目的 评价瑞马唑仑对老年胸腔镜手术患者术后谵妄（postoperative delirium， POD）发生率的影响。 方法 选取徐州医科大学附属医院2024年10月至2025年4月择期行单侧胸腔镜手术的114例肺癌患者。将患者随机分配至瑞马唑仑组（R组）和丙泊酚组（P组），分别采用瑞马唑仑和丙泊酚进行麻醉诱导与维持。其中R组患者在手术结束时静脉注射氟马西尼0.5 mg进行特异性拮抗。比较两组术后第1天、第3天的谵妄发生率与术后恢复质量（Quality of Recovery-15，QoR-15）评分，并记录术中血流动力学参数、阿片类用药总量及液体平衡情况；术毕气管导管拔除时间及不良反应的发生率等。 结果 两组POD发生率、QoR-15评分差异均无统计学意义（P > 0.05）。与P组比较，R组患者气管拔除时间缩短（P < 0.05），患者术中的血流动力学更平稳，术中低血压、心动过缓发生率及血管活性药物使用量更少（P < 0.05）。R组在麻醉后监护病房（post anesthesia care unit， PACU）中抢救性镇痛药使用率更低（P < 0.05）。两组术中舒芬、瑞芬太尼用量、液体出入量、脑电双频指数（bispectral index， BIS）及恶心呕吐等不良反应发生情况差异无统计学意义（P>0.05）。 结论 在择期行胸腔镜手术的老年患者中，与丙泊酚相比，应用瑞马唑仑麻醉诱导及维持，POD的发生率未观察到显著差异且术后恢复质量无差异。此外，接受瑞马唑仑的患者血管活性药物需求更少，患者拔管时间更短。

Objective To evaluate the effect of remimazolam on the incidence of postoperative delirium （POD） in elderly lung cancer patients undergoing thoracoscopic surgery. Methods A total of 114 elderly patients who underwent unilateral thoracoscopic lung surgery at the Affiliated Hospital of Xuzhou Medical University from October 2024 to April 2025 were recruited in this trial. Patients were randomly assigned to remimazolam group （group R） and propofol group （group P）. Anesthesia induction and maintenance were performed with remimazolam and propofol， respectively. In R group， 0.5 mg of flumazenil was intravenously injected at the end of the surgery for specific antagonism. The incidence of delirium and Quality of Recovery-15 （QoR-15） scores on the first and third postoperative days were compared between the two groups. Intraoperative hemodynamic parameters， total opioid dosage， fluid balance， tracheal tube extubation time after surgery， and the incidence of adverse reactions were recorded. Results There were no statistically significant differences between the groups in the incidence of POD or in QoR-15 scores （P > 0.05）. Compared with group P， patients in group R had a shorter extubation time （P < 0.05）， more stable hemodynamics， lower incidences of intra-operative hypotension and bradycardia， and reduced requirement for vasoactive drugs （P < 0.05）. The consumption of rescue analgesics in the post-anesthesia care unit （PACU） was also lower in group R （P < 0.05）. No significant differences were observed between the groups in the amounts of sufentanil and remifentanil administered， fluid balance， bispectral index （BIS） values， or the incidence of nausea and vomiting （P > 0.05）. Conclusion In elderly patients undergoing elective thoracoscopic surgery， remimazolam-based induction and maintenance of anesthesia did not significantly alter the incidence of POD or compromise postoperative recovery quality compared with propofol. However， the remimazolam group required fewer vasoactive agents and exhibited a shorter tracheal extubation time.