目的:评估肺不张区域容积（NILT）与中-重度急性呼吸窘迫综合征（ARDS）患者预后的关系。方法:收集2016年3月至2019年6月天津市第三中心医院重症监护病房（ICU）收治的131例中-重度ARDS患者的临床资料。收集患者的性别、年龄、体重指数（BMI）、ARDS原因、急性生理学与慢性健康状况评分Ⅱ（APACHEⅡ）、序贯器官衰竭评分（SOFA）和氧合指数（PaOn 2/FiOn 2）等基本资料；收集患者入ICU 1 d和7 d的CT影像学资料，根据CT值分为过度通气区域（-1 000～-900 HU）、正常通气区域（-899～-500 HU）、通气欠佳区域（-499～-100 HU）、肺不张区域（-99～100 HU），计算总肺容积和NILT占总肺容积百分比（NILT%）；同时收集患者机械通气时间、ICU住院时间和总住院时间。根据28 d随访的生存情况分为存活组和死亡组，采用多因素Logistic回归分析来确定ARDS患者28 d死亡的危险因素；绘制受试者工作特征曲线（ROC），计算NILT%预测ARDS患者28 d预后的ROC曲线下面积（AUC）及其95%可信区间（95%n CI），确定诊断的准确性，并根据NILT阈值对患者进行亚组分析。n 结果:131例中-重度ARDS患者中，剔除诊断ARDS超过48 h、非首次因ARDS入ICU、ICU住院时间<7 d、入ICU 72 h内死亡、存在慢性间质性肺病或充血性心力衰竭、入ICU后7 d内未进行胸部CT检查、入ICU后2 h内未进行标本采集的患者，最终共53例患者纳入分析。53例患者中，28 d存活31例，死亡22例，28 d病死率为41.5%。与存活组比较，死亡组患者年龄更大（岁：65.32±11.29比55.77±14.23），SOFA评分更高（分：11.68±3.82比8.39±2.23），差异均有统计学意义（均n P<0.05）；而两组性别、BMI、ARDS原因、APACHEⅡ评分和PaOn 2/FiOn 2比较差异均无统计学意义。入ICU 1 d时两组CT检查获得的CT值、总肺容积和NILT%比较差异均无统计学意义；入ICU 7 d时死亡组NILT%明显高于存活组〔（28.95±8.40）%比（20.35±5.91）%，n P<0.01〕，而CT值和总肺容积与存活组比较差异仍无统计学意义。多因素Logistic回归分析显示，ARDS患者28 d预后与年龄、SOFA评分和7 d NILT%独立相关〔年龄：优势比（n OR）＝0.892，95%n CI为0.808～0.984，n P＝0.023；SOFA评分：n OR＝0.574，95%n CI为0.387～0.852，n P＝0.006；7 d NILT%：n OR＝0.841，95%n CI为0.730～0.968，n P＝0.016〕。ROC曲线分析显示，7 d NILT%可以预测中-重度ARDS患者的28 d预后，AUC为0.810（95%n CI为0.678～0.952，n P15.50%定义为临床预后高风险，≤15.50%为低风险；与低风险患者（7例）比较，高风险患者（46例）机械通气时间、ICU住院时间和总住院时间均明显延长〔机械通气时间（d）：9.37±6.14比4.43±1.72，ICU住院时间（d）：12.11±5.85比7.57±1.13，总住院时间（d）：18.39±5.87比11.29±2.22，均n P15.50%与不良预后相关。“,”Objective:To assess the impact of not inflated lung tissue (NILT) volume on the prognosis of patients with moderate-to-severe acute respiratory distress syndrome (ARDS).Methods:The clinical data of 131 patients with moderate-to-severe ARDS admitted to the intensive care unit (ICU) of Tianjin Third Central Hospital from March 2016 to June 2019 were collected. The basic data of patients, including gender, age, body mass index (BMI), causes of ARDS, acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score, sequential organ failure assessment (SOFA) score and oxygenation index (PaOn 2/FiOn 2), were collected. The CT imaging data of patients on the 1st and 7th day in the ICU were collected. According to the CT value, they were divided into hyperventilated areas (-1 000 to -900 HU), normal ventilation areas (-899 to -500 HU), poorly ventilated areas (-499 to -100 HU), and atelectasis area (-99 to 100 HU). The total lung volume and the percentage of NILT to the total lung volume (NILT%) were calculate. At the same time, duration of mechanical ventilation, length of ICU stay, total length of hospital stay were collected. According to the 28-day follow-up, they were divided into survival group and death group. Multivariate Logistic regression analysis was used to determine the risk factors for 28-day death in ARDS patients. The receiver operating characteristic (ROC) curve was drawn, the area under ROC curve (AUC) and 95% confidence interval (95%n CI) were calculated to determine the accuracy of NILT% in predicting the 28-day prognosis of ARDS patients, and the NILT% threshold was used for subgroup analysis of patients.n Results:Among the 131 patients with moderate-to-severe ARDS, patients were excluded for more than 48 hours after ARDS diagnosis, repeated admission to ICU due to ARDS, the ICU duration less than 7 days, death within 72 hours of admission, chronic interstitial lung disease or congestive heart failure, no chest CT examination within 7 days of admission to ICU, and no specimen collection within 2 hours of admission to ICU. Finally, a total of 53 patients were enrolled in the analysis. Of the 53 patients, 31 patients survived and 22 patients died. The 28-day mortality was 41.5%. Compared with the survival group, patients in the death group were older (years old: 65.32±11.29 vs. 55.77±14.23), and had a higher SOFA score (11.68±3.82 vs. 8.39±2.23) with significant differences (both n P < 0.05), while there were no significant differences in gender, BMI, ARDS cause, APACHE Ⅱ score and PaO n 2/FiOn 2 between the two groups. There was no significant difference in CT value, total lung volume and NILT% between the two groups at 1st day after admission to ICU; NILT% on day 7 after admission to ICU in the death group was significantly higher than that in the survival group [(28.95±8.40)% vs. (20.35±5.91)%, n P < 0.01], but there was no significant difference in CT value and total lung volume between the two groups. Multivariate Logistic regression analysis showed that the 28-day prognosis of ARDS was related to age, SOFA score and NILT% independently [age: odds ratio ( n OR) = 0.892, 95%n CI was 0.808-0.984, n P = 0.023; SOFA score: n OR = 0.574, 95%n CI was 0.387-0.852, n P = 0.006; NILT%: n OR = 0.841, 95%n CI was 0.730-0.968, n P = 0.016]. ROC curve analysis showed that 7-day NILT% could predict the 28-day prognosis of patients with moderate-to-severe ARDS, and AUC was 0.810 (95%n CI was 0.678-0.952, n P 15.50% was defined as a high-risk clinical prognosis, and ≤ 15.50% was a low-risk. Compared with low-risk patients ( n n = 7), the duration of mechanical ventilation, the length of ICU stay and total length of hospital stay in high-risk patients (n n = 46) were significantly prolonged [duration of mechanical ventilation (days): 9.37±6.14 vs. 4.43±1.72, length of ICU stay (days): 12.11±5.85 vs. 7.57±1.13, total length of hospital stay (days): 18.39±5.87 vs. 11.29±2.22, all n P 15.50% of patients with moderate-to-severe ARDS after ICU admission is related to poor prognosis.