seeretary Stenographer Stores Mgr. Supe~-isor Supervisor Materi~|. Sup.ervlsor Plant Protection Te chni ci an Technician Advisor Technician Assoc. Technician Mgr, Technologist Tool.. Expeditor Training; Rep. Treasury M'~r. Treasury Super~isor Typist Utilities Ass't. Mgro Ya rdma~ Production-related* N N N ? N N N N N N N N N N N N N y * A production-related job was one that involved routine or regular (at least 20% of the time) potential exposure to chemical substances in the refinery. For job titles with questionable exposure, decisibn was made or~ the ba,~is of the department in which the worker was employed. v59- SHELL-LEBLANC-O 1110' APPENDIX B Record Abstract Form -6;0- SHELL-LEBLANC-O1110: Type of record 2. Na~e [|ast, first, 3. Soclal security number Wood River number WOOD RIVER RECORD ABSTRACT L)a~e Box No. 6. Race (W=l, 8=2, Oth=3, U=9) 7. Date of BIRIII (U=9) Date of FIRST E~-IPLOYI4ENT 9. EQUATED EI4PLOYHEI~T date %0. Date of most tecent IERHI~IATICII (U=9) 11. Emp]oyment status (term=l, teL=2, ]aId-off=3,deceased~4,transf=8,U=9) Last payrol] cIassification (Ilourly=l~ Staff=2, U=9) Ever hourly (Y=I, N=0, U=9) 14. Last date HOURLY (Never hourly=8, U=9) 15. Type of staff position (Exp=1~ Not exp=0~ U--9~ Never staff=B) 16. Date of DEATH (Nane=O~ U=9) 17. State of death 18. Harita| status (H=I, 5=2, W:3, D=~, U=9) -61- SH ELL-LEBLANC-O 1110: Name of most recent spouse 20. Last address 21. Last te;ephone number 22. Other tracing information (veteran status, new employment~ physician, etc.) -62- SHELL-LEBLANC-OI I I O, APPENDIX C Documentation of Merging Procedures Used to Combine Computer Files -63- SHELL-LEBLANC-O 1110( Mer@inq o[ Abstract and Previous RFS* Records Variable Record type Name Wood River No. Source of the merqed record Aostract Abstract Previous RFS Race Birth date Hire date t Equated employment date Previous RFS Previous RFS Previous RFS Abstract Abstract Termination date Previous RFS Employment status "Termination reason" in Previous RFS Last payroll...classification Abstrac~ Last date hourly Type of staff position Date of death Abstract Abstract Pre~ious RFS * Previous RFS, previous retrospective follow-up study of the Wood River Refinery workers (i). - 6 4- SHELL-LEBLANC-01110( APPENDIX D Vital Status Ascertainment -65- SHELL-LEBLANC-01110~ VITAL STATUS FOI.LOW-UP Social Security Administration Vital Status Follo~.;-um The SSA has a Summary Earnings Record (SER) on file for each individual to whom a SSN has been issued since ]936 The SER contains anm]al taxable earnings as well as demo- graphic data. It is to this and other SSA record systems t~lat death information is posted when it is reported to SSA. Reporting is motivated by the existence of survivors ' benefits and lump sum death payments It has been estimated that reporting.of deaths to the SER is about 91% complete. Since 1962, the SSA has also maintained the ~aster Beneficiary Records file (MBR). The MBR contains a record for each person entitled to receive retirement, survivors or disability insurance. Mortality coverage is higher on the MBR than cn the SER. It is estimated that less than 2% of decedents who ]lad received benefits during their life had no death indication on the MBR. ,.., The numi.denh file is a numerical file by SSN of the information contained on the