Document 44LVdpkLEzJ6BpXEgJKv2JVvx

PLAINTIFF'S EXHIBIT Occupational Cancer Epidemiology in the People's Republic of China Weiai Wu,MD The practice of occupational cancer epidemiology in China began in the 1970s. Prior to t h a t time, occupational diseases were evaluated only by industrial hygiene practitioners and through rudimentary surveillance, and there was little awareness t h a t t h e epidemiologic approach was applicable to occupational problems. Most of t h e industries in China have been developed since 1949. and during t h e early stages of development, when scientific and technologic experience was limited, t h e effects of occupational agents on worke r s were not well recognized. By t h e mid-1970s. however, t h e problems of occupational hazards in China gradually emerged. Reports of excesses in cancer incidence among workers in specific industries followed one after another: these attracted the attention of governments, health professionals, managers, and workers. To determine t h e extent of occupational hazards in China and to develop preventive measures. Chinese health professionals began using epidemiologic method8 to study occupational hazards among workers exposed to various occupational agents. With the implementation of a more open policy, the scientific and technologic exchanges between China and other countries will increase: therefore, a brief introduction to t h e current s t a t u s of epidemiologic studies of occupational cancer in China is warranted. In China. most practitioners of occupational epidemiology a r e associated with one of t h e following types of agencies: (1) national and provincial (or city) institutes of health under t h e Ministry of Public Health. such as the Institute of Occupational Medicine in'Beijing or the Shanghai Institute of Labor Hygiene; (2) From t h e Inatitute of Occupational Medicine of t h e Chinese Academy of Preventive Medicine. 29 N a n Wei Rd. People'. Republic of China. Beijing. China 1oa)sO. ~1736/88/3012~.~/0 Capyright C by A m a r i u n College of & c u p . t l o n d M e d ~ c l n s 968 departments of public health in the national medical schools, such as the School of Public Health, Beijing Medical University; (3)institutes of occupational health under ministries other than Public Health, such as t h e Institute of Industrial Health, under the Anshan Iron and Steel Company of t h e Ministry of Heavy Industry; and (4) t h e antiepidemic stations under provinces or cities, such as t h e Zhejiang Antiepidemic Station. Currently in China there a r e 170 institutes of occupational health and antiepidemic stations, and about 2,500 professionals for occupational health. Four regulations on occupational health, 124 hygiene standards (maximum allowance concentration [MAC]), and 20 diagnostic criteria have been promulgated. Annually, approximately $3,OOO,ooOis spent for the prevention and treatment of occupational diseases. Some Characteristics of Current Methodology As in other countries, the major methods used for occupational epidemiologic study of cancer in China a r e the retrospective cohort mortality study and the casecontrol study. Some prospective cohort studies have been b d on retrospective cohort studies; however. no results of prospective cohort studies had been reported in China until recently. Airhougn basic epidemiologic methods used in China a r e in theory similar to those used in other countries. the identification of cohort members and follow-up of vital status. as well as the collection of death certificates, a r e quite different. In China. identification of cohort membera is based on payroll records (i.e.. pay sheets) on which names of all persons who have worked at a factory a r e listed monthly according to departments and job categories. These records are maintained by the Department of Labor and Capital in the facility and a r e a very convenient means of identifying cohort members. In addition, there is a cumulative personnel system in which detailed records for every worker from initial work experience t o death a r e recorded and transferred to a master file maintained by the government. This Cancer Epidemiologyin China/Wu Frovides a source from which detailed data on v i t a l s t a t u s of cohort members can be obtained. Also. the vital s t a t u s of cohort