Comments on the NCRP Statement

Abstract

It is reassuring to know that NCRP continues to support the notion of Probability of Causation (PC) as a means of assessing whether a specified malignancy occurring in an individual was caused by a specific previous exposure to ionizing radiation. So long as we have no medical means of determining the cause of a cancer that has been diagnosed in someone, the concept of PC represents the best available basis for attributing causation. As the authors and others (1) note, however, there may be considerable uncertainty involved in making a PC calculation because of model assumptions, appropriateness of the population from which epidemiological information was obtained, and the sparsity of data. Thus, in using PC tables, it is important to attempt to document the various sources of uncertainty and, ideally, give an idea of how much these might affect the resulting calculations. I wish to raise three rather technical points on which I disagree with the comments in NCRP Statement No. 7. 1. In its most general form, the formula for PC is not that shown on page 2 of the report. Rather, it is (2)