Registering clinical trials

摘要

Clinical trials should be registered so that essential details are made public from a trial's inception, rather than from publication many years later. Openness about trials in progress reduces the impact of publication bias, prevents duplication of effort, promotes collaboration, and can save lives. Hundreds of trial registers already exist, but the information on them is not standardised and is incomplete; most contain only a subset of trials, often in high profile areas such as cancer or AIDS; and there are few incentives for researchers to register trials. Initiatives have begun to unify the existing web of registers, but they are at an early stage. Registration of clinical trials should be compulsory, either by law or by linking it to ethical approval of research.