TABLE OF CONTENTS In many clinical trials, the main response variable is often the time to the occurrence of a particular event. In a cancer study, for example, surgery, radiation and chemotherapy might be compared with respect to the time from randomisation and the start of therapy until death. In this case the event of interest is the death of a patient, but in other situations it might be the end of a period spent in remission from a disease, relief from symptoms, or the recurrence of a particular condition. Such data are generally referred to by the generic term survival data even when the endpoint or event being studied is not death but something else. Questions of interest for such data involve comparisons of survival times for different treatment groups and the identification of prognostic factors useful for predicting survival times. Since these do not differ from the questions usually posed about other response variables used in clinical trials, it might be asked why survival times merit any special consideration, in particular a separate chapter? There are a number of reasons. The first is that survival data are generally not symmetrically distributed they will often be positively skewed, so assuming a normal distribution will not be reasonable. A more important reason for giving survival data special treatment is the frequent occurrence in such data of censored observations. These arise because at the completion of the study, some patients may not have reached the endpoint of interest (death, relapse,...
