A Perspective from US Environmental Protection Agency (EPA) Scientists: How Your Epidemiologic Analyses Can Inform the Human Health Risk Assessment Process.

Abstract

Background: Environmental epidemiologists strive to conduct research that will lead to actions that improve public health outcomes. The risk assessment process is the bridge between scientific research and policies that can impact public health. Historically, epidemiologic studies have not frequently been used to inform US Environmental Protection Agency (EPA) assessments outside of the context of air pollution. There are certain practices that the epidemiology community can adopt to facilitate the integration of epidemiologic studies into policy-relevant assessments.

Objectives: The central objective of this commentary is to provide guidance to epidemiologists that will enhance the value of their studies for US EPA assessments. First, we provide an overview of the US EPA dose-response and toxicity value derivation to increase literacy about these processes across the environmental epidemiology community. Second, we provide suggestions for modeling and reporting to facilitate the use of epidemiologic studies in US EPA dose-response assessments that form the basis for decision-making.

Discussion: Epidemiologic research can be used in all aspects of dose-response assessment, which involves identifying a point of departure followed by specific adjustments and extrapolations to identify a toxicity value intended to prevent adverse effects across the population. To facilitate the integration of epidemiologic research into the dose-response assessment process, we provide specific recommendations for additional modeling (e.g., modeling in the low exposure range; exploring nonlinearity) and reporting (e.g., sufficient information to conduct study evaluation; more details on exposure levels in the population) in published epidemiologic research. Many of these suggestions require only additional reporting in the final manuscript or associated appendixes but would have substantial impact on the contribution of the published work to the assessment process. https://doi.org/10.1289/EHP15203.