Abstract: Background　Airway driving pressure (DP)is a recognized independent predictor of mortality in mechanically ventilated patients with acute respiratory distress syndrome (ARDS). Although pathophysiological differences between pulmonary and extrapulmonary ARDS are well-established, it remains unclear whether the prognostic value of DP varies by ARDS etiology. Objective　This prospective cohort study evaluated whether ARDS etiology modifies the association between DP and 90-day all-cause mortality. Methods　We enrolled 172 mechanically ventilated ARDS patients aged>18 years who met the Berlin definition at Nanjing Gaochun People's hospital from December 2021 to October 2023. Patients were classified as having pulmonary ARDS (n=82) or extrapulmonary ARDS (n=90)based on clinical features. We collected data on Day 1, including Sequential Organ Failure Assessment (SOFA)score, Acute Physiology and Chronic Health Evaluation II (APACHE II) Score, the ratio of arterial oxygen partial pressure to fractional inspired oxygen (PaO2/FiO2), and respiratory mechanics including DP, plateau pressures (Pplat), and respiratory system compliance (Crs). The primary outcome was 90-day all-cause mortality. Cox proportional hazards regression models were employed to assess the interaction between DP and ARDS etiology (pulmonary versus extrapulmonary) on 90-day mortality. Sensitivity analyses using inverse probability of treatment weighting (IPTW) based on propensity scores were conducted to account for potential confounders. Results　Among the 172 ARDS patients mean age (57±18) years, 63(36.3%) died during follow-up. Non-survivors had higher SOFA scores, APACHE II scores, DP, Pplat, and lower PaO2/FiO2 ratios and Crs than survivors(all P<0.05). The association between DP and 90-day mortality (HR=1.134, 95%CI =1.067-1.206, P<0.001) varied by ARDS etiology (P=0.04 for interaction) after adjustment for age, PaO2/FiO2 ratio, and SOFA score. In pulmonary ARDS (n=82), both PaO2/FiO2(P=0.04) and DP (HR=1.253，95%CI =1.146-1.370，P<0.01) were associated with 90-day mortality. Conversely, in extrapulmonary ARDS (n=90), DP (HR=1.083，95%CI=0.957-1.227，P=0.21) was not associated with mortality，while only the SOFA score remained an independent risk factor(P<0.01). These findings were consistent in the IPTW analysis (P=0.009 for interaction between ARDS etiology and DP on mortality). Conclusion　While DP is a strong predictor of mortality in ARDS overall, its prognostic significance differs by etiology. Specifically, in extrapulmonary ARDS, elevated DP is not significantly associated with poor prognosis, suggesting that ventilation strategies should be aligned with the underlying ARDS etiology.

Key words: Respiratory distress syndrome, Airway driving pressure, Precision therapy, Prognosis, Prospective studies

摘要： 背景　在接受机械通气急性呼吸窘迫综合征（ARDS）患者中，过高的气道驱动压（DP）是全因死亡风险的独立危险因素。尽管不同病因（肺内源性和肺外源性）ARDS在病理生理学方面具有较大的差异已得到充分证实，但是不同病因是否影响DP在ARDS中的预后价值，目前尚不清楚。目的　本研究旨在确定在ARDS中，DP与90 d全因死亡风险的关联是否受不同病因的影响。方法　选取2021年12月—2023年10月南京市高淳人民医院重症监护室收治的>18岁、符合柏林共识诊断标准的ARDS患者172例。根据临床资料将患者分为肺内源性ARDS（n=82）或肺外源性ARDS（n=90），入选首日，记录年龄、性别、序贯性器官功能衰竭评分（SOFA评分）、急性生理与慢性健康状况评分（APACHEⅡ评分）、氧合指数（PaO2/FiO2）以及呼吸力学参数包括DP、气道平台压（Pplat）、呼吸系统顺应性（Crs），研究终点为90 d全因死亡，随访结束时根据患者结局分为死亡组和存活组。采用Cox比例风险回归模型评估DP与病因在90 d死亡风险中的交互作用。敏感性分析包括使用基于倾向评分的逆概率加权（IPTW）方法，以控制潜在混杂因素的影响。结果　纳入172例ARDS患者，年龄（57±18）岁，随访结束死亡63例，病死率为36.3%。死亡组SOFA评分、APACHE Ⅱ评分、DP和Pplat高于存活组，PaO2/FiO2、Crs低于存活组（P<0.05）。调整年龄、SOFA评分、PaO2/FiO2后，过高的DP（HR=1.134，95%CI=1.067~1.206，P<0.001）与90 d全因死亡风险独立相关。交互作用分析显示：DP与90 d全因死亡风险的关联在不同ARDS病因人群中的差异有统计学意义（P=0.04）；在肺内源性ARDS患者（n=82）中，PaO2/FiO2降低（P=0.04）和过高的DP（HR=1.253，95%CI=1.146~1.370，P<0.01）均为90 d全因死亡风险的独立危险因素；而在肺外源性ARDS患者中（n=90），高SOFA评分（P<0.01）是全因死亡风险唯一的独立危险因素，DP（HR=1.083，95%CI=0.957~1.227，P=0.21）不再是90 d全因死亡风险的危险因素。IPTW后加权Cox分析的结果与主分析结果高度一致（不同ARDS病因与过高的DP对死亡风险影响存在显著交互作用，P=0.009）。结论　虽然DP是ARDS全因死亡风险的强预测因子，但受不同病因的影响，在肺外源性ARDS，其与不良预后并不显著相关。提示在机械通气策略的制定中应结合ARDS的病因特征，实施个体化通气管理。

关键词: 呼吸窘迫综合征, 气道驱动压, 精准治疗, 预后, 前瞻性研究