Residency · Residency · Anesthesiology
Processed EEG Monitoring (BIS, SedLine, Entropy) and Awareness Under Anesthesia
Introduction
Intraoperative awareness with recall is a devastating complication that affects approximately 1 to 2 per 1,000 general anesthetics. Processed electroencephalographic (EEG) monitoring was developed to provide an objective, real-time measure of anesthetic depth, complementing clinical signs and end-tidal agent monitoring. This chapter examines the technology, clinical evidence, and practical application of these monitors.
The Raw EEG in Anesthesia
EEG Frequency Bands
The EEG frequency bands relevant to anesthesia include beta (13 to 30 Hz), which reflects the awake, alert state; alpha (8 to 13 Hz), which reflects relaxed wakefulness and light sedation, with prominent frontal alpha being a hallmark of propofol and volatile anesthetic effect; theta (4 to 8 Hz), which indicates moderate anesthetic depth; and delta (0.5 to 4 Hz), which indicates deep anesthesia with slow-wave activity. Burst suppression, consisting of alternating periods of high-amplitude activity and isoelectric silence, indicates very deep anesthesia or cerebral injury.
EEG Signatures of Common Anesthetic Agents
Propofol produces a dose-dependent progression from beta activation to frontal alpha predominance to slow-delta with burst suppression at high doses. Sevoflurane and isoflurane produce patterns similar to propofol but with more theta activity at surgical depth. Ketamine produces paradoxical high-frequency gamma and beta activity and does not produce the classical alpha-delta pattern. Nitrous oxide reduces alpha power and can mask other EEG signatures. Dexmedetomidine produces spindle-like activity resembling natural sleep (Stage 2 NREM). Opioids have minimal direct EEG effect at clinical doses and do not reliably produce unconsciousness alone.
Processed EEG Monitors
Bispectral Index (BIS)
The BIS, manufactured by Medtronic, is the most widely studied processed EEG monitor. It derives a dimensionless number from 0 to 100 using a proprietary algorithm that incorporates time-domain analysis (burst suppression ratio), frequency-domain analysis (power spectral analysis), and bispectral analysis (phase coupling between frequency components). The target range for general anesthesia is 40 to 60; values below 40 suggest excessive depth, and values above 60 suggest light anesthesia. It uses a single-channel frontally placed sensor.
SedLine (Patient State Index - PSI)
The SedLine, manufactured by Masimo, calculates the Patient State Index (PSI) on a range of 0 to 100. It uses 4-channel bilateral frontal EEG, which allows detection of hemispheric asymmetry. The algorithm is based on power spectral analysis, phase relationships, and coherence. The target range for general anesthesia is 25 to 50. It displays the density spectral array (DSA), a color-coded spectrogram showing frequency content over time, which aids pattern recognition and drug-specific EEG interpretation.
Entropy (GE Healthcare)
The Entropy monitor calculates two indices from the frontal EEG. State Entropy (SE), ranging from 0 to 91, is based on EEG frequencies from 0.8 to 32 Hz and reflects cortical state. Response Entropy (RE), ranging from 0 to 100, includes higher frequencies up to 47 Hz that capture frontal EMG activity. An RE minus SE gap greater than 10 suggests inadequate analgesia or impending arousal, as EMG activation precedes EEG change. The target SE for general anesthesia is 40 to 60.
| Feature | BIS (Medtronic) | SedLine/PSI (Masimo) | Entropy (GE Healthcare) |
|---|---|---|---|
| Index range | 0–100 | 0–100 (PSI) | SE: 0–91; RE: 0–100 |
| GA target range | 40–60 | 25–50 | SE: 40–60 |
| EEG channels | 1 (single frontal) | 4 (bilateral frontal) | 1 frontal |
| Hemispheric asymmetry | No | Yes | No |
| Spectrogram (DSA) display | No (standard) | Yes | No |
| Unique feature | Most studied; widest adoption | Bilateral monitoring; DSA aids interpretation | RE-SE gap detects inadequate analgesia |
| EMG sensitivity | Susceptible to artifact | Susceptible to artifact | RE specifically captures EMG |
| EEG Frequency Band | Frequency (Hz) | Anesthetic Significance |
|---|---|---|
| Beta | 13–30 | Awake, alert; also paradoxical activation with ketamine |
| Alpha | 8–13 | Relaxed wakefulness; frontal alpha = hallmark of propofol/volatile effect |
| Theta | 4–8 | Moderate anesthetic depth |
| Delta | 0.5–4 | Deep anesthesia; slow-wave activity |
| Burst suppression | Variable | Very deep anesthesia or cerebral injury; reduce dose |
Clinical Evidence and Controversy
Evidence Supporting Processed EEG Monitoring
The B-Aware Trial (Myles 2004) demonstrated that BIS-guided anesthesia reduced awareness in high-risk patients by 82%. The B-Unaware Trial (Avidan 2008) found no significant difference between BIS-guided and end-tidal anesthetic concentration (ETAC)-guided protocols. The BAG-RECALL Trial (Avidan 2011) found that ETAC-guided protocol was non-inferior to BIS-guided protocol for preventing awareness. Meta-analyses suggest BIS monitoring may reduce awareness incidence, but ETAC-guided anesthesia is a reasonable alternative.
