Objectives: To compare divalproex sodium and valproic acid for therapeutic patterns, persistence rates and predictors of hospitalization among bipolar patients on monotherapy in the Veterans Affairs (VA) healthcare system.
Methods: Using VA administrative data bases, we conducted a retrospective inception cohort study of VA patients'>or= 18 years of age who had at least one outpatient diagnoses of bipolar disorder and two continuous prescription records for the study drugs in the VA PBM pharmacy database during the study period of 1st April 2001 to 30th September 2003. Persistence for the comparative drugs was reported as continuous variable and compared using t-tests. Logistic regression models were used to examine the risk of hospitalization whereas Cox proportional hazard regression models were used to evaluate the time to hospitalization and time to interruption of therapy for the two drug groups.
Results: We identified 4624 bipolar patients on monotherapy with valproic acid (n = 4036) and divalproex sodium (n = 588) during the study period. The descriptive statistics included sociodemographics, disability and comorbidity status and were similar for the two groups. For the crude persistence rates there were no statistically significant differences between divalproex sodium (120 days) and valproic acid (110 days). The logistic regression model for risk of hospitalization showed no statistically significant difference between the two comparators [odds ratio = 1.06, 95% confidence interval (CI) = 0.787-1.444]. The Cox model for time to interruption of therapy showed an insignificant hazard ratio (HR) for divalproex sodium vs. valproic acid (HR = 0.928, 95% CI = 0.844-1.020) and for time to hospitalization also no statistically significant difference in the HR for the two drugs (HR = 0.984, 95% CI = 0.784-1.295).
Conclusion: The study showed a comparable profile of generic valproic acid with divalproex sodium for persistence and predictors of hospitalization for bipolar patients on monotherapy in the VA. Results have important healthcare implications for treatment and costs.