members can be traced through the local poiice station from which every citizen must get foad coupons, tu p a r t of a current system of grain rationing. Birth and death records, as well as the relocation of persons, a r e c u r r e n t and kept in good order in the local police station. All salaried persons in China may avail themselves of public health services and a r e assigned to health c a r e facilities, from the small clinic located in their place of employment to the general and specialized hospitals of the provinces and cities. Therefore, death certificates a n d medical histories of deceased workers can be ob- tained from the medical recorda at hospitals at various levels of government. The mobility of Chinese citizens is not aa great as t h a t of citizens in other countries with different social systems and customs. Workers and staff reside close to previous employment locations after retiring, because these a r e the ~ourcesof their monthly retirement pay. As a result, follow-up rates of more t h a n 95% can be reached with little effort. Some ImportantOccupational EpidemiologicStudies in China Most occupational epidemiologic research in China haa been carried out since 1970. Eight major studies have been performed and will be summarized. All a r e retrospective cohort mortality studies. Vinyl Chloride The cohort is composed of 5,958 workers who worked at one of 14 major production facilities for at least 1 year. between 1958 a n d 1981.' Of t h e 14 vinyl chloride facilities. most were built in the 1960s. with the earliest in 1958 and the latest in 1970. Prior to 1970, the concentration of vinyl chloride monomer in these facilities waa high (Table 1). Since the late 1970s. the quality of industrial hygiene at these facilities has been distinctly improved: however, the vinyl chloride concentration in air at several work places has remained fairly high due to manual operations in several of t h e produciion processes, rapeciaily iu iiie claauing of the reactor. Table 1 shows the vinyl chloride concentrations at one of these facilities after t h e control technology was improved.' These levels are still greater than the standard of 1 ppm recommended by the National Institute for Occupational Safety and Health for t h e United States. T h e results of this study showed no excess risk of mortality from all deaths, all malignancies, liver cancer. or any other malignant neoplasms among workers exposed to vinyl chloride. Also. no angiosarcoma was diagnosed among vinyl chloride/polyvinyl chloride workers. nor was t h e r e a n association of incidence and mortality with job categories, extent of exposure, or latency. The authors suggested t h a t the results may be due t o a short latency period (85% of cases had latencies of 10 years or less) since most of the facilities were built in the 19608. The prevalence rates for hepatomegaly and splenomegaly among workers in this study has been found to be markedly increased. In one of these vinyl chloride facilities. the prevalence rate due to splenomegaly was 9%. The prevalence rate for hepatomegaly was 10% in 1973,20% in 1975. and 25% in 1980.' More recently, a prospective cohort study of workers exposed to vinyl chloride has been conducted in China.' Chioromethyl Methyl Ether (CMME)and Bis(chloromethyl)Ether (BCME) Most of t h e plants producing CMME and BCME in China were built between t h e late 1950~and the early 1970s. Exposure data maintained in the plants were incomplete: however, according to the recordn obtained from one of Shanghai's CMME plants in 1977 to 1978, the exposure of CMME level ranged from 1.2 to 59 ~ p m . ~ Before 1977,t h e air CMME concentration at t h e workshops in t h e plants may have been much higher. The retrospective cohort study included workers exposed to CMME for at least 1 year in any of the 10 plants producing CMME." The cohort included 915 workers (534 men and 381 women). From the beginning of operations through 1981.15 deaths due to lung cancer were observed and 0.97 were expected (using Shanghai death rates aa a reference), yielding a standardized mortality ratio (SMR) of 1,546 (95% confidence interval (CI) = 944.2.531). Excesses in mortality from all causes and all malignancies were also observed due to the excess of lung cancer deaths. The study indicated t h a t mortality