Comparing Treatments, Adjusting for Competing Risks

摘要: Often some concomitant information about the experimental units in a study accompanies usual survival time information. This may identify with two or more treatment groups (in which case, indicator variables are usually employed) and/or enter certain well-defined characteristics need to be considered evaluating functions. paper is concerned incorporating such into one’s analysis when there k(>1) competing causes of failure. The theoretical lifetimes associated each cause assumed independent. Two different approaches - viz., i) assuming hazard rates arbitrary and ii) underlying life distributions completely specified apart from unknown parameters. methods discussed this treated detail monograph by David Moeschberger (1978).