Document 3EDX3ZRvZmJLgpMrjeeG8bYD

Occuparwnal and Environmental Medicine 1997;54:443-451 443 Surveillance of nasal and bladder cancer to locate sources of exposure to occupational carcinogens Kay Teschke, Michael S Morgan, Harvey Checkoway, Gary Franklin, John J Spinelli, Gerald van Belle, Noel S Weiss Abstract Traditional disease surveillance has used vital Objective-To locate sources of occupa- statistics data bases to examine patterns of tional exposure to nasal and bladder mortality in occupations and industries and carcinogens for surveillance follow up in generate hypotheses for further epidemiologi- British Columbia, Canada. cal study.' M e r the definition by Rutstein et al? Methods-Incident cases of nasal cancer of occupational sentinel health events, many (n=48),bladder cancer (n=105), and population based controls (n=159) matched for sex and age, were interviewed about their jobs, exposures, and smoking histories. Odds ratios (ORs) were calculated for 57 occupational groups with stratified exact methodsto control for age, sex, and smoking. ResuZts-Occupational groups at increased risk ofnasal cancer included: textile workers (six cases, OR 7.6); miners, drillers, and blasters (six cases, OR 3.5); welders (two cases, OR 3.5); pulp and paper workers (three cases, OR 3.1); and plumbers and pipefitters (two cases, OR 3.0). Nasal cancer ORs were not increased in occupations exposed to wood dust, pos- recent surveillance efforts have aimed to provide a more direct impetus to action, including treatment of patients, screening for disease in coworkers, and investigations into industrial hygiene at associated work sites.' Few such studies have investigated cancer outcomes,4d although 13 cancers are listed as occupational sentinel health events.2 There may be several reasons. Cancers have long induction and latent periods, so sources of exposure may no longer exist by the time the case is identified. Although this will often be true, recent studies of mesothelioma led to the discovery of previously unrecognised current sources of exposure to asbe~tosT.~his is especially important evidence of the potential value of cancer surveillance, as asbestos is one of the best sibly due to low exposures in local wood known industrial hazards of this century. industries. Strongly increased risks of Sources of other carcinogens are likely to be bladder cancer were found for sheet metal less well recognised. workers (four cases, OR 5.3), miners (19 Studies that followed the sentinel event Department of Health Care and Epidemiology, University of British Columbia,Vancouver, Canada K Teschke J J Spinelli School of Public Health and Community Medicine, University of Washington, Seattle, USA cases, OR 4.9, gardeners (six cases, OR 3.7), and hairdressers (three cases, OR 3.2). Among occupations originally considered at risk, the following had increased risks of bladder cancer: painters (four cases, OR 2.8); laundry workers (five cases, OR 2.3); chemical and petroleum workers (15 cases, OR 1.8); machinists (eight cases, OR 1.6); and textile workers (three cases, OR 1.5). ConcZusions-Occupational groups with increased risks and three or more cases with similar duties were selected for surveillance follow up. For nasal cancer, these included textile workers (five were garment makers) and pulp and paper workers (three performed maintenance approach of Rutsein et a1 have identified case series, then used case by case evaluations to assess the relation of a disease event to The ability to assess aetiology on an individual basis was a criterion for selecting diseases for the United States sentinel event notification system for occupational risks (SENSOR).' For cancer, and most other diseases on the sentinel events list, it would be difficult to attribute cause on an individual basis, as most cancers with occupational aetiologies also occur in the absence of such exposures. Surveillance would require analytical methods which compare groups. We conducted a surveillance study of three cancer types, mesothelioma, nasal cancer, and bladder cancer, with case-control methodology.It was designed to locate sources M S Morgan H Checkoway G Franklin G van Belle N S Weiss tasks likely to entail stainless steel welding). For bladder cancer, these included miners (12 worked underground), machinists (five worked in traditional machining), hairdressers (three had applied of occupational exposure to known or probable carcinogens in the province of British Columbia, Canada, with the aim of alerting regulatory agencies and industrial health professionals about occupations that warranted occupational Correspondence to: K a y Teschke, Deparunent of Health Care and Epidemiology, University of hair dyes), and laundry workers (three hygiene exposure measurement, and control. were drycleaners). This paper reports the results for nasal and bladder cancer, both of which have several British Columbia, Vancouver, BC, V6T 123, Canada. (Occup Envimn Med 1997;54443451) known occupational aetiologies, as well as non-industrial aetiologies, in particular, smok- I Keywords: textile industry; paper; wood; mining;occu- ing. The results for mesothelioma are reported Accepted 18 December 1996 pational exposure elsewhere.' Materials and methods IDENTIFICATION OF CASES All incident cases with histologically confirmed primary malignant tumours of the nasal cavity or sinuses (ICD-0 topography codes 160.0 and 160.2 to 160.9), and urinary bladder (ICD-0 188) age 2 19 and registered by the British Columbia Cancer Agency (BCCA) were considered eligible for the study. The registry of the BCCA is population based and for case ascertainment uses reports from all provincial pathology and haematology laboratories, death notices from the British Columbia Division of Vital Statistics, reports from other Canadian registries about cancers in residents of British Columbia, and admission records from BCCA clinics. A BCCA pathologist, not blinded to the initial diagnosis, reviewed tissue samples for every case initially identified as eligible. The study period was 1 September 1990 to 31 August 1992 for nasal cancer, to allow accrual of about 50 cases, and 1 September 1990 to 31 May 1991 for bladder cancer, sufficient for accrual of at least 100 cases. As people with an industrial background have often been found to develop bladder cancer at younger ages than those with other backgrounds: '" cases of bladder cancer were restricted to those born after 31 December 1915. SELECTION OF CONTROLS Controls were randomly selected from fiveyear age and sex strata of the provincial voters list updated to 17 July 1991. Those eligible included all residents of British Columbia who were Canadian citizens aged 2 19 and who were not in prison