Limitations of Processed EEG Monitors
Several limitations exist. The proprietary algorithms make the exact processing a "black box." EMG contamination from frontalis muscle activity can falsely elevate BIS/PSI values. Ketamine and N2O produce paradoxical increases in BIS values despite adequate anesthesia. Neuromuscular blockade eliminates EMG artifact, which may cause an artifactual drop in index values. Patient variability means the same index value may represent different levels of consciousness in different patients. The monitors are not validated for specific populations, including children under 1 year, patients with neurological disease, and patients on psychoactive medications.
The Argument for Raw EEG and Spectrogram Interpretation
Experts increasingly advocate for direct spectrogram interpretation rather than reliance on a single index number. Spectrograms allow the clinician to identify drug-specific EEG signatures and titrate accordingly. Burst suppression is readily visible on the spectrogram and warrants dose reduction. Spectrogram interpretation also enables recognition of artifacts (EMG, electrical interference) that would mislead processed indices.
Awareness Under Anesthesia
Definition and Incidence
Accidental awareness during general anesthesia (AAGA) is defined as unintended consciousness with or without explicit recall. The incidence is 0.1 to 0.2% (1 to 2 per 1,000) in the general surgical population. Higher risk populations include cardiac surgery (up to 1%), obstetric cesarean section under GA, trauma, and patients receiving TIVA without processed EEG monitoring.
Risk Factors
Patient factors include female sex, younger age, obesity, chronic opioid/benzodiazepine/alcohol use, and prior awareness. Surgical factors include cardiac surgery, cesarean section, trauma surgery, and rigid bronchoscopy. Anesthetic factors include the use of neuromuscular blockers (which remove the ability to move), TIVA without monitoring, light anesthesia technique, and drug errors.
Consequences
Post-traumatic stress disorder occurs in up to 70% of awareness victims. Other consequences include sleep disturbances, anxiety, depression, and fear of future anesthetics. Awareness claims are among the most common in anesthesia malpractice litigation.
Prevention Strategies
Prevention includes maintaining an end-tidal volatile agent concentration of 0.7 MAC or greater (or age-adjusted equivalent), using processed EEG monitoring in high-risk patients (ASA Practice Advisory), avoiding gaps in drug delivery by checking IV lines, pumps, and vaporizer levels, administering benzodiazepines (midazolam) in high-risk cases as amnestic adjuncts, and using a structured checklist before incision to confirm delivery of the hypnotic and verify EEG/ETAC values. During TIVA, the pump must be functioning and the IV patent, and processed EEG monitoring should be used as standard.
Management of Suspected Awareness
If awareness is suspected intraoperatively, anesthesia should be immediately deepened and midazolam 2 to 4 mg IV administered. Postoperatively, the patient should be visited within 24 hours using a modified Brice questionnaire that asks: What is the last thing you remember before going to sleep? What is the first thing you remember upon waking up? Do you remember anything between going to sleep and waking up? Did you dream during your procedure? What was the worst thing about your operation? The conversation should be documented, and psychological support and referral offered if needed. The event should be reported through institutional quality improvement channels.
Clinical Pearls
No processed EEG monitor reliably prevents awareness in all circumstances; they are adjuncts to clinical judgment, end-tidal agent monitoring, and vigilance. Learning to read the raw EEG spectrogram is valuable: frontal alpha power is the hallmark of adequate hypnotic drug effect with propofol and volatile agents, and its disappearance signals either too-light or too-deep anesthesia. Neuromuscular blockers increase the risk of undetected awareness by abolishing the patient's ability to move, and processed EEG monitoring should be used whenever NMBs are administered. A structured postoperative interview should always be performed if awareness is suspected, as early recognition and psychological support significantly reduce the incidence of PTSD.
References
- Mashour GA, Avidan MS. Intraoperative awareness: controversies and non-controversies. Br J Anaesth. 2015;115(Suppl 1):i20-i26.
- Purdon PL, Sampson A, Pavone KJ, Brown EN. Clinical electroencephalography for anesthesiologists. Part I: Background and basic signatures. Anesthesiology. 2015;123(4):937-960.
- Myles PS, Leslie K, McNeil J, Forbes A, Chan MT. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet. 2004;364(9444):1757-1763.
- Pandit JJ, Andrade J, Bogod DG, et al. 5th National Audit Project (NAP5) on accidental awareness during general anaesthesia. Br J Anaesth. 2014;113(4):549-559.