SSA application form plus claim-related information. Information on the SSA applic.ation form includes name, SSN, birth date, mother's maiden name and father's surname. -66- SH ELL-LEBLANC-O 1110~ deceased, the NDI provides the name of the state in which the death occurred, the date of death and the co[responding death certificate number. A copy of the death certificate can then be purchased directly from the state office. The matching of study records with the NDI uses the followin~ in formation: an individual's first name, middle initial, last name, father's surname (for married women) , SSN, birth date, sex and race. The criteria for matching include at least one of the following seven combinations of data items: I. SSN and first name 2. SSN and last name 3. SSN and father's surname 4. If-the subject is a woman~ ssN, last name (submitted) and father's surname (NDI record) 5. Month and exact year of birth, first and last name 6. Month and exact year of birth, first name, father's surname 7. If ~he subject is a woman: month and exact year of birth, first name, last name (submitted) and father's surname (NDI record) 8. Month and exact year of birth, first and middle initials, last name. 9. Month and + i year of birth, first and middle initials, last name -67- SHELL-LEBLANC-01110! By submitting the SSN and the first six letters of the surname to the Records Use and Services Branch of the SSA, it is possible to identify an individualas belonging o one of the following categories: I. Death - State where death claim was filed 2. Death - State of last residence 3. Death - Funeral Director's notice 4. Assumed living 5. Status unknown 6. SSN Unmatched 7. Invalid SSN .- First, the computerized search is done on the numident file to identify cases belonging to the "Death - State where death claim was filed" and the "Unmatched" categories. Then it continues with the MBR search identifying cases belonging in the "Death - State of Last Residence" and some of the "Assumed Living". The last step is the SER search to identify cases belonging in the "Death- Funeral Director's Notice", "Assumed Living", "Status Unknown" , and "Invalid SSN ' s" The National Death Index (NDI) The NDI is a computerized index to the death certificates filed in each state vital records office and is capable of identi lying each person dying in the United States, starting in 1979, using a standard set of information. When a name submitted is identif].ed by the NDI to be -68- SHELL-LEBLANC-O1111( i0. Month and + lyear of birth, first and last name II. Month and day of birth, first and last name 12. Month and day of birth, first and middle initials, last name. Upon receiving the NDI search results, an individual judgment is then made on each possible record match to verify a true match. The results of the NDI search .