Outcome prediction during an ICU surge using a purely data-driven approach: A supervised machine learning case-study in critically ill patients from COVID-19 Lombardy outbreak. (August 2022)

• Record Type: Journal Article
• Authors: Greco, Massimiliano; Angelotti, Giovanni; Caruso, Pier Francesco; Zanella, Alberto; Stomeo, Niccolò; Costantini, Elena; Protti, Alessandro; Pesenti, Antonio; Grasselli, Giacomo; Cecconi, Maurizio
• Is Part Of: International journal of medical informatics. Volume 164(2022)
• Publication Date: 2022-08
• Subjects: COVID-19 -- Machine learning -- ICU management -- Emergency organization -- Outcomes -- Epidemiology; Medical informatics -- Periodicals; Information science -- Periodicals; Computers -- Periodicals; Medical technology -- Periodicals; Medical Informatics -- Periodicals; Technology, Medical -- Periodicals; Computers; Information science; Medical informatics; Medical technology; Electronic journals; Periodicals; Electronic journals; 610.285
• DOI: 10.1016/j.ijmedinf.2022.104807
• Languages: English
• ISSNs: 1386-5056

Abstract

Highlights: Machine learning (ML) performs robustly in predicting mortality for COVID patients. ML can be used to build robust models despite being based on emergency data. ML can provide clinical insights to optimize resource allocation in COVID upsurges. Abstract: Purpose: COVID-19 disease frequently affects the lungs leading to bilateral viral pneumonia, progressing in some cases to severe respiratory failure requiring ICU admission and mechanical ventilation. Risk stratification at ICU admission is fundamental for resource allocation and decision making. We assessed performances of three machine learning approaches to predict mortality in COVID-19 patients admitted to ICU using early operative data from the Lombardy ICU Network. Methods: This is a secondary analysis of prospectively collected data from Lombardy ICU network. A logistic regression, balanced logistic regression and random forest were built to predict survival on two datasets: dataset A included patient demographics, medications before admission and comorbidities, and dataset B included respiratory data the first day in ICU. Results: Models were trained on 1484 patients on four outcomes (7/14/21/28 days) and reached the greatest predictive performance at 28 days (F1-score: 0.75 and AUC: 0.80). Age, number of comorbidities and male gender were strongly associated with mortality. On dataset B, mode of ventilatory assistance at ICU admission and fraction of inspired oxygen were associated with an increase in prediction performances. Conclusions: Machine learning techniques might be useful in emergency phases to reach good predictive performances maintaining interpretability to gain knowledge on complex situations and enhance patient management and resources.