Accounting for uncertainty in environmental health risk assessment

This dissertation explores several common sources of uncertainty in environmental health risk assessment and proposes new methods to account for uncertainty in a systematic manner. The relationship between prenatal polychlorinated biphenyl (PCB) exposure and early childhood cognitive development serves as an exemplary backdrop for studying uncertainty. Although the effects of large, toxic exposures to PCBs are well-documented, studies examining low-level, chronic PCB exposure have produced seemingly conflicting results. Researchers have proposed various explanations for observed discrepancies, citing several sources of uncertainty in estimating effects of PCBs on cognitive development. The methods described in this dissertation help reconcile past study inconsistencies and provide a framework to formally address uncertainty in both human and animal studies on the health effects of environmental pollutants. Such an integrated approach to uncertainty analysis is necessary to ensure that risk-managers have the means to weigh the strengths and limitations of available research and prioritize regulatory actions to effectively protect the public. The Benchmark Dose (BMD), a common regulatory risk measure, is utilized to demonstrate how quantitative risk assessment summary measures can be modified to explicitly account for uncertainty.