ABSTRACT
There are many aspects of the design of a medical research programme that need to be considered when the response variable of interest is a survival time. These include factors such as the inclusion and exclusion criteria for study participants, the unambiguous denition of the time origin and the end-point of the study, and the duration of patient follow-up. In a clinical trial, the specication of treatments, the method of randomisation to be employed in allocating patients to treatment group, and the use of blinding must also be specied. Consideration might also be given to whether the study should be based on a xed number of patients, or whether a sequential design should be adopted, in which the study continues until there is a sucient number of events to be able to distinguish between treatments. The need for interim analyses, or adaptive designs that allow planned modications to be made to the sample size or allocated treatment as data accumulate, also needs to be discussed.
