Abstracts

Rationale: Epilepsy is a chronic disorder characterized by recurrent seizures, and is among one of the primary conditions associated with dysphagia. The successful treatment of epilepsy with antiepileptic drugs (AEDs) can depend on the patient's adherence to medication and literature has shown that non-adherence is a significant issue in this population. This study estimates the incremental healthcare burden of dysphagia and other comorbidities which may impact adherence in individuals with epilepsy. Methods: This retrospective cohort study uses data from MarketScan?(R) Commercial Claims and Medicare Supplemental database (Truven Health Analytics, New York, NY) from 2010 Q2 through 2015 Q2. Eligible patients met the following criteria: 1] at least two visits with a record of epilepsy, 2] at least six months of medical and pharmacy enrollment prior to the first epilepsy diagnosis, and 3] at least two years medical and pharmacy enrollment following the first epilepsy diagnosis. A dysphagia cohort and a non-dysphagia cohort were established from the epilepsy patient group. To be categorized into the dysphagia cohort, patients must have a record of a diagnosis of dysphagia or a record of a swallowing study. To be categorized into the non-dysphagia cohort, none of the following conditions can be present: 1] a diagnosis of dysphagia, 2] a record of a swallowing study, or 3] for those old enough to take pill form, one or more outpatient pharmacy fills of the liquid form of AED medication (where solid form is also available). Adherence to the AED medication is estimated from the outpatient pharmacy fills using the Medication Possession Ratio, defined as total days' supply divided by (last fill date ?" first fill date + last fill's day supply). Adherence to AED medication and annual expenditures are modeled by cohort, while adjusting for patient demographics and comorbidities. Results: Patients meeting all inclusion criteria are displayed in an attrition diagram. Patient demographics, comorbidities, clinical events, annual expenditures and healthcare utilization are summarized by cohort. Results of the multivariable expenditure models estimating the annual healthcare burden for epilepsy patients with and without dysphagia are reported separately for commercial and Medicare patients. Statistically significant model covariates are also reported. Results of multivariable adherence models estimating the difference in adherence rates for epilepsy patients with and without dysphagia as well as relevant comorbidities are reported. Conclusions: Currently the literature has shown that dysphagia in epilepsy can lead to non-adherence to AED medication, therefore type and size of oral preparation bears consideration in these patients. This research contributes to the current body of evidence by providing the estimated adherence rates of AED medication for epilepsy patients with and without dysphagia and other comorbidities as well as the estimated annual incremental burden of dysphagia in epilepsy patients. Funding: This study was sponsored by Aprecia Pharmaceuticals Company.