In a recent study examining accelerated approvals between 2009 and 2013, less than half (42%) showed efficacy in post approval trials which were completed at least 3 years after approval, and studies done before and after approvals had limitations in study designs and end points used, (Huseyin Naci, Katelyn R. Smalley, Aaron S. Kesselheim, "Characteristics of Preapproval and Postapproval Studies for Drugs Granted Accelerated Approval by the US Food and Drug Administration," JAMA 318, no. 7 (August 15, 2017): 626-636, doi:10.1001/jama.2017.9415). Accelaterd approvals are granted for drugs purported to treat serious or life-threatening conditions. From 2009 to 2013, 22 drugs received accelated approval for 24 indications. 14 of the 24 indications had not completed all the postapproval requirements. Some comfirmatory studes showed either no benefit, were terminated or delayed by over a year. In fact, five years after approval, 8 indications were still lacking evidence in confirmatory studies. Clinical trials that the FDA accepted were not blinded, meaning people knew if they were being treated or not, and were not randomized, meaning the subjects were not randomly put into different groups to compare treatment. In fact, comparator groups were not required. Furthermore, studies that go through the accelerated approval process only require surrogate outcome measures, that may not necessarily be predictive of better outcomes. For example, body mass index, as a surrogate measure for body fat, might be inaccurate, as muscle and bone mass also contribute to body mass index. In addition, for standard approvals, the FDA requires psychiatry medications to be tested against a placebo, and will not accept using an active comparator instead. Of course, it would be unethical for individuals with life-threatening behaviors to be placed on placebos. Therefore, results might not be able to be generalized to individuals with severe behaviors.