[or the present study are su~,arized in the Figure. -69- SHELL-LEBLANC-01111' reiiminary cohort = 13, U n known vital .st utus N-- 10,272 I IA~_" C iv ~': workers and active perts i Ol'~e FS N = I ,93g N = 2,479 Assumed livlngI Dead~ located Dea th cer#i [ D.:nte$ not located N = 226 before t979 N = 178 Deceased I,t~ or afte[ 1979 N = 48 S t~'~t~s I,I = 889 I Deaths: confirmed, N=44 not conf Irmcd N=4 Additional deaths identified N = 28 Assumed i iv I nZ N = 29 Four_ Vital status ascertaicment -70- 1 Unknown status N = ,],o SHELL-LEBLANC-O 1111; APPENDIX E Nosologic Coding for Leukemia -71- SHELL-LEBLANC-01111: Appendix Nosologic Coding_ of Leukemia Nosologic coding of causes of death used in this study was based on the Eighth Revision of the International Classification of Diseases, (ICD) (25) which was adopted in 1968. In the Sixth Revision of the ICD, the four-di~]it codes allowed for a distinction to be made between lymphatic and mye]oid leukemia and between acute and chronic forms of the disease. In the Seventh Revision, all acute leukemias were grouped together, and it was thus impossible to distinguish among acute leukemia cell types (table Eli. According to the Eighth Revision, however, lymphocytic, mye locytic and monocytic leukemias are three separate categories based on three-digit codes and each category is subdivided into. acute, chronic and unspecified (table This is basically the same classification system used in the Ninth Revision, except for the code for other and unspecified cell types of leukemia. In the Ninth Revision, code 208 (leukemia of unspecified cell type) incl~des leukemia, not otherwise specified (NOS), acute leukemia NOS, chronic leukemia NOS, stem cell leukemia, and blast cell leukemia. In the Eighth Revision, these leukemia cell types were included under code 207 together with subleukemic, myeloproliferative, and megakaryocytic leukemia, etc. -72- SHELL-LEBLANC-01111, Table E] Classification Of Selected Leukemia Cell Types In The 6th-gTH Revisions Of The International Classification of Diseases (23-26) Leukemia cell type 6th Revision (1948-1958) Lymphatic -acute -chronic -unspecified 204.0 Myeloid -acute -chronic -unspecified 204.1 ~ Monocytic -acute -chronic -unspecified 204.2 Acute leukemia, unspecified cell type 204.3 All acute leukemia Other and unspecified leukemia 204.4 Other specified leukemia Leukemia of unspecified cell type ICD Codes 7th Revision (1959-1967) 8th Revision 9th Revision (1968-1978) (1979-Present. 204.0 2040 204.0 204.1 204.9 204.0 204.1 204.9 204.1 204.1 204.2 205.0 205.1 205.9 206.0 206.1 206.9 205.0 205.1 205.9 206.0 206ol 20.6.9 204.3 204.9 207 207 2O8 -73- SH ELL-LEBLANC-O 1111{ : Table E2 Classification Of Leukemia In The 8th Revision, ICDA i25) Code 204 LYMPHATIC LEUKEMIA Includes: aleukemia, legkemia (aleukemic) : .lymphatic, lymphoblasticr lymphogenous, lymphoid, splenolymphatic 204.0 Acute 204.1 Chronic .... 204.9 Unspecified Code 205 MYELOID LEUKEMIA Ihclude~: aleukemia, leukemia (aleukemic) : granulocytiC, mye!oblastic, myelocytic, myel0genous, myelosclerotic, myelosis (aleukemic), monomyelocytic, eosinophilic, basophilic, mixed cell, splenomyelogenous 205.0 Acute 205.1 Chronic 205.9 Unspecified -74- SHELL-LEBLANC-01111~ Code 206 Code 207 MONOCYTIC LEUKEMIA Includes: aleukemia, leukemia (aleukemic) : histiocytic, monoblastic, monocytic, ~eticular ceil. 206.0 Acute 206.1 Chronic 206.9 Unspecified OTHER AND UNSPECIFIED LEUKEMIA Includes : aleukemia, leukemia (aleukemic} : NOS, giant ce].l, megakaryocytic, stem ceil, plasma cell, plasmacytic, blast cell, hemoblastic, mast cell, Rieder cell 207.0 Leukemia, acute 207.1 Leukemia, chronic 207.2 Acute erythremia 207.9 Leukemia (unspecified) , subleukemic, myelopro~if~rative, eosinophilic -75- SHELL-LEBLANC-O 1111", APPENDIX