from lung cancer among workers exposed to CMME and BCME w m related to t h e extent of CMME expasure. and no aseociation between lung cancer mortality and cigarette smoking was found. Cigarette smoking w s s assessed by interwew of kin and c o - w o r k e r ~ . ~ Of t h e 15 cases of lung cancer, 11 had a histologic diagnosis. Eight of these 11 were confirmed to be undifferentiated cell cancers and three were squamous cell cancers. The a g e a t death of workers who died from lung cancer ranged from 32 to 64years, with an average of 50 years. which is 10 years younger than would have been expected if the mortality rates of Shanghai applied to the ~ o h o r t . ~ Benzidine Benzidine has been manufactured in China since 1956: prior to t h a t time i t was imported.' In 1977. the use and production of benzidine was stopped due to i t s carcinogenicity. The retrospective cohort study of benzidine manufacturing plants included 2.525 workers (1.860 men and 665 women) exposed to benzidine for a t least 1 year. from 1972 through 1981. The control workers were from plants with no exposure to benzidine. naphthylamines. 4-aminobiphenyl. or other aromatic amines. A relative Journal of Occupational Medicine/Volume 30 No. 12/December 1988 969 Sampling L o c a m Synthetlccontrd room Synthetc r m at building Compressor room 1 Compressor room 2 polymenzlngPolymenzmgantrd r m Basc processmg Packing TABLE 1 Concecrtratm of Vinyl Chlonda in a Production Facility 1970 Samples 13 1 8 20 (0.5-214 0) 13 76.9 22 (14 2-367 0) 12 469.4 23 (1 44.0-4600.0) 12 84.9 22 (3.9-4900.0) 13 60.8 23 (8.7-244.0) 12 1.3 23 (0.5-75.1) 10 9.4 23 (3.5-145.0) 12 5.7 23 (0.5-76.5) 12 4.8 16 (0.5-17.5) 1981 3.9 (0.5-68.6) 56.2 (1.6-636.0) 221.9 (40.8-1 540.0) 35.2 (5.5-380.1 ) 24.8 (0.5-163.0) 3.3 (0.5-61.6) 15.4 (0.5-262.8) 15.5 (0.5-235.1) 3.1 (0.5-20.4) TABLE 2 IntecsctlonBetween BenzdineExposue and Ctgarette Sr0ku-g Y)me Oevebqnent of B W Cancer Groups B.nzidii Contrd Case (cohort) -- + + Smoking -+ - + lntidmcr p.rlO. 1.36 8.45 86.15 207.13 RR 1.00 6.21 63.35 152.30 P - .05 C.05 CO.01 risk (RR) was calculated as the ratio of the crude mortality rate (per 100,OOO) in t h e exposed cohort divided by t h e crude mortality rate in a control cohort. Among male workers exposed to benzidine, 102 deaths from all causes were observed with a RR of 1.4 (90% CI = 1.1. 1.82);52 deaths and 71 incident casea due to all malignancies resulted in a RR of 2.07 (90% CI = 1.4, 3.0) for mortality, and a RR of 2.7 (90% CI = 1.9.39) for incidence.' Speciflc attention waa given to t h e mortality from and incidence of bladder cancer among workers exposed to benzidine. Twelve deaths and 30 Incident caaes due to bladder cancer were ftfllrid, yielding a FtX of IZ..CS (90% CI = 3.4,48.9) for mortality and a RR of 32.20 (90% CI = 11.2, 99.6) for incidence. Thirty cases of bladder cancBr observed vs 1.15 cases expected based on the experience of Shanghai's general population, standardized by age,sex, and calendar year, resulted in a standardized incidence ratio of 2,610(90% CI = 1.878. 3,538).In addition, mortality from a n d incidence of lung cancer and stomach cancer were also slightly in excess. The 39 cases of bladder cancer observed in t h e study were all confirmed as transitional epithelial cancer. Of these, 35 cases were diagnosed as papillary and four as infiltrating. The study also suggested a synergistic effect of benzidine and cigarette smoking upon t h e development of bladder cancer in workers (Table 2). After analyzing t h e incidence of bladder cancer among workers exposed to benzidine, by calendar year, the authors 970 noted t h a t it seemed to be increasing and has not peaked: Asbestos The retrospective cohort study of asbe~tosincluded nine asbesta product manufacturem and five ssbestos minm located in varioua districts in China Findings for manufacturing and mining will be dimusad mparately. Nine brbmtaa Roduct Manufacturers. Moat of the nine asbe8ta product manufacturing planta were built 30 years a g 0 . O The principal asbestos used w ~ cdhrysotile. Although exporure levels for asbeeta dust in these manufacturing p l a n b have decreased markedly in the past t h m decades due to improvements in technology and engineering, the asbestos dust concentration a t the workshops until 1982,when this study was