Non-Linear Association Between Intraoperative Remifentanil Infusion and Postoperative Opioid Consumption After Uniportal VATS: A Propensity-Matched Retrospective Cohort Study.

Background and Objective: Remifentanil may be associated with increased postoperative opioid requirements, but the dose-response pattern in uniportal video-assisted thoracoscopic surgery (U-VATS) is uncertain. We investigated the association between the time-averaged intraoperative remifentanil infusion rate and 24-hour postoperative opioid consumption and explored a potential non-linear transition point. Methods: This single-center retrospective cohort study included 356 patients who underwent elective U-VATS lung resection at Shanghai Chest Hospital in 2025. The primary outcome was cumulative intravenous morphine milligram equivalents (MME) within 24 hours postoperatively. Propensity score matching (PSM), multivariable linear regression, doubly robust estimation, and restricted cubic splines (RCS) were used to assess adjusted linear and non-linear associations. Results: PSM yielded 89 matched pairs (N = 178). In the fully adjusted matched-cohort model, each 0.1 μg/kg/min increase in the remifentanil infusion rate was associated with 7.03 mg higher 24-hour MME consumption (95% CI, 4.10 to 9.96; P < 0.001). As a secondary behavioral outcome, the group receiving ≥0.20 μg/kg/min had more 24-hour PCA pump attempts than the group receiving <0.20 μg/kg/min (median, 5 vs 3; P < 0.001), despite similar resting pain scores. The RCS analysis indicated a non-linear association (P for non-linearity = 0.032), with the curve crossing the reference line near 0.20 μg/kg/min. Conclusion: Higher time-averaged intraoperative remifentanil exposure was associated with greater postoperative opioid consumption after U-VATS. The observed transition near 0.20 μg/kg/min should be interpreted as a hypothesis-generating inflection point rather than a definitive safety threshold and requires prospective validation.