F Mortality Patterns Among Always Staff Employees in PR and NPR Jobs -76- SH ELL-LEBLANC-O 1111~ Table F1 Observed and expected numbers of deat~.s and SMRs for selected causes among white men e,'npoyed at t~.e Wood River Refinery according to payroll c!asszfication All causes All cancer Infectlve and perasltlc Diseases Zver hourly Obs Ex~ St4R (95% CI ) 2,588 3,126.4 0.8 (0.8-0.9] 585 636.3 0,9 (0.8-1.0) 12 53.1 0.2 (0.1-0.4) A lwa~,s Staff P rod uc t i~Oi~i related ~on-pro~uct ion related Obs ~.~ ,St.:R (95% CI ! Obs S;.~R 195% CI ) 127 228.7 0.~6 (0.5-0.71 136 i90.0 0.7 (0.6-0.9) 32 48.3 0,7 (0.5-0.9) 37 38:9 1.0 1 3~5 0,3 (0.0-1.6) 1 3.0 0.3 (0.0-1,8) B~nlgn neoplasms 7 8.4 0.8 (0.3-1.7) ~ 0.7 1,4 (0.0-8.0) Diabetes me!litus 44 44.5 1.0 (0.7-1.3) 2 3.2 0,6 (0,1-2,3) Diseases of the blood i ann blood forming organs 9 8.2 i.I (0.5-2.1) 0 0.6 0.0 ~0.0-6.6) 0,5 0,0 (0,0-6.8) 2.7 0.8 (0,1-2.7) 0.5 0.0 (0.0-7.6) Diseases of the nervous system 13 27.3 0,5 (0.3-o.s) ~ 2.4 0.~ (0.~-2.~ 0 !.B 0.0 (0.0-2.0) Diseases of the circulatory syste~ 1,316 1,638.5 0.8 (0.8-0.9} 50 106.7 0.5 (0.4-0.6) 67 94.,I.0,7 Diseases of the resplratory system i~9 199.1 0.7 (0.5-0.8} 2 12.7 0,2 (0.0-0.6~ 4 11.5 0,4 (0.I-0.9) Diseases of the digestlve system 93 146.1 0.6 {0.5-0.~) 7 12,3 ~.6 {0.2-1.2) 8 Diseases of the genlto-urlnary system 25 47,7 0,5 (0,3-o.sl 0 2,~ 0,0 Io,o-~.3) 3 9.5 0.9 {0.4-1,7) 2.7 I.I (0.2-3.3) Senility and ill ~efined conditmons 19 30.8 0~6 0,4-1,0) 0 2.7 0.0 (0.0-1,41 0 2.0 0.0 All et~rna.[ causes 199 249.0 0,8 0.7-0.9) 18 29.6 0.6 (0.4-1.0) 7 20.0 0.4 (0,1-0.7) r- Table. F2 Observed and expected numbers of deaths and SS!Rs for selected cancers among white men employed at the Wood River Refinery according to joD class~ficaczon Cancer Slte All cancer Obs. 584 Ever hourly ~xo SMR (95% CI ~ 636.5 0.9 (0.8-1.0) Alwe~,s staff Produc~ioD related OOs ~xp SMR (95% CI ~ 32 48.3 0.7 (0.5-0.9) Non-production related Obs St, JR (95% CZ ) 37 38.9 1.0 (3.~-1.3) Bucca! cavlty and pharynx Dlgest!ve system 5 19.6 0.3 (0.1-0.6) 0 ! ,5 0.0 (0.0-2,4) ! 1.2 0,8 (0.0-4.6) 157 181.4 0.9 :(0~7-I,0) 8 12.4 Q.6 (0.3-1.3) 9 10.6 0.9 (0,4-1.6) Esophagus Stomach 13 15.0 0.9 (0.5-1.5) 22 34.4 0.6 (0.4-I.0) 0 2 i.i 0.0 (0,0-3.21 2,1 1.0 (0.1-3.4) 0 2 0.9 0.0 (0.0-4,1) 1.9 i.! (0.i-3.8) Large intestine and reczum 84 79.1 i.i (0.9-1.3) 3 5.4 0.6 (0.i-!.4) 5 4.6 i.i (0,3-2,3) Liver .6 13.7 0.4 (0.2-i.0) Pancreas 26 34.5 0.8 (o. S-] .]) Larynx 5 9.5 0.5 (0,2-1.2) Lung 188 204 1 0.9 (0.8-I.I) Skln 13 10.3 1.3 (0.7-2.2) Prostate 41 46.4 0,9 (0,6-i.2) Bladder 17 20.1 0.9 (0.5-1.4) Kidney 16 15.6 1.0 (0.6-t.7) (CNS) 14 17,4 0.8 (0.4-1.4} Lymphosarccma 13 11.9 i.i (0.6-I,9) Hodgkln's disease Leukemza Other lymphatic and hematopolecc ~ssue 8 35 19 6,5 24,5 !5.4 1.2 (0.5-2.4) 1,4 (1.0-1.91 1.2 (0.7-1,9~ 0 3 0 6 2 1 0 1 2 0 L 4 0.9 2.5 0.7 16.7 !0 24 2 13 l7 10 07 i9 13 0,0 1.2 0.0 0.4 2.0 0,4 0.0 0.8 ',1.2" 0.0 1.5 2,! 0.8 (0.0-4,1 (0.3-3.5 (0.0-5.2 01-0.8 02-7.1) 00-2.3) 00-3.2) 00-4.4) 01-4.2) 00-3.7) 00-8.i) 06-5.3) 00-4.4) 0 1 0 13 0 2 0 1 2 ! 