Abstract

Clinical trials play a prominent role today in medicine, but are not without controversy. These issues start from the day physicians begin their specialization process in medical school and continues onto their day-to-day practice as attendings with referral patterns and resulting financial incentives. This combined with the lack of training in basic issues of epistemology and statistics, allows poor interpretations of clinical trials to reign free. A proposal to integrate the notion of severity to help remedy these issues are made by ensuring that studies are tested “As Severe As Reasonably Possible” will put P-values, multiplicity adjustments, and model checking in their rightful place. In light of this, multiple clinical trials are examined with an in-depth analysis of erroneous statistical inferences taken from phase III randomized trials, non-inferiority, early phase, as well as observational studies. Severe testing in the setting of medicine will help with alleviate some of the issues with the current misinterpretation of clinical studies.