Biasing effect of the electromyogram on BIS: a controlled study during high-dose fentanyl induction

A biasing effect of the electromyogram (EMG) on the Bispectral Index (BIS) may explain discrepancies in previous studies assessing BIS in the presence of neuromuscular activity. Our aims were: to evaluate variations of BIS in the presence of high EMG activity associated with muscular rigidity after administration of high-dose fentanyl; to compare muscular rigidity, as measured by the EMG variable of the BIS monitor, in patients who were administered two different dosages of fentanyl at induction of cardiac anaesthesia.

26 patients undergoing CABG surgery, after premedication with morphine 0.15 mg/kg, were randomized to receive either fentanyl 50 mcg/kg (group F) or fentanyl 10 mcg/kg plus etomidate 0.2 mg/kg (group EF). The induction dose was administered over 2 minutes. Patients were manually ventilated with O2 via face mask. Five minutes after induction was complete, patients were clinically assessed using the Responsiveness portion of the Observer's Assessment of Alertness/Sedation scale (OAAS). Haemodynamic data were recorded and arterial blood samples obtained at the time of OAAS observation. Patients were administered a neuromuscular blocking agent only after the OAAS assessment. BIS (3.4) was recorded from an A-2000 EEG monitor (Aspect Medical Systems) using disposable sensors (BIS Sensor, Aspect Medical Systems) applied per manufacturer's instructions. Data were recorded on a PC for off-line analysis.

At the time of OAAS observation, mean (95% CI) BIS in group F was 85 (77-92) compared to 67 (56-79) in group EF (p = 0.01). Similarly, mean (95% CI) EMG was 50 dB (45-56) in F and 41 dB (35-47) in EF (p = 0.01). Correlation between BIS and EMG was very high (r2 = 0.88). OAAS scores were significantly higher in group F (p = 0.03). Non significant correlation was observed between BIS and OAAS scores (r2 = 0.32, p = 0.1). Backward stepwise multiple regression analysis including EMG, pH, CO2, O2 and OASS scores showed EMG as strong predictor of BIS (p < 0.0001, r2 = 0.7). Regression of EMG against BIS yielded the equation: BIS = 3.7 + (1.6 x EMG).

During fentanyl-induced muscular rigidity BIS recordings reflect EMG variations. When assessing BIS in the absence of neuromuscular blockade, it is necessary to evaluate the effect of EMG on BIS before making conclusions about depth of sedation. Fentanyl-induced rigidity appears to be a dose-related phenomenon which the EMG variable of BIS 3.4 is able to quantify.