Probabilistic risk assessment's use of trees and distributions to reflect uncertainty and variability and to overcome the limitations of default assumptions
Conference Paper
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
Probabilistic risk assessment is an emerging approach to exposure assessment and quantitative cancer and non-cancer risk characterizations. The approach is easily extended to other types of risks and outcomes. A tree, like a decision tree or probability tree, encourages the evaluation of not only the default assumptions but also alternatives to those defaults, and reflects the uncertainty in the current state of knowledge. Trees are used in both the characterization of the dose received by individuals in a potential exposure situation and the characterization of the dose-response relationship for a specified response of concern. Probability distributions are used to reflect the variability in exposure, dose, and dose-response relationships among individuals and over time within individuals. Distributions incorporating variabilities, uncertainties, subjective probabilities, and expert judgements are used to characterize the probabilities of observing an individual in a population with a specified dose from exposure, with a specified probability of a certain adverse health effect for a designated dose, and with a specified probability of a certain adverse health effect (i.e., a specified risk). Some suggestions are given on how a risk manager can incorporate a distributional risk characterization into decision making. Some discussion is included concerning sensitivity analyses and path analyses. The major finding is methodology to explicitly incorporate variability, uncertainty, and alternatives to defaults into exposure, dose-response, and risk characterizations.