or confined to a mental health institution by court order. Controls were frequency matched to the age and sex distribution of cases of all three types of cancer included in the study. All study controls within two years of age of the cases were included in the analyses for each cancer site. INTERVIEWS The methods of contacting subjects in this study were reviewed and approved by the University of British Columbia Behavioural Sciences Screening Committee. All subjects living within about six hours surface travel time of Vancouver (one way) were interviewedeither in person or by telephone, at their convenience. The remaining interviews were conducted by telephone (21% for cases and 23% for controls). If a subject did not speak English well or had trouble recalling events in his or her life, a relative chosen by the subject was asked to translate or help with the interview. If the subject was dead, the survivingnext of kin who had most recently lived with the subject was contacted for interview. An attempt was made to frequency match next of kin control interviews to the number of next of kin case interviews within age-sex strata. However, when no next of kin was availableor the request for a next of kin interview would have resulted in a refusal, the interview was conducted with the control instead. Interviews were conducted with a standardised questionnaire based on a prototype originally developed by the United States National Cancer Institute to investigate occupational associations with mesothelioma.'' It included occupational, residential, smoking, and medical histories, and an exposure history aimed at identifying exposures considered by the International Agency for Research on Cancer (IARC)'z-'6to be known or probably carcinogenic. Interviews were conducted with a standardised questionnaire. For this study, questions about specific exposures and work processes relating to nasal and bladder cancer were added. An interview guide was used to aid recall of exposure; it listed types and trade names of products in which carcinogens were used. Data for the interview guide was obtained from the Canadian Centre for Occupational Health and Safety trade names data base, the Canadian Pharmaceutical Society, Health and Welfare Canada, and texts." All interviews were conducted by a registered nurse who was aware of the case-control status of the subjects, but was not aware of aetiologies for the specific cancer sites. All interviews were reviewed for completeness by an industrial hygienist (KT) who was not aware of the case-control status of the subjects. Additional information requested by the hygienist (missing data, or further information to clarify tasks or exposures) was sought by the interviewer. OCCUPATIONAL GROUPS All occupations and industries listed were coded to the fourth digit of the standard occupational and industrial classifi~ation.''O~ In total, 3951 separate jobs were listed by the interviewees, representing over 300 occupational and 500 industrial codes. For analysis, grouping of jobs was necessary. However, collapsing to the third or second digit of the standard classifications was likely to group people with very different exposures (all construction trades together, production and management employees together within an industry), who might be kept separate in a grouping scheme specific to the study. Such a scheme was therefore developed without knowledge of the case or control status of the people whose jobs were being grouped. When occupation was considered most likely to determine exposures, it was used as the basis for assignment to a group (managers, teachers, truck drivers), and when the industry was judged more likely to determine exposures, such as production and labouring jobs, it was used to assign groups (sawmill workers, railway transport workers). Where both industry and occupation would be expected to determine potential exposures (industrial mechanics, welders, and other tradespeople), occupation was used as the basis for classification. After the initial assignment to occupational groups, the descriptions of duties and exposures accompanying each job listing were reviewed to ensure the best possible classification-for example, a retail store "manager"who stocked shelves and acted as a 1 Surveillance of'nasal and bladder cancer to locate sources of exposure to occupational carcinogens __ 445 T@le 1 Response characteristics of eligible cases and controls, and characteristics of inrerviemees Nasal cancer - Total eligible: Physician refusal Unable to contact Case or control refusal Interview incomplete Interview complete (% of eligible) Men (I%) Age (Y, mean(SEM)) Next of kin interviewees (%) Never smoked &-smokers Current smokers Pack-years (mean(SEM)) Cases 54 1 0 5 0 48 (88.9) 33 (68.8) 66.3 (1.9) 7 (14.6) 11 31 6 32.3 (5.5) Controls 195 NA 3 33 0 159 (81.5) 128 (80.6) 65.5 (0.8) 22 (13.8) 36 88 35 27.3 (2.8) Bladder cancer Cases Conrrols* 119 2 7 8 1 105 (88.2) 88 (83.8) 64.8 (0.8) 19 (18.1) 14 56 35 38.9 (3.4) 173 NA 3 31 0 139 (80.3) 112 (80.3) 63.4 (0.8) 19 (13.7) 31 75 33 26.7 (3.0) Controls used in the bladder cancer analysis are a subset of those used in the analysis of nasal cancer; 22 were not included because they were born before 31 December 1915. NA = not applicable. cashier was grouped with sales clerks instead of managers; a motor body shop worker who painted cars was grouped with painters, rather than service station attendants or vehicle mechanics. Finally, occupational groups with c 20 listings were reviewed to find whether they could be combined with others. In a few instances, this was not considered possible (hairdressers and barbers, laundry personnel, firefighters). In the final classification, all job listings were assigned to one of 57 occupational groups (table 2). The ORs and 95% CIS were also calculated for occupational exposures to specific agents and processes, and for other factors-for example, diseases, or drugs-considered to be possibly related to nasal and bladder cancer. When non-occupational factors originally suspected were shown to be positively associated, unconditional logistic regression was used to find whether effect estimates for occupational groups with increased relative risks remained the same after adjustment for these additional potential confounders. SELECTION OF OCCUPATIONS FOR SURVEILLANCE FOLLOW UP A standard framework was used to select which occupations were of interest for surveillance follow up. An occupational group was designated for notification of occupational health personnel if (a) it had an OR of > 3.0 in the ever employed analysis, or an OR > 1.2 if it was originally considered to be at risk based on a literature review, and (b) at least three cases