Clinical trials are a crucial step in translating fundamental medical research into improved healthcare. Many hundreds of trials are conducted every year, each involving hundreds, sometimes thousands, of patients. But they are expensive to conduct, and making changes to a trial while preserving statistical validity is difficult.
Research carried out at the University has helped to make decisions on when to stop a study, and to allow a broader range of adaptations to be made during the course of a trial. The results of this research have made clinical trials faster and more efficient while maintaining safety.
Clinical trials are expensive to conduct, with costs as high as £30,000 per patient. Results can be slow to obtain, and it can be challenging to get large enough sample sizes. There can be good reasons to make changes to the trial once it is underway, or to stop a trial early with a positive result and so bring a new medicine to market sooner. However, it is essential to account for such adaptation properly so that the conclusions remain statistically valid.
Research carried out at the University has developed sequential methods which are used to monitor trials and make decisions on when to stop them early, and designs for adaptive clinical trials which allow a broader range of decisions to be made during the course of a trial, such as treatment selection or a re-definition of the target population. These research insights are summarised in the book Group Sequential Methods with Applications to Clinical Trials by Christopher Jennison and Bruce Turnbull (Chapman and Hall / CRC Press, 1999), which has become the standard text on the topic and is widely used by medical statisticians.
Establishing better ‘stopping rules’ for the termination of a clinical trial benefits huge numbers of future patients by making effective new treatments available sooner, or by halting a poorly performing treatment early to release resources for studies of other promising therapies.