Many oncology studies have time-to-event endpoints of interest, such as overall or progression-free survival. Frequently, time-to-event endpoints are subject to censoring, where an event time is only known to have occurred in a particular interval, mostly commonly right-censoring occurs, where it is only known that the event has not occurred up to some particular time”. In general, specialized techniques are required for the appropriate estimation and comparison of quantities related to time-to-event data, such as force of mortality, median time-to-event, or long-term survival probability.
This chapter provides an introduction to several of the most common problems related to the analysis of time-to-event data, and outlines a few typical approaches to handling these problems, as well as related analytic techniques. Topics discussed include time-to-event data and right-censoring, estimation of the survival function and related quantities, the log-rank test and Cox proportional hazards modeling, and power and sample size when the hazard ratio is the metric of comparison. An example “toy” randomized controlled trial example is used throughout the chapter to illustrate the techniques, as well as the interpretation of results and the output of statistical software.
Log-rank test Cox proportional hazards Kaplan-Meier Right-censored data Survival analysis Time-to-event data
This is a preview of subscription content, log in to check access
Kleinbaum DG, Klein M. Survival analysis: a self-learning text. New York: Springer Science & Business Media; 2006.Google Scholar