Linear Low Dose Extrapolation for Cancer Risk Decisions: Sources of Uncertainty and How They Affect the Precision of Risk Estimates
The Agency poses the following questions to the SAP regarding Linear Low Dose Extrapolation for Cancer Risk Decisions: Sources of Uncertainty and How They Affect the Precision of Risk Estimates.
The cancer risk assessment using the upper bound estimates contains inherent uncertainties due to the limitations of scientific knowledge, available data, or choice of models. As discussed in the book, Science and Judgment in Risk Assessment (National Research Council, 1994), some claim EPA's risk assessments reflect "cascading conservatism", layering an overly conservative assumption upon another. However, the Council also cautions against underestimating the true risk by stating,
"[T]his mathematical truism that the more uncertainty, the greater the level of conservatism required not to underestimate the mean,... If each of a series of uncertain quantities is distributed in such a way that a reasonable conservative estimator (say, the 95th percentile) approximates or even falls below the mean of that quantity, then the more steps in the cascade the less conservative the output becomes with respect to the correct risk-neutral estimator". (p. 608 in Science and Judgment in Risk Assessment)
- Does the presentation of data in the Examples in this
paper adequately define the potential impact of uncertainties
on the risk? Given that the uncertainties may be qualitatively
defined with respect to the direction and magnitude of the risk
estimate, but may be unquantifiable, is the presentation of the
uncertainties in the example risk characterizations adequate.
- Could the Panel comment of the appropriateness of presenting
the rounded numeric cancer risk estimate (e.g. 10 -6) as a tool
to the risk manager as part of the risk characterization?
- Could the Panel give guidance on how uncertainty should be handled
in the risk assessment?
- Could the Panel give guidance on alternative methods of expressing cancer risk estimates (e.g. rounding to half orders of magnitude; 10-6; 10-6.5; 10-7)?
Questions for Session 2: