Benzodiazepine-Free Cardiac Anesthesia for Reduction of Postoperative Delirium: A Cluster Randomized Crossover Trial.

Summary

Across 20 centers and nearly 20,000 cardiac surgeries, an institutional policy restricting intraoperative benzodiazepines did not significantly reduce delirium within 72 hours (aOR 0.92, P=0.07) compared with a liberal policy. No increase in patient-reported intraoperative awareness was observed.

Key Findings

• Delirium within 72 hours: 14.0% (restricted) vs 14.9% (liberal); aOR 0.92 (95% CI 0.84–1.01), P=0.07
• High adherence to assigned policy: 90.9% (restricted) vs 93.2% (liberal)
• No patient spontaneously reported intraoperative awareness
• Pragmatic, patient- and assessor-blinded, cluster randomized crossover across 20 centers (n=19,768)

Clinical Implications

Routine restriction of intraoperative benzodiazepines should not be expected to meaningfully reduce postoperative delirium at the population level; individualized decisions remain reasonable and further patient-level trials are warranted.

Limitations

• Delirium was detected via routine clinical screening, potentially introducing variability
• Cluster policy intervention may dilute patient-level effects; effect estimate borderline (P=0.07)

Future Directions

Conduct patient-level RCTs targeting benzodiazepine avoidance or dose minimization, identify high-risk subgroups, and harmonize delirium assessment with research-grade tools.