Abstract Details

Title Validation of a semi-automated method for tracking quality metrics in benzodiazepine-resistant status epilepticus

Topic Epilepsy/Clinical Neurophysiology (EEG)

Presentation(s) S7 - Epilepsy and Clinical Neurophysiology (EEG) 1 (4:42 PM-4:54 PM)

Poster/Presentation
Number 007

Background Timely treatment is essential for reducing morbidity in BRSE. Well-accepted guidelines describe the treatment timeline for status epilepticus including target time to second-line ASD administration. In practice, these metrics are difficult to track using electronic health record (EHR) data without resource-intensive chart review.

Objective

This study aimed to evaluate a semi-automated method of (1) identifying inpatient encounters with benzodiazepine-resistant convulsive status epilepticus (BRSE) and (2) calculating time to second-line anti-seizure drug (ASD) administration.

Design/Methods We implemented a neurology consult note template for seizure-related inpatient encounters at a tertiary care pediatric center. The template was developed using user-centered design principles and features discrete fields requesting seizure-related treatment information. Template responses were used in an EHR query designed to identify BRSE encounters, termed the “semi-automated” method. Time from first- to second-line ASD was calculated among the BRSE encounters with a target time of less than 20 minutes. Sensitivity and positive predictive value (PPV) of the semi-automated method were calculated using manual chart review as the gold standard.

Results Among 227 seizure-related inpatient encounters from July and August 2021, 58 cases of BRSE were identified by manual review. Locations of initial BRSE treatment included emergency department (N=42), general floors (N=9), and intensive care units (N=7). The note template was used in 45 (78%) of these encounters. The semi-automated method had sensitivity of 62% and PPV of 92% for detecting BRSE. When used to identify cases of BRSE where time to second-line ASD was >20 minutes, the semi-automated method had sensitivity of 54% and PPV of 94%.

Conclusions The semi-automated approach used in this study identifies BRSE cases across a variety of inpatient settings and calculates seizure-related quality metrics with high PPV and moderate sensitivity. The method can be used to track key metrics over time without the need for resource-intensive chart review with each Plan-Do-Study-Act cycle.

Authors/Disclosures
Benjamin Siegel, MD PRESENTER	Dr. Siegel has nothing to disclose.
Mohammed Shahnawaz, MD	Dr. Shahnawaz has nothing to disclose.
Kathryn C. Elkins, MD (Emory University)	Dr. Elkins has nothing to disclose.
Ammar Kheder, MD (Penn Epilepsy Center, Hospital of the University of Pennsylvania)	Dr. Kheder has nothing to disclose.
	No disclosure on file