Automated Tool for Identification of Patients with Anti-seizure Medication Reduction in the Long-term Monitoring Unit
Abstract number :
1.411
Submission category :
7. Anti-seizure Medications / 7C. Cohort Studies
Year :
2024
Submission ID :
1316
Source :
www.aesnet.org
Presentation date :
12/7/2024 12:00:00 AM
Published date :
Authors :
Presenting Author: Olivia Mezheritsky, BA – Boston Children’s Hospital, 300 Longwood Ave, Boston, MA 02115, USA
Xingyan Liu, BS – Boston Children's Hospital
Fatemeh Mohammad Alizadeh Chafjiri, MD – Boston Children’s Hospital, 300 Longwood Ave, Boston, MA 02115, USA
Emily Peter, BA – Western University of Health Sciences, Pomona, CA 91766, USA
Michele Jackson, BA – Boston Childrens Hospital
Tobias Loddenkemper, MD – Boston Children's Hospital, 300 Longwood Ave, Boston, MA 02115, USA
Rationale:
Patients with epilepsy may require long-term monitoring (LTM) for presurgical evaluations (Phase 1). During Phase 1 LTM admission, anti-seizure medications (ASMs) may be tapered to record seizures. Changes in ASM dosage are logged in the electronic health record (EHR). However, it is difficult to evaluate ASM changes over time. We developed an automated tool to identify reduced ASM periods of patients who wore wearables in the LTM. With this initiative, we aim to advance neurophysiological biomarker development for non-invasive ASM monitoring.
Methods:
ASM release type and intake dosage for one 24 hour timeframe prior to LTM admission and all 24 hour timeframes during LTM admission for patients enrolled in a wearable device (Empatica, Milan, Italy) study at Boston Children’s Hospital from 2015 to 2021 were extracted from the EHR and entered into REDCap (Nashville, TN). The automated analysis tool performs 6 processing steps (Figure 1): 1) identifies the enrollment periods during LTM admission, 2) classifies LTM enrollments as Phase 1 (patients with reduced ASM dose from their Pre-LTM ASM regimen) and non-Phase 1 (patients without reduced ASM dose from their Pre-LTM ASM regimen), 3) identifies time periods [morning: 3:00 am to < 12:01 pm), afternoon (12:01 pm to < 5:00 pm), evening (5:00 pm to < 3:00 am)] that present with a reduced ASM dose as compared to their Pre-LTM ASM regimen, 4) quantifies ASM dose reduction/period (morning, afternoon, and evening), 5) identifies lowest dose window/period (1 hour prior to each ASM intake), 6) identifies highest ASM dose window/period (ASM peak window from last ASM intake). ASM peak window was calculated as the mean of all ASM peak concentrations/period and patient. We excluded patients on continuous infusion and neurostimulation.
Results:
The automated analysis pipeline classified 403 enrollments (361 patients) into 299 Phase 1 and 104 non-Phase 1 enrollments. 63 patients had > 1 enrollment. Mean duration of LTM enrollment was 3.6 days (1.0-5.0). A mean of 2.7 ASMs (1.0-9.0) were administered per enrollment with 0% of enrollments with no ASM administered, 17.4% with 1 ASM administered and 82.6% with > 1 ASM administered. Mean ASM dose reduction was -41.3% in 47.4% of morning periods, -61.6% ASM reduction in 46.7% of afternoon periods, and -57.8% in 48.2% of evening periods. In non-Phase 1 patients, mean ASM peak concentration was 2.5 (0.25-24.0) hours after morning ASM intake, 1.9 (0.25-7.0) hours after afternoon ASM intake and 2.6 (0.25-24.0) hours after evening ASM intake. In Phase 1 patients, mean ASM peak concentration was 2.6 (0.02-20.0) hours after morning ASM intake, 2.3 (0.02-10.5) hours after afternoon ASM intake and 2.7 (0.02-20.0) hours after evening ASM intake.
Conclusions:
Our automated tool classified patients with and without ASM reduction during LTM enrollment and quantified percentage of dose reduction in Phase 1 patients. ASM dose quantification may serve as an additional assessment tool for biomarker evaluation of seizure risk and prediction.
Funding:
The Epilepsy Research Fund supported this study.
Anti-seizure Medications