目的 分析腹腔镜妇科手术全身麻醉患者围手术期睡眠障碍对麻醉苏醒和术后疼痛的影响。 方法 回顾性地选取2024年3月至2025年3月医院收治的160例腹腔镜妇科手术患者作为研究对象，按照术前匹兹堡睡眠质量指数（PSQI）分为NSD组（PSQI ≤ 5分，n = 80）和SD组（PSQI > 5分，n = 80）。比较两组麻醉苏醒［苏醒时间、改良Aldrete评分、定向力恢复时间、苏醒期不良反应、瑞芬太尼总量、丙泊酚总量、七氟醚MAC值、停药至拔管时间、术中平均动脉压（MAP）、术中心率（HR）、补救镇痛使用率］、疼痛指标［视觉模拟评分（Visual Analogue Scale， VAS）、阿片类药物用量、自控镇痛（patient-controlled analgesia， PCA）按压次数］，并采用多元线性回归、决策树模型及中介效应分析睡眠质量、麻醉苏醒和术后疼痛之间的关系。 结果 NSD组患者的苏醒时间和定向力恢复时间短于SD组（P < 0.05），改良Aldrete评分更高（P < 0.05）。SD组的丙泊酚总用量、术中HR及补救镇痛使用率高于NSD组（P < 0.05）。NSD组不良反应发生率、术后各时点VAS评分、阿片类药物用量和PCA按压次数均低于SD组（P < 0.05）。多元回归显示PSQI每增加1分，苏醒时间延长0.63 min、定向力恢复延迟0.55 min、Aldrete评分降低0.05分、术后VAS评分升高0.20 ~ 0.22分、阿片用量增加0.87 mg、PCA按压次数增加0.98次（P < 0.05）。扩展模型显示，纳入术中因素后苏醒时间（R2 = 0.456）和术后疼痛（R2 = 0.524）的预测效能显著提升，停药至拔管时间、瑞芬太尼用量和BIS值是苏醒时间的关键预测因子；七氟醚MAC值和MAP波动对术后疼痛预测贡献显著，PSQI评分在两组模型中均保持独立预测作用（β = 0.421/0.312）。决策树分析确认PSQI是预测麻醉苏醒和术后疼痛的核心因素（重要性51%），PSQI < 5.5患者苏醒时间显著短于PSQI ≥ 5.5者（18.68 min vs. 23.29 min）。含麻醉苏醒指标的疼痛预测模型表现优于不含该指标的模型（R2 = 0.391 vs. 0.336）。中介效应分析显示，改良Aldrete评分在PSQI与术后疼痛间表现出轻微中介作用（中介比例5.98%），而PSQI对疼痛的直接效应占总效应的93.8%，提示睡眠障碍主要通过其他机制影响术后疼痛。 结论 睡眠障碍通过延长麻醉苏醒时间、降低改良Aldrete评分影响术后疼痛体验，其中直接效应占主导地位，提示临床应关注睡眠质量评估和干预以改善术后疼痛管理。

Objective To analyze the influence of perioperative sleep disorders on anesthesia recovery and postoperative pain in patients undergoing laparoscopic gynecological surgery under general anesthesia. Methods A total of 160 patients who underwent laparoscopic gynecological surgery at the hospital from March 2024 to March 2025 were retrospectively selected as research subjects. According to the preoperative Pittsburgh Sleep Quality Index （PSQI）， they were divided into the non-sleep-disorder （NSD） group （PSQI ≤ 5 points， n = 80） and the sleep-disorder （SD） group （PSQI > 5 points， n = 80）.The following aspects were compared between the two groups： anesthesia recovery indicators ［recovery time， modified Aldrete score， orientation recovery time， adverse reactions during the recovery period， total dosage of remifentanil， total dosage of propofol， minimum alveolar concentration （MAC） value of sevoflurane， time from drug discontinuation to extubation， intraoperative mean arterial pressure （MAP）， intraoperative heart rate （HR）， and rescue analgesia utilization rate］； pain indicators ［Visual Analogue Scale （VAS） score， opioid dosage， and number of patient-controlled analgesia （PCA） presses］. Moreover， multivariate linear regression， decision tree modeling， and mediation effect analysis were employed to explore the relationships among sleep quality， anesthesia recovery， and postoperative pain. Results The recovery time and orientation recovery time in the NSD group were shorter than those in the SD group （P < 0.05）， and the modified Aldrete score was higher （P < 0.05）. The total propofol dosage， intraoperative HR， and rescue analgesia utilization rate in the SD group were higher than those in the NSD group （P < 0.05）. The incidence of adverse reactions， VAS scores at various postoperative time points， opioid dosage， and number of PCA presses in the NSD group were all lower than those in the SD group （P < 0.05）. Multivariate regression analysis indicated that for every 1-point increase in PSQI， recovery time was prolonged by 0.63 minutes， orientation recovery time was delayed by 0.55 minutes， the modified Aldrete score decreased by 0.05 points， postoperative VAS score increased by 0.20 ~ 0.22 points， opioid dosage increased by 0.87 mg， and the number of PCA presses increased by 0.98 （P < 0.05）. The extended model demonstrated that after incorporating intraoperative factors， the predictive performance for recovery time （R² = 0.456） and postoperative pain （R² = 0.524） was significantly enhanced. Time from drug discontinuation to extubation， remifentanil dosage， and bispectral index （BIS） value were key predictive factors for recovery time； sevoflurane MAC value and MAP fluctuations made significant contributions to postoperative pain prediction. PSQI score maintained an independent predictive role in both models （β = 0.421/0.312）. Decision tree analysis confirmed that PSQI was the core factor for predicting anesthesia recovery and postoperative pain （importance： 51%）. Recovery time in patients with PSQI < 5.5 was significantly shorter than in those with PSQI ≥ 5.5 （18.68 minutes vs. 23.29 minutes）. The pain prediction model incorporating anesthesia recovery indicators exhibited better performance than the model without such indicators （R² = 0.391 vs. 0.336）.Mediation effect analysis revealed that the modified Aldrete score exerted a mild mediating effect between PSQI and postoperative pain （mediation proportion： 5.98%）. The direct effect of PSQI on pain accounted for 93.8% of the total effect， suggesting that sleep disorders mainly affect postoperative pain through other mechanisms. Conclusions Sleep disorders affect postoperative pain experience by prolonging anesthesia recovery time and reducing the modified Aldrete score. The direct effect is dominant， indicating that clinical practice should focus on sleep quality assessment and intervention to improve postoperative pain management.