Machine learning for patient risk stratification for acute respiratory distress syndrome

• Published in: PloS one Vol. 14; no. 3; p. e0214465
• Main Authors: Zeiberg, Daniel, Prahlad, Tejas, Nallamothu, Brahmajee K, Iwashyna, Theodore J, Wiens, Jenna, Sjoding, Michael W
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 28.03.2019; Public Library of Science (PLoS)
• Subjects: Adult respiratory distress syndrome; Adults; Aged; Artificial intelligence; Biology and Life Sciences; Clinical medicine; Computer and Information Sciences; Computer science; Critical care; Data mining; Electronic health records; Electronic medical records; Electronic records; Feasibility studies; Feature extraction; Female; Health; Health care policy; Hospitalization; Hospitals; Humans; Hypoxia; Illnesses; Internal medicine; International conferences; Knowledge discovery; Learning algorithms; Machine Learning; Male; Medical personnel; Medical research; Medicine; Medicine and Health Sciences; Middle Aged; Model testing; Models, Statistical; Pathogenesis; Patients; People and Places; Physical Sciences; Physicians; Physiology; Prediction models; Regression models; Research and Analysis Methods; Respiratory distress syndrome; Respiratory Distress Syndrome, Adult - epidemiology; Risk; Risk Assessment - methods; Ventilators; United Kingdom; Ann Arbor Michigan; United States > US; Michigan
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0214465

Existing prediction models for acute respiratory distress syndrome (ARDS) require manual chart abstraction and have only fair performance-limiting their suitability for driving clinical interventions. We sought to develop a machine learning approach for the prediction of ARDS that (a) leverages electronic health record (EHR) data, (b) is fully automated, and (c) can be applied at clinically relevant time points throughout a patient's stay. We trained a risk stratification model for ARDS using a cohort of 1,621 patients with moderate hypoxia from a single center in 2016, of which 51 patients developed ARDS. We tested the model in a temporally distinct cohort of 1,122 patients from 2017, of which 27 patients developed ARDS. Gold standard diagnosis of ARDS was made by intensive care trained physicians during retrospective chart review. We considered both linear and non-linear approaches to learning the model. The best model used L2-logistic regression with 984 features extracted from the EHR. For patients observed in the hospital at least six hours who then developed moderate hypoxia, the model achieved an area under the receiver operating characteristics curve (AUROC) of 0.81 (95% CI: 0.73-0.88). Selecting a threshold based on the 85th percentile of risk, the model had a sensitivity of 56% (95% CI: 35%, 74%), specificity of 86% (95% CI: 85%, 87%) and positive predictive value of 9% (95% CI: 5%, 14%), identifying a population at four times higher risk for ARDS than other patients with moderate hypoxia and 17 times the risk of hospitalized adults. We developed an ARDS prediction model based on EHR data with good discriminative performance. Our results demonstrate the feasibility of a machine learning approach to risk stratifying patients for ARDS solely from data extracted automatically from the EHR.