Assessing excess lifetime risk for disease after radiation exposure.

In this paper, we define the excess lifetime risk for a disease due to radiation exposure, and we evaluate the excess lifetime risk for an exposed group with the same age and with a mixture of ages. We propose two methods to estimate the excess lifetime risk: the standard maximum likelihood approach and a Monte-Carlo approach. Both approaches allow us to construct confidence intervals for the excess lifetime risk and perform hypothesis testing to compare two excess risks. The maximum likelihood approach is straightforward and computationally simple; however, it may not provide good approximations when the normality assumption is violated. The Monte-Carlo approach is always reliable, but often involves heavier computations and additional programming. These two approaches are illustrated and discussed based on a pooled analysis of five cohorts exposed to external radiation and followed up for subsequent thyroid cancer incidence.