Purpose: Problems with the availability of standard EEG monitoring in the intensive care unit have led to the use of recordings that have limited spatial coverage. We studied the performance of limited coverage EEG compared with more traditional full-montage EEG.
Methods: Continuous EEG recordings were performed on 170 patients using the full-montage 10-20 placement of electrodes as a reference recording and an abbreviated montage of electrodes applied below the hairline (subhairline). Recordings were reviewed independently, with the identity of the patients concealed.
Results: Seizures were found in 8% of patients. Sensitivity for detecting patients with seizures using the subhairline system was 0.54 [95% confidence interval (95% CI), 0.29-0.77] with specificity of 1.00 (95% CI, 0.97-1.00) and positive predictive value of 1.00 (95% CI, 0.65-1.00). For detecting interictal epileptiform activity, we found sensitivity to be 0.60 (95% CI, 0.46-0.74), specificity to be 0.94 (95% CI, 0.88-0.97), and positive predictive value to be 0.81 (95% CI, 0.65-0.91). Performance was poor for triphasic waves, alpha/theta/spindle coma, and suppression.
Conclusions: The subhairline montage shows excellent specificity for detecting patients with seizure activity but has limited sensitivity. It has relatively poor performance for other EEG phenomena, but further applications in trending and assessing reactivity should be assessed in further studies.