An overview of methods for studying events and their timing
Andrew R. Pickles, Bianca Lucia De Stavola · Oxford University Press eBooks · 2007
© Oxford University Press, 2014. All Rights Reserved. This chapter explains the concepts of censoring and time scales, and introduces proportional hazards, accelerated failure time, and proportional odds models. More complex models for competing events, for multiple events of the same and of different types, and for multistate processes are considered. Errors in the dating of events are discussed. For profiles of event occurrence over an extended period, the chapter illustrates a typological analysis. Finally, the problem of time-varying confounding is described and ways in which it can be addressed using marginal structural models and G-estimation considered.