A Machine Learning Prediction Model of Respiratory Failure Within 48 Hours of Patient Admission for COVID-19: Model Development and Validation.
Siavash Bolourani,Max Brenner,Ping Wang,Thomas McGinn,Jamie S. Hirsch,Douglas P. Barnaby,Theodoros P. Zanos +6 more
TLDR
In this paper, a machine learning model was used to predict respiratory failure within 48 hours of admission based on data from the emergency department of patients with COVID-19 who were admitted to Northwell Health acute care hospitals.Abstract:
Background: Predicting early respiratory failure due to COVID-19 can help triage patients to higher levels of care, allocate scarce resources, and reduce morbidity and mortality by appropriately monitoring and treating the patients at greatest risk for deterioration. Given the complexity of COVID-19, machine learning approaches may support clinical decision making for patients with this disease.
Objective: Our objective is to derive a machine learning model that predicts respiratory failure within 48 hours of admission based on data from the emergency department.
Methods: Data were collected from patients with COVID-19 who were admitted to Northwell Health acute care hospitals and were discharged, died, or spent a minimum of 48 hours in the hospital between March 1 and May 11, 2020. Of 11,525 patients, 933 (8.1%) were placed on invasive mechanical ventilation within 48 hours of admission. Variables used by the models included clinical and laboratory data commonly collected in the emergency department. We trained and validated three predictive models (two based on XGBoost and one that used logistic regression) using cross-hospital validation. We compared model performance among all three models as well as an established early warning score (Modified Early Warning Score) using receiver operating characteristic curves, precision-recall curves, and other metrics.
Results: The XGBoost model had the highest mean accuracy (0.919; area under the curve=0.77), outperforming the other two models as well as the Modified Early Warning Score. Important predictor variables included the type of oxygen delivery used in the emergency department, patient age, Emergency Severity Index level, respiratory rate, serum lactate, and demographic characteristics.
Conclusions: The XGBoost model had high predictive accuracy, outperforming other early warning scores. The clinical plausibility and predictive ability of XGBoost suggest that the model could be used to predict 48-hour respiratory failure in admitted patients with COVID-19.read more
Citations
More filters
Journal ArticleDOI
Comparing machine learning algorithms for predicting ICU admission and mortality in COVID-19.
Sonu Subudhi,Ashish Verma,Ankit B. Patel,C. Corey Hardin,Melin J. Khandekar,Hang Lee,Dustin McEvoy,Triantafyllos Stylianopoulos,Lance L. Munn,Sayon Dutta,Rakesh K. Jain +10 more
TL;DR: In this paper, the authors compared the performance of 18 machine learning algorithms for predicting ICU admission and mortality among COVID-19 patients and showed that ensemble-based models performed better than other model types at predicting both 5-day ICU admissions and 28-day mortality from COVID19.
Journal ArticleDOI
Machine Learning Based Clinical Decision Support System for Early COVID-19 Mortality Prediction
TL;DR: In this article, the authors proposed a machine learning (ML) method based on blood tests data to predict COVID-19 mortality risk using a powerful combination of five features: neutrophils, lymphocytes, lactate dehydrogenase (LDH), high-sensitivity C-reactive protein (hs-CRP), and age.
Journal ArticleDOI
Potential applications and performance of machine learning techniques and algorithms in clinical practice: A systematic review
TL;DR: In this article , a systematic review examines the performance of machine learning algorithms and evaluates the progress made to date towards their implementation in clinical practice, concluding that the XGBoost (extreme gradient boosting) algorithm showed the highest potential for clinical applications.
Journal ArticleDOI
Diagnosis and prediction of COVID-19 severity: can biochemical tests and machine learning be used as prognostic indicators?
Alexandre de Fátima Cobre,Dile Pontarolo Stremel,Guilhermina Rodrigues Noleto,Mariana Millan Fachi,Monica Surek,Astrid Wiens,Fernanda S. Tonin,Roberto Pontarolo +7 more
TL;DR: In this article, the authors used Artificial Neural Networks (ANN), decision trees (DT), partial least squares discriminant analysis (PLS-DA), and K nearest neighbor algorithm (KNN) models to predict positive and negative patients with COVID-19.
Journal ArticleDOI
Early Prediction of COVID-19 Ventilation Requirement and Mortality from Routinely Collected Baseline Chest Radiographs, Laboratory, and Clinical Data with Machine Learning
Abdulrhman Aljouie,Ahmed Almazroa,Yahya A. Bokhari,Mohammed Alawad,Ebrahim Mahmoud,Eman Alawad,Ali Alsehawi,Mamoon Rashid,Lamya Alomair,Shahad S. Almozaai,Bedoor Albesher,Hassan Alomaish,Rayyan Daghistani,Naif Khalaf Alharbi,Manal Alaamery,Mohammad Bosaeed,Mohammad Bosaeed,Hesham Alshaalan +17 more
TL;DR: In this paper, the authors presented a predictive model for COVID-19 ventilatory support and mortality early on from baseline (at the time of diagnosis) and routinely collected data of each patient (CXR, CBC, demographics, and patient history).
