The design of adaptive clinical trials

Clinical trials are research studies which involve both healthy people and patients. The aim of these trials is to assess the efficacy of a new treatment or to compare the efficacy of new treatments with the current treatment. In this thesis, we focus on comparing the efficacy of a new treatment with that of an existing treatment. The objective of a clinical trial design is to obtain a correct conclusion as well as to be concerned with the economic issues. Since this trial involves human beings, importantly, it should address ethical concerns. In a traditional (equal randomisation, ER) design, half of the participants will be assigned to an inferior treatment. When comparing the efficacy of treatments, one of the disadvantages of the ER design is that although the evidence may be strong that one treatment is superior to others, a trial cannot finish early. As a result of the issues discussed above, ethical problems and economic issues arise. In order to cope with these problems, an adaptive design should be considered. An adaptive design is one in which a trial can be changed (adapted) during its progress. These changes are based upon the accrued data. In this thesis, the designs developed involve two specific areas of adaptive designs: adaptive randomisation and interim analyses. Particular attention is paid to response-adaptive and covariate-adaptive randomisations.