Characteristics associated with first anti-seizure medication prescribed in a cohort of adults with newly diagnosed epilepsy

Abstract

Background

Anti-seizure medication (ASM) is the primary treatment modality in epilepsy. There exist evidence-based recommendations published by the American Academy of Neurology and American Epilepsy Society for ASM selection in epilepsy, but these medication recommendations are inconsistently followed. We sought to examine predictors of recommended first ASM in newly diagnosed adults with epilepsy.

Methods

We conducted a retrospective cohort study of adults (≥18 years) newly diagnosed with epilepsy in New York, identified using validated ICD-CM codes, for the period 2011 to 2019. The primary outcome of interest was use of an AES/AAN guideline-informed ASM, and exposures of interest included patient characteristics (e.g. age), provider characteristics (e.g. specialty) and structural characteristics (e.g., practice setting). Multivariable Poisson regression for risk ratios modeled the probability of being prescribed a neutral/recommended or non-recommended ASM, adjusting for covariates.

Results

2340 adults with newly diagnosed epilepsy were prescribed an ASM within 1-year. The most frequently prescribed ASM was levetiracetam (45.5%), which aligns with recommendations. However, 39% were prescribed a non-recommended medication. The prescription of a recommended ASM was associated with older age at diagnosis (relative risk (RR) 1.01; 95% confidence interval (CI) 1.00–1.01), Black race (RR 1.12; 95% CI 1.03–1.23), being in a relationship (vs. divorced/separated/single or widowed) (RR 1.13; 95% CI 1.06–1.22) and a history of stroke (RR 1.19; 95% CI 1.05–1.34). Prescriptions from physician trainees (vs. non-trainees) were more likely to align with recommended ASMs (RR 1.15; 95% CI 1.06–1.25), with no differences by physician specialty. Inpatient setting was associated with fewer recommended ASM prescriptions (vs. outpatient setting) (RR 0.82; 95% CI 0.75–0.89). There was no difference between emergency department and outpatient prescriptions.

Conclusions

Understanding where/why less favorable ASM prescription may occur is important to target potential prescribing interventions. In this study, recommended ASM prescriptions were associated with patient, prescriber and setting characteristics. Notably, trainees prescribed recommended/neutral ASM more often, which underscores the importance of prescriber education in improving prescribing practices.