Abstract

An incidence study of malignant melanoma of the skin (MMS), conducted previously among the workers of four plants of a large telecommunications industry located in Montreal, Canada, showed a standardised incidence ratio of 2.7 (95% confidence interval (95% CI) 1.3-5.02) for the years 1976 to 1983. To describe more precisely the magnitude of the problem a mortality study was started among the same population (n = 9590) for the same period (1976-83). At the end of 1983, 9180 workers were alive, 261 were dead, and 149 (1.5%) were not traced. Standardised mortality ratios (SMRs) for all causes of death were surprisingly low for men (SMR = 0.57; 95% CI 0.50-0.64) and women (SMR = 0.56; 95% CI 0.37-0.82). The SMRs for major causes of death were also less than expected. These results may be explained by a pronounced selection bias (healthy worker effect) and by the short duration of follow up (eight years). For MMS, two deaths occurred among men (SMR = 2.00; 95% CI 0.24-7.22) and one among women (SMR = 4.81; 95% CI 0.12-26.78). A third man who died of MMS was miscoded as having a primary pulmonary melanoma. Including this case increased the SMR for MMS to 3.00 (95% CI 0.62-8.77; p = 0.08). Polyvinyl chloride and polychlorinated biphenyls were used in the plants and some of the workers did soldering. A planned case-control study will investigate other possible exposures at work.