Abstracts

Rationale: Poor medication compliance is widespread and problematic in epilepsy, due in part to complex dosing regimens for some antiepileptic drugs (AEDs). The impact of AED compliance on seizure recurrence has not been widely investigated using administrative data from a managed care setting. As a context for the introduction of newer AEDs with simpler dosing regimens, we analyzed retrospective claims from the PharMetrics database (1/1/00–12/31/05) to evaluate the effect of AED compliance on the risk of serious breakthrough seizure within 1 year following AED initiation.Methods: Inclusion criteria were: age 21-64 years, at least 1 diagnosis of epilepsy (ICD-9 345.xx) or non-febrile convulsions (ICD-9 780.3, 780.39) by a neurologist, at least 2 AED prescriptions, and continuous health plan enrollment for at least 6 months prior to and at least 60 days following AED initiation. Serious breakthrough seizure was defined as a hospital or emergency department admission with a primary discharge diagnosis of epilepsy or non-febrile convulsions. Observation time began 7 days after AED initiation and ended with the first occurrence of serious breakthrough seizure, end of plan enrollment, or end of study period (AED initiation + 365 days). Compliance was measured using the medication possession ratio (MPR), a time-dependent variable assessed cumulatively at the end of each month following AED initiation through follow-up end. MPR was defined as the ratio of AED days supplied to total days within each cumulative interval following AED initiation. Patients with an MPR of at least 0.8 at the end of an interval were classified as compliant for that interval. Risk of serious breakthrough seizure was assessed using an extended Cox proportional hazards model in which time to first seizure was estimated as a function of the time-dependent compliance measure (continuous MPR or compliance status), patient demographics, Charlson comorbidity score, and indicators for initiation of newer or older AED and subsequent AED switch.Results: 18,073 patients met all inclusion criteria, of whom 2,467 (14%) had a serious breakthrough seizure post-AED initiation. The majority of patients were female (59%), between ages 30 and 64 (82%), and initiated an older AED (63%). Mean follow-up time for patients with an event was 133 days compared to 305 days for censored patients. The risk of serious breakthrough seizure following AED initiation was significantly lower for compliers (hazard ratio [HR]=0.84; P=0.001). Incremental improvements in the MPR were also found to significantly reduce the risk of serious seizure (HR=0.78; P=0.002).Conclusions: The risk of serious breakthrough seizure appears to be lowered by achievement of an 80% compliance threshold for AED therapy and with incremental improvements in the percentage of days with an AED on hand. Efforts in managed care to promote AED compliance, possibly through prescribing of newer AEDs with simpler dosing regimens, may lead to reduced risk and delayed onset of serious breakthrough seizures.