Psychological issues in clinical trial design

Randomised controlled trial (RCT) designs are widely regarded by the medical establishment as the trial design of choice, believed to offer greater internal validity than non-random trials. However, where patients have preferences among treatments to be compared, randomisation can create differences between groups in a trial. The limitations of conventional RCTs are considered in the context of treatments for chronic conditions where patients often have strong treatment preferences. Precautions required in selection and recruitment of patients into RCTs are recommended together with strategies for evaluating any effects of preferences. Alternative trial designs which take account of patients' and/or doctors' preferences when recruiting patients and allocating treatments are reviewed, including Brewin and Bradley's increasingly widely used partially-randomised preference trial (PRPT) design. Recommendations are made for future use of trial designs which take account of preferences and provide interpretable results of value to clinicians and patients in routine clinical practice.