conducted. greatly exceeded the current standard (2 mg/m3) in China. Fig. 1 display8 the average ssbestos dust levels between 1956 and 1981 as determined from 297 samples obtained from five raw material d e p a r t m e n k 5 The cohort membem for the retrospective cohort study of nine esbestos product manufacturers were lc),C?95s o r f e n who had been exposed to mbeatos for a t least 1 year before 1972. The follow-up of vital status began in 1972 and continued through 1981: The resulta showed t h a t t h e workers exposed to asbestos at manufacturing plants had a significant excess risk of death compared with control workers who had not been exposed to asbestos. T h e mortality rates (per lOO.OO0) for all causes. all malignancien, and lung cancer were 522 (752men, 238 women), 170 (240 men: 83 women), and 67.4 (93 men: 36 women), respectively. Malignancies were observed to be the leading cause for all deaths, and lung cancer was the most common malignancy. The SMR for lung cancer is 633 (90% CI = 474,830).The latency for lung cancer ranged from 8.0 to 34.0 years with an average of 21.7years. The excess risk of lung cancer among workers exposed to asbestos was found to be related to job, latency, and extent of Cancer Epidemiology in China/Wu -80 -7 0 c) E 60. : -\ 50 40 -3 0 20 -10 1ma0 1e70 1aao Fig. 1. Average asbestas dust bevels n five plants between 1956 and 1981. TABLE 3 RelabonsnipBetween Cgareno smoking and Lung Cancer Among Ww(ters Exposed to Asbestos -- - - 1 2.030 1 .o + 4 3.1 10 2.6 + 6 1.044 11.7 + + 26 2,148 24.6 exposure. Table 3 shows a clear association between cigarette smoking and an excesll risk of lung cancer among workers exposed to asbestos. Only one meso- thelioma was observed in this retrospective cohort study.' Five A s b a b e Minea. These five mines a r e the largest asbestos mines in China. with an operating history of more than 30 years. In t h e past. manual work in these mines w m very common and the exposure level ex- tremely high. Although t h e a i r concentration of asbestos dust at t h e mines h a s declined remarkably since 1955 due to a wet operating method instead of a dry operating method. t h e exposure level substantially exceeded t h e current standard (2 mg/m3) in China, sometimes a 100- fold. Airborne asbestoa dwt in t h e environment of some residential districts around the mines was also high.3 The retrospective cohort study of five asbestos mines included 6.193 worheis Zsd Sccr. e r ~ s e dto asbes- tos for at least 1 y e a r before Jan 1, 1972. Vital s t a t u s was observed from 1972 through 1981. The results were similar to those observed in the study of asbestos product manufacturers. t h a t is. excess risk due to all causes. all malignancies. and lung cancer. The SMR for lung cancer was 940 (90% CI = 610, 1.416). Again. t h e excess risk of lung cancer w a s found to be related to job categories. latency, extent of exposure. and cigarette smoking hab- i t s of these workers.' Benzene In this retrospective cohort study, 28.460 cohort members (15.643 men. 12.817 women) who had been exposed to benzene for at least 1 year from 1972 through 1981. were identified from 233 plants located in 12 cities in China.' The results indicate that mortality for all causes. all malignancies. and leukemia among the workers exposed to benzene is increased significantly (Table 4). Twenty five deaths observed for leukemia vs 4.36 deathsexpected based on unexposed control workers. or 5.45 deaths expected based on the population of the specific city, resulted in a SMR of 574 (90% CI = 398, 800)o r 459 (90% CI = 318.640).respectively. Exposure levels for 25 leukemia patients varied from 3.3 to 313.5 ppm according to records kept in the plants. Latency for leukemia among workers exposed to benzene ranged from 0.8 to 48.5 years with en average of 11.4 years. The a g e of leukemia patients ranged from 24 to 62 years. with a n average of 39.7 years.* An excess risk due to leukemia among workers exposed to benzene was found to be related to the extent of exposure, job title, latency, and history of chronic benzene intoxication. Of 30 leukemia c w s . moat (76.7%) were diagnosed 89 acute leukemia. In thie study, excess risk for lung cancer among workers ex- posed to benzene was also observed (SMR = 231. 