Previously submitted to: Journal of Medical Internet Research (no longer under consideration since Jun 21, 2021)
Date Submitted: Jan 2, 2021
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Prediction of the risk of severe outcome in patients infected with COVID-19: Based on XGBoost method
ABSTRACT
Background:
The coronavirus disease (COVID-19) has spread worldwide and posed a great threat to human health.
Objective:
To establish a new prediction model for the prognosis of severe patients with COVID-19, so as to provide more comprehensive, accurate and timely indicators for early and dynamic monitoring of severe patients.
Methods:
Based on the patient's admission indicators, the severity of the initial classification of COVID-19, dynamic changes of admission indicators (the difference between indicators of two measurements) and other input variables, a prediction model was established to evaluate the risk of serious outcomes in COVID-19 patients after admission using the XGBoost method.
Results:
Prediction model included 372 subjects screened six predictors with higher scores. The high-risk range of the predictor variables was calculated as: blood oxygen saturation < 94%, peripheral blood leukocyte count > 5.77× 109, systolic blood pressure change > -28 mmHg, heart rate > 102.5 beats / min, multiple patchy shadows, age > 46.5 years old. Predictive sensitivity and accuracy of the model based on the training set and test set were 77.6%, 70.1% and 47.1%, 70.2%, respectively.
Conclusions:
Prediction model based on XGBoost method was more comprehensively and accurately to predict the risk of severe outcomes in patients with COVID-19, as well as reducing the rate of missed diagnosis in critically ill patients compared with the traditional prediction model.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.