Document dQ6kb272zax7brJj5BpkZ0NN0

notice Inly " ."-jrial may be pro lei, i,id by copyri/f"' lav; (Title 17 U.&. Code) A Mortality Study of Oil k Refinery Workers Gilles Theriault, M.D., Dr.PH., and Lise Goulet, M.D. Survival status of 1205 men employed for more than five years in a Canadian oil refinery in East Montreal, from 1928 through 1976, was assessed and death certificates were reviewed. Expectednumbers of deaths were estimat ed based upon age- and cause-specific death rates for the Province of Quebec applied to person-years at work. Oil refinery workers showed a standard mortality ratio lower than expected for all causes of death (SMR = 78.43). Three cancers of the brain were found among young people who died less than 20 years since start of exposure. This was statistically higher than expected. Cancers of the digestive system, though not significantly higher than expected, re mained suspect of being associated with work. There is a need to expand this research to other oil refinery workers. ough processing in an oil refinery is carried out in airtight compartments, the presence of hydrocarbons of all kinds and the possibility of worker exposure to these chemicals at the time of sampling and maintenance rais es the possibility that cancers of occupational origin'may appear in this environment. Few epidemiological studies dealing with the health of oil refinery workers have been reported. Mortality rates from genital cancers, lymphomas,1 cancers of the diges tive system and lung cancers11 were described as con spicuous and appeared to be exposure-related, but their excess was neither large nor statistically significant and could have been a chance occurrence. Occupations other , than those encountered in an oil refinery in which workers were exposed to oil derivatives were associated with the appearance of cancer. Benz pyrene was associated with an irfbrease in lung cancer deaths.4 1 * Benzene was suspected of inducing leuke mia.' A high incidence of urinary bladder cancer among workers in the rubber industry led to confirmation of a causal association between exposure to betanaphthylamine and urinary bladder cancer in man.* From Ihe Department ft* Social and Preventive Medicine, Laval Unaerntv, Qurhec. Canada C IK 7P4 This study wa* supported m part by a contract between Shell Oil Canada and the Department of Social and Preventive Medicine, Laval Journal of Occupational Medicine/Vol. 21, No. 5/May 1979 This mortality study was carried out in a Canadian oil refinery plant which has been in operation since 1928 and which has had an average work force of 750rnen over the years. The purpose of the study was to determine if can cer mortality among oil refinery workers was higher than expected. Methods Study Population. -- The study population comprised all men who worked at the oil refinery for more than five years between the start of operations in 1928 and Decem ber 31,1975. Survival Sfatus. -- The survival status was determined as of December 31, 1976.' The company furnished the name, date of birth, date of employment, work termina tion date, last known address, the social insurance and unemployment insurance numbers, when known, as well as a classification of workers according to type of work (operation, maintenance, or office work). A detailed ser vice file shoyving in chronological order the occupations and the departments in which each person worked during his years of employment was made available upon re quest so that references could be made to it when further details were needed. People who no longer were in. the company's employ were traced through several means: telephone calls; in quiries at the Motor Vehicle Bureau, the Quebec Health Insurance Board, the Pension Board, and die Unemploy