Randomized clinical trials have limitations because they focus on small samples of highly selected patients. Observational studies, which follow large cohorts of typical patients receiving pharmacological treatments, should overcome some of these trial limitations and provide information that cannot be generated with clinical trials. The present study aimed to compare experimental and observational studies of clozapine-treated subjects with treatment-resistant schizophrenia. A systematic review of experimental and observational studies evaluating clozapine-treated subjects in treatment-resistant schizophrenia was carried out. We identified 50 studies that met the inclusion criteria. Less than one-third of clinical trials enrolled more than 50 patients compared to 44% of prospective and nearly 90% of retrospective studies. In addition, 78% of prospective and 89% of retrospective observational studies lasted more than 12 weeks, while the majority of trials lasted less than 8 weeks. Most clinical trials defined treatment-resistant schizophrenia according to Kane's criteria, while the majority of observational studies adopted implicit criteria. In comparison with clinical trials, observational studies provided a higher weighted mean rate of clozapine-responders and a lower weighted mean rate of clozapine-dropouts. This literature survey suggests that the role of observational studies in the evaluation of medicines should be reconsidered. A new generation of observational studies should be developed to provide evidence on patient outcome in typical settings and under real-world circumstances, and on variables which may affect outcome.
