Reinforcement Learning Automated Anesthesia System Matches Anesthesiologists on Hypoxemia in Multicenter Gastrointestinal Endoscopy Trial

Bing and colleagues conducted a multicenter randomized trial comparing their reinforcement learning automated anesthesia system AAS-GE with manual management by anesthesiologists for ciprofol infusion in 418 gastrointestinal endoscopy patients. The primary safety endpoint, hypoxemia incidence, was similar between groups (14.42% vs 14.29%, P=0.968), but body movement was more frequent with the automated system (17.79% vs 6.67%, P=0.001), and induction time was shorter (1.55 vs 1.90 min, P<0.001).