in the occupational group had a pattern of similar job duties or exposures. Before notification of local regulatory agencies and hygienists, consideration was given to whether the designated jobs were already recognised as associated with cancer in British Columbia. As this final selection process was mainly of local interest, it is not described in detail here. Results CASE-CONTROL ANALYSIS Odds ratios (ORs) and 95% confidence intervals (95% CIS) were calculated for "ever" (defined as six months of employment or longer) versus "never" employed in each occupational group. Latency analyses were conducted for all occupational groups, with the most recent 20 years of employment removed. Odds ratios were also calculated for 5, 10, and 15 year latencies for all occupational groups COMPARABILITY OF CASES AND CONTROLS There were six cases of nasal cancer and 18 cases of bladder cancer initially reported by the BCCA that were found on pathology review not to be eligible for the present analysis. Table 1 indicates the response rates of those considered eligible for the study after pathology review. Information was obtained for 80%-90% of both cases and controls, and in most instances from the subjects themselves. which were originally suspected or which had Table 1 also indicates selected descriptive ORs of 32.0 in the initial ever or never analy- characteristics of the interviewed subjects. ses. Effect estimates were also calculated for Because the cases of bladder cancer and two duration of employment categories: 6 controls were restricted to those born after months to < 10 years; and 2 10 years. These 1915,22 of the controls selected for the overall are reported when they influence the study were not eligible for the bladder cancer interpretation of the results. analysis. The mean ages among cases were All OR estimates were adjusted for sex, age similar to controls. There were differences in (in three strata: c 60; 60 to 69; and sex distribution between the cases of nasal can- 3 70), and cigarette smoking. Strata for smok- cer and controls, because the frequency ing were based on associations within this matching of controls to cases was based on the dataset for each cancer site, resulting in two distribution for all three cancer sites combined strata for nasal cancer (0-19; and 2 2 0 and nasal cancer had a higher proportion of pack-years), and three strata for bladder cancer female cases than did mesothelioma or bladder (0-9; 10-39; and 240 pack-years). Sex-specific cancer. analyses (men only) were conducted for ever versus never employed in each occupational NASAL CANCER group. When these showed notable differences Table 2 shows the ORs and 95% CIS of compared with the sex adjusted analyses, associations between the 57 occupational further investigations were carried out. groups and cancers of the nasal cavity and Because of the small size of some occupa- sinuses. Textile workers had a significantly tional groups, exact methods were used to increased risk in the ever employed analysis. summarise ORs across all strata and calculate Only one exposed case and two exposed 95% CIS.^' *' All summary ORs and 95% CIS controls were men. Sex specific analyses gave were calculated with Egret (Statistics and Epi- ORs of 3.5 (95% CI 0.1 to 89.3) for men and demiology Research Corporation, Seattle, WA, 12.9 (95% CI 1.3 to 663) for women ever 1993). employed. All five female cases and the one Table 2 Odds ratios* showing associations between occupational groups and cancer of the nasal caviry or sinuses: all cases ( ~ 4 8 a)nd controls (n=159) included Ewr employed Caseslcontrols n Ever employed OR 95% CI Mosr recenf 20yean removed Caseclconrrols n Ever emplayed OR 95% CI Occupational groups with OR Z 3.0: Textile workers Miners, drillers, and blasterst Welderst Pulp and paper mill workerst Plumbers and pipefitterst Occupational groups originally suspect: Painters Machistst Carpenters and wood workers, nect Smelter and foundry workerst Chemical and biological laboratory personnel Forestry and logging workerst Sawmill workerst Shoe and leather workers Other occupational groups: Administrators, managers Travellingmanagers and salesmen Accountants, book-keepers Engineers, designerst Teachers, librarians Sales clerks Office clerks and secretaries Healthcare workers Radio operators Surveyors,prospectors, trapperst Warehouse clerks and labourers Delivery personnel, unrnotorised Firefighterst Guards, policet Armed forces personnel, nect Janitors Housekeepers Handymen and apartment caretakers Laundry personnel Hairdressers, barbers cooks Food service personnel Food processors Farmers and farm labourers Gardenerst Fishermen* Sheet metal workerst Chemical and petroleum workers Construction forement Consauction labourerst Bricklayers, plasters, and cement workerst Elecmcians and elemical equipment installerst Heavy equipment operatorst Motor vehicle operators Pilots, aircraft C r e e Railway transport workers, nect Ship transport workers, nect Shipbuilding workers, nect Transport engineers and fuement Stationary engineers, boilermakers? Service station attendants and managerst Vehicle mechanicst Small equipment repairerst Industrial mechanicst 613 617 214 313 212 214 318 6/27 219 218 2113 3/22 016 5/22 4/33 7121 017 5/15 1214 1 12/27 419 019 118 9/35 7137 215 318 10137 3111 8/13 2/16 014 111 4/13 6/24 7/22 10144 314 014 1/2 1/10 us 5/23 117 019 117 10128 217 318 4/12 015 116 2111 4/19 5/20 3/10 2/10 7.6 3.5 3.5 3.1 3.0 2.2 1.9 1.o 0.9 0.7 0.5 0.4 0 0.7 0.4 1.4 0 0.9 0.9 1.2 1.3 0 0.5 0.9 0.6 1.8 1.6 1.o 1.4 1.6 0.4 0 2.5 0.9 0.7 1.2 0.9 2.9 0 2.3 0.4 0.9 0.9 0.5 0 0.5 1.5 1.1 1.8 1.4 0 1.o 0.8 0.8 0.9 2.9 1.o 1.4-56.6 0.9-14.6 0.2-53.7 0.4-25.4 0.2-49.1 0.2-17.9 0.3-10.6 0.3-2.9 0.1-5.3 0.1-4.0 0.1-2.4 0.1-1.6 0-2.8 0.2-2.2 0.1-1.4 0.5-4.0 0-4.0 0.2-3.0 0.4-2.1 0.4-3.1 0.3-5.6 0-1.3 0-42 0.42.3 0.2-1.5 0.2-13.4 0.2-8.0 0.42.4 0.2-6.3 0.46.7 0.0-2.0 0-13.9 0-225 0.2-3.0 0.2-2.0 0.4-3.4 0.3-2.2 0.4-19.8 0-6.6 0-67.2 0-2.8 0.1-5.0 0.2-2.8 0-4.8 0-3.6 0-4.8 0.6-3.7 0.1-7.1 0.3-9.1 0.3-5.3 0-6.0 0-13.7 0.1-4.4 0.2-2.8 0.2-2.9 0.418.5 0.1-5.5 413 617 213 313 212 213 317 4/25 217 217 2112 -212 1 3/16 4/26 -6/13 5/10 11/34 11/25 -418 11% 7128 7/35 115 312 10137 118 7/12 015 - 111 2/10 6/19 6/22 10143 214 - 112 117 215 4/19 115 - 017 10124 217 317 4/12 - 116 2111 4/15 5/20 319 119 5.0 3.5 3.9 3.1 3.0 2.6 2.1 0.7 1.4 0.9 0.5 0.3 - 0.66 0.57 1.88 - 1.49 1.06 1.24 1.53 - 0.48 0.96 0.63 1.02 9.19 1.0 0.5 1.5 0 - 2.5 0.7 1.1 1.o 0.9 2.0 - 2.3 0.6 1.4 0.7 0.7 - 0 1.9 1.1 2.1 1.4 1.o 0.8 1.0 0.9 3.8 0.5 * Adjusted for sex, age (in 3 strata: < 60,60-69, and Z 70 years), and cigarette smoking (in two strata: 0-19 and 2 20 pa&-years). t All exposed cases and controls were men. -No cases, no