90% CI = 196, 271). In addition, a matched case-control study was conducted to study further the relationship between worke r s exposed to benzene and leukemia. Cases were 88lected from three large hospitals, and controis were patients from the same hospitals who had no hematopoietic disease. A threefold excess exposure to benzene for cases of leukemia was observed.' Anenic Cohort members were identified from the following facilities: (1) a copper ore-smelter, (2) a secondary copper smelter: (3) an arsenic smelter: and (4) an arsenic-bearing ore mine. These facilities have a long operating history and were built before the 19309.' In the early years. working conditions were primitive and simple: manual work was very common. and the air at work sites was severely polluted by arsenic. Although control technologies in these plants and mines had been improved since the 19509. arsenic levels remained high. The situation was so bad a t the arsenic smelting plants t h a t fhP Ruthorities closed down some of the operations in 1977.' I t is noteworthy t h a t both workers and miners. as well as family members. had drunk mineral spring water containing arsenic from natural sources for a long time. The retrospective cohort study included workers exposed to arsenic from 1972 to 1981-18.795 Personyears at risk. The results indicated that malignancies were t h e leading cause of death, and that lung cancer w a s the leading malignancy. The forty deaths observed from lung cancer resulted in an S M R of 678 (90% CI = 511. 882). based on control workers' death rates. directly standardized by age. sex, and calendar years. This study found a strong dose-response relationship between lung cancer and air arsenic concentration. The latency period for lung cancer among workers exposed Journal of Occupational Medicine/Volume 30 No. 12/0ecernber 1988 97 1 All causes Men Women Total All malignances Men WomeO Total Leukemia Men W0rnet-l Total TABLE 4 Mmality from All Causes, All Malqnances. and Leukemld Among Workerr E x m W to &nZene PYAR' C a s e Subject Death, Death Rates Ilo-') P Y AR Conlrol Subpcf Deaths Death Rater (lo-') 100.025 78.531 178.556 100.025 78,531 178.556 100.005 78,531 178.556 394 80 474 187 33 220 17 8 25 393.90 101.87 265.46 187.0 42.0 123.0 17.0 10.2 14.0 121 268.5 77.932.5 199.201.O 121.268.5 77.932.5 199,201.O 121,268.5 77,932.5 199.201.O 236 41 277 91 18 109 3 1 4 194 61 52.61 139.06 75.04 23.10 54.74 2.47 1.28 2.01 AR P 2.02 1.94 191 2.49 1.82 2.25 6.88 7.24 6.97 < 01 c.01 <.Ol <.Ol <.OS C.01 C.01 c.05 c.01 to arsenic ranged from 6 to 31 years. with an average of 22.9 years.' Chromate The chief raw materials used for producing chromate a r e chromite, sodium carbonate, limestone, and dolomite. The products include sodium bichromate, potassium dichromate. chromic acid, and chromic sulfate. Based on incomplete exposure data recorded in the plants. the exposure level8 from 1963 through 1980 varied from 0.020 to 21.33 mg/m3. The air chromate levels at some work sites at times have been much higher.' In the retrospective cohort study, 2,982 workers exposed to chromate for at lee& 1 y e a r before 1981 were identified from seven chromate manufacturers in seven large cities. Excess risk for all causes. all malignancies. and, specifically, lung cancer among workers exposed to chromate was observed. Malignancies were the leading causes of death. T h e d e a t h rates for male and female workers due to all malignancies were 141.08/100.000 and 88.05/100.000. respectively; these figures were much higher t h a n those of control workers (99.9/ 1OO.OOO and 8.4/100.000. respectively). The author attributes this finding for female workers to employment in more dangerous jobs.' Eleven d e a t h from iung cancer were observed amang male workera exposed to chromate, and resulted in a RR of 2.7 (90% CI = 1.3. 5.5) when compared with the rates for the control w o r k e n not exposed to chromate. The latency period for l u n g cancer among workers exposed to chromate ranged from 2.2 to 20.3 years, with a n average of 11.7 years. T h e average age of workers diagnosed as having lung cancer was 55.6 years: this was 8.2 years less t h a n t h e a v e r a g e for control workers.' The study also found t h a t excess risk due to lung cancer among workers exposed to chromate was related to latency, extent of exposure a n d smoking habits of workers. In addition. t h e r e was an increasing tendency for cancer of t h e esophagus among male workers exposed to chromate, and for cancer of the breast among the female workers.' 