An interesting analysis on how the fact that many negative scientific trials are not published can affect the accuracy of research on the safety and efficacy of drugs.http://www.scientificamerican.com/artic ... ts&page=13