further analyses done. nec = Not elsewhereclassified 0.8-43.0 0.9-14.6 0.2-63.4 0.425.4 0.2-49.1 0.2-24.8 0.3-13.4 0.2-2.3 0.1-9.8 0.1-5.3 0.1-2.7 0-1.4 - 0.1-2.7 0.1-1.9 0.5-6.1 - 0.45.4 0.4-2.6 0.5-3.2 0.3-6.6 - 0-4.2 0.3-2.6 0.2-1.7 0-11.1 0.9-139 0.42.4 0-4.4 0.46.1 0-3.1 - 0-225 0.1-3.5 0.3-3.2 0.3-3.0 0.3-2.3 0.2-16.1 - 0-67.2 0-5.1 0.1-10.1 0.2-2.3 0-7.7 - 0-2.6 0.7-4.9 0.1-7.1 0.3-11.8 0.3-5.3 - 0-13.7 0.1-4.4 0.2-3.7 0.2-2.9 0.5-30.1 0-4.3 female control were employed in garment making. Men in this group had varying jobs, none in garment making. Most of the cases mentioned exposure to fabric dust, and some remarked on the fabric finish, cleaning solvents, and sewing machine oil. None of the textile worker cases had been employed in other occupations found to be at risk in this study or considered beforehand to be at risk. Four of the cases and none of the controls were non-smokers. Risk estimates 23.0 were measured in four other occupational groups in the ever employed analysis: pulp and paper mill workers; miners, drillers, and blasters; welders; and plumbers and pipefitters.All exposed cases and controls in these occupations were men. For all except the pulp and paper mill group, risks increased with 2 10 years of employment. An examination of job duties among these occupational groups showed an interesting pattern related to pulp mill employment. Among the welders, one case had worked in pulp and paper mill construction. Many pulp mill piping systems are stainless steel, and welding this metal produces fumes containing nickel and hexavalent chromium.19Two other cases may have had similar exposures: a pulp mill pipefitter (included in both the plumberpipefitter and pulp-paper mill worker occupational groups), and a pulp mill instrument fitter (in the pulp and paper mill group). The i; Surveillance of nasal and bladder cancer to locate sources of exposure to occupational carcinogens 447 remaining case who had worked at a pulp mill had no specific job details given, because the interviewee was a next of kin. All of the pulp and paper mill worker controls were involved in production rather than maintenance job:, and none of the welder or plumber-pipefitter controls reported working in pulp and paper mills or other industries likely to involve stainless steel welding. The cases in the miners, drillers, and blasters group had widely varying work descriptions, which did not suggest exposures in common. Two worked underground and four above ground. Among occupational categories originally considered suspect, those involving work with wood or leather had ORs consistently <1.0: sawmill workers; forestry and logging workers; carpenters and woodworkers; and shoe and leather workers. Consistent with the results obtained in the analysis of these occupational groups, ORs were <1.O for occupational exposures to the following agents and processes which have been strongly associated with nasal cancer in other studies: hardwood dust (OR 0.6, 95% CI 0.1 to 3.0); softwood dust (OR 0.7, 95% CI 0.3 to 1.6); furniture or cabinet making (OR 0.7, 95% CI 0.1 to 3.8); boot or shoe manufacture or repair (no cases); and leather tanning or processing (no cases). Chemical and biological laboratory personnel were also not at increased risk of nasal cancer. Smelter and foundry workers, who might have been exposed to nickel or chromium, showed slightly increased risks only with a 20 year latency period taken into account. Painters had increased risks that increased with latency. Machinists also had raised risks that increased with both duration of exposure and latency. No pattern of duties or exposures was found within any of the occupational groups originally suspected of being at high risk. The histologies of the nasal cancers in this study included 23 squamous cell carcinomas, seven melanomas, seven lymphomas, two adenocarcinomas, two adenoid cystic carcinomas, and seven other histologies with one case each. A histology specific analysis was judged possible only for squamous cell carcinoma (n=23). One control who was more than two years older than the remaining cases was excluded from the reanalysis (n=158). The following occupational groups had ORs 23.0 in the ever employed analysis:textile workers (OR 5.3, 95% CI 0.2 to 107); sheet metal workers (OR 5.3,95% CI 0.1 to 470); machinists (OR 4.3, 95% CI 0.6 to 29.2); railway transport workers not elsewhere classified (OR 4.2, 95% CI 0.6 to 26.7); hairdressers and barbers (OR 4.1, 95% CI 0 to 502); and welders (OR 3.1, 95% CI 0 to 267). Only the railway workers and machinists had at least three cases, but no pattern of duties or exposures was found within these occupational groups. All of the occupational groups that had ORs >3.0 in the analysis of all histologies had decreased risks for those ever employed in this analysis, indicating that the risks were not specific to squamous cell carcinoma. All the remaining occupational groups originally suspected of being at high risk had ORs c 1.0 in this analysis. There were no pulp and paper mill workers or painters with this histology. Adjustment for nasal polyps and decongestant use did not change the results sufficientlyto alter the interpretation of the occupational analyses. BLADDER CANCER Table 3 shows the associations between occupation and bladder cancer. Plumbers and pipefitters had an infinite relative risk, based on only one case. Sheetmetal workers had an OR of 5.4. Two of the four sheetmetal worker cases (but not the control) had worked in aircraft assembly, a job which has been reported to have exposure to carcinogenic aromatic amines.?' Miners, drillers, and blasters had an OR of 4.5 in the ever employed analysis. The relative risk estimate increased with duration of employment ( 310 years, OR 8.1). All but one case in this occupational group were men. Twelve cases and four controls mined underground where diesel exhaust fumes and oil mist exposures are often high. The most commonly named types of mines were gold (six cases, three controls), nickel (four cases, one control), and coal (three cases, one control). Strongly increased ORs were measured in two other occupational groups in the ever employed analyses: hairdressers and barbers; and gardeners. All three cases in the hairdressers and barbers group recalled applying hair dyes. The control, who was a barber, did not. Gardeners had increased risk with increased duration of employment (210 years, OR infinity). There was no pattern of duties or exposures in this occupational group. Among occupations shown to have increased risks in other studies, several had increased ORs in the ever employed analysis: painters; laundry personnel; chemical and petroleum workers; machinists; and textile workers. Except for