Increases in Electroencephalogram and Electromyogram Variability Are Associated with an Increased Incidence of Intraoperative Somatic Response

BACKGROUND: sBIS, the variability of the Bispectral Index (BIS), sEMG, the variability of facial electromyogram power (EMG), and the Composite Variability Index (CVI) are 3 new measures of electroencephalogram and EMG variability. CVI is a single measure of the combined variability in BIS and EMG. We investigated whether increases in these variables are associated with intraoperative somatic responses. METHODS: This multicenter study included 120 patients undergoing elective, noncardiac surgery from 4 different sites. General anesthesia was maintained using propofol and remifentanil at 2 of the sites and sevoflurane and remifentanil at the 2 other sites. Propofol or sevoflurane was adjusted to maintain BIS between 45 and 60. Clinicians were blinded to CVI (v2.0) at all times, and remifentanil infusions were adjusted at the discretion of the clinician. The times of all intraoperative somatic events, defined as movement, grimacing, or eye opening, were recorded. Offline, the maintenance phase of each case was divided into consecutive, nonoverlapping, 10-minute segments. Segments were identified as containing a somatic event or containing no events. For each segment, mean sBIS, sEMG, and CVI and the heart rate (HR) range and mean arterial blood pressure range were calculated. To quantify how effectively each variable discriminated between somatic event segments and nonevent segments, we computed the area under the receiver operating characteristic (ROC) curve for each variable. Finally, we observed the time course of sBIS, sEMG, CVI, and the HR range before each somatic event and characterized the earliest time before the somatic event at which each variable was able to discriminate between the somatic events and a specified set of nonevents. RESULTS: The analysis included 33 somatic event segments and 829 nonevent segments from 105 surgical cases. The areas under the ROC curve (±SE) for sBIS, sEMG, and CVI were 0.83 ± 0.04, 0.92 ± 0.02, and 0.89 ± 0.03, respectively. The areas under the ROC curve for HR range and mean arterial blood pressure range were 0.77 ± 0.03 and 0.68 ± 0.05, respectively. CVI, sBIS, and sEMG all demonstrated higher average values before upcoming somatic events when compared with nonevents. HR range only showed a difference within a few seconds before the somatic event. CONCLUSION: sBIS, sEMG, and CVI, measures of electroencephalogram and EMG variability, increased when intraoperative somatic events occurred. sBIS, sEMG, and CVI discriminated between 10-minute segments that contained a somatic event and those segments that did not contain an event better than changes in HR and mean arterial blood pressure. Furthermore, CVI increases before somatic events began earlier than HR changes and may provide caregivers with an early warning of potentially inadequate antinociception.