Background: Partial seizure disorder is typically 1026 treated by monotherapy with antiepileptic drugs (AEDs). However, when the condition is refractory to the initial treatment regimen, patients may be switched to monotherapy with another AED or to combination therapy with the initial AED plus a second AED. Objectives: The purpose of this study was to examine1026 the economic costs associated with treatment-refractory partial seizure disorder and to compare the costs of 2 alternative approaches: a switch to oxcarbazepine (OXC) monotherapy or the addition to the regimen of another AED (AED add-on). Methods: Adult patients with a diagnosis of partial 1026 seizure disorder who received initial AED monotherapy between January 1, 2000, and March 31, 2003, were identified from the PharMetrics Patient-Centric Database, a health plan administrative claims database. The medical and pharmacy history of these patients was analyzed from 6 months before a change to either OXC monotherapy or AED add-on therapy through 12 months after the change in treatment. Total health care resource utilization and the associated costs were compared within each cohort before and after the change, as well as between cohorts, with statistical differences tested using Wilcoxon rank sum tests. Multivariate econometric analyses were performed to examine the impact of age, sex, geographic location, Charlson Comorbidity Index, and the presence of specific comorbidities. Results: Demographic and clinical characteristics 102 were similar between the OXC monotherapy cohort (n = 259) and the AED add-on cohort (n = 795). Annual direct treatment costs increased in both groups in the period after the failure of initial monotherapy, increasing from $10,462 to $11,360 in the OXC cohort and from $10,137 to $12,201 in the AED add on cohort ( P < 0.01). Increased pharmacy costs were the primary driver behind cost increases in both cohorts. Patients in the AED add-on cohort were significantly more likely to have an emergency department visit during the period after the failure of initial monotherapy compared with the OXC monotherapy cohort (odds ratio = 1.52; P < 0.05). Conclusion: Despite limitations, the results of thiConclusion:026 retrospective analysis of claims data suggest that the care of patients with treatment-refractory partial seizure disorder is costly and may vary significantly based on the pattern of care.