Poor allocation concealment methods are associated with heterogeneity in age and statistical significance of the primary outcome: Review of recent trials published in four general medical journals.

There is evidence of an association between poor allocation concealment methods and statistical significance of the primary outcome. Trials that use inadequate allocation concealment methods are more likely to have statistically significant P values compared with trials using good or adequate allocation concealment methods.