Reducing publication bias of prospective clinical trials through trial registration.

Publication bias is the selective publishing of favorable or statistically significant results. This practice, over time, distorts the medical literature by depicting inordinately optimistic outcomes for treatments and interventions. Sources of publication bias include preferential publishing by journals and preferential submission by researchers. Mandatory trial registration, as instituted by the International Committe of Medical Journal Editors (ICMJE), should reduce publication bias by improving the ability to identify all trials pertaining to a specific intervention. Contraception endorses the views of the ICMJE and will now require registration of all prospective trials.