Publication bias and dissemination of clinical research

Abstract

Publication bias is a widely recognized phenomenon that occurs because of the influence of study results on the chances of publication. Usually, studies with positive results are more likely to be published than studies with negative results, which leads to a preponderance of false-positive results in the literature. Empiric studies have demonstrated that the induced bias is large and can have a serious impact on meta-analyses, in which data from several studies are aggregated, as well as on informal reviews. The problem is deeply embedded in current research practice, which encourages demonstration of statistical significance to "prove" theories, and one of its causes is the pressure to publish extensively that is an integral part of the competition for academic promotion. Serious efforts to reduce this problem will involve restructuring the process by which study results are disseminated, changing editorial policies, and altering the style and methods of statistical analysis.