Abstract
A Review of the Impact of Exclusion Criteria on the Generalizability of Schizophrenia Treatment Research
Author(s): Keith HumphreysTreatment research studies employ criteria that determine which patients are eligible to participate and which are not. When such exclusion criteria produce a treatment research sample that is a small and unrepresentative subset of all patients with a particular disease, clinicians may be hesitant to apply the research results in front-line clinical practice. Accordingly, the present paper reviews the English-language literature on exclusion criteria in schizophrenia treatment research and draws initial conclusions about their impact. Empirically derived estimates of the rate of exclusion vary widely (31.0–98.2%), but the best available evidence suggests that about 4 in 5 patients with schizophrenia would be ineligible to enroll in a typical treatment research study. Women are particularly likely to be excluded from schizophrenia treatment research, which is problematic from both a clinical and social justice viewpoint. Excluded patients also tend to be older than eligible patients, and, though it has been examined in only a few studies, they also tend to have more severe problems at baseline and different outcomes over time than patients who are allowed to participate in research. More limited use of exclusion criteria in schizophrenia treatment research would be beneficial in terms of increasing generalizability, but would also potentially involve costs, particularly a need for larger samples. More modest steps that would improve treatment outcome research reports include requiring a full description of the rationale for, and nature of, any exclusion criteria, and, having a designated place in the discussion section which draws attention to the proper scope of generalization.