0 5 0 0.8 2.1 0,6 12.9 2.5 i.1 1,0 1.2 0.5 1.6 i.o 0.0 (0.0-4.7) 0.5 (0 0-2.7} 0.0 (6 0-6.5) 1.0 10 5-1.7) 0,0 (0 0-5.0~ 0.8 (0 I-2.91 0.0 (0 0-3.3} 1.0 (0 0-5.71 1.6 (0 2-5.9) 1.3 (0 0-7.3) 0.0 (0 0-7.7) 3.2 (~.0-7.5} 0,0 (0.0-3,7) FCNS, central nervous system, APPENDIX G Wood River Refinery Employees Who Worked for Fewer than Six Months -79- SH ELL-LEBLANC-01112' Mortality patterns.. amgn.q white men: There were 2,042 white men who had worked for fewer than sx months at the refinery. Vital status ascertainment was complete ~.or 1,883 (92%) of these men. Table G] displays the observed and expected numbers of deaths from selected causes among white men employed for fewer than six months. There were non-statistically significant excesses of deaths from all causes (507/486.9; SMR=I.0) , all (113/102.4; SMR=I.I) , diabetes mellitus (10/6.7; SMR=!.5) , diseases of the respiratory system (34127.9; SMR=I.2) r and benign neoplas{ns (4/1.5; SMR:2.7). These short-term workers also experienced a statistically significant excess of lung cancer (50/35.0; S~R.=].4, i.I-].9) and slight excesses of cancers of several other organs inc]~ding the liver (3/2.0; SMR=I.5), bladder (5/2.6; SMR:I.9) and leukemia (?/4.0; SMR=I.7) (table G2) ~ None of the latterexcesses was statistically significant. Of the seven men who died of leukemia, all were hourly, and all but one were hired in the 1940s (table G3) Three of the seven had acute myelocytic leukemia, two had cbzonic myelocytic leukemia, one had acute lymphocytic leukemia and one had unspecified lymphocytic leukemia. Two of these men (cases 53 and 55) actually worked for less than one day at the refinery. Mortalit~z patterns amonq white women: There were 286 white women who had worked for fewer -80- SHELL-LEBLANC-01112: than six months at the refinery. Vital status was determined for 230 (80%) of them. There were 18 observed compdred with 27.9 expected deaths, with SMRs of 0.6 for all caL~ses and 0.7 (6/8.7) for all cancers. No striking excesses of specific diseases were observed among these women, and there were no deaths from leukemia. -81- SHELL-LEBLANC-01112: Table G1 Observed and expected numbers of deaths and SMRs for selected causes among 2,042 white men employed for fewer than 6 months at the Wood River Refinery Cause of death All causes All cancers Infective and parasitic diseases Benign neoplasms Diabetes mellitus Diseases of the blood and blood forming organs Diseases of the nervous system Diseases of the circulatory system Diseases of the respiratory system Diseases of the digestive system Diseases of the genitourinary system Senility and ill defined conditions All external causes Other known causes Obs 507 i19 6 4 I0 0 1 222 34 20 7 5 49 2 Exp 486.9 ]02.4 7.6 1.5 6.7 1.2 4.8 235.5 27.9 25.4 6.2 5.4 55.6 5.4 SMR 1.0 ].I O. 8 2.7 1.5 0.0 0.2 0.9 1.2 0.8 I.i 0.9 0.9 04 (95% CI (i. 0-] .i) (0.9-1.3) (0.3-1 .'7) (0.7.-7.0) (0.7-2.7) (0.0-3.1) (0.8-1.1) (0.8-1. 7) (0.5-1.2] '(0.5-2.3) (0.3-2.2) (0.7-1.2) (0.0-1.0) Includes 33 deaths for whom the death certificate and cause of. death were not available. -82- SHELL-LEBLANC-01112, Table G2 Observed and expected mlmbers of deaths and SMRs for selected cancers among 2,042 white men employed for fewer than 6 months at the Wood River Refinery Cancer site Obs All sites Buccal cancer & pharynx 113 1 Esophagus 2 Stomach 3 Large intestine and rectum 12 Liver 3 Pancreas 4 Larynx 2 Lung 50 Skin 1 Bladder 5 Kidney 1 CNS 1 Lymphosarcoma 3 Hodgkin's disease 1 Leukemia 7 Other lymphatic and hematopoietic [issue 1 Other and unknown sites 16 Exp i02~4 3.3 2.4 4.7 11.8 2.0 5.4 1.5 35.0 2.0 2.6 2.6 3.5 2.1 1.4 4.0 2.6 15.5 SMR I.i 0.3 0.8 0.6 1.0 i.