The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
consider a variety of possible strategies for meeting goals, to favor actions that are robust to uncertainties, to favor actions that are informative, to probe and experiment, to monitor results, to update assessments, and to modify policy accordingly and favor actions that are reversible (Ludwig and others 1993).
To make an exposure assessment consistent with such an approach, both sensitivity and uncertainty analyses should be incorporated directly into an iterative process in which premises lead to measurements, measurements lead to models, models lead to better premises, better premises lead to additional but better-informed measurements, and so on. In 1996, the EPA Risk Assessment Forum held a workshop on Monte Carlo analysis. Among the many useful discussions at the meeting was a call for a "tiered" approach to probabilistic analysis, which is iterative and progressively more complex. The need for formal uncertainty analysis and a tiered approach will require the development by the exposure-assessment community of new methods and will put greater demands on the number and types of exposure measurements that must be made. At least three tiers are needed, as follows:
First, the variances of all input values should be clearly stated, and their effect on the final estimates of risk assessed. At a minimum, that can be done by listing the estimation error or the experimental variance associated with the parameters when these values or their estimation equations are defined. It would help to define and reduce uncertainties if a clear summary and justification of the assumptions used for each aspect of a model were provided. In addition, it should be stated whether the assumptions are likely to result in representative values or conservative (upper-bound) estimates.
Second, a sensitivity analysis should be used to assess how model predictions are affected by model reliability and data precision. The goal of a sensitivity analysis is to rank input parameters on the basis of their contributions to variance in the output.
Third, variance-propagation methods (including but not limited to Monte Carlo methods) should be used to map how the overall precision of risk estimates is tied to the variability and uncertainty associated with the models, inputs, and scenarios.
The Committee's Evaluation Of Uncertainties In Risk Assessment Of Radon In Drinking Water
Uncertainties in Molecular Biology of Cancer Induction by Radiation
As discussed in chapter 6, the exposure of human cells to the high-LET radiation from the decay of radon and its progeny initiates a series of events that can lead to lung and other cancers. This series of events is now thought to be well outlined, but the quantitative link between radon concentration in tissues and