972 Coking, Gas, Carbon,and Oil Refinery Plants Coking Plurb. Nineteen coking plants. which includeo almost all of the coking plants in t h e metallurgical industry in China. were studied. Three of t h e plants were built prior to the 19309, seven prior to t h e 1950s. and nine prior to the 19609.' The cohort cornprisea 21,995 workers who had worked at any of t h e 19 coking plants before 1966. Follow-up of vital status of cohort members began in 1971 and continued through 1982. a total of 12 years. A comparison group wp85 chosen from workers in a primary rolling mill (steel company). An excea8 risk of death d u e to lung cancer among coking plant workers wan observed (standardized risk ratio [SRR] = 2.55. 90% CI = 2.13. 3.03). There waas also an increased risk for all causes of death and all malignancies among coking plant workers (Table 5). The excess risk of lung cancer among coking plant workers was significantly related to job title. Among workers in the coking plants. t h e highest mortality from lung cancer w- observed in the coking departments. especially among workers at t h e top or side of the coking oven. The differences in risk of lung cancer observed among workers with different job titlea were consistent with extent of exposure to t h e emission from the coking oven. mainly with exposure to benzo(a)pyrene. In addition, mortality from lung czncer among cokiag workers wc9 a!xer*:ed ocly after 15 y e a n of latency. The peak of the death r a t e for lung cancer was reached after a latency of 30 years. The authors concluded that the excess risk of lung cancer among coke oven workers in the steel company was related to t ~ n z o ( a ) p y r e n e . ~ Coal G u P l ~ b T. he study cohort comprised 3,107 workers from six coal gas plants and t h e comparison group was from workers in a steel rolling mill: 234 deaths for all causes (SRR = 1.29.90% CI = 1.16, 1.44) were identified among workers who were employed in 1971 and who died during the next 10 years. One hundred nine of those deaths were for all malignancies (SRR = 1.73, 90% CI = 1.46. 2.02). The study showed that. from 1971 through 1981. the standardized risk ratio for lung cancer among workers who worked 7 the Cancer Epidemiology in ChinalWu F TABLE 5 CI = 1.27.2.99). especially among workers who worked I Rlsk of Lunq Cancer for &kina Plant workers - ~~~~~ Causes o l Oeath Observed Expected SRR P in the fuel oil-producing department (SRR = 3.88.90% CI = 1.99. 6.68). No excess risk of lung cancer was I All causes 837 628.77 1.33 4.01 found. All rnalignames I Lung cancer 291 198.52 1.47 4.01 93 36.44 2.55 4.01 The third retrospective cohort study, which included 3.774 workers, was conducted in a synthetic oil plant Coking department Coking oven Top of Oven Side of Oven 53 12.03 4.41 4 01 40 8.05 4.97 <.01 13 0.52 8.59 4.01 15 3.05 4.91 c.01 built in t h e 19409. SRRs were calculated aa in the previous study. In this plant, oil was made from coal gas in a process t h a t involved a cobalt catalyst contam- Noncoking oven 13 3.98 3.27 c.01 inated with radioactive elements. During the early years Department of Prepanng W i2 4.84 2.48 <.oi of the operation, t h e environment in this plant w'88 severely polluted by emissions from the coal gas ovens ! gas departments of t h e coal gas plants was very high and by radioactive thorium and radon daughters emit- I (SRR = 3.66, 90% CI = 2.36, 5.48), and related to ted during t h e production process. A total of 200 demth latency. Among t h e w workers t h e r e w t a~lso an excess from all c a w s were observed, including 74 d e a t h from risk of cancers of t h e digestive tract.' all malignancies, of which 20 were from lung cancer. Carbon. The raw materials used in producing carbon Excess risb of all malignancies. lung cancer, and diges- ', are smokeleaa coal, coke, and pitch. T h e air pitch con- tive tract cancer were found, with SRRs of 1.35 (90% centration at work sites in the carbon plants has been CI = 1.10, 1.64). 