Abstract

Doctors need reliable information about the efficacy and hazards of treatments which they can apply to their own practice. This is what a good clinical trial supplies. Unfortunately, despite the increasing rigour of editors and referees, published clinical trials often have faults in design and analysis which make their conclusions unreliable and sometimes misleading.^1,2 Most doctors have to rely on reading review articles or promotional literature, rather than the original reports. They may then be misled, as some reviewers may unwittingly base their conclusions on invalid trials, and some manufacturers seem to do so almost when it suits them.