Analysis of treatment effects and event rates that change over time in clinical trials

When designing and analysing clinical trials with time-to-event outcomes, the most commonly applied methods assume that treatment effects and event rates are constant over time and apply equally to all patients receiving the treatment. Newly developed treatments can have effects that vary in beneficial value over time, and can vary for different patient populations. This thesis illustrates the increased insight and clinical understanding that can be achieved through the use of new analysis approaches and my proposed presentation of complementary graphs should provide the impetus to more fully assess and report clinical trial findings.