In this era of genetic enlightenment, we now know that patients are just like their fingerprints: There are no two alike. This also applies to a patient’s response to a drug.8 Consequently, I would propose that the data from every clinical trial for PAH prespecify an analysis of the efficacy response by displaying the probability distribution of the primary outcome variable (with a histogram) so that whether the responses are unimodal, multimodal, or skewed on the basis of an identifiable variable could be determined. The current standard for regulatory approval requires proof that the treated group has a significantly greater improvement in the primary end point than the control group, even (in the case of PAH) if the overall difference is only 1 m in the 6MWD. I can easily construct a hypothetic trial in which the overall improvement in 6MWD was 40 m, which was uniformly distributed among the patients on active therapy, or a trial in which the overall improvement was 40 m but the benefit was confined to only young, white women, with other phenotypes being made worse by the treatment. Because identifying a subgroup that has greater or lesser benefit in a clinical trial is not required by regulatory authorities, and might ultimately hurt the marketing of the drug to physicians, there has been no incentive for industry to conduct such analyses.