浙江医学2025，Vol.47Issue(3)：243-248,6.DOI:10.12056/j.issn.1006-2785.2025.47.3.2024-1338
基于肺部超声分数的急性呼吸窘迫综合征撤机结局列线图预测模型的构建和验证
Construction and verification of the nomogram prediction model for the extubation outcome of acute respiratory distress syn-drome based on lung ultrasound score
摘要
Abstract
Objective To explore the correlation between the change of lung ultrasound score(LUS)and the extubation outcome of acute respiratory distress syndrome(ARDS),and construct the nomogram prediction model for predicting the extubation outcome.Methods A total of 130 patients who were diagnosed with ARDS and received invasive mechanical ventilation at Ningbo Medical Center Lihuili Hospital from January 2019 to December 2022 were selected in this prospective observation study.The data of LUS immediately after intubation and before extubation were collected,and the difference between the LUS before extubation and the LUS immediately after intubation(ΔLUS)was calculated.All patients were divided according to the quartiles of ΔLUS into Q1 group(≤-7 points,n=37),Q2 group(-6--4 points,n=32),Q3 group(-3--1 points,n=31)and Q4 group(≥0 points,n=30).The correlation of the change of LUS and extubation outcome was analyzed.logistic regression was used for trend test.Univariate and multivariate logistic regression were utilized to explore the influencing factors affecting the extubation outcome to establish the nomogram prediction model.The ROC curve was used to analyze the efficacy of the nomogram prediction model in predicting extubation outcome of ARDS.The Hosmer-Lemeshow test and calibration curve using a Bootstrap resampling method(1 000 Bootstrap resamples)were applied to evaluate the consistency of the nomogram.The decision curve was used to evaluate the net benefit of the nomogram.Results The success rates of extubation of Q1,Q2,Q3 and Q4 groups were 81.1％(30/37),78.1％(25/32),83.9％(26/31)and 36.7％(11/30),respectively.The success rate of extubation tended to increase as ΔLUS levels increased(Ptrend＜0.01).Multivariate logistic regression showed LUS immediately after intubation(OR=0.765,95％CI:0.639-0.916),ΔLUS[Q1 group(OR=175.493,95％CI:10.970-2 807.528),Q2 group(OR=70.007,95％CI:6.014-811.264),Q3 group(OR=53.081,95％CI:6.455-436.507)],and Sequential organ Failure Assessment(SOFA)Score(OR=0.384,95％CI:0.252-0.585)were independent influencing factors of extubation outcome(all P＜0.05).The area under the curve(AUC)of the nomogram constructed based on variables screened by multivariate analysis for predicting the successful weaning from mechanical ventilation in patients with ARDS was 0.917(95％CI:0.865-0.968).The model showed good prediction accuracy and goodness of fit,and had a net benefit within a certain probability threshold range.Conclusion The success rate of ARDS extubation tends to increase as ΔLUS decline increases.The nomogram prediction model established based on LUS immediately after intubation,ΔLUS and SOFA Score has good prediction efficacy,consistency and clinical practicality,which can help clinicians make the correct extubation decision of ARDS.关键词
肺部超声分数/急性呼吸窘迫综合征/撤机结局/列线图预测模型Key words
Lung ultrasound score/Acute respiratory distress syndrome/Extubation outcome/Nomogram prediction model引用本文复制引用
周银超,刘盼盼,伍慧珊,应唯楠,董绉绉..基于肺部超声分数的急性呼吸窘迫综合征撤机结局列线图预测模型的构建和验证[J].浙江医学,2025,47(3):243-248,6.基金项目
浙江省医药卫生科技计划项目(2019KY612) （2019KY612）
