Abstracts

Rationale: Limited data exist regarding the impact of patient characteristics on treatment discontinuation in epilepsy populations. This analysis examined factors associated with time to antiepileptic drug (AED) treatment discontinuation in an adult epilepsy population concurrently receiving 2 or 3 AEDs. Methods: A retrospective analysis was conducted using medical and pharmacy claims data for a commercially-insured U.S. population from the MarketScan Database of Thomson Reuters. Selected patients were aged ≥18 years, had prescription claims for 2 or 3 different AEDs overlapping for ≥30 days between 7/1/2004 and 12/31/2005 (first of such combinations was designated the index regimen; date of claim for second or third AED in regimen was designated the index date), had an epilepsy diagnosis (ICD-9-CM 345.xx) ≤6 months prior to or on the index date, and had continuous eligibility for 6 months pre- and 12 months post-index date. The number of days for which the index regimen was continued was evaluated. Patients were categorized into two groups based on Continuation (no >30 day gap in index regimen) or Discontinuation/Addition (>30 day gap in index regimen or addition of another AED to index regimen) of their index regimen during 12 months post-index date (follow-up). Factors associated with time to Discontinuation/Addition of the AED index regimen were evaluated using Cox proportional hazards regression. Independent factors considered included age, gender, geographic region, health plan capitation status, provider type, epilepsy type, comorbidity burden (as measured by the Charlson Comorbidity Index score), depressive disorders, hypertension, and number of AEDs within the index regimen. Statistical significance testing was done at the 5% level. Results: Of 5,134 included patients (mean age 49.3 years; 55.7% female), 38% continued while 62% discontinued or added to their index regimen during the 12-month follow-up period. Of the factors evaluated in the model, greater comorbidity burden (Charlson Comorbidity Index HR 1.05; 95% CI 1.03-1.08; P<0.0001), presence of depressive disorders (HR 1.22 CI 1.07-1.38; P=0.0033), and presence of hypertension (HR 1.26 CI 1.12-1.43; P=0.0002) were associated with shorter time receiving the AED index regimen while male gender (HR 0.89 CI 0.83-0.95; P=0.0011), unspecified epilepsy type (HR 0.88 CI 0.79-0.97; P=0.015), and the north central geographic region (HR 0.81 CI 0.73-0.89; P<0.0001) were associated with longer time receiving the index regimen. Conclusions: In this analysis, comorbidity burden, depressive disorders, and hypertension were associated with shorter time receiving the index regimen of 2 or 3 AEDs while male gender, unspecified epilepsy type, and north central geographic region were associated with longer time receiving the index regimen. Further research to better understand these findings and the impact of patient characteristics on treatment patterns in epilepsy is warranted. Supported by Ortho-McNeil Janssen Scientific Affairs, LLC.