This article is freely available to all

Article Abstract
Sir: Gómez et al. are to be commended for attempting a large observational study intended to advance the knowledge of the use of atypical antipsychotics under conditions that closely approximate those in the clinic, as opposed to the tightly proscribed circumstances dictated by a clinical study. Paradoxically, the authors undermine their own efforts by standardizing data collection around selected parameters, consequently introducing an ascertainment bias.