Abstracts

Rationale: Lennox-Gastaut Syndrome (LGS) is a severe, chronic form of epilepsy characterized by intractable seizures, cognitive impairment and regression, and abnormal electroencephalogram (EEG). Clobazam (CLB), a 1,5-benzodiazepine, is indicated as adjunctive therapy for seizures associated with LGS in patients ≥2 years old. A retrospective healthcare claims analysis was conducted to examine healthcare resource utilization (HRU) among patients with likely LGS treated with clobazam. Methods: Patients with probable LGS were identified as having ≥2 claims for epilepsy [ICD-9-CM 345.xx] or unspecified epilepsy [ICD-9-CM 780.39] in 6 state Medicaid databases using a random forest machine-learning algorithm. Patients with ≥2 prescription claims for CLB and ≥1 year of continuous enrollment before CLB initiation (index date) were included. HRU of patients receiving CLB was analyzed by month over 1 year before index date to end of insurance eligibility, data availability, or death. Generalized estimating equations (GEE) models with repeated measures were conducted to characterize time trends in all-cause and epilepsy-related outpatient claims and all-cause home care claims. Predicted monthly HRU before and after CLB initiation was assessed; HRU before CLB initiation was extrapolated beyond the index date to predict HRU use if CLB had not been initiated. Areas under the curve (AUCs) quantified the difference between extrapolated (without CLB) vs predicted (with CLB) monthly claims. Results: A total of 1,301 LGS patients were included. Mean (SD) age at CLB initiation was 15.65 (11.92) and 52.81% were male. Before CLB initiation, patients were receiving a mean (SD) of 3.44 (1.55) distinct antiepileptic drugs and had annual incidence rates of 22.62, 12.49, and 4.79 for all-cause outpatient visits, epilepsy-related outpatient visits, and home care, respectively. Mean (SD) duration of observation post-CLB initiation was 1.60 (0.86) years. All-cause outpatient visits increased by 0.02 visits/month before CLB and decreased by 0.02 visits/month post-CLB (Figure). The model generally predicted observed outpatient visits pre- and post-CLB (Figure). Similarly, epilepsy-related outpatient visits increased by 0.02 visits/month and decreased by 0.01 visits/month pre- and post-CLB. Home care increased before CLB treatment by 0.02 visits/month but stabilized at 0.58 visits/month after CLB initiation. CLB was associated with decreased all-cause and epilepsy-related outpatient visits and home care (all, P < 0.001). The difference in AUC between extrapolated vs predicated annual all-cause outpatient visits was equivalent to a reduction of 4.56 visits/year (17.97%; Figure). For epilepsy-related outpatient visits and home care, reductions were 3.93 (25.65%) and 2.92 visits/year (29.61%), respectively. Conclusions: Patients with LGS experienced an increase in HRU as measured by outpatient visits (all-cause and epilepsy-related) and home care. CLB treatment was associated with a reduction or stabilization in otherwise increasing HRU associated with LGS. Funding: Lundbeck LLC