Abstracts

Rationale: There is some evidence that generic substitution of antiepileptic drugs (AEDs) may adversely impact clinical effectiveness. Recent research from Canada has shown that generic lamotrigine use was associated with increased medical resource utilization and economic costs compared to brand use. The objective of this study was to evaluate the impact of switching from branded to generic for five widely-used AEDs in the US on the associated medical resource utilization in patients diagnosed with epilepsy. Methods: Health insurance claims from the PharMetrics Patient-Centric Database, representing over 75 managed care organizations, between 01/2000 and 10/2007 were analyzed. Patients >18 years with ≥1 claim for epilepsy or non-febrile convulsions (ICD-9 codes 345, 780.3 or 780.39) and continuously treated (≥60 of the first 90 days of AED treatment) with primidone (generic entry date 12/1978), carbamazepine (8/1986), phenytoin (9/1992), gabapentin (12/2003), or zonisamide (12/2005) were selected. An open-cohort design was used to classify patients’ observation period into mutually-exclusive periods of brand versus generic use. One-year switching rates of brand-to-generic and switchback-to-brand were calculated using Kaplan-Meier estimations. Periods of branded versus generic use of AEDs were compared for medical resource utilization, including hospitalizations, outpatient-related physician and other services, and prescription drugs. Multivariate Poisson regression analysis was further conducted to adjust for demographics, treatment characteristics, and comorbidities in estimating the differences in medical resource utilization between periods of branded versus generic use of AEDs. Results: 37,929 users of the five studied AEDs were observed for a total of 31,320 and 21,437 patient-years of branded and generic use, respectively. Compared to users of older AEDs (primidone, carbamazepine, phenytoin), those of newer AEDs (gabapentin, zonisamide) were more inclined to switch to a generic drug (20.2% vs. 11.6%), and had a lower switchback rate to the branded compound (10.9% vs. 29.0%). During generic periods, significant increases in health care utilization were observed for both older and newer AEDs. Compared to brand-use periods, generic-use periods were associated with 33% higher hospitalizations rates (adjusted rate ratio [ARR]=1.33; p<.001) and 36% longer hospital length of stay (2.3 vs. 1.6 days per patient-year; ARR=1.36; p<.001). The rate of outpatient-related services and outpatient prescription drugs were 19% and 13%, respectively higher during generic periods than in brand periods (ARR=1.19, p<.001, and ARR=1.13, p<.001, respectively). Conclusions: Generic AED use was associated with significantly higher inpatient, outpatient, and pharmacy utilization, compared to brand use. These findings in the US further confirm that the use of generic AEDs is associated with increased resource utilization in patients with epilepsy.