Rapid Response EEG Shows Class II Evidence for Reliably Detecting Status Epilepticus
A study published in Neurology has provided Class II evidence demonstrating the reliability of automated seizure burden estimator (ASBE) software associated with rapid response EEG (rr-EEG) used in screening for electrographic status epilepticus (ESE) without the need for expert review. The findings suggest that rr-EEG may be effective and accurate for detecting ESE in limited-resource settings.
In the retrospective observational study, 3 blinded reviewers evaluated all consecutive clinical rr-EEG procedures performed at Yale New Haven Hospital, an affiliated community hospital, and an affiliated inner-city regional hospital between November 2019 and February 2021. A reference standard of 2/3 agreement was used to assess the 3 reviewers’ analysis. The coprimary endpoints in the study were the positive predictive value (PPV) and the negative predictive value (NPV) of the ABSE in detecting ESE or possible ESE (ESE/pESE).
At a >10% seizure burden cutoff:
- In the first 2 hours, 86% of studies with ESE alone (95% CI, 42% to 100%) and 88% with ESE/pESE (95% CI, 68% to 97%) were detected.
- The NPV was 99% (95%, CI, 97% to 100%) for detecting ESE and 98% (95% CI, 95% to 100%) for detecting ESE/pESE.
- The PPV was 11% (95% CI, 4% to 23%) for detecting ESE and 39% (95% CI, 26% to 53%) for detecting ESE/pESE.
- The specificity was 79% (95% CI, 73% to 84%) for detecting ESE and 84% (95% CI, 79% to 89%) for detecting ESE/pESE.
At a >90% seizure burden cutoff:
- The PPV was 33% (95% CI, 7% to 70%) for detecting ESE and 78% (95% CI, 40% to 97%) for detecting ESE/pESE.
- The specificity was 97% (95% CI, 94% to 99%) for detecting ESE and 99% (95% CI, 97% to 100%) for detecting ESE/pESE.
- The sensitivity was 43% (95% CI, 10% to 82%) for detecting ESE and 29% (95% CI, 13% to 51%) for detecting ESE/pESE.
The findings suggest that, at a >90% seizure burden threshold, the ABSE’s high specificity, good PPV, and low-to-moderate sensitivity enable applications for triaging patients to receive immediate treatment, urgent expert review, or additional continuous EEG. At a >10% threshold, the ABSE’s high sensitivity is suited for use as a screening tool for excluding ESE.