References
More filters
Proceedings ArticleDOI
XGBoost: A Scalable Tree Boosting System
Tianqi Chen,Carlos Guestrin +1 more
TL;DR: XGBoost as discussed by the authors proposes a sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning to achieve state-of-the-art results on many machine learning challenges.
Journal ArticleDOI
Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area.
Safiya Richardson,Safiya Richardson,Jamie S. Hirsch,Jamie S. Hirsch,Mangala Narasimhan,James M. Crawford,Thomas McGinn,Thomas McGinn,Karina W. Davidson,Karina W. Davidson,Douglas P. Barnaby,Douglas P. Barnaby,Lance B Becker,John Chelico,John Chelico,Stuart L. Cohen,Stuart L. Cohen,Jennifer Cookingham,Kevin Coppa,Michael A Diefenbach,Andrew J. Dominello,Joan Duer-Hefele,Louise Falzon,Jordan Gitlin,Negin Hajizadeh,Negin Hajizadeh,Tiffany G. Harvin,David Hirschwerk,Eun Ji Kim,Eun Ji Kim,Zachary Kozel,Lyndonna Marrast,Lyndonna Marrast,Jazmin N. Mogavero,Gabrielle A. Osorio,Michael Qiu,Theodoros P. Zanos +36 more
TL;DR: This case series provides characteristics and early outcomes of sequentially hospitalized patients with confirmed COVID-19 in the New York City area and assesses outcomes during hospitalization, such as invasive mechanical ventilation, kidney replacement therapy, and death.
Journal ArticleDOI
Clinical predictors of mortality due to COVID-19 based on an analysis of data of 150 patients from Wuhan, China.
TL;DR: This study conducted a retrospective multicenter study of 68 death cases and 82 discharged cases with laboratory-confirmed infection of SARS-CoV-2 and confirmed that some patients died of fulminant myocarditis, which is characterized by a rapid progress and a severe state of illness.
Journal ArticleDOI
Missing value estimation methods for DNA microarrays.
Olga G. Troyanskaya,Michael N. Cantor,Gavin Sherlock,Patrick O. Brown,Trevor Hastie,Robert Tibshirani,David Botstein,Russ B. Altman +7 more
TL;DR: It is shown that KNNimpute appears to provide a more robust and sensitive method for missing value estimation than SVDimpute, and both SVD Impute and KNN Impute surpass the commonly used row average method (as well as filling missing values with zeros).
Journal ArticleDOI
A study of the behavior of several methods for balancing machine learning training data
TL;DR: This work performs a broad experimental evaluation involving ten methods, three of them proposed by the authors, to deal with the class imbalance problem in thirteen UCI data sets, and shows that, in general, over-sampling methods provide more accurate results than under-sampled methods considering the area under the ROC curve (AUC).
Related Papers (5)
Utilization of machine-learning models to accurately predict the risk for critical COVID-19.
Dan Assaf,Dan Assaf,Ya’ara Gutman,Ya’ara Gutman,Yair Neuman,Gad Segal,Gad Segal,Sharon Amit,Sharon Amit,Shiraz Gefen-Halevi,Shiraz Gefen-Halevi,Noya Shilo,Noya Shilo,Avi Epstein,Avi Epstein,Ronit Mor-Cohen,Ronit Mor-Cohen,Asaf Biber,Asaf Biber,Galia Rahav,Galia Rahav,Itzchak Levy,Itzchak Levy,Amit Tirosh,Amit Tirosh +24 more
Machine learning in predicting respiratory failure in patients with COVID-19 pneumonia - challenges, strengths, and opportunities in a global health emergency
Davide Ferrari,Jovana Milic,Roberto Tonelli,F. Ghinelli,Marianna Meschiari,Sara Volpi,Matteo Faltoni,Giacomo Franceschi,Vittorio Iadisernia,Dina Yaacoub,Giacomo Ciusa,Erica Bacca,Carlotta Rogati,Marco Tutone,Giulia Burastero,Alessandro Raimondi,Marianna Menozzi,Erica Franceschini,G. Cuomo,Luca Corradi,G. Orlando,Antonella Santoro,M. Di Gaetano,Cinzia Puzzolante,Federica Carli,Andrea Bedini,Riccardo Fantini,Luca Tabbì,Ivana Castaniere,Stefano Busani,Enrico Clini,Massimo Girardis,Mario Sarti,Andrea Cossarizza,Cristina Mussini,Federica Mandreoli,Paolo Missier,Giovanni Guaraldi +37 more
Prediction of COVID-19 deterioration in high-risk patients at diagnosis: an early warning score for advanced COVID-19 developed by machine learning.
Carolin Jakob,Ujjwal M. Mahajan,Marcus Oswald,Melanie Stecher,M. Schons,Julia Mayerle,Siegbert Rieg,Mathias W. Pletz,Uta Merle,Kai Wille,Stefan Borgmann,Christoph D. Spinner,Sebastian Dolff,Clemens Scherer,Lisa Pilgram,Maria Rüthrich,Frank Hanses,Martin Hower,Richard Strauß,Steffen Massberg,Ahmet Görkem Er,Norma Jung,Jörg Janne Vehreschild,Jörg Janne Vehreschild,Hans Stubbe,Lukas Tometten,Rainer König +26 more