Diagnostic accuracy of quantitative EEG to detect delayed cerebral ischemia after subarachnoid hemorrhage: A preliminary study.

Objective: Delayed cerebral ischemia (DCI) is the most important and preventable morbidity cause after subarachnoid hemorrhage (SAH). Therefore, DCI early detection is a major challenge. Yet, neurological examination can be unreliable in poor grade SAH patients. EEG provides information from most superficial cortical area, with ischemia-related changes. This study aims at defining an alpha-theta/delta (AT/D) ratio decrease thresholds to detect DCI.

Methods: We used EEG with a montage matching vascular territories (right and left anterior central and posterior) and compared them to follow-up brain imaging.

Results: 15 SAH patients (Fischer ≥ 3, World Federation of Neurological Surgeons scale ≥4, 9 DCI) were monitored during 6.4 [4-8] days (min = 2d, max = 13d). AT/D changes could follow three different patterns: (1) prolonged or (2) transient decrease and (3) no decrease or progressive increase. A regional 30% decrease outlasting 3.7 h reached 100% sensitivity and 88.9% specificity to detect DCI. Only 22.6% were in a zone of uncertain diagnosis (3.7-8.04 h). These prolonged decreases, with a loss of transient changes, started in cortical areas evolving toward DCI, and preceded intracranial changes when available.

Conclusion: Although this study has a small sample size, prolonged AT/D decrease seems to be a reliable biomarker of DCI.

Significance: cEEG changes are likely to precede cerebral infarction and could be useful at the bedside to detect DCI before irreversible damage.