目的 探讨镇痛药剂量对俯卧位通气（prone position ventilation，PPV）患者肠内喂养不耐受（enteral feeding intolerance，EFI）的影响，明确其剂量-反应关系，为危重患者镇痛与营养管理的平衡提供依据。方法 回顾性收集2020年1月—2024年12月入住南京市某三级甲等医院ICU的630例行肠内营养支持的PPV患者的临床资料。根据是否发生EFI分为耐受组与不耐受组，收集患者的人口学资料、疾病与治疗相关资料、肠内营养支持相关资料。通过Logistic回归探究镇痛药剂量与EFI的独立相关性，采用限制性立方样条图（restricted cubic splines，RCS）分析镇痛药与EFI的剂量-反应关系，并通过森林图可视化各亚组中镇痛药剂量与EFI的关系。结果 630例患者中肠内营养耐受366例，不耐受264例，EFI发生率为41.90%。264例EFI患者中有70例出现≥2种EFI症状，共发生336例次EFI，发生率最高的症状为胃潴留137例次（40.8%，137/336），喂养开始前3 d EFI发生率较高，分别为19.0%、20.8%和19.6%。多因素回归分析结果显示，瑞芬太尼单日累积剂量（OR=1.032，95%CI：1.005~1.061）、使用≥2种镇痛药（OR=2.963，95%CI：1.557~5.640）、接受体外膜肺氧合治疗（OR=3.352，95%CI：1.819~6.175）、使用抗生素种类多（OR=1.605，95%CI：1.329~1.939）及幽门前喂养（OR=1.564，95%CI：1.114~2.196）是EFI的独立危险因素；模型拟合优度良好（Hosmer-Lemeshow检验P=0.259）。RCS分析显示，瑞芬太尼单日累积剂量与EFI风险呈正向线性剂量-反应关系（整体模型P=0.026，非线性趋势P=0.379），当单日累积剂量>10.77 mg时，EFI风险显著增加（OR>1.0）。亚组分析表明，瑞芬太尼单日累积剂量对EFI的影响在不同年龄、性别、急性生理学与慢性健康状况评分Ⅱ、改良危重症营养风险评分、喂养途径及镇痛药种类的患者中均保持一致，交互效应均P>0.05，但其效应在幽门后喂养、女性患者和使用≥2种镇痛药亚组中差异无统计学意义。结论 瑞芬太尼剂量是PPV患者EFI的独立危险因素，且存在剂量依赖性效应，该研究为PPV患者镇痛-营养平衡管理提供了瑞芬太尼剂量阈值参考数据。临床应遵循最低有效镇痛剂量原则，并对高剂量镇痛用药需求患者加强EFI监测与预防，以改善营养支持效果。

Objective To investigate the effect of analgesic dose on enteral feeding intolerance（EFI） in patients undergoing prone position ventilation（PPV），to determine the dose-response relationship，with aiming to provide evidence for balancing analgesia and nutrition management in critically ill patients. Methods We retrospectively collected clinical data from 630 patients receiving enteral nutrition（EN） support and PPV in the intensive care unit of a tertiary hospital in Nanjing between January 2020 and December 2024. Patients were divided into a tolerance group and an intolerance group based on the occurrence of EFI. Demographic data，disease and treatment-related information，and EN support data were collected. Logistic regression was used to explore the independent relationship between analgesic dose and EFI. Restricted Cubic Splines（RCS） were used to analyze the dose-response relationship between the daily cumulative dose of analgesic and EFI. A forest plot was used to visualize this relationship across subgroups. Results Among the 630 patients，366 tolerated EN and 264 did not，resulting in an EFI incidence of 41.90%. Among the 264 patients with EFI，70 presented with ≥2 symptoms of feeding intolerance，with a total of 336 episodes of EFI symptoms recorded. The most frequently reported symptom was gastric retention，occurring in 137 instances（40.8%，137/336）. The incidence of EFI was relatively higher within the first three days of initiating feeding，with rates of 19.0%，20.8%，and 19.6%，respectively. Multivariable regression analysis showed that the daily cumulative dose of remifentanil（OR=1.032，95%CI：1.005~1.061），use of ≥2 analgesics（OR=2.963，95%CI：1.557~5.640），extracorporeal membrane oxygenation（ECMO） therapy（OR=3.352，95%CI：1.819~6.175），higher number of antibiotic types used（OR=1.605，95%CI：1.329~1.939），and post-pyloric feeding（OR=1.564，95%CI：1.114~2.196） were independent risk factors for EFI. The model demonstrated good calibration （Hos-mer-Lemeshow test，P=0.259）. RCS analysis revealed a positive linear dose-response relationship between the daily cumulative dose of remifentanil and the risk of EFI（overall P=0.026；nonlinear P=0.379）. A daily cumulative dose exceeding 10.77 mg significantly increased EFI risk（OR>1.0）. Subgroup analysis showed that the effect of the remifentanil dose on EFI was consistent across patients of different ages，gender，APACHE Ⅱ scores，modified Nutrition Risk in the Critically Ill（mNUTRIC） scores，feeding routes，and types of analgesic agents（all interaction effects P>0.05）. However，the association was not statistically significant in the post-pyloric feeding，female，or ≥2 analgesics subgroups. Conclusion The dose of remifentanil is an independent risk factor for EFI in PPV patients，showing a dose-dependent effect. These findings provide a dose-threshold reference to support the balance between analgesia and nutrition management. Clinical practice should follow the principle of the lowest effective analgesic dose，and intensify EFI monitoring and prevention for the patients requiring high-dose analgesics to improve nutrition support outcomes.