Cohort methods: the CANadian RADiation workers (CANRAD) cohort

The long-term follow-up of workers with protracted exposure to low dose occupational ionizing radiation is important to understand the health effects and risks of radiation. This paper describes the creation and record linkage of the CANadian RADiation workers (CANRAD) cohort, including the work to resolve data issues and linkage challenges to outcome databases.

The CANRAD cohort includes all workers with occupational exposure to ionizing radiation (1951-2018) in Canada's National Dose Registry (NDR) and three facility-based sub-cohorts: Eldorado uranium workers (1932-1980), Ontario uranium miners (1954-1996), and Atomic Energy of Canada Limited nuclear workers (1956-1985). The sub-cohorts were added to the NDR to update workers' employment histories and radiation exposures to 2018. Deterministic and probabilistic record linkage methods linked the CANRAD cohort to national population-based mortality (1950-2020), cancer incidence (1969-2019), and tax (1981-2020) data.

Of 957,580 CANRAD workers, 883,150 (92.2%) successfully linked to Statistics Canada vital statistics databases. The analytical cohort includes 860,650 (89.9%) workers, including 427,920 (49.7%) females. At the end of follow-up, 751,520 (87.3%) of the cohort were alive and 94,410 (11.0%) were dead. Neoplasms and circulatory disease represented 63.8% (57,030) of all known causes of death (89,330). Prostate, breast, lung & bronchus, and colorectal cancer represented 49.9% (40,780) of first primary cancers (81,670). The mean cumulative effective dose was 17.1 mSv; but most cumulative doses were <10 mSv (∼85.0% of males and 98.1% of females). Females (350,510, 82.4%) dominated the medical sector.

The CANRAD cohort is a large and valuable cohort of Canadian occupationally exposed radiation workers with employment histories, radiation exposures and long-term follow-up of tax, mortality, and cancer incidence. Analysis of this cohort will support the radiation protection framework.