ABSTRACT
This chapter covers analysis of trials with a survival endpoint, including censoring, the Kaplan-Meier survival curve, the logrank test, and the hazard rate and Cox proportional hazards model. Another variant of the ‘O minus E except after V’ principle provides a simple, accurate estimate of the log hazard ratio. We also discuss an alternative method of conducting survival analysis while accounting for competing risks, and we illustrate the idea through a trial of COVID-19. Lastly, we describe parametric survival methods, focusing on a popular method called the conditional binomial procedure. We apply this technique to a trial of a Moderna vaccine for COVID-19.
