Abstracts

Assessing Economic Burden Across Distinct Epilepsy Patient Segments Using Machine Learning in a Commercial Population

Abstract number : 1.103
Submission category : 13. Health Services (Delivery of Care, Access to Care, Health Care Models)
Year : 2025
Submission ID : 76
Source : www.aesnet.org
Presentation date : 12/6/2025 12:00:00 AM
Published date :

Authors :
Presenting Author: Pallavi Mudumby, MS – STATinMED

Jensy Rodriguez, MS – STATinMED
Maxine Fisher, PhD – STATinMED
Keshia Maughn, MPH – STATinMED

Rationale: In the United States, epilepsy affects about 3 million adults (1.1%). This study uses machine learning cluster analysis to identify key epilepsy patient groups based on comorbidities and their impact on healthcare utilization, focusing on economic burden.

Methods: A retrospective claims analysis using the STATinMED RWD Insights database evaluated epilepsy patients from January 1, 2014, to May 31, 2021. The index date was the first antiseizure prescription after an epilepsy diagnosis. Eligible patients were ≥18 years, with at least one inpatient claim or two outpatient claims ≥ 30 days apart, and 12 months of continuous medical and pharmacy benefits before and after index. Machine learning cluster analysis identified clusters using the Elbow method and hierarchical clustering.

Results: A total of 35,351 commercially insured epilepsy patients met the eligibility criteria. Cluster analysis identified four segments based on dominant comorbidities: 1) Gastrointestinal Cluster (n=1,841, 5.2%), with a high prevalence of gastrointestinal disorders, particularly GERD (100% vs. 13.88% overall); 2) Mental Health Cluster (n=5,230, 14.8%), with high rates of depression, anxiety, and schizophrenia (92.77% vs. 35.31% overall); 3) Neurological Cluster (n=11,794, 33.3%), with high rates of neurological disorders, including multiple sclerosis, cerebral palsy, and ADHD (100% vs. 47.61% overall); 4) Low-Risk (LR) Cluster (n=16,486, 46.7%), with low rates of comorbidities. At the 12-month follow-up, the Gastrointestinal Cluster incurred the highest total healthcare costs compared to the other clusters and the overall population, followed by the Mental Health Cluster. The Gastrointestinal Cluster of epilepsy patients had total costs of $108,576; 37.4% higher than the overall epilepsy population ($79,063), 13.9% higher than the Mental Health Cluster ($95,256), 46.6% higher than the Neurological Cluster ($74,094), and 49.5% higher than the LR Cluster ($72,616). Inpatient costs for the Gastrointestinal Cluster were $38,074; 38.5% higher than the overall population ($27,489), 16.0% higher than the Mental Health Cluster ($32,819), 54.2% higher than the Neurological Cluster ($24,745), and 48.2% higher than the LR Cluster ($25,692). Prescription costs for the Gastrointestinal Cluster were $11,487, 21.4% higher than the overall population’s $9,468, 0.6% higher than the Mental Health Cluster ($11,415), 31.8% higher than the Neurological Cluster ($8,749), and 26.9% higher than the LR Cluster ($9,116).

Conclusions:

This study highlights the impact of gastrointestinal and mental health comorbidities on healthcare costs in epilepsy patients, with both segments showing significantly higher utilization and economic burden compared to other segments. These findings emphasize the need for integrated management strategies in epilepsy care to improve outcomes and reduce overall healthcare costs.



Funding: The authors received no financial support for the research, authorship, or publication of this article, which was funded by STATinMED.

Health Services (Delivery of Care, Access to Care, Health Care Models)