Abstracts

Rationale: Epilepsy centers provide connect a population of patients to a set of specialized care providers. The size, demographics and epilepsy severity in this population may vary over time. Automated techniques to longitudinally review clinic population size and disease complexity may provide objective metrics to equitably assign personnel and other resources, as well as enable remote surveillance strategies. Here, we utilized data from an EMR reporting tool to examine how such metrics varied over a 11-year period at a single epilepsy center.


Methods: We included patients seen at the Baylor Comprehensive Epilepsy Center between 1/1/2013 and 12/31/2023 by a set of 14 providers (13 epileptologists, 1 nurse practitioner). Using EPIC Reporting Workbench tools, we generated: (i) a list of all individual prescriptions for ASMs (including seizure rescue medications), (ii) a list of date/times of all office or telemedicine visits, and (iii) CPT entries for the interrogation/programming of RNS, DBS and VNS devices.


Results: Over the study period, the clinic served a total of 5135 patients, of which 1502 were seen only once. Clinic volumes increased with time: in 2013, 905 unique patients were seen at a rate of ~100-200 visits/month, while in 2023, 1746 patients were seen at 200-300 visits/month. Telemedicine visits introduced in April 2020 made up ~20-30% of all visits in 2022-2023. 4230 patients received at least one ASM prescription. Levetiracetam (20.1% [2023]), lamotrigine (13.3%), zonisamide (8.3%) and oxcarbazepine (7.7%) were the most frequently utilized ASMs across all years. Shares of cannabidiol, cenobamate, lacosamide, perampanel and clobazam increased, with concomitant reductions in the use of phenytoin, carbamazepine. Similar ASM landscapes were observed in females aged 18-45 years of age, where the use of valproic acid (2-3%) and topiramate (5-6%) remained constant. The proportion of patients on ASM monotherapy decreased with time (43.7 to 35.1%). In parallel, we observed a gradual increase in the proportion of patients receiving a liquid ASM (5.8% to 11%) or a rescue benzodiazepine (12.4% to 20.9%). We designed a dashboard of all patients that received at least one ASM prescription in 2023, ranked by descending rates of ASM polytherapy. As expected, patients ranked at the “top” of this dashboard also featured high rates of adjunct neuromodulation, as well as both liquid and rescue ASM use.


Conclusions: EMR reporting tools can provide epilepsy centers with metrics that pertain to clinic engagement, ASM use and disease severity. Our results reveal that between 2013 and 2023, (i) the population of patients that we serve has almost doubled in size, (ii) their seizure severity has increased and (iii) rates of valproate and topiramate use have remained constant in female patients of reproductive age. We design a “clinic dashboard” prototype that can seamlessly sort and identify patients by polytherapy burden and visit compliance.

Funding: VK is supported by NINDS K08NS110924, NINDS R01NS13199 and seed funding from the BCM Office of Research.