A Brief History of Time…at Risk: A Review of Competing Risk Methods. Or, Who’s Right, Who’s Wrong, and Who Cares?
Peter Rebeiro, PhD,MHS Vanderbilt University School of Medicine
Survival analysis may not be so straightforward if a study participant’s time at risk for the outcome can be truncated by some other event (often death). How do we deal with this sort of censoring? And why do we use the methods we do? I’ll review the motivation for thinking of censoring due to a competing event a little differently than other sorts of censoring, the parsing of risk sets in the presence of a competing event, and the assumptions underlying a few different competing risk methods. I’ll also provide a brief example of these methods applied in a few different contexts.