the chemical workers, all of these had increased risks with 2 10 years of exposure. Only laundry workers and machinists showed consistent patterns of duties or exposures. Three of the laundry worker cases worked in dry cleaning, versus one control. Five machinist cases and three controls worked in traditional machine shops, doing metal grinding, drilling, and lathe work. All reported using cutting oils. There were also some occupational groups, considered potentially at risk based on evidence from other studies, which did not have increased ORs in the ever employed analysis: smelter and foundry workers; motor vehicle operators; cooks; vehicle mechanics; and shoe and leather workers. Sex specific analyses did not show any important differences from the sex adjusted analyses. The cases of bladder cancer had the following histologies: 55 transitional cell carcinomas; 43 papillary transitional cell carcinomas; four squamous cell carcinomas; one adenocarcinoma; one colloid carcinoma; and one carcinoma not otherwise specified. As the papillary, Table 3 Odds ratios * showing associations between occupational groups and cancer of the urinary b included - Ever empbyed u , all cases (n=lO5) and controh (n=139) ~ Most r e ~ g n r20 yean removed ~ ~~ Caseslmmlr n Ever empLyed OR 95% CI Caseslconmlr n Ever e m p w OR 95% CI Occupational groups with OR 3 3.0: Plumbers and pipefitterst Sheet metal workerst Miners, drillers, and blasters Hairdressers, barbers Gardeners Occupational groups originally suspect: Painters Laundry personnel Chemical and petroleum workers Machinistst Textile workers Smelter and foundry workerst Motor vehicle operators Cooks Vehicle mechanicsf Shoe and leather workers Other occupational groups: Administrators, managers Travellingmanagers and salesmen Accountants, book-keepers Engineers, designerst Teachers, librarians Sales clerks Office clerks and secretaries Healthcare workers Chemical and biological laboratory personnel Radio operatorst Surveyors, prospectors, trapperst Warehouse clerks and labourers Delivery personnel, unmotorised Firefighterst Guards, policet Armed forces personnel, nec Janitors Housekeepers$ Handymen and a p a w e n t caretakers Food service personnel Food processors Farmers and farm labourers Fisherment Forestry and logging workerst Sawmill workerst Pulp and paper mill workerst Welderst Construction foremen-) Construction labourers Bricklayers, plasterers, and cement workerst Electricians and electrical equipment installerst Carpenters and wood workers, n e e Heavy equipment operatorst Pilots, aircraft crewt Railway transport workers, n e e Ship transport workers, n e e Shipbuilding workers, n e 4 Transport engineers and firemen* Stationary engineers, boilermakerst Service station attendants and managerst Small equipment repairerst Industrial mechanicst 1IO 4/1 1916 311 613 413 514 15/10 817 313 618 21/26 7/12 12/17 216 22/23 19/32 13/18 415 10112 28/38 19/23 619 618 717 716 26/31 16/33 114 1018 38/32 6111 6111 7/14 17/24 13/19 39/37 4/4 15/12 15/18 212 4/4 815 15121 717 219 18/24 917 4/7 318 9111 615 316 6111 6/18 7/10 1019 m 5.4 4.5 3.2 3.7 2.8 2.3 1.8 1.6 1.5 1.o 0.9 0.8 0.7 0.4 1.3 0.7 1.1 1.3 1.2 1.o 1.3 0.9 1.1 1.2 1.4 0.9 0.6 0.4 1.7 1.5 0.8 1.0 0.6 1.o 0.9 1.7 0.9 1.6 1.o 1.5 0.9 1.5 0.8 1.4 0.3 0.9 1.3 0.8 0.5 0.9 1.5 0.6 0.7 0.3 1.0 1.4 0.1-m 0.5-283 1.6-14.7 0.2-176 0.7-25.0 0.4-21.3 0.4-13.9 0.7-4.9 0.5-5.7 0.2-13.5 0.3-3.3 0.4-1.9 0.2-2.3 0.3-1.7 0-2.6 0.6-2.7 0.3-1.5 0.5-2.7 0.3-6.6 0.4-3.5 0.5-1.9 0.6-2.9 0.2-3.3 0.3-3.9 0.4-4.3 0.4-5.5 0.5-1.9 0.3-1.1 0-4.6 0.6-5.1 0.8-3.0 0.2-2.6 0.1-6.5 0.2-1.6 0.42.2 0.4-2.2 0.9-3.1 0.2-5.3 0.6-4.0 0.42.3 0.1-22.6 0.2-5.6 0.46.0 0.3-1.7 0.45.0 0-1.6 0.41.9 0.44.5 0.2-3.2 0.1-2.1 0.3-2.7 0.4-6.6 0.1-2.7 0.2-2.1 0.1-0.99 0.3-3.2 0.5-4.2 1IO 411 1816 211 513 u2 414 1417 716 313 616 21/22 519 11117 215 14/16 14/25 13/17 313 9/7 24131 15/21 518 517 717 516 22/24 15/31 141 712 38/32 318 4/10 413 14/19 11/19 39/36 4/3 14111 13/17 212 313 614 14/17 715 217 11/22 717 417 317 911 1 514 316 6111 5/14 519 1018 W 5.4 3.9 2.6 2.7 2.0 1.8 2.2 1.6 1.5 1.3 1.1 0.7 0.7 0.4 1.o 0.7 1.2 1.6 1.7 1.o 1.o 0.8 0.9 1.2 1.o 1.0 0.5 0.4 4.6 1.5 0.5 0.4 0.9 1.0 0.7 1.7 1.4 1.7 0.9 1.5 0.9 1.5 0.9 1.9 0.4 0.6 1.o 0.8 0.6 0.9 1.8 0.6 0.7 0.4 0.8 1.5 * Adjusted for sex, age (in 3 strata: < 60,60-69, and Z 70 years), and cigarette smoking (in three strata: 0-9, 10-39, and 3 40 pack-years). t AU exposed cases and controls were men. $ All exposed cases and controls were women. nec = Not elsewhere classified. 0.1-03 0.5-283 1.412.5 0.1-159 0.5-19.8 0.1-33.0 0.3-11.3 0.8-6.8 0.4-6.3 0.2-13.5 0.3-5.1 0.5-2.4 0.2-2.7 0.3-1.6 U-2.9 0.4-2.4 0.3-1.4 0.5-2.9 0.2-12.5 0.5-5.9 0.5-2.0 0.42.3 0.2-3.2 0.2-3.6 0.4-4.3 0.2-4.3 0.5-2.2 0.2-1.1 0-4.6 0.846.4 0.8-3.0 0.1-2.5 0-2.9 0.1-5.9 0.42.4 0.3-1.8 0.9-3.2 0.2-10.0 0.4-4.5 0.42.2 0.1-22.6 0.1-7.2 0.3-8.1 0.3-2.1 0.5-8.1 0-2.0 0.2-1.4 0.3-3.6 0.2-3.2 0.1-2.6 0.3-2.7 0.4-6.3 0.1-2.7 0.2-2.1 0.1-1.1 0.2-3.0 0.5-4.8 transitional, and squamous cell carcinomas all arise from the transitional cell line, and made up 97% of the tumours, no separate analysis by histology was performed. Additional adjustment for consumption of coffee, tea, and diet soft drinks, and a history of bladder infections, abdominal injuries, and chemotherapy did not change results sufficiently to alter the interpretation of the occupational analyses. Discussion NASAL CANCER Increased risks have been measured in textile workers elsewhere.2czT8 he specific exposure of interest is not known, although fabric dust has been considered a possibility.25Certain fibres are made from plant materials (cotton, linen, rayon), and may produce exposures similar to furniture workers and cabinet makers. Another similarity is the potential for concurrent exposure to formaldehyde. This chemical is used in permanent press applications and is considered probably carcinogenic by LARC." Sewing machine oils may also be a carcinogenic exposure. Among four cases who had worked in the pulp and paper industry, three were known to have been maintenance workers and were likely to have welded or brazed stainless steel. Exposures to nickel and hexavalent chromium are likely in these operations; these metals have Surveillance of nasal and bladder cancer to locate sources of exposure to occupational carcinogens 449 been consistently associated with increased Hairdressers and barbers have shown in- risks of nasal cancer and are considered to be creased risks of bladder cancer in numerous known carcinogens.I6Cohort studies of pulp Stu&eS.R38 39 4247 The IARC recently classified and paper mill workers in the United States the occupations, hairdresser and barber, as and Finland have not found nasal cancers, probably carcinogenic