Abstracts

Rationale: Thirty to 40% of adults with epilepsy respond inadequately to anti-seizure medications (ASMs). Relative to those in whom ASMs are effective, patients with drug-resistant epilepsy (DRE) have poorer quality of life and incur almost twofold higher mean annual healthcare costs. Three types of neuromodulation devices (i.e., vagus nerve stimulation [VNS], responsive neurostimulation [RNS], deep brain stimulation [DBS]) are approved by the United States Food and Drug Administration for patients with seizures refractory to ASM. While all three have demonstrated efficacy in DRE, there is a lack of comparative effectiveness data on these devices.

Methods: Relevant procedure codes were used to identify patients with epilepsy in a large US healthcare claims database who underwent a neurostimulation procedure any time between 2012 and 2019. The earliest date on which neuromodulation was identified during this period was designated the index date. Patients without a diagnosis of epilepsy on index date, those without claims for ASMs in the year prior, and those not enrolled for the 2-year period prior to index were excluded. Patients were allocated to one of two cohorts based on index procedure (i.e., VNS Cohort, RNS/DBS Cohort). Propensity scores were generated using characteristics (e.g., demographics, comorbidities, healthcare utilization and costs) during the 2 years prior to index, and used to match the cohorts. Kaplan-Meier methods and log rank tests were used to contrast incidence of emergency department (ED) visits and hospitalizations during the 24-month follow-up period; corresponding rates of these events were reported per 100 person-months (PMs) and compared using Poisson regression. All-cause and epilepsy-related healthcare utilization were assessed; the latter defined as ED visits/admissions with a diagnosis of epilepsy.

Results: A total of 836 patients met selection criteria (310 VNS; 526 RNS/DBS), and 179 VNS patients were matched to an equal number of RNS/DBS patients. VNS patients were significantly less likely than RNS/DBS patients to experience the composite outcome (hospitalization/ED visit) during follow-up on both an all-cause (55.9% vs. 96.6%) and epilepsy-related (49.2% vs. 95.0%) basis (both p< 0.001). Rates of the composite outcome were numerically, albeit not significantly, lower for VNS patients (18.5 per 100 PMs vs. 22.3 per 100 PMs [all-cause, p=0.193]; 14.5 per 100 PMs vs. 18.2 per 100 PMs [epilepsy-related, p=0.134]). Median time to the composite outcome was 12.6 months for VNS vs. 8.3 months for RNS/DBS (Figure).