Time-segmented, feedback-guided tiered warming during radical esophagectomy: a single-center retrospective cohort study.

Objective: To evaluate whether a protocolized intraoperative temperature-management strategy based on time-segmented warning thresholds and tiered, feedback-guided warming interventions reduces hypothermia and improves early postoperative recovery in patients undergoing radical esophagectomy. Methods: This single-center retrospective comparative cohort study included 116 adults who underwent elective radical esophagectomy at The affiliated Hospital of Southwest Medical University between August 2024 and January 2026 and were managed with either a time-segmented, feedback-guided tiered warming protocol (n = 58) or conventional warming (n = 58). Core temperature was monitored continuously with a lower esophageal probe until the start of esophageal mobilization and with a nasopharyngeal probe thereafter, and was recorded every 15 min. All warming adjustments were implemented manually by the clinical team according to predefined thresholds. The primary endpoint was intraoperative hypothermia, defined as a core temperature below 36 °C at any time from incision to skin closure. Secondary endpoints included severe hypothermia, nadir temperature, hypothermia burden, postoperative shivering, time to extubation, post-anesthesia care unit (PACU) recovery time, coagulation indices, 30-day complications, length of stay, and hospital cost. Results: Intraoperative hypothermia occurred in 53 of 58 control patients (91.4%) and 5 of 58 intervention patients (8.6%) (risk ratio 0.09, 95% CI 0.04-0.22; P < 0.001). Severe hypothermia below 35.5 °C developed in 27 control patients (46.6%) and in none of the intervention patients. Median nadir core temperature was 35.55 °C in the control group and 36.27 °C in the intervention group (P < 0.001). Compared with conventional warming, the protocol reduced postoperative shivering (5.2% vs 51.7%), shortened time to extubation by 8.9 min, lowered the rate of any postoperative complication (34.5% vs 62.1%; P = 0.005), shortened hospital stay (12.2 vs 13.5 days; P = 0.010), and reduced mean in-hospital cost by 8,911 CNY (P < 0.001). The treatment effect was materially unchanged in adjusted analyses using Firth penalized logistic regression (adjusted odds ratio 0.004, 95% CI 0.000-0.024) and modified Poisson regression (adjusted risk ratio 0.10, 95% CI 0.04-0.23). Conclusion: In adults undergoing radical esophagectomy, a time-segmented, feedback-guided tiered warming protocol substantially reduced intraoperative hypothermia and was associated with faster early recovery, fewer postoperative complications, shorter hospitalization, and lower cost. Because the intervention was a multicomponent warming bundle, the contributions of its individual components cannot be separated, and the associations with complications, hospital stay, and cost should be regarded as hypothesis-generating. The protocol uses standard perioperative equipment and can be implemented without specialized hardware.