to humans.15The sus- although these cohorts and the expected pected carcinogenic exposures include hair number of cases of nasal cancer (based on dyes, as well as coal tar derivatives and oils.l5 population rates) were quite small.2931 Laundry personnel have been shown to have Although exposures to chromium, nickel, increased risks in many epidemiological and formaldehyde were queried in the expo- stu&es.3944 47 4n The chlorinated aliphatic hy- sure section of the questionnaire, they were not drocarbons, including carbon tetrachloride and recalled by the pulp and paper maintenance perchloroethylene, have been postulated as worker or the textile worker cases. This is likely aetiological agents; carbon tetrachloride was to be because these agents are constituents of classified as possibly carcinogenic and perchlo- composite materials and therefore not known roethylene as probably carcinogenic by to employees who are IARC. 49 Several other occupational groups identified Increased risks of bladder cancer among in this study have had increased risks of nasal machinists have been reported else- cancer reported in the scientific literature: 4 0 4 2 44 50 The carcinogenic exposures of miners, drillers, and blaster~;'27~ machinists cutting and grinding metals may plumbers;26 painter^;'^ 33 and machinists.26" include polycyclic aromatic hydrocarbons in We found no associations with occupations used mineral oils and aromatic amines in exposed to wood dust or wood, in contrast to synthetic cutting oils." reported results." The woods used in this Occupational groups with increased relative region are almost exclusively SoftWOods, which risks, but without clear patterns of job duties or have been associated with increased risks of exposures in this study have also been reported squamous cell carcinomas of the nose and previously in the epidemiological literature as nasal cavities in previous investigations." 34 35 Earlier studies conducted in British Columbia and nearby Washington State accrued cases from 1939 to 1977 and 1979 to 1987 being associated with bladder cancer: wtpraooilrnektueerrmss;;89w31o60r33k96e44r1sW.4834-145401 513-6533t7aa3in9l4od4r5sc0 h52e-5m4ainPcdaailnatentrdesxuptiselee- re~pectively".~T~he different result found in various pigments, extenders, and solvents; this this study may reflect evidence that more job has been classified as carcinogenic to recent exposures are relatively low (means <1 humans by IARC,55largely based on stronger mg/m3) in the major wood industries of this evidence for lung cancer than bladder cancer. region (construction carpentry and production Tailors and textile workers may also have expo- of primary products from 37 sure to pigments and dyes. Several occupations No associations were found with other occu- with exposures to inks, aromatic amines, poly- pational groups originally suspected of being at cyclic aromatic hydrocarbons, and coal tars high risk. Increased risks of nasal cancer have were included in the occupational group been found elsewhere among nickel refining chemical and petroleum workers: printers, hot workersI6;however, the smelters and foundries tar roofers, rubber processors, coal workers, in British Columbia are not involved in nickel and oilfield manual workers. refining or alloy production. Chemical and Three other occupations with increased risk biological laboratory personnel may have exposures to f0rrna1dehyde.I~Although the OR estimate for the occupational group was not increased, both laboratory workers with nasal cancer reported this exposure, whereas half the controls did not. Boot and shoe manufacture and repair have been classified by LWC" as carcinogenic to humans. The lack of association in this study may reflect the fact that shoe estimates in this study have had fewer reports of positive associations with bladder cancer. Increased risks have been detected among plumber^,'^ 3' " sheet metal and gar- deners or nurserymen.4n42 Several occupations which were originally suspected of being at high risk had no association with bladder cancer in this study. Consistently increased risks of bladder cancer and leather work are not usual occupations in have been found for aluminum production; the British Columbia; only one of the controls had process has been classified by IARC as this occupation in this province. carcinogenic to humans." The group smelter and foundry workers included one case and three controls who had worked at an aluminum BLADDER CANCER smelter. The lack of an association in this study Increased risks of bladder cancer among may be related to the case definition (see limi- miners have been reported elsewhere.' 3843 tations). I Diesel equipment is commonly used in above Motor vehicle operators have often had ground and underground operations, as is increased risks of bladder cancer reported, drilling equipment with cutting oils (Tom although relative risks have generally been Carter, British Columbia Ministry of Energy, low.37 38 4Q 42 44 50 51 53 54 56 Our results are consist- Mines, and Petroleum Resources, personal ent with several other studies which have found communication, 8 August 1994). Diesel ex- no excess risk.' 26 41 53 A possible explanation for haust has been designated as probably carcino- the lack of association in this study may be the genic by IARC, and mineral oils have been rural and wilderness settings for much western designated as known human carcinogen^.'^-^' Canadian vehicle traffic, unlike the urban Y3V , settings which may predominate elsewhere. (suspect exposures: chromium and nickel from Most studies which have examined vehicle stainless steel welding or brazing in mainten- mechanics as an occupational category have ance jobs). These jobs were held by 10 cases, found increased risk^.'^ 40-43 54 " Our study had representing 21 % of the cases of nasal cancer. results similar to others which did 53 None had been recognised previously in Although cooks and kitchen workers have compensation claims or penalty actions by the shown increased risks in some studies,' 42 43 our local regulatory agency despite evidence link- results are consistent with those from several ing the job or its exposures to nasal cancer in other recent studies which found no the scientific literature. association." 