床旁肺部超声对ARDS的诊断价值和严重程度的评估

Value of bedside lung ultrasound for diagnosis of acute respiratory distress syndrome and for assess-ment of the severity

目的评价床旁肺部超声对ARDS的诊断价值和严重程度的评估.方法选择疑似ARDS(氧合指数<300 mmHg)需进行胸部CT检查,且因病情需要行脉波指示剂连续心排出量监测(PiCCO)的患者50例,年龄18~80岁,性别不限.于入ICU 24 h时行胸部CT、超声检查和动脉血气分析,记录血管外肺水指数(EVLWI)和B线数目,计算肺损伤超声评分和氧合指数.将胸部CT和肺部超声均诊断为ARDS的患者按照严重程度分为3组:轻度组(200 mmHg<氧合指数≤300 mmHg)、中度组(100 mmHg<氧合指数≤200 mmHg)和重度组(氧合指数≤100 mmHg).采用Kappa一致性检验分析肺部超声与胸部CT诊断ARDS的一致性.绘制B线数目、EVLWI及肺损伤超声评分诊断重度ARDS的受试者工作特征(ROC)曲线,计算曲线下面积及其95%可信区间、临界值、敏感度和特异度.结果共有46例患者胸部CT和肺部超声均诊断为ARDS.肺部超声与胸部CT诊断ARDS具有一致性(Kappa值0.648,P<0.01);肺部超声与胸部CT诊断肺实变具有一致性(Kappa值0.788,P<0.01);肺部超声与胸部CT诊断后背部区域肺实变具有一致性(Kappa值0.825,P<0.01).与轻度组比较,中度组肺损伤超声评分升高,B线数目增加,重度组肺损伤超声评分和EVLWI升高,B线数目增加(P<0.05);与中度组比较,重度组肺损伤超声评分和EVLWI升高,B线数目增加(P<0.05).B线数目诊断重度ARDS的曲线下面积(95%可信区间)为0.915(0.905~0.937),临界值为15.5条,敏感度及特异度分别为78.9%、85.2%;肺损伤超声评分诊断重度ARDS(95%可信区间)的曲线下面积为0.856(0.833~0.878),临界值为25.5分,敏感度及特异度分别为73.7%、82.5%;EVLWI诊断重度ARDS(95%可信区间)的曲线下面积为0.907(0.888~0.933),临界值为15.5,敏感度及特异度分别为73.7、92.6%.结论肺部超声可用于ARDS的诊断及严重程度的评估.

Objective To evaluate the value of bedside lung ultrasound for diagnosis of acute re-spiratory distress syndrome ( ARDS) and for assessment of the severity. Methods Fifty patients of both se-xes suspected of having ARDS ( oxygenation index<300 mmHg) and required lung CT tests and Pulse Indi-cator Continuous Cardiac Output because of their condition, aged 18-80 yr, were selected. At 24 h after entering ICU, chest CT, lung ultrasound and arterial blood gas analysis were performed to record Extravas-cular Lung Water Index ( EVLWI) and the number of B lines, and lung injury ultrasound score and oxygen-ation index were calculated. The patients diagnosed with ARDS by chest CT and lung ultrasound were divid-ed into 3 groups: mild group ( 200 mmHg<oxygenation index≤300 mmHg) , moderate group ( 100 mmHg<oxygenation index≤200 mmHg) and severe group ( oxygenation index≤100 mmHg) . Kappa consistency a-nalysis was used to assess the consistency between lung ultrasound and chest CT in diagnosis of ARDS. The receiver operating characteristic curves of th number of B lines, EVLWI and lung injury ultrasound score in assessing the severity of ARDS were drawn, and the area under the curve and 95% confidence interval ( CI) , critical value, sensitivity and specificity were calculated. Results Forty-six patients were diag-nosed as having ARDS by both chest CT and lung ultrasound. There was good consistency ( Kappa value 0. 648, P<0. 01) between chest CT and lung ultrasound in diagnosis of ARDS. There was good consistency ( Kappa value 0. 788, P<0. 01) between lung ultrasound and chest CT in diagnosis of pulmonary consolida-tion. Lung ultrasound and chest CT were in good agreement ( Kappa value 0. 825, P<0. 01) with each oth-er in diagnosis of pulmonary consolidation in the posterior region. Compared with mild group, the lung inju-ry ultrasound score was significantly increased, and the number of B lines was increased in moderate group, and the lung injury ultrasound score and EVLWI were significantly increased, and the number of B lines was increased in severe group ( P<0. 05) . Compared with moderate group, the lung injury ultrasound score and EVLWI were significantly increased, and the number of B lines was increased in severe group ( P<0. 05) . The area under the curve ( 95% CI ) of the number of B lines in diagnosing severe ARDS was 0. 915 ( 0. 905-0. 935 ) , and the critical value, sensitivity and specificity were 15. 5, 78. 9% and 85. 2%, respectively. The area under the curve ( 95% CI) of lung injury ultrasound score in diagnosing severe ARDS was 0. 856 (0. 833-0. 878), and the critical value, sensitivity and specificity were 25. 5, 73. 7% and 82. 5%, respectively. The area under the curve (95% CI) of EVLWI in diagnosing severe ARDS was 0. 907 ( 0. 888-0. 933) , and the critical value, sensitivity and specificity were 15. 5, 73. 7%and 92. 6%, respectively. Conclusion Lung ultrasound can be used for diagnosis of ARDS and for evalu-ation of the severity of ARDS.