Reporting of clinical trials of drugs shows bias

Results from clinical studies sponsored by the pharmaceutical industry that reach the public domain are subject to bias, according to a study by Melander and colleagues (p 1171). They investigated 42 placebo controlled studies of selective serotonin reuptake inhibitors submitted to the Swedish drug regulatory authority in 1983-99 and compared the published results with the results in the full reports of the studies. A major cause of bias was the tendency to publish only the more favourable per protocol results, rather than intention to treat analyses. Other causes were extensive duplicate publication and a preference for publishing only studies showing significant effects of the drug.