44 45 Leather workers have shown Four occupations had increased risks of increased risks of bladder cancer in many bladder cancer and showed shared patterns of 96 40 42 43 51 56 although several reports job duties or exposures among three or more have shown no excess.' '7 44 As for nasal cancer, cases: mining (suspect exposures: diesel ex- the lack of association for shoe and leather haust and oil mist); machining (suspect expo- workers may simply reflect the paucity of sures: polycylic aromatic hydrocarbons and leather working occupations within British aromatic amines in cutting oils); hairdressing Columbia. (suspect exposures: hair dyes as well as coal tar derivatives and oils); and drycleaning (suspect LIMITATIONS exposures: volatile chlorinated aliphatic hydro- The main limitation of this study was its small carbons). These jobs were held by 33 cases, size, resulting in imprecise OR estimates. representing 31% of the cases of bladder Looking for patterns of duties or exposures cancer. Two of these occupations had not been among at least three cases in an occupational previously recognised by the local regulatory group, as well as increased relative risks, acted agencies, in either penalty actions or compen- as a check for selecting occupations of interest sation claims: hairdressers, and machinists. for surveillancefollow up. The value of disease surveillance in this Grouping of jobs meant that people with dif- study will only be known after workplace follow ferent work duties and exposures were in- up determines whether exposures to carcino- cluded in the same occupational group, usually gens continue in the targeted occupations. The expected to result in non-differentialmisclassi- responsible British Columbia regulatory agen- fication and bias of risk estimates to the null.57 cies, as well as occupational hygienists in the This may have been a special issue for trades pulp and paper industry have been informed of occupations the exposures of which would also the study results to target exposure measure- be influenced by their industry of employment. ments and control strategies. Some of the As a check, we looked for patterns of exposure targeted occupations are found in small shops among occupations with increased ORs, and if which are unlikely to employ professional they suggested an industry of interest, all the health and safety personnel. Without attention job listings for that industrial code were exam- from regulatory agencies, exposures to carcino- ined descriptively. gens in these work sites are unlikely to be rec- This study excluded carcinomas in situ. The ognised. natural history of many bladder cancers may include a prolonged period (years) of transitional cell carcinoma in situ before development of invasive carcinoma. Difficulty in classifying a tumour as invasive or in situ was responsible for most of the changes in histological eligibility of bladder cancer in this We thank the following people for their assistance during the design, conduct, or analysis of this study: Jody Slakov, Dr Paul Demers, Ruth Herder, Dr Clyde Hertzman, Bob H i d e , Dr Steve Marion, Dr Anne Milner, and Hazel Sarginson. We are grateful to all the cases, controls,and next of kin who gave their time to take part in this study. The study was supported by the British Columbia Health Research Foundation. study. Some epidemiologists have recommended that in situ carcinomas be included in case definitions'' 51; however, comparisons of occupational associations with invasive and in situ cancers in the United States national bladder cancer study indicated that risks were very similar.58Therefore exclusion of in situ tumours should not bias results, except when at risk occupational groups are screened to allow early detection. This may have been the case for aluminum smelter workers in British Columbia, who were made aware of their risk in a recent cohort study." SURVEILLANCE FOLLOW UP Results from these studies were used to select occupations which warranted follow up investigations in British Columbia work sites. Two occupational groups with increased risks of nasal cancer and three or more cases with similar duties or exposures were selected: textile workers (suspect exposures: formaldehyde, oils, fabric dust); and pulp and paper workers 1 Registrar General. Occupational monaliry, decennial supplementfor Great Britain, 1982-3. London: Her Majesty's Sta- tionery Office, 1986. 2 Rutstein DD, Mullan RJ, Frazier TM, Halperin WE,Melius JM? Sestito p. Sentinel health events (occupational): a basis for physician recognition and public health surveil- lance.Am 3l'ublic Hedth1983;73: 1054-61. 3 Matte TD,Baker EL, Honchar PA. The selection and defi- nition of targetted work-related conditions for surveillance under SENSOR. Am 7 Public Health 1989;79(suppl):21-5. 4 WeinsteinAL, Howe HL, BurnettWS. Sentinel health event surveillance:skin cancer of the scrotum in New York State. Am J Public Health 1989;79:1513-5. 5 Paci E, Dlni S, Buiatti E, Costantini A S , Lenzi S, Zappa M. Malignant mesothelioma in non-asbestos textile workers in Florence.Am3lnd Med 1987;11:249-54. 6 Veys CA. Bladder tumors and occupation:a coroner's notification scheme. BrJ Ind Med 1974;31:65-71. 7 TeschkeK, Morgan MS, Checkoway H, Franklin G, Spinelli JJ, van Belle G, Weiss NS. Mesothelioma surveillance to locate sources of exoosure to asbestos. Cut1 7 Public Health 1997; (in press). 8 Anthony HM, Thomas GM. Tumors of the urinarybladder: an analvsis of the occuoations of 1030 Datients in Leeds, EnGankr Natl Cancerhrt 1970;45:879:95. 9 Cole P,Hoover R,Friedell GH. Occupation and cancer of the lower urinary tract. Cancer 1972;29:1250-60. 10 Morrison AS, Ahlbom A, Verhoek WG, Aoki K, Leck I, Ohno Y, Obata K. Occupation and bladder cancer in Boston, USA, Manchester, UK, and Nagoya, Japan. Epiukmwl Communtry Heulzh 1985;39:294-300. 1 1 McDonald AD, McDonald JC. Malignant mesothelioma in North America. Cancer 1980;46:1650-6. I r I. Surveillance of nasal and bladder cancer to Iocare sources of exposure IO occupational carcinogens 451 12 International Agency for Research on Cancer. Monographs on rhe evaluation of rhe carcinogenic risk of chemicals IO humans. over011 evaluarions of carcinogenicity:an updating of LARC monographs.Vols 1 to 42 (suppl 7). Lyon: IARC Sci Publ, 1987. 13 International Agency for Research on Cancer. Monographs on rhe evaluarion of carcinoEenic risks IO humans. W,-d dusz and formaldehyde. Vol 62. Lyon: IARC Sci Publ, 1995. 14 Internauonal Agency for Research on Cancer. Monographs on rhe evaluarion of carcinogenic risks IO humans. Diesel and gasoline engine exhausrs and some nirroarenes.Vol 46. Lyon: IARC Sci Publ, 1989. 15 International Agency for Research on Cancer. Monographs on the evaluation of carcinogenic nsks IO humans. Occupariaal exposures of hairdressers and barbers and personal use of hair colorants;some hair dyes, cosmetic coloranrs, indusrrial dyesruffs and aromaric amines.Vol 57. Lyon: IARC Sci Publ, 1993. 16 International Agency for Research on Cancer. Monographs on rhe evaluation of carcrnogenic risks w humans. Chromium, nickel and welding.Volume 49. Lyon: IARC Sci Publ, 1990. 