ment Insurance Commission; and circular letters to fellow workers. Persons whose survival status could not be as certained by a direct personal contact or through their family or.friends, even though a current address could be found in the registries, were considered lost to follow-up. The list of untraced workers was linked to the death cer tificates filed at the Quebec Population Registry to fur ther increase the number of traced people. Death certif icates were consulted and causes of death were recoded according to the 8th Revision of the International Classi fication of Disease. Expected Number of Deaths. -- A chart of person-years worked was calculated* A man entered the chart after five years of work in the oil refinery. He remained under observation to the time of his death or December 31, 367 ucc 053982 Tabid 1. --Distribution of 1205 Oil Rofinory Workers According to Survival Status as o( December 31,1976. Survivil Stilus Alive Deed Cause known Cause unknown Lost to tollow-up Total Milt workers No. % 907 108 104 4 190 1,205 75 9 16 100 1976, whichever came first. For the years 1928 to 1962,, the expected number of deaths was based upon the ageand cause-specific death rates for the Province of Quebec, 1951, while for the years 1963 to 1976, those of 1971 were used. The influence of. time evolved between the start of ex posure and the cancer appearance was studied by divid ing the population into two subgroups: more than 20 years and less than 20 years of elapsed time between the date of employment and the time of death. Expected numbers of deaths were determined and the standard mortality ratios (SMR) were calculated for each sub group. The Bailar-Ederer test for the ratio of a Poisson vari able to its expectation was used to test the differences Table 2.--Distribution of 1205 Oil Refinery Workers According to Survival Status'and Birth Oato. Date of Birth 1880-1884 1885-1889 1890-1894 1895*1899 1900-1904 1905-1909 1910-1914 1915-1919 1920-1924 1925-1929 1930-1934 1935-1939 1940-1944 1945-1949 1950-1954 Total Alive NO. % -- 2 0.2 1 0.1 6 0.7 19 2.1 38 4,2 44 4.9 83 9.2 154 17.0 141 15,5 134 14.8 75 8.3 99 10.9 103 11.4 8 0.9 907 100 Dead No. % 2 1.9 5 46 5 4.6 12 11.1 9 8.3 18 16.17 17' 15 7 13 12.0 14 13.0 6 5.6 5 4.6 1 0.9 1 0.9 ---- ---- 108 400 Lost It Fellow-up He. % ___ ---- 2 1.1 1 0.5 2 1.1 5 2.6 17 8.9 ' 13 6.8 37 19.5 48 25.3 36 18.9 13 6.8 10 5.3 6 3.2 ---- 190 100 Table 3.--Distribution of 1205 OK Refinery Workers According to Survival Statui and Date of Employment. Oateef Employment 1928-1932 1933-1937 1938-1942 1943-1947 1948-1952 1953-1957 1958-1962 1963-1967 1968-1971 Total Alive Ns. % 19 27 46 142 202 135 99 95 142. 907 2.1 30 5.1 15.7 22.3 14.9 10.9 10.5 15.7 100 Dead NO. % 17 15.7 18 16.7 20 18.5 31 28.7 12 11 1 5 4.6 2 1.9 3 2.8 00 108 100 LostloFaltw-Up No. % 3 1.6 8 4.2 18 9.5 26 13.7 59 31.1 36 18.9 20 10.5 13 6.8 7 3.7 190 100 .between the observed and the expected numbers of deaths.10 Results The Study Population. -- Table 1 shows the survival" status of the people of the study population. There were 1205 male employees who worked for more than five years at the refinery. Of these, 190 men were considered lost to follow-up. The studied group comprised 1015 men or 84% of the total study population. Tables 2 and 3 show the distribution of the employees alive, dead or lost to follow-up according to their birth dates and their dates of employment. The distribution of the lost-to-follow-up men according to age and date of employment was similar to that of the regrouped alive and dead men. The lost-to-follow-up group is therefore expected not to generate any bias in the overall result of the study. Mortality Among Oil Refinery Workers. -- There were 108 deaths in the employee population during the 49-year observation period. Ninety-one percent of the deaths (98 men) occurred in the group of employees who were over 40 years of age. The person-years of exposure chart gen erated from the study cohort is shown in Table 4. Table 5 shows the observed and expected numbers of deaths and standard mortality ratios according to the time elapsed since start of exposure. Whereas the overall mortality is lower than expected among the refinery workers (SMR--78.43), the SMR's for specific causes, though sometimes differing from 100, were not statisti cally significant. When mortality is studied according to whetber more than 20 years or less than 20 years have elapsed between Table 4. --Person-Years Worked by Tima Period and Ago (1015 Man who Worked lor Over 5 Years at the 01 Raflntry). _ Time ftriod 1928-32 1933-37 1938-42 1943-47 1948-52 1953-57 1958-62' 1963-67 1968-72 1973-76 Total 20-24 25-29 8 l' Age 30-34 35-39 40-44 45-49 50-54 55-59 60-64 85-89 70-74 75-79 80+ Total 5.0 8.6 14.0 17.7 32 3 15.5 12.1 8f 08 114 2 5.1 6.1 45.7 85.3 67-. 5 96.4 179.3 216.9 243 2 448.9 293.2 537 7 178.5 560 9 162.3 281.4 241.2 338.5 1416.6 2572.1 8.6 71.3 150.9 205.4 421.0 J16.1 715.4 621.7 304.4 3214.8 3.4 36.3 109.2 221.8 281.8 510.5 777.7 725.5 527.9 3194.1 5.6 21.0 59.5 150.8 241.8 286.5 511.5 ' 774.4 581.4 2632 5 3.9 22.1 31.3 83.8 156.8 237.1 263.0 494.8 601.1 1893.9 9.a 27.9 33.8 78.5 151.9 223.2 252.5 358.4 11360 1.2 10.1 27.7 31.8 80.3 1456 205.9 179.2 682.0 1.2 9.2 27.8 29.8 73.4 133.9 153.6 428.9 1.2 4.3 25.3 17.8 54.8 79.4 182.8 0.8 14.3 9.9 32.5 57.5 0.8 9.1 10.3 20.2 37.7 301.3 568.0 1147.6 1968.2 2885.3 3494.4 3734.4 3408.7 17545.6 368 Mortality Study of Oil Refinery Workers/TWriault and Goulet UCC 053983 > Table 5.--Observed end Expected Number el Deaths and Stindard Mortality Ratio (SMR) Among OK Retinary Workers (1928-1976) According to the Him Elapsed Since Start o( Exposure. Disease Category (ICDA*) Malignant neoplasms (140-209) Central nervous system diseases (320-349) Cardiovascular diseases (390-458) Respiratory system diseases (460-5t9) Digestive system diseases (520-577) Genitourinary system diseases (580-607) Acpidents and violence (800-999) Otttrs Alt causes (001-999) 20 Years Since 1st Expet. Obs/Exp.* SMR 7/8 36 1/0.70 12/19.07 0/1.57 0/2.70 0/1.27 12/11.44 1/7.77 33/52.28 83.73 142.85 62.92 -- -- -- 104.69 12.87* 63.12* 4-20 Years Since 1 it Expos. Obs/Exp. SMR 18/20.IS 0/0.65 40/42.56 3/4.29 7/3.92 0/1.27 1/5.82 6/6.73 75/85.42 89.19 -- 93.98 69.93 178.57 -- 17.18* 89.15 87.80 Total Obs/Exp. SMR 25/28.54 1/1.35 52/61.63 3/5.86 7/6.62 0/2.54 13/17.26 7/13.90 108/137.70 89 29 74,07 84.37 -51.19 105.74 -- 75.31 50.36 78.43* `international Classification of Diseases, Eighth Revision tExpected deaths based on age specific death rates of Quebec males ^Statistically significant Table 6.--Observed and Expected Number of Deaths One to Malignant Neoplasms and Standard Mortality Ratio (SMR) Among OK Refinery Workers According to Time Elapsed Slnco Start ot Exposura. Disease Category (ICDA*) Nose, longue, orbit (140-149) Digestive system (150-159) Larynx, lung, pleura (160-163) Bone, skin (170*173) Kidney, genital (185-189) Brain, central nervous sytem (190-192) Leukemia, lymphoma (200-209) Other cancers All neoplasms (140-209) -20 Years Slnco 1st Expn. Obs/Exp.* SMR 0/0.21 3/3.10 0/1.80 0/0.23 0/0.56 3/0.46 1/0.90 0/1.08 7/8.36 -- 96.77 -- -- -- 652.17* . 111.11 -- 83.73 4- 20 Years Since 1st Expos. Obs/Exp, SMR 0/0.59 9/7.12 3/6.6B 1/0.19 3/1.97 0/0.31 2/1,46 ,0/1.87 18/20.18 -- 126.40 44.91 526.31 152.28 -- 136.98 -- 89.19 Total Obs/Exp. SMR 0/0.79 12/10.22 3/8.48 1/0.42 3/2.55 3/0.77 3/2.36 0/2.95 25/28.54 117,41 35.37 238 09 "117.64 389.61 127.11 -- 89.29 `International Classification of Diseases, Eighth Revision tExpected deaths based on age specific death rates of Quebec males ^Statistically significant ,, Table 7. - Ust of Worker* Who DM el Cancer (Among 1019 Mon Who Worked lor Over 5 Yoars at the 01 Refinery). No. Type of Cancer Age st Ocsth Time Bspsed Betweea Occupaben it the Refinery Startleg Work* Death 1 Stomach (151) 2 Stomach (151) 3 Stomach (151) 4 Stomach (151) 5 Rectum (154) 6 Rectum (154) 7 Pancreas (157) a Pancreas (157) 9 Intestinal (159). 10 Brain (191) 11 Lung (162) 12 Lung (162) 13 Liposatcoma ol the legs (170) ' 14 Prostate (185) 15 Prostate (185) 16 Bladder (188) 17 General lymphoma (200) 18 Myelogenous leukemia (205) 19 Intestine (153) 20 Brain (192) 21 Lymphosarcoma (200) 22 Pancreap (157) 23 Lung (ft62) 24 intestinal (153) 25 Brain (191) 51 Maintenance 55 Maintenance 43 Laboratory 50 Office 57 . Laboratory 67 Operations 65 Maintenance 71 Operations 45 Maintenance 43 Operations 77 Maintenance and others SO Office 79 Maintenance 74 Office 66 Maintenance 51 Maintenance 68 Maintenance 43. Operations 72 Operations and office 41 Maintenance 56 Office and maintenance 43 Maintenance 46 Maintenance 73 Office 41 Maintenance-operations 24 33 21 23 13 18 29 30 14 18 25 42 36 29 23 21 25 13 36 13 29 25 20 40 19 No. el Years ol Sendee 15. 29" 21 23 13 ' 15 23 8 14 6 15 19 19 17 19 21 20 13 28 12' 29 19 8 28 18 Journal ol Occupational Medicine/Voi. 21, No. S/May 1979 369 ucc 053984 the first exposure and the date of death, the "less than 20 years since start of exposure group" has statistically sig nificant lower SMR's for all causes of death and for the category "other diseases" when compared with the Quebec population. In the "more than 20 years since start of exposure" subgroup, the observed numbers of deaths for all causes do not differ from the expected numbers. The SMR for diseases of the digestive system (520-577) is elevated but not enough to be statistically significant. The SMR for accidents and violence ($00-899) is statistically signifi cantly lower than expected when compared with the Quebec population. Cahcer Mortality. -- Taldfle 6 shows mortality due to malignant neoplasms. There is no significant difference in the total cancer mortality death rate between the refinery workers and the Quebec population (SMR = 89.29). Three observations stand out of the detailed study of the various cancer types. (1) Lung cancer was less fre quent than expected (3 vs 7.42). (2) Three brain cancers appeared in workers within an interval of less than 20' years since the start of exposure. This was unusual and showed an SMR of 652.17. (3) Though cancers of the di gestive system were not statistically higher than expected (12 vs 10.22), an SMR of 117 combined with an elevated number of deaths due to diseases of the digestive system deserve further attention. Table 7 gives the list of workers who died of cancer by type of cancer; age at death, occupation, time elapsed between start of work and death, and number of years of service. Table 8 gives the distribution of workers who died of cancer by-occupational categories. The occupa tional group "others" includes individuals who worked in more than one of the preceding categories. There is no statistically significant difference in the cancer mortality between the three subgroups: operators, office employ ees and maintenance workers. Discussion The finding of an overall mortality rate for oil refinery workers better than that expected from the general popu lation agrees with what other authors have found in simi lar occupational mortality studies.1111 There seems to be a selective process by which healthy persons constitute the employed portion of the population. This is rein forced by the fact that it is the "less than 20 