Arsenic in drinking water and bladder cancer: Comparison between studies based on cancer registry and death certificates

Associations between arsenic in drinking water and bladder cancer in an area along the southwest coast of Taiwan have been documented for decades. Several ecologic studies were conducted to assess the dose-response relationships. Some of them used the National Cancer Registry Program to identify cancer cases, and some used death certificates. Whereas the cancer registry collects information on all patients no matter if they died of bladder cancer or not, the case ascertainment might be incomplete due to the fact that reporting of cases is not mandatory. Reporting of death, on the other hand, is strictly enforced by law, but patients who did not die of bladder cancer might not be identified. In order to assess the problems with both approaches, we conducted a study using both case identification mechanisms. A total of 243 townships with measurements of arsenic in drinking water were included in the analysis of cancer registry data, and death certificates were collected from 10 of those townships. In both analyses, the same measurements of arsenic made by the mercuric bromide stain method were adopted. Due to limitation of the method, all levels below 0.04 mg L^-1 were combined as a single exposure category. The results were very much alike; both approaches detected statistically significant associations between high arsenic levels in drinking water (above 0.64 mg L^-1) and occurrence of bladder cancer but did not find such associations for arsenic exposures at lower levels.