17 International Agency for Research on Cancer. Cross index of momrms and made names in volumes 1 IO 26. Monopraohs on he kaluation of rhe carcinogenic risk of chemicals himans (sup@ 3). Lyon: IARC Sci Publ, 1982. 18 Gosselii RE. Smith RP. Hodce HC. Clinical w x i c o l o ~o~f commercial producrs. 5;h ed; Baltimore: Williams -and Wilkins, 1984. 19 Statistics Canada. Srandard indusm'al classification.Ottawa: Minister of Supply and Services, 1980. 20 Statistics Canada. Srandard occupational classification. Ottawa: Minister of Supply and Services, 1980. 21 Breslow NE, Day NE.Srarictical method in cancer research. Lyon: IARC Sci Publ, 1980:140. 22 Mehta CR, Pate1 NR, Gray R. Computing an exact confidence interval for the common odds ratio in several 2x2 contingency tables. JOUIW~ of rhe American Srarizrical Association 1985;80:969-73. 23 Sieber WK, Sundin DS, Young RO. Development of a job exposure matrur. The changing nature ofwork and workforce, ' rhird 7mnr US - Finnish Science S-ymp_osium-, Frankfort. KenNcky: October 22-24,1986. 24 Acheson ED, Cowdell RH,Rang E. Adenocarcinoma of the nasal cavirv and sinuses in Enn-land and Wales. Br 7 Ind Med 1972>9:21-30. 25 Brinton L,Blot W, Fraumeni J. Nasal cancer in the textile and clothing indusmes. BrJ Ind Med 1985;42:469-74. 26 Maker H, McLaughlim J, Blon W, WeinerJ, Maker B, Ericson J, Stone B. Nasal cancer and occupation in Sweden, 1961-79. AmJIndMed 1986;9477-85. 27 Comba P, Bamsta G, Belli S,de Capua B, Merler E, Orsi D, ez al. A case-control study of cancer of the nose and paranasal sinuses and Occupational exposures. A m J Ind Med 1992;22:511-20. 28 Luce D, Leclerc A, Morcet J-F, Casal-Lareo A, Gerin M, Brugkre J, et nl. Occupational risk factors for sinonasal cancer: a case-control studv in France. Am 7 Ind Med 1992;21: 163-75. 29 Henneberger PK, Ferns BG, Monson RR. Mortality among pulp and paper workers in Berlin, New Hampshire. BrJInd Med 1989;46:658-64 30 Jappmen P; Hakulmen T, Pukkala E, Tola S, Kurppa K. Cancer incidence of workers in the F h s h uulu and DaDer industry. ScandJ mrk E m n m H d r h 1987,131197-iOi. 31 Robinson CF, Waxweiler RJ, Fowler DP. Mortality among production workers in pulp and paper mills. ScandJ Wrk E n v i m Healrh 1986;12:552-60. 32 Teschke K, Kennedy SM, Olshan AF. Effect of different questionnaire formats on reporting of occupational exposures. A m y Ind Med 1994;3:327-37. 33 Luce D, GCrin M, Leclerc A, Morcet J-F, Brugere J, Goldberg M. Sinonasal cancer and occupational exposure to formaldehyde and other substances. Inr J Cancer 1993;53: 224-3 1. 34 Vauhan T. Scott D. Wood dust emosure and sauamous cell cancers of the upper respiratory tract. Am ' J Epidemiol 1991:133:560-4. 35 Elwood JM. Wood exposure and smoking: association with cancer of the nasal cavity and paranasal sinuses in British Columbia. Can MedAssocJ 1981;1241573-7. 36 Teschke K, Hertzman C, Morrison B. Level and distribu- tion of employee exposures to total and respirable wood dust in two Canadian sawmills.Am Ind Hyg Assoc J 1994; 59245-50. 37 Veda1 S, Chan-Yeung M, Enarson D, Fera T, Maclean L, Tse KS, Langille R. Symptoms and pulmonary function in western red cedar workers related to duration of employ- ment and dust exposure. Arch Emron Healrh 1986;41: 179-83. 38 Wynder E, Onderdonk J, Mantel N. An epidemiological investigation of cancer of the bladder. Cancer 1963;16: 1388-407. 39 Silverman D, Hoover R, Albert S, Graff KM. Occupation and cancer of the lower urinary Wac1 in Detroit. J Narl Cancer Insr 1983;70:237-45. 40 Brownson R, Chang J, Davis J. Occupation, smoking, and alcohol in the epidemiology of bladder cancer.Am J Public Health 1987;77:1298-300. 41 D o h P. A descriptivestudy of occupation and bladder cancer in England and Wales. BrJ Cancer 1992;65:476-8. 42 Howe GR, Burch JD,Miller AB, Cook GM, Morrison B, Gordon P, ez al. Tobacco use, occupation, coffee, various numents, and bladder cancer. J Nad Cancer Insr 1980;64: 701-13. 43 Cordier S, ClavelJ, Limasset J-C, Boccon-Gibod, Le Mom1 N, Mandereau L,HCmon D. Occupational risks of bladder cancer in France: a multicentre case-control study. Inr J Epidemiol 1993;22:403-11. 44 Silverman DT, Levin LI,Hoover RN,Hartge P. Occupa- tional risks of bladder cancer in the United States: I White men.3 Natl Cancer Insr 1989;81:1472-9. 45 Schoenberg JB, Stemhagen A, Mogielnicki AI', Alman R, Abe T, Mason TJ. Case-control study of bladder cancer in New Jersey. 1 Occupational exposures in white males. J Nad CancerInst 1984;72:973-81. 46 Claude J, Frenael-Beyme R, Kunze E. Occupation and risk of cancer of the lower urinary tract among men. A case-control study. Inr J Cancer 1988;41:37 1-9. 47 Silverman D, Levin L, Hoover R. Occupational r i s k s of bladder cancer among white women in the United States. AmfEpidemw11990;132;453-61. 48 Weiss NS. Cancer in relation to occupational exposure to perchloroethylene.Cancer Causes Connol1995;6257-66. 49 International Agency for Research on Cancer. Monographs on rhe evaluation ojcarcinogenicrisks w humans. Drydecming, some ChIorinated solvenu, and orher industrial chemicals. Vol 63. Lyon: IARC Scientific Publications, 1995. 50 Schiwcrs E, Jamart J, Renard V.Tobacco and occupation as risk factors in bladder cancer: a case-control study in southern Belgium. Inr J Cancer 1987;39:287-92. 51 Jensen OM, Wahrendorf J, Knudsen JB, Sorensen BL. The Copenhagen case-referent study on bladder cancer. Scand 3 Envim Healrh 1987;13:129-34. 52 Burns P, Swanson G. fisk of urinary bladder cancer among blacks and whites: the role of cigarette use and occupation. Cancer Causes Conno1 1991;2:371-9. 53 Risch H, Burch J, Miller A, Hill G, Steele, Howe G. Occu- pational factors and the incidence of cancer of the bladder in Canada. BrJInd Med 1988;45:361-7. 54 Bonassi S, Merlo F,Pearce N, Puntoni R. Bladder cancer and occupational exposure to polycyclic aromatic hydrocarbons. InrJ Cancer 1989;44:648-51. 55 International Agency for Research on Cancer. Monographs on rhe evaluorirmof carcinogenic risks w humans. Some organic solvenrs, resin monomers and related compoundc, p'gmentc and occupational exposures in paint manufactureand painring. Vol 47. Lyon: IARC Sci Publ, 1989. 56 Vieis P, Magnani C. Occupation and bladder cancer in males: a case-control study. InrJ Cancer 1985;35:599-606. 57 Copeland KT, Checkoway H, McMichael AJ, Holbrook RH.Bias due to misclassification in the estimation of rela- tive risk. AmJEpidemiol1977;105:48&95. 58 Sturgeon SR, Hartge P, SilvermanDT, Kantor AF, Linehan WM, Lynch C, Hoover RN.Associations between bladder cancer risk factors and tumor stage and grade at diagnosis. Epidemiology 1994;5:218-25. 59 Spinelli JJ,Band P, Svirchw L, Gallagher R. Mortality and cancer incidence in aluminum reduction plant workers. 3 Occup Med 1991;33:1150-5.