~ 0.7 1.3 1.4 0.5 1.9 0.4 0.3 1.4 0.7 1.7 0.4 1.0 (95% CI (0.1-2.0) (o. 1-.[.. 9] (o.5-3..7) (0.3-4.4) (o.2-1.9) (0.2-4.7) (!.~-]_.9) (o.o-2.8) (o.6-4.4) (o.o-.2.1] (.o.o-1.5) (0.3-4.2) (o.o-4.1.) (0.7-3.6) (0.0-2.1) (0.6-1.7] CNS, central nervous system. -83- SHELL-LEBLANC-01112{ Table G3 Characteristics of leukealia deaths among white men employed for fewer than six months at the Wood River Refinery Case no. 53 52 51 50 49 54 55 Year of Death 1950 1955 1965 1976 1979 1981 1984 Age at Death 31 50 52 79 63 81 65 ICD* Code 2050 205]. 2040 2051 2050 2049 2050 Employment dates 1943-]943 ]943-1943 1951-1952 1945-1945 1942-1942 1944-1945 1942-1942 Job% EH EH EH EH EH EH EH * ICD, Internation.al Classification of Diseases, Adapted for use in the United States, Eighth Revision (25). Job, job classification: EH:ever hourly. -84- SHELL-LEBLANC-01112( APPENDIX H Detailed Listing of Leukemia Deaths Among Wood River Refinery Employees -85..... SHELL-LEBLANC-01112] Characteristics of Wood River Refinery employees with leukemia Case no. Year of Dea th Age at Death ICD* Code Employment dates Jcb~ I. White men, leukemia reported as u~IderJylng cause of death 42 9 1 44 13 16 39 5 .40 41 34 22 35 32 12 J 8 33 24 7 10 6 26 14 25 29 17 19 30 36 27 31 37 2 4 21 23 38 28 20 18 43 Ii 15 1947 1"950 1952 1953 1954 1954 1955 1956 1957 1958 1959 1961 1962 1963 1965 1966 ]966 1966 1967 1968 1968 1969 1973 1974 1974 1974 1975 1975 1926 1976 1977 1977 1977 ~978 1978 3978 19?9 1979 1981 1982 1983 1983 -:1984 1984 39 42 58 46 62 62 42 59 41 65 38 52 66 70 72 71 45 58 43 71 61 69 ' 51 53 62 54 85 54 70 77 75 76 62 70 65 59 74 81 73 64 70 65 71 69 2079 2043 2043 2070 2049 2041 2071 2049 2079 2079 2069 2050 2060 206~ 2041 2041 204~ 2060 2051 2049 2041 2049 2050 204~ 2051 2059 2050 2059 2059 2060 2050 2050 2069 2041 2041 2050 2050 2070 2050 2050 2050 2070 2041 2041 1939-1947 1933-1950 1923-1949 ].935-1947 1927-1952 1938-1950 1939-1955 1922-!952 1939-1957 1934-1953 1947-!954 1941-1961 1923-1958 ]922-]949 1925~1952 1928-1956 1953-]966 1939-1963 1947-1949 ~940-1944 1926-1968 ]9~9-1944 1956-1973 194~-1974 1937-1973 1953-1955 1929-1951 1942-1946 1935-1965 Iq~7-1943 1923-1963 1942-1961 1941-1970 1944-1945 1943-1974 1940-1977 1942-1951 ]976-1957 1942-1967 1953-1977 1951-1975 1939-1955 1942-1975 1945-1946 El! Ell NPR EH EH Eli EH EH EH EH Ell EH EH PR EH NPR ~I~ NPR El| PR EH EH EH NPR EH NPR White men, !eukem~a repo~ted as contributing coafusdeeath# 46 1953 61 2049 1927-1946 EH 45 1973 67 2079 1941-1942 EH 47 1976 51 2079 1947-1975 48 1980 82 2041 1934-1945 57 1968 49 2050 1941-1946 EH 58 1968 59 2079 1936-1947 IIl. ~]ite woman 56 198~ 52 2051 1952-1952 NPR ICD, International Classification of Diseases, Adapted for use in the United States, ~ighth Revision (25). Job, job classiicati(n~: EH = ever hourly; PR = staff, production-related; NP~ = staff, nonproduction-re|.ated. /- Under].ying cause of duath (ICD code] : case ~o. 46 "= [030 (tuberculous meningitis) case no. 45 ~ 5321 (perforated duodenal ulcer) cnse no.. 47 - 5820 (uremia) case no~ 48 = 4123 (Arteriosclerotic heart disease) case no.. 57 = 2090 (chronic myt~l~)fibrosis) case no. 58 = 2090 (myelofibfosis). SHELL-LEBLANC-011121