Purpose: To assess rates and predictors of medication nonadherence and hospitalization among patients with bipolar I disorder.
Methods: This was a retrospective cohort analysis of Medicaid patients who were aged ≥ 18 years, had ≥ 1 inpatient or ≥ 2 outpatient medical claims indicating bipolar I disorder (ICD-9-CM codes 296.0x-296.1x, 296.4x-296.7x), and filled ≥ 1 prescription for antipsychotic medication between January 1, 2004, and December 31, 2006. Patients were followed for 1 year from the date of first (index) antipsychotic prescription. Patients were required to be continuously eligible for Medicaid without dual Medicare eligibility from 1 year before (baseline) through 1 year after (follow-up) index, and were required to receive ≥ 1 additional antipsychotic during follow-up. Descriptive statistics and predictors of medication nonadherence (medication possession ratio <0.8) and hospitalization were generated.
Results: A total of 9410 patients met study eligibility criteria with a mean age of 38 years; 74% were female and 75% were white. Approximately 31% and 57% had baseline diagnoses of substance abuse and other psychiatric conditions, respectively. During follow-up, roughly 60% of patients were nonadherent and 40% of patients were hospitalized for any reason (37% psychiatric-related). Multivariate analysis showed that new antipsychotic starts, younger patients, those with a baseline concomitant substance abuse diagnosis, those taking a baseline antidepressant, and those with a baseline psychiatric hospitalization had significantly higher risk of nonadherence. Baseline psychiatric hospitalization, baseline substance abuse or other psychosis diagnosis, baseline use of an anxiolytic, anticholinergic, or anticonvulsant, and nonadherence to therapy in the follow-up period were significant predictors of increased risk of hospitalization.
Limitations: This analysis did not attempt to evaluate the complex relationships among treatment type, adherence, hospitalization, and other variables.
Conclusions: Study results showed that the risk of nonadherence is relatively high and confirmed that nonadherence is associated with a greater risk of hospitalization.