New studies highlight opportunities to improve pediatric anesthesia safety, comfort

Editor's Note

Artificial intelligence (AI) and evidence-based fasting practices could significantly enhance safety and comfort for children undergoing surgery, according to research presented at the ANESTHESIOLOGY® 2025 annual meeting that took place on October 10–14.

One study found AI systems outperform standard methods in key pediatric anesthesia tasks, including selecting breathing tube size, tracking oxygen levels, and assessing pain. In studies spanning tens of thousands of cases, AI tools predicted oxygen desaturation up to a full minute before standard alarms and gauged postoperative pain in toddlers with 95% accuracy. Researchers described AI as a “co-pilot” that supports, rather than replaces, anesthesiologists by continuously analyzing real-time data to flag complications earlier and tailor care to each child’s physiology.

In a separate large-scale review of more than 70,000 pediatric surgeries, investigators discovered that most children—and nearly 80% of infants—fasted from clear liquids for at least twice as long as guidelines recommend. The median fasting time was three to four times longer than the 2-hour American Society of Anesthesiologists guideline, which permits clear liquids such as water or juice up to two hours before anesthesia. Prolonged fasting was linked to dehydration, heightened stress, nausea, and delayed recovery. Researchers attributed the issue to outdated “nothing after midnight” routines and limited awareness of updated recommendations.

Together, the studies show two practical pathways for improving pediatric perioperative care: integrating AI-assisted monitoring to enhance precision and safety, and encouraging adherence to modern fasting guidelines to reduce discomfort and promote recovery. Both, experts said, depend on well-informed perioperative teams and parent engagement before surgery.