Abstracts

Rationale: Subarachnoid hemorrhage (SAH) affects 30,000 people in the U.S. annually. Symptomatic vasospasm is thought to be a major cause of morbidity occurring in 46% of patients. Standard monitoring methods for vasospasm include clinical exam, transcranial Doppler ultrasonography(TCD), brain imaging, and conventional angiography. Continuous EEG (CEEG) may be performed at the bedside and has excellent spatial and temporal resolution, allowing for detection of ischemia in real time. The optimal quantitative EEG methods for ischemia detection in SAH have not been established. Methods: Twenty patients with SAH had CEEG monitoring over a year-long period at The Hospital of the University of Pennsylvania. CEEG was analyzed using three calculations: alpha delta ratio, alpha variability, and EEG asymmetry index. Daily clinical exam, TCDs, angiography, and imaging results were collected. Time of CEEG change was compared with time of first change in any standard monitoring methods for ischemia: clinical exam change, TCD, HCT or MRI results, or angiography results. The patients were divided into two groups for analysis, those with radiographic evidence of cerebral infarction and/or angiographic vasospasm, and those with neither. An electroencephalographer blinded to patient group reviewed graphical displays of 12 hour EEG segments and determined whether the trend showed focal changes suggestive of ischemia. Results: Of the 20 patients, 7 had radiographic evidence of brain infarction, 6 had angiographic vasospasm, and 2 had both. There was a total of 12 abnormal angiography or head imaging studies consistent with focal infarction or vasospasm in 11 patients. Three patients with infarction received CEEG monitoring after the time of infarct and were not included in statistical analysis. In these 3 patients, the CEEG was focally abnormal from the onset of monitoring in the region corresponding to the infarct. Six infarction or vasospasm events were detected by CEEG an average of 75.25 hours prior to standard monitoring methods. One patient with a small deep cingulated gyrus infarct was not detected by CEEG. One patient with vasospasm that was not detected by the alpha delta ratio or alpha variability was evident using the EEG asymmetry index. One patient with bilateral severe vasospasm had no focal abnormality on CEEG. In comparison to standard monitoring methods for ischemia, the sensitivity of the alpha delta ratio and alpha variability for detection of angiographic vasospasm or infarction was 66.67%, and the specificity was 100%. With the addition of the EEG asymmetry index, the sensitivity was 77.78% and the specificity was 100%. Conclusions: CEEG monitoring is a useful adjunct for the detection of ischemia after SAH. It may detect cerebral ischemia prior to other currently used modalities. More widespread use of quantitative EEG may allow early detection and treatment of reversible ischemia and may improve neurologic outcomes.