Recently, I read an interesting article:
Mather et al. «Intraoperative Frontal Electroencephalogram Alpha Power Is Associated with Postoperative Mortality and Other Adverse Outcomes.» Anesthesiology 2025; 142:500–10.
The study suggests that lower intraoperative alpha power in the EEG is independently associated with increased postoperative mortality. While this is an intriguing hypothesis, I believe we must be cautious with these conclusions.
Critical Points to Consider
1. Lack of Information on Consciousness State Titration
The study reports anaesthetic doses, including
- Propofol median total dose: 200 mg (IQR 150-260 mg)
- Sevoflurane mean end-tidal concentration: 1.00% (IQR 0.00-1.39%)
- Nitrous oxide use: 60.3% of patients (a lot)
- Propofol mean total dose: 200 mg (IQR 150-260 mg)
- Sevoflurane mean end-tidal concentration: 1.00% (IQR 0.00-1.39%)
- Use of nitrous oxide: 60.3% of patients.
However, it is not specified how the anaesthesia was titrated for each patient. - No mention of burst suppression, a key indicator of excessive anaesthetic titration.
- No BIS, Entropy, Narcotrend, Conox monitoring, making it impossible to compare alpha power with other validated EEG monitoring data.
- No analysis of mean arterial pressure (MAP), so we cannot rule out cerebral hypoperfusion as a cause of reduced alpha power.
Without data on titration strategies, we cannot determine whether lower alpha power reflects patient frailty or simply excessive anesthesia administration.
2. Retrospective Study Design and Selection Bias
- This was a retrospective observational study, meaning it cannot establish causality, only statistical associations.
- Only patients with artifact-free EEG recordings were included, introducing selection bias.
- Other critical confounders (e.g., opioid use, neuromuscular blockade, ventilatory parameters) were not considered.
A prospective study is needed to validate these findings before drawing clinical conclusions.
3. Failure to Differentiate Between Patient Frailty and Over-Titration of Anesthesia
Low intraoperative alpha power could result from:
1. Pre-existing neurological frailty ? reduced cortical connectivity and baseline EEG activity.
2. Inadequate titration of consciousness state ? deeper hypnotic states leading to EEG suppression.
3. Cerebral hypoperfusion ? low MAP causing decreased cortical activity.
The study does not provide enough data to distinguish between these scenarios.
4 Unusual Anesthetic Practice: Predominance of Volatile Agents & Nitrous Oxide
One striking aspect of this study is the low use of total intravenous anesthesia (TIVA) and the high prevalence of volatile agents and nitrous oxide:
- 98.9% of patients received propofol, but only as an induction agent.
- 72.1% received sevoflurane as the primary anesthetic.
- 60.3% received nitrous oxide, a practice that has largely declined in LATAM and Europe due to concerns about environmental impact and neurotoxicity.
The study did not investigate whether the choice of anaesthesia influenced the EEG findings, and it could be questioned whether the findings would be the same in a cohort with predominantly intravenous anaesthesia.
5 No Comparison with Established Risk Prediction Models
- Several validated models predict postoperative mortality (ASA, POSSUM, ACS-NSQIP).
- This study does not compare alpha power with these models, making it unclear whether it adds predictive value.
Final Thoughts
At this stage, it is premature to consider intraoperative alpha power as an independent biomarker for mortality.
More rigorous, prospective studies are needed, incorporating Raw EEG/DSA, MAP, burst suppression analysis, and direct comparisons with existing risk models.
Additionally, the high use of volatile agents and nitrous oxide raises the question of whether these findings would hold true in a TIVA-based anesthesia protocol.
What are your thoughts on this? Have you encountered similar discussions in your practice?