Previous chapters have described ways in which “real-world” patients usually present with a diversity of psychiatric, medical, psychosocial, and other features that make a “one-size-fits-all” approach to treatment problematic. Large-scale randomized trials typically favor diagnostic uniformity so that all enrolled subjects more or less display the same kinds of symptoms under study. Consequently, the controlled trials literature that informs evidence-based practice largely comes from rarified, homogeneous study groups with rigidly defined diagnostic criteria. As a result, such studies trade off optimal outcomes (“efficacy”) for generalizability (“effectiveness”) under more ordinary conditions. This is why so-called “effectiveness” studies such as the Clinical Antipsychotics Treatment Intervention Effectiveness trial (CATIE; see Chapter 15) strive to enroll representative patients with comorbidities, imperfect treatment adherence, and issues with drug tolerability, adopting “bottom line” primary outcome measures such as “all-cause dropout.” No matter how well a treatment can work, the pragmatic concern remains how well it actually does work in real-life settings.