Can Big Data and Machine Learning Improve Our Understanding of Acute Respiratory Distress Syndrome?

Predicting mortality and risk factors of sepsis related ARDS using machine learning models

Sepsis related acute respiratory distress syndrome (ARDS) is a common and serious disease in clinic. Accurate prediction of in-hospital mortality of patients is crucial to optimize treatment and improve prognosis under the new global definition of ARDS. Our study aimed to use machine learning models to develop models that can effectively predict the in-hospital mortality of patients with sepsis related ARDS, calculate the mortality, and to identify related risk factors under the new global definition of ARDS. Based on MIMIC database, our study included 3470 first-time admission records of patients with sepsis related ARDS. After excluding 4 patients under the age of 18, 75 patients with less than 24 h stay in ICU, and 5 cases with missing indicators > 30%, finally 3386 cases were retained. The variance inflation factor (VIF) analysis was used to test the collinearity of the explanatory variables. The data were divided into the training set and the test set according to the ratio of 7:3. Six models, extreme gradient boosting (XGBoost), light gradient boosting (LightGBM), random forest (RF), classification and regression tree (CART), naive bayes (NB) and logistic regression (LR), were designed for training and testing. In the training set, XGBoost (AUROC = 0.951, 95% CI 0.942-0.961), LR (AUROC = 0.835, 95% CI 0.817-0.854), RF (AUROC = 1.0, 95% CI 1.0-1.0), LightGBM (AUROC = 1.0, 95% CI 1.0-1.0), CART (AUROC = 0.831, 95% CI 0.811-0.852), NB (AUROC = 0.793, 95% CI 0.772-0.814). In the test set, XGBoost (AUROC = 0.833, 95% CI 0.804-0.861), LR (AUROC = 0.82695% CI 0.796-0.856), RF (AUROC = 0.846, 95% CI 0.818-0.874), LightGBM (AUROC = 0.827, 95% CI 0.798-0.856), CART (AUROC = 0.753, 95% CI 0.718-0.787), NB (AUROC = 0.799, 95% CI 0.768-0.831). The RF model has the best performance on the test set. Further analyze the feature importance ranking and partial dependence plots of random forest model. Acute physiology and chronic health evaluation III (APACHE III), bicarbonate, anion gap and non-invasive blood pressure systolic were identified as the four most important risk characteristics. In this study, a variety of machine learning models have been successfully constructed to predict the in-hospital mortality of patients with sepsis related ARDS, among which the RF model performs well. Key risk factors identified include APACHE III, bicarbonate, anion gap and non-invasive blood pressure systolic. The identification of these factors helps clinicians to assess patients' conditions more accurately and develop personalized treatment plans, thereby improving the survival rate and prognosis quality of patients under the new global definition of ARDS.

Can adverse childhood experiences predict chronic health conditions? Development of trauma-informed, explainable machine learning models

Introduction

Decades of research have established the association between adverse childhood experiences (ACEs) and adult onset of chronic diseases, influenced by health behaviors and social determinants of health (SDoH). Machine Learning (ML) is a powerful tool for computing these complex associations and accurately predicting chronic health conditions.

Methods

Using the 2021 Behavioral Risk Factor Surveillance Survey, we developed several ML models-random forest, logistic regression, support vector machine, Naïve Bayes, and K-Nearest Neighbor-over data from a sample of 52,268 respondents. We predicted 13 chronic health conditions based on ACE history, health behaviors, SDoH, and demographics. We further assessed each variable's importance in outcome prediction for model interpretability. We evaluated model performance via the Area Under the Curve (AUC) score.

Results

With the inclusion of data on ACEs, our models outperformed or demonstrated similar accuracies to existing models in the literature that used SDoH to predict health outcomes. The most accurate models predicted diabetes, pulmonary diseases, and heart attacks. The random forest model was the most effective for diabetes (AUC = 0.784) and heart attacks (AUC = 0.732), and the logistic regression model most accurately predicted pulmonary diseases (AUC = 0.753). The strongest predictors across models were age, ever monitored blood sugar or blood pressure, count of the monitoring behaviors for blood sugar or blood pressure, BMI, time of last cholesterol check, employment status, income, count of vaccines received, health insurance status, and total ACEs. A cumulative measure of ACEs was a stronger predictor than individual ACEs.

Discussion

Our models can provide an interpretable, trauma-informed framework to identify and intervene with at-risk individuals early to prevent chronic health conditions and address their inequalities in the U.S.