/practice/journal-club / 2019
ENGAGES: EEG-Guided Anesthesia + Postop Delirium
Wildes TS et al. Effect of Electroencephalography-Guided Anesthetic Administration on Postoperative Delirium Among Older Adults Undergoing Major Surgery (ENGAGES). JAMA 2019;321:473-483.
BIS · EEG · delirium · geriatric
Hook
EEG-guided anesthesia did NOT reduce postop delirium in older adults — a clean negative result.
Population, Intervention, Comparison, Outcome
- Population
- 1,232 adults ≥60 yo undergoing major surgery expected to require general anesthesia and ICU/floor admission.
- Intervention
- EEG-guided anesthesia: target processed-EEG (BIS) 40-60, avoid burst suppression.
- Comparison
- Usual practice — anesthesiologist could view EEG but no specific target.
- Outcome
- Postoperative delirium days 1-5, assessed by trained CAM-ICU/CAM-S.
Methods
Single-center RCT (Washington University). Both arms had EEG monitor visible. Anesthesiologists in EEG-guided arm received structured feedback on hypnotic depth.
Findings
- Delirium: 26% EEG-guided vs 23% usual care (NS).
- Volatile usage: lower in EEG-guided arm (0.69 vs 0.80 MAC) — protocol followed.
- Burst suppression episodes: shorter in EEG-guided arm.
- BUT: no translation to clinical delirium outcome.
Clinical takeaway
Avoiding burst suppression alone doesn't prevent postop delirium — pre-existing cognitive vulnerability dominates. The CODA + AwaRE-2 trials suggest depth + AGE + comorbidity matter more than BIS targeting in isolation. Clinical implication: don't rely on BIS as a delirium-prevention tool. DO continue to use EEG to identify burst suppression (which IS associated with worse cognitive outcomes in observational data) + to confirm adequate depth especially in TIVA. Multifactorial delirium prevention bundles (sleep hygiene, mobilization, family presence, multimodal opioid-sparing) outperform any single intervention.
Limitations
- Single-center, urban academic — limited generalizability.
- Delirium is multifactorial; small effect of one intervention may have been masked by noise.
- Did not stratify by frailty score, which strongly predicts post-op delirium.
- Did not examine 90-day cognitive outcomes (longer-term effects could differ).
Discussion questions
- How does ENGAGES change (or not change) your use of EEG monitoring in geriatric anesthesia?
- What multifactorial delirium-prevention strategies does your facility implement, and which ones have the best evidence?
- Should burst suppression duration become a documented quality metric even though ENGAGES didn't tie it to clinical outcome?