Uncertainties in Recorded Doses in the Nuclear Industry: Identification, Quantification and Implications for Epidemiology Studies

Publisher Website
Google Scholar
Abstract
Although radiation exposures in the nuclear industry have been assessed much more precisely than exposure to most other occupational carcinogens, individual dose estimates are based on recording and monitoring practices which vary both over time and between facilities. The accuracy of these estimates is therefore a function of time, placed and radiation quality. Accuracy of estimates may not be essential in the first steps of epidemiological studies aimed at identifying the existence of a carcinogenic risk; it is, however, of prime importance when comparing risk estimates across studies and time, especially where the risk is expected to be small and the magnitude of the errors (often dose related) is large compared with that of the doses themselves. A retrospective study of monitoring and recording practices in nuclear facilities, currently underway as part of an IARC collaborative study of cancer risk among nuclear industry workers, is described. Sources of biases and errors, including dosemeter design, radiation energy and dose estimation, and steps taken to quantify biases and uncertainties are discussed. Recently proposed statistical methods for taking account of errors in dose estimates and assessing their effects on risk estimates are presented.

This publication has 0 references indexed in Scilit: