Predictive model for acute respiratory distress syndrome events in ICU patients in China using machine learning algorithms: a secondary analysis of a cohort study

Abstract

Background To develop a machine learning model for predicting acute respiratory distress syndrome (ARDS) events through commonly available parameters, including baseline characteristics and clinical and laboratory parameters. Methods A secondary analysis of a multi-centre prospective observational cohort study from five hospitals in Beijing, China, was conducted from January 1, 2011, to August 31, 2014. A total of 296 patients at risk for developing ARDS admitted to medical intensive care units (ICUs) were included. We applied a random forest approach to identify the best set of predictors out of 42 variables measured on day 1 of admission. Results All patients were randomly divided into training (80%) and testing (20%) sets. Additionally, these patients were followed daily and assessed according to the Berlin definition. The model obtained an average area under the receiver operating characteristic (ROC) curve (AUC) of 0.82 and yielded a predictive accuracy of 83%. For the first time, four new biomarkers were included in the model: decreased minimum haematocrit, glucose, and sodium and increased minimum white blood cell (WBC) count. Conclusions This newly established machine learning-based model shows good predictive ability in Chinese patients with ARDS. External validation studies are necessary to confirm the generalisability of our approach across populations and treatment practices.

Machine Learning Methods to Predict Acute Respiratory Failure and Acute Respiratory Distress Syndrome

“…Ding et al ( 2019 ) conducted a secondary analysis of data from 296 patients in six ICUs in Beijing, China, to predict ARDS using the Berlin definition. The authors used a random forest approach to predict development of ARDS using baseline characteristics, clinical variables, and predisposing conditions collected at admission.…”

“…Ding et al (2019) conducted a secondary analysis of data from 296 patients in six ICUs in Beijing, China, to predict ARDS using the Berlin definition. The authors used a random forest approach to predict development of ARDS using baseline Predictive model for acute respiratory distress syndrome events in ICU patients in china using machine learning algorithms: a secondary analysis of a cohort study (Ding et al, 2019) 296…”

Novel Perspectives Regarding the Pathology, Inflammation, and Biomarkers of Acute Respiratory Distress Syndrome

“…Since the pathophysiology of ARDS is complex and heterogeneous, current research suggests that combinations of biomarkers that reflect different aspects of ARDS (such as epithelial and endothelial injury, inflammation or infection) are more likely to be use in a clinical context. Indeed, the best approach will probably combine clinical predictors with several biomarkers as has been suggested and tested with varying degrees of success in quite a few studies by now, including cohorts based on several RCTs [8][9][10]65,92,113,114]. However, none of these candidates have been used clinically in patients with ARDS.…”

Prediction of blood culture outcome using hybrid neural network model based on electronic health records

“…For example, Zou et al [ 21 ] used decision tree, random forest and neural network to predict diabetes mellitus based on 14 clinical attributes. Ding et al [ 22 ] applied a random forest model for predicting acute respiratory distress syndrome events in ICU patients based on 42 clinical variables. Yin et al [ 23 ] used preprocedural clinical variables to develop a model for prediction of contrast-induced nephropathy (CIN) before radiological procedures among patients administered contrast media.…”