Comparative Effectiveness of Adaptive Treatment Strategies for Schizophrenia

This retrospective cohort study will use data from national (45-state) Medicaid Analytic Extracts data (2001-2010). The cohort will consist of adults who are 18 to 64 years old and diagnosed with schizophrenia who initiate a new psychotropic medication after a period of stable antipsychotic treatment. The eligibility criteria select a cohort of individuals diagnosed with schizophrenia, have received antipsychotic monotherapy prior to the index date, and are still experiencing problems for which a new psychotropic medication strategy was initiated. A 1-year period of eligibility prior to follow-up initiation ensures sufficient time to collect service use related covariates to characterize cohort members. Five clinical subgroups will be defined based on the presence of psychiatric diagnoses during the 30 days prior to and inclusive of the treatment change under study (index date) and who have this same diagnostic code from more than one provider to increase the validity of the diagnostic groups. The subgroups will be defined by codes to capture 1) uncomplicated schizophrenia; 2) schizoaffective disorder; 3) depression; 4) mania; and 5) anxiety. These subgroups are defined to reflect the reason for the change in treatment. Pharmacological treatment options for patients with schizophrenia who are nonresponsive to antipsychotic monotherapy will include (1) initiation of a second antipsychotic, (2) initiation of an antidepressant, (3) initiation of a mood stabilizer and (4) initiation of a benzodiazepine. The primary effectiveness outcome will be time to psychiatric hospitalization. Secondary measures include time to index treatment discontinuation, time to introduction of another psychotropic medication, psychiatric emergency department visits, all-cause hospitalization, and death.