9: IMPLEMENTATION OF 2HELPS2B SEIZURE RISK SCORE: A HIGH-VALUE APPROACH TO SEIZURE DETECTION IN THE ICU

Abstract

Introduction: The use of continuous electroencephalograms (cEEG) has increased over the last decade with guidelines recommending at least 24 hours of monitoring to detect nonconvulsive seizures (NCS). However, recent studies suggest it may be unnecessary to monitor low-risk patients beyond a few hours. Since cEEGs are resource-intensive, we sought to optimize utilization by reducing nonvalue-added time on cEEG, applying a seizure risk score, 2HELPS2B, based on initial EEG findings and patient clinical factors. Methods: We included patients aged >18-years, who required ICU level care and underwent cEEG at a tertiary center between June 2020 and June 2022 (n=252). We excluded patients undergoing cEEG post-cardiac arrest or for ICP or status epilepticus management. Pre-intervention, seizure incidence, cEEG monitoring duration, and 2HELPS2B score were documented via chart review, including EEG reports. For the intervention, we convened a multidisciplinary group of neurointensivists, epileptologists, nurses, and technologists to educate clinical teams on use of the 2HELPS2B seizure risk score. We also implemented a daily touchpoint between the primary team and the reading epileptologist to communicate the seizure risk score and aid in decision-making. Post-intervention data evaluated the impact of utilizing 2HELPS2B on cEEG duration and seizure incidence. Results: Of 252 patients, most had a low risk of seizures with a 2HELPS2B score of 0 (n= 149, 59%), followed by moderate-risk (score: 1; n=79, 31%) and high-risk (score: >2; n=24, 9.5%). Prior to the intervention, monitoring duration for low-, moderate-, and high-risk groups was similar (33 hours vs 32 hours vs 38 hours) and seizure incidence was low except for the high-risk group (0% vs 6.7% vs 50%). Following 2HELPS2B implementation, monitoring duration for all groups decreased with significant reduction in the low-risk group (33 vs 22 hours; p= 0.009), whereas seizure detection rates did not change significantly for any risk group (0% vs 5% vs 50%). Conclusions: Most patients had low or moderate seizure risk based on the 2HELPS2B score, which correlated with a low incidence of recorded seizures on cEEG. Implementing the risk score identified low-risk patients and reduced cEEG monitoring duration without compromising seizure detection rates.