Delays in Access to EEG and Its Impact on Hospital Length of Stay: Conventional EEG and Point-of-care EEG Cases from the SAFER-EEG Trial

Abstract number : 2.015
Submission category : 3. Neurophysiology / 3B. ICU EEG
Year : 2023
Submission ID : 922
Source : www.aesnet.org
Presentation date : 12/3/2023 12:00:00 AM
Published date :

Authors :
Presenting Author: Mariel Kalkach Aparicio, MD – University of Wisconsin-Madison

Venkatraman Thulasi, MD – Department of Neurology – Rutgers University; Masoom Desai, MD – Assistant Professor, Department of Neurology, University of New Mexico; Aaron Struck, MD – Assistant Professor, Department of Neurology, University of Wisconsin-Madison

Delays in non-convulsive seizure and status epilepticus (NCS/NCSE) diagnosis and treatment have been linked to poor health outcomes (Hill et al., 2017 Ann of Neurology 82(2) 155 – 165). These delays are often due to challenges in access to electroencephalographic (EEG) monitoring, which is necessary for diagnosing NCS/NCSE. Even large academic hospitals with 24/7 coverage of conventional EEG have reported a median delay of four hours to start EEG monitoring (Vespa et al., 2020 Critical Care Medicine 48(9)). Point-of-care EEG (POC-EEG) allows for rapid access to EEG monitoring at the bedside and might potentially impact patient health outcomes by facilitating faster detection and management of NCS/NCSE. We set out to contrast the impact of POC-EEG compared to conventional EEG (cEEG) in seizure-positive patients.

Data was collected as part of a muti-center retrospective study (SAFER-EEG trial) across four different academic hospitals. The patients were divided into two cohorts, according to the EEG that they received: cEEG only or POC-EEG (alone or followed by cEEG). Patients included in the trial were adult subjects with at least 1 hour of POC-EEG or 4 hours of cEEG recording who had a seizure diagnosed by EEG. A sub-analysis included those who had their first seizure within one minute of the start of EEG. We calculated the time from admission to EEG start, and hospital stay from EEG start to discharge. We also identified how many had SE diagnosis during their stay.

A total of 198 seizure-positive patients were selected for analysis across four sites (cEEG = 138, POC-EEG = 60). Overall, the seizure diagnostic yield was similar across the groups: 21% (138 of 668) for cEEG cases, and 15% (60 of 392) for POC EEG. As a group, the seizure-positive cEEG patients waited a median of 5.9 [2.7, 33.7] hrs from admission to the EEG start, while the seizure-positive POC EEG patients’ wait time was a median of 3.6 [0.57, 9.1] hrs.

Of the seizure-positive cEEG patients, 37 (27%) were seizing at the start of recording, and 25 of these patients (68%) were diagnosed with SE during their stay. The median time to cEEG was 5.7 [2.1, 40.1] hrs, and the median hospital stay after cEEG start was 8.4 [4.6, 15.4] days. In comparison, 10% (6 of 60) of POC EEG patients were seizing at the time the POC EEG was started and 5 (83%) of them were diagnosed with SE during their stay. For POC-EEG patients, the median time to POC-EEG was 0.6 [0.09, 1.3] hrs, and their median stay after POC-EEG start was 2.8 [1.7, 6.7] days.

We found that 27% of patients who only received cEEG were actively seizing when cEEG recording was initiated. Compared to the cEEG-only group, the POC-EEG cases had shorter time to EEG and shorter hospital stays after the EEG. These observations highlight the issue of delayed access to EEG where seizures are left untreated while waiting for cEEG and show initial evidence of the impact of these delays in length of hospital stay.

Funding: Study funding provided by Ceribell, Inc.