Machine Learning-Based Decision Model to Distinguish between COVID-19 and Influenza: A Retrospective, Two-Centered, Diagnostic Study

Background: Considering the current situation of the novel coronavirus disease (COVID-19) epidemic control, it is highly likely that COVID-19 and influenza may coincide during the approaching winter season. However, there is no available tool that can rapidly and precisely distinguish between these two diseases in the absence of laboratory evidence of specific pathogens. Methods: Laboratory confirmed COVID-19 and influenza patients from Zhongnan Hospital of Wuhan University (ZHWU) and Wuhan No.1 Hospital (WNH) between December 1, 2019 and February 29, 2020, were included for analysis. A machine learning-based decision model was developed using the XGBoost algorithms. The specificity, sensitivity, positive and negative predictive values (PPV/NPV), accuracy and area under the receiver operating characteristic curve (AUC) were calculated to evaluate the model performance. Findings: The data from 357 COVID-19 and 1893 influenza patients from ZHWU were divided into a training and a testing set in the ratio 7:3. The external test used the data of 308 COVID-19 and 312 influenza patients from WNH. In the training and testing sets, the model achieved good performance in identifying COVID-19 from influenza with an accuracy of 0.968 (AUC, 0.943 (95%CI 0.925, 0.962)) and 0.960 (AUC, 0.928 (95%CI 0.897, 0.959)), respectively. Our decision tree suggested that older age (>16 years), higher hsCRP (>14.2 mg/L) and lower monocyte (≤0.68×109/L) drive the prediction towards COVID-19. In addition, the external test determined a COVID-19 prediction accuracy of 0.839 (AUC, 0.839 (95%CI: 0.811, 0.868). Interpretation: Machine learning