Reviewing the occurrence data used in the revised Arsenic Rule

Large sampling initiatives are integral to the process of developing regulations, yet best practices for the data acquisition and analysis required to quantify potential health risks posed by contaminants are not well-defined. Despite differing database sizes and analytical approaches, previous studies predicting water systems likely to be out of compliance with the maximum contaminant levels (MCLs) in the proposed revision of the Arsenic Rule were in close agreement with the predictions of the US Environmental Protection Agency's (USEPA's) Arsenic Occurrence and Exposure Database and the current study's analysis of the agency's second six-year review of enforced MCLs. The epidemiologic analyses used in the final revision of the Arsenic Rule cite the differences in proposed concentrations leading to a reduction in potential incidences of cancer; however, for this type of analysis, population-based occurrence data are necessary. When the data were analyzed on the basis of population served, the summary statistic generated an approximate 10% difference in the population exposed to arsenic concentrations > 10 mu g/L. The research described in this article furthers the discussion of what information is needed to accurately predict nationwide arsenic occurrence, exposure, and health outcomes.