Effect Of Renal Failure On Weaning Outcome In Patients Requiring Prolonged Mechanical Ventilation

At least 5% of all intensive care unit patients require prolonged respiratory support. Multiple factors have been suggested as possible predictors of successful respiratory weaning so far. We sought to verify whether the Acute Physiology and Chronic Health Evaluation II (APACHE II) can predict freedom from prolonged mechanical ventilation (PMV) in patients treated in a regional weaning centre. The study group comprised 130 consecutive patients (age; median (interquartile range): 71 (62–77) years), hospitalized between 1 January 2012, and 31 December 2013. APACHE II score was assessed based on the worst values taken during the first 24 hours after admission. Glasgow coma scale was excluded from calculations due to the likely influence of sedative agents. The outcome was defined as freedom from mechanical ventilation, with or without tracheostomy on discharge. Among survivors ( n = 115), 88.2% were successfully liberated from mechanical ventilation and 60.9% from tracheostomy. APACHE II failed to predict freedom from mechanical ventilation (area under the receiver–operating characteristic curve [AUROC] = 0.534; 95% confidence interval [CI]: 0.439–0.628; p = 0.65) and tracheostomy tube removal (AUROC = 0.527; 95% CI: 0.431–0.621; p = 0.63). Weaning outcome was unrelated to the aetiology of respiratory failure on admission ( p = 0.41). APACHE II cannot predict weaning outcome in patients requiring PMV.

Download Full-text

Explainable Machine Learning to Predict Successful Weaning Among Patients Requiring Prolonged Mechanical Ventilation: A Retrospective Cohort Study in Central Taiwan

Frontiers in Medicine ◽

10.3389/fmed.2021.663739 ◽

2021 ◽

Vol 8 ◽

Author(s):

Ming-Yen Lin ◽

Chi-Chun Li ◽

Pin-Hsiu Lin ◽

Jiun-Long Wang ◽

Ming-Cheng Chan ◽

...

Keyword(s):

Machine Learning ◽

Mechanical Ventilation ◽

Prediction Model ◽

Prolonged Mechanical Ventilation ◽

Number Of Patients ◽

Extreme Gradient Boosting ◽

Feature Importance ◽

Successful Weaning ◽

Central Taiwan ◽

Weaning Outcome

Objective: The number of patients requiring prolonged mechanical ventilation (PMV) is increasing worldwide, but the weaning outcome prediction model in these patients is still lacking. We hence aimed to develop an explainable machine learning (ML) model to predict successful weaning in patients requiring PMV using a real-world dataset.Methods: This retrospective study used the electronic medical records of patients admitted to a 12-bed respiratory care center in central Taiwan between 2013 and 2018. We used three ML models, namely, extreme gradient boosting (XGBoost), random forest (RF), and logistic regression (LR), to establish the prediction model. We further illustrated the feature importance categorized by clinical domains and provided visualized interpretation by using SHapley Additive exPlanations (SHAP) as well as local interpretable model-agnostic explanations (LIME).Results: The dataset contained data of 963 patients requiring PMV, and 56.0% (539/963) of them were successfully weaned from mechanical ventilation. The XGBoost model (area under the curve [AUC]: 0.908; 95% confidence interval [CI] 0.864–0.943) and RF model (AUC: 0.888; 95% CI 0.844–0.934) outperformed the LR model (AUC: 0.762; 95% CI 0.687–0.830) in predicting successful weaning in patients requiring PMV. To give the physician an intuitive understanding of the model, we stratified the feature importance by clinical domains. The cumulative feature importance in the ventilation domain, fluid domain, physiology domain, and laboratory data domain was 0.310, 0.201, 0.265, and 0.182, respectively. We further used the SHAP plot and partial dependence plot to illustrate associations between features and the weaning outcome at the feature level. Moreover, we used LIME plots to illustrate the prediction model at the individual level. Additionally, we addressed the weekly performance of the three ML models and found that the accuracy of XGBoost/RF was ~0.7 between weeks 4 and week 7 and slightly declined to 0.6 on weeks 8 and 9.Conclusion: We used an ML approach, mainly XGBoost, SHAP plot, and LIME plot to establish an explainable weaning prediction ML model in patients requiring PMV. We believe these approaches should largely mitigate the concern of the black-box issue of artificial intelligence, and future studies are warranted for the landing of the proposed model.

Download Full-text