Quantifying the Invisible
Risk characterization is the final step of risk assessment — combining hazard identification, dose-response analysis, and exposure assessment into a single number: the probability that exposure will cause harm. For carcinogens, this probability is expressed as excess lifetime cancer risk; for non-carcinogens, as the hazard quotient. These numbers drive billion-dollar cleanup decisions, chemical regulations, and public health policy.
The Linear Low-Dose Model
EPA's cancer risk assessment assumes no safe threshold for carcinogens — any exposure carries some risk. At low doses, the dose-response is assumed linear, allowing simple multiplication: Risk = LADD x SF. This conservative assumption, derived from the linearized multistage model, may overestimate risk but is designed to protect public health. The slope factor represents the upper 95% confidence bound on carcinogenic potency.
Acceptable Risk
No human activity is risk-free, so regulators must define 'acceptable' risk levels. EPA's cancer risk range of 10⁻⁶ to 10⁻⁴ balances protection against practical feasibility. A 10⁻⁶ risk (one excess cancer per million exposed) is roughly equivalent to the risk of being struck by lightning. At 10⁻⁴, the risk becomes comparable to other accepted voluntary risks like recreational activities.
From Numbers to Action
Risk characterization informs but does not dictate decisions. Risk managers weigh the calculated risk against costs, technical feasibility, community values, and environmental justice concerns. A 10⁻⁵ cancer risk might trigger different responses depending on whether it affects a wealthy suburb or a disadvantaged community. Uncertainty analysis — characterizing what we don't know — is equally important as the point estimate, because conservative assumptions compound through the assessment chain.