Establishment and validation of a predictive model for the 30-day mortality risk of patients with the first invasive mechanical ventilation based on the ultrasound texture features of rectus femoris at the early stage
Wang Liang
Wang Haibo
Zhai Menglong
Cui Shaoshuai
Li Wenjuan
Wang Guandong
Abstract：Objective To explore the establishment and validation of a predictive model for the 30-day mortality risk of patients with the first invasive mechanical ventilation based on the ultrasound texture features of rectus femoris at the early stage.Methods The patients from the Respiratory Intensive Care Unit of Henan Provincial People′s Hospital by using the convenient sampling method from January 2022 to December 2023.The patients were randomly assigned to the training set(n =304)and the validation set(n =76)in a ratio of 8∶2.The bedside ultrasound was used to collect texture feature parameters on the 1st,3rd,5th and 7th day after intubation.The training set was divided into the death group(n =80)and the survival group(n =224)based on the endpoint event of death or survival on the 30th day.A predictive model for the 30-day mortality risk of patients with the first invasive mechanical ventilation was established by generalized estimating equations.The predictive model was evaluated by the area under receiver operating characteristic(ROC)curve(AUC),calibration curve and Hosmer-Lemeshow.Results Influencing factors of 30-day mortality of patients with the first invasive mechanical ventilation were age,oxygenation index,kidney disease,steroid hormone use,rate of change in the thickness of rectus femoris,mean,and angular second moment(P<0.05).The AUC of the training set and the validation set was 0.832(95%CI 0.767-0.898)and 0.819(95%CI 0.722-0.916),indicating good discrimination of the model.The Hosmer-Lemeshow values of the training set and the validation set were 5.969(P =0.651)and 4.336(P =0.826)respectively.The calibration curve had a high degree of coincidence with the diagonal line,indicating good calibration of the prediction model.Conclusions The model can predict the 30-day mortality risk of patients with the first invasive mechanical ventilation by incorporating ultrasound texture features of rectus femoris at the early stage.
Keywords：UltrasoundRectus femorisTexture analysisMechanical ventilationPrediction modelRate of change in the thickness of rectus femorisAngular second moment
Publication Date：2024-04-10
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 330-336 )