Multivariate Survival Analysis

Abstract

In the previous chapters of this book, we have examined a variety of techniques for analyzing survival data. With few exceptions, these techniques are based on the assumption that the survival times of distinct individuals are independent of each other. Although this assumption may be valid in many experimental settings, it may be suspect in others. For example, we may be making inferences about survival in a sample of siblings or litter mates who share a common genetic makeup, or we may be studying survival in a sample of married couples who share a common, unmeasured, environment. A third example is where we are studying the times to occurrence of different nonlethal diseases within the same individual. In each of these situations, it is quite probable that there is some association within groups of survival times in the sample.