A Simple Weaning Model Based on Interpretable Machine Learning Algorithm for Patients With Sepsis: A Research of MIMIC-IV and eICU Databases

Abstract

BackgroundInvasive mechanical ventilation plays an important role in the prognosis of patients with sepsis. However, there are, currently, no tools specifically designed to assess weaning from invasive mechanical ventilation in patients with sepsis. The aim of our study was to develop a practical model to predict weaning in patients with sepsis.MethodsWe extracted patient information from the Medical Information Mart for Intensive Care Database-IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD). Kaplan–Meier curves were plotted to compare the 28-day mortality between patients who successfully weaned and those who failed to wean. Subsequently, MIMIC-IV was divided into a training set and an internal verification set, and the eICU-CRD was designated as the external verification set. We selected the best model to simplify the internal and external validation sets based on the performance of the model.ResultsA total of 5020 and 7081 sepsis patients with invasive mechanical ventilation in MIMIC-IV and eICU-CRD were included, respectively. After matching, weaning was independently associated with 28-day mortality and length of ICU stay (p < 0.001 and p = 0.002, respectively). After comparison, 35 clinical variables were extracted to build weaning models. XGBoost performed the best discrimination among the models in the internal and external validation sets (AUROC: 0.80 and 0.86, respectively). Finally, a simplified model was developed based on XGBoost, which included only four variables. The simplified model also had good predictive performance (AUROC:0.75 and 0.78 in internal and external validation sets, respectively) and was developed into a web-based tool for further review.ConclusionsWeaning success is independently related to short-term mortality in patients with sepsis. The simplified model based on the XGBoost algorithm provides good predictive performance and great clinical applicablity for weaning, and a web-based tool was developed for better clinical application.

Predicting Successful Weaning from Mechanical Ventilation by Reduction in Positive End-expiratory Pressure Level Using Machine Learning

“…al. [ 16 ] defined weaning as 48h without ventilation or death and they used MIMIC-IV data as training set and test set and eICU collaborative research database (eICU-CRD) data for external validation to predict a full weaning. One of the latest studies by Zhao et al [ 20 ] used data from MIMIC-IV to train and evaluate a CatBoost algorithm to predict extubation failure as the need for re-intubation or death within 48-hour intervals following the planned extubation.…”

Explainable Machine Learning to Predict Successful Weaning of Mechanical Ventilation in Critically Ill Patients Requiring Hemodialysis

“…An essential concern in critical care is the process of weaning from mechanical ventilation. Recent studies, including our own, have developed a weaning prediction model using the machine learning (ML) approach [ 3 , 4 , 5 ]. A number of clinical studies have found that acute kidney injury, particularly kidney injury that requires hemodialysis, is a substantial determinant of difficulty weaning that leads to the prolonged use of mechanical ventilation [ 6 , 7 , 8 , 9 ].…”

Research trends and global productivity on mechanical ventilation with the impact of COVID-19: a bibliometric analysis in the period 1980-2021

“…The primary purpose of MV is to normalize and/or stabilize the patient's gas exchange by increasing ventilation and oxygenation in patients with respiratory failure (1). MV plays a very important role in the management of patients with acute respiratory distress syndrome (ARDS), chronic obstructive pulmonary disease (COPD), acute severe asthma, sepsis, hypoxemia, COVID-19, and newborn babies with respiratory problems (2)(3)(4). If MV is performed through an instrument such as an endotracheal tube or a tracheostomy tube, it is called "invasive ventilation" (1).…”

“…However, ongoing clinical uncertainty regarding the optimal separation strategy still continues (7). It has been suggested to switch from simple maneuvers to more complex methods such as multivariate scoring systems and computerized decision support models to identify patients ready for extubation (4,7). The search for an accurate way of predicting success in weaning a patient from MV continues (8).…”

“…The search for an accurate way of predicting success in weaning a patient from MV continues (8). Various clinical scores have been developed to predict the weaning success of MV in different samples (4,(8)(9)(10).…”