多脏器超声联合临床指标预测有创机械通气患者脱机后结局

Multi-organ ultrasound combined with clinical indicators for predicting weaning outcomes of patients after invasive mechanical ventilation

目的基于多脏器超声联合临床指标构建预测有创正压通气(IMV)患者脱机结局列线图,评估其临床价值。方法前瞻性收集357例接受IMV的呼吸衰竭患者,依脱机后结局分为生存组(n=200)和死亡组(n=157);以单因素分析和单、多因素logistic回归分析IMV患者多脏器超声参数及临床指标,筛选脱机后死亡相关独立危险因素,以之构建预测IMV患者脱机结局的列线图。将患者随机分为训练集(n=178)和验证集(n=179),绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC),评估列线图预测脱机结局的效能;基于验证集应用校准曲线评价列线图的拟合优度,以决策曲线分析(DCA)评估其临床价值。结果左心室射血分数、每搏输出量、三尖瓣环收缩期位移、下腔静脉塌陷率、肺超声B线评分、右膈肌浅快呼吸指数,以及合并神经系统疾病、休克、IMV时间、急性生理学与慢性健康状况评分系统Ⅱ(APACHEⅡ)评分均为IMV患者脱机后死亡的独立危险因素(P均<0.05)。列线图预测训练集及验证集IMV患者脱机后死亡的AUC分别为0.86、0.99,其预测性能良好,且概率阈值为0.2~0.8时临床净获益较高。结论基于多脏器超声联合临床指标的列线图可有效预测IMV患者脱机后结局。

Objective To construct a nomogram based on multi-organ ultrasound combined with clinical indicators for predicting weaning outcomes of patients after invasive mechanical ventilation(IVM),and to evaluate its clinical value.Methods Totally 357 respiratory failure patients who underwent IMV were prospectively enrolled and divided into survival group(n=200)and death group(n=157)according to the weaning outcomes.Univariate analysis,univariate and multivariate logistic regression analyses were used to explore multi-organ ultrasonic parameters and clinical indicators of IMV patients,and independent risk factors related to death after weaning were screened.Then a nomogram was constructed to predict outcomes of IMV patients after weaning.The patients were randomly divided into training set(n=178)or validation set(n=179).Receiver operating characteristic(ROC)curves were drawn,and area under the curve(AUC)was calculated to evaluate the predictive efficiency of the nomogram.Calibration curve and decision curve analysis(DCA)were used to evaluate the goodness of fit based on verification set and the clinical value of this nomogram,respectively.Results Left ventricular ejection fraction,stroke volume,tricuspid annular plane systolic excursion,collapse rate of inferior vena cava,lung ultrasonic B-line score and right diaphragmatic rapid shallow breathing index,as well as complicated with nervous system disease,shock,IMV time,acute physiology and chronic health status scoring systemⅡ(APACHEⅡ)score were all independent risk factors for death of IMV patients after weaning(all P<0.05).AUC of the nomogram for predicting weaning outcomes of IMV patients in training set and validation set was 0.86 and 0.99,respectively,which had good predictive performances and high clinical net benefit at the threshold of 0.2 to 0.8.Conclusion Nomogram based on multi-organ ultrasonic parameters and clinical indicators could be used to effectively predict the weaning outcomes of IMV patients.