ABSTRACT
Objectives: To analyze, from a payer perspective, the net pharmaceutical and medical costs of prescribing divalproex sodium extended-release (DVPX-ER) versus valproic acid (VPA) in patients with bipolar disorder.
Methods: The study used a decision analytic framework to compare the total costs associated with DVPX-ER relative to VPA in the treatment of bipolar disease. The decision model incorporated two primary outcomes: GI side effects and treatment success. Levels of health service utilization and probability values were obtained from an expert panel comprised of 15 psychiatrists. Unit costs were obtained from an academic medical center. Two-way sensitivity analysis and a Monte Carlo simulation were conducted to examine the stability of the results.
Results: The average probability of GI side effects associated with VPA and DVPX-ER estimated by the expert panel was 0.36 and 0.10 respectively. The average probability of treatment success for VPA and DVPX-ER was estimated to be 0.45 and 0.58 respectively. In the base case analysis, the expected total cost per patient was $34 208.84 and $25 336.13 for VPA and DVPX-ER respectively, a difference (incremental cost) of –$8188.87. Probabilistic sensitivity analysis using Monte Carlo simulation indicated a negative incremental cost of using DVPX-ER relative to VPA in a majority (approximately 70%) of cases.
Conclusions: The results from this decision analysis, based on probabilities of major events and associated utilization from an expert panel, suggest that divalproex sodium extended-release results in lower total costs than valproic acid in the treatment of bipolar disorder. Limitations of this study primarily involved the sources of data used in the model, including that derived from expert opinion.