Estimation of population cancer risks from the DS86 bomb survivor data

Abstract

The Japanese bomb survivor data, with the latest dosimetry (DS86) and recently extended follow-up, are analysed using a hybrid relative/absolute risk model. A number of ways are investigated of transporting the estimated low LET excess risk to the current England and Wales population, as well as to a UK radiation workforce. In one particular approach the authors follow the methodology of Land (1980, 1984) and BEIR III and transport the absolute excess risk. Since the base cancer rates in the youngest age group at the time of the bombing (those 20 years old or less at exposure) are only available on average up to the age of 50, the assumption has been made that the absolute excess risk in this category remains constant from the age of 50 until the end of life (for which there is some evidence in the latest RERF analyses). The resulting risk obtained for the England and Wales population is 0.05 excess cancer deaths/Gy (with a 95% confidence interval of (0.04, 0.06) excess deaths/Gy). Another approach adopted is to transport the relative excess risk, which yields a figure for the general population of 0.12 excess deaths/Gy (with a 95% confidence interval of (0.09, 0.15)).