2.02 (90% CI = 1.34,2.93), and 1.56 as high aa 40 mg/ms. During t h e early years of opera- (90% CI = 1.13, 2.09), respectively. Resulb showed tion, environmental contamination at t h e work sites was that the mortality from lung cancer was significantly extraordinarily related to job title and extent of exposure.' I In this retrospective cohort study, 5,122cohort mem- bers in two carbon plants built in t h e 19509were studied. Prospectsand ProMems I Workers in a steel rolling mill were used as a control j group. A total of 227 d e a t h from all causes (SRR = 1.24,90%CI = 1.1 1.1.39)were observed; these included As China itself is developing, so is ita practice of occupational epidemiology. These retrospective cohort 80 deaths from all malignancies (SRR = 1.33. 90% CI mortality studies covered eight categories of lc~~oam I = 1.09. 1.60), and 24 deaths &om l u n g cancer (SRR = carcinogens in 460 factories o r mines in various geo- F! 2.60, 90% CI = 1.79. 3.60). L u n g cancer was observed to be a leading caw of death for carbon plant workers. graphic are- of China. The studies involved 93,500 cohort members (988,572person-years) and 63,000 con- 1 The excess risk of lung cancer observed among cohort trols (646.744person-years). The findinga presented in workers was m i a t e d with extent of exposure to pitch. thin paper generally are consistent with findinga of and displayed an exposure-response relationship. In the studies performed in other countries.'&'' An exception study, there was an excess risk for diseases of the is t h e vinyl chloride retrospective cohort study; how- circulatory, respiratory, and digestive systems.' ever, t h e latency period for angiosarcoma in t h a t study Oil Refining and Synthetic Oil Plants. One retrospec- was relatively short. These studies of Chinese industries tive cohort study w m conducted in shale oil refining formed t h e basis for making and maintaining lists of I plant built in the 1920s.Data obtained in 1975 indicated that air benzo(a)pyrene levels at the plant ranged from 2.0 to 12.40 pg/lW m3. The air benzo(8)pyrene level carcinogens, 90 t h a t in the future effective protective intervention measures can be taken. and strategies can be developed for other occupational epidemiologic stud- I on t h e top of a n oven at one of t h e departments reached 33.69 pg/lOO m'. In among t h e 6.285 members of t h e ies. In addition. these studies have been training experiences in occupational epidemiology for many health 1 study cohort. there were 657 deaths from all causes: professionals. i 182 deaths from all malignancies. a n d 46 deaths from Today, based on these retrospective cohort studies. ! lung cancer. Based on age- aiit aex-a-@ific death rates so1114 new epidemioiogic studies and updating of past for t h e local population, the SMRs were 141 (90% CI = cohort studies a r e being conducted. As mentioned. most ~ ! 124, 159) for all malignancies, including lung cancer of the industries in China were built in the 1950s. when and cancer of the esophagus. 240 (90% CI = 185, 307) working conditions at the plants or mines were unsat- for lung cancer. and 244 (90% CI = 157,311)for cancer isfactory, environmental pollution in the air was a se- of t h e esophagus. The excess risk of cancers of the lung rious problem. and exposure fevels for workers were too and esophagus was found to be related to the title and high. After 30 years. the severity of occupational haz- latency.' ards in t h e plants and mines has been gradually re- The second retrospective cohort study was conducted vealed. and is now being controlled to some extent. in three oil refining plants buiit in the 1960s.The cohort There are still many occupational hazards and a great comprised 12,422workers. T h e r e were 313 deaths from need for occupational epidemiologic studies. and, com- all causes, and 80 deaths from malignancies. of which pared with other countries. occupational epidemiology 26 were from liver cancer. SRRs were calculated based in China is underdeveloped. For example, there a r e no on t h e experience of a population of workers in a steel up-to-date national population death rates specific by rolling mill. In this study, an excess risk of mortality sex. age. race. and calendar years. Advanced computer was observed for liver cancer only (SRR = 1.99. 90% techniques a r e not used extensively in epidemiologic Journal of Occupational Medicine/Volume 30 No. 12/December 1988 i 973