years since start of exposure group" which showed the lowest stan dard mortality ratios. The occupational history of the three young workers who died of brain cancer revealed that one of them was assigned to the fire safety department, the second was a maintenance man in the furnace room, and the third, an operator in the heavy oil unit. No common exposure of these men to a single or a class of chemicals could be identified which would have given clues as to what was . associated with the appearance of these cancers. No obvious reason could account for the low lung cancer death rate among oil refinery workers either. Other authors have described elevated lung cancer rates Table 8. --Cancar Distribution According to Occupational Categories. - Occupational' Categories Operators Office employees Maintenance Others (belong to several groups) Total No. at Workers 299 168 321 No. af Ne.ef Cancan Cancan Obsarved Expecttd* Obs/Exo. 4 5.5 0.73 4 3.8 .1.05 11 99 1.11 227 1015 6 25 5.8 1.03 25 'Expected numbers based on age adjusted death rates in the study population X1- 0.55,01 - 3,p - .91 among workers exposed to hydrocarbons.14 * The smok ing history of the present workers was not known. The high number of deaths caused by disease and can cer of the digestive system organs deserves particular at tention. It is worth noting that similar observations were made in two previous mortality studies of oil refinery workers.11 A review of the detailed occupational history of the persons deceased from these diseases did not lead to the identification of a common exposure to a chemical within the refinery. Poorly documented data on the men's alcohol consumption could not account for the five cases of cirrhosis of the liver among seven deaths due to dis eases of the digestive system. The questions raised in this study as to the presence of an association between work in an oil refinery, and the development of brain cancers and diseases of the diges tive system will only find an answer by extending the study to larger groups of oil refinery workers. The authors wish to acknowledge the hdp of Or. JeccgiHine Fabia and Hilda Iturra, Laval university, and Dr. jean-Marc Fredette and Mrs. Gervafee Thebault, Quebec Population Registry. References 1. Caffey W, Weaver NK, and Tabershaw IR: Mortality study of petroleum refinery workers. The American Petroleum Institute, Wash ington D.C., Medical Research Report, No. EA 7402,1974. 2. Hanis N and Stavraky K. Ten-years mortality study of employees in a petroleum company. Society for Epidemiologic Researchj-Abstracts Am I Epidemiol 106:244,1974. 3. Weil CS, Smyth HF, and Nale TW: Quest for a suspected industrial carcinogen. AMA Arch/ndHyg 5:535-547,1952 4. Hammond EC, Selikoff I], Lawther PL, and Seidman H: Inhalation of benzpyrene and cancer in man. Ann NYAcad SO 271:116-124,1976. 5; Doll R, Vessey MP, and Beasley RMP: Mortality of gas-workers. Final report of a prospective study. Br/lnd Med 29:394-406,1972. 6. Uoyd JW: Long-term mortality study of steelworkers. Respiratory cancer in coke plant workers. / Occup Med 13:53-68,1971. 7. Vtglianl EL: leukemia associated with benzene exposure. Ann NY Acad So 271:143-151,1976. 6 Case RAM and Hocker ME: Tumour of the urinary bladder as an occupational disease in the rubber industry in England and Wales. Br / PrevSoc Med 0:39-50,1954. 9. Hill ID: Computing man years at risk, fir / ftev Soc Med 26:132134,1972. 10. BaRar jC and Ederer F: Significance factors for the ratio of a Poisson variable to its expectation Biometncs 20:639643,1964. 11. Andjetkovk D, Taulbee J, and Symons M: Mortality experience of * cohort of rubber workers, 1964-1973.1 Occup Med 18:387-394,1976. 370 Mortality Study of Oil Rflflnory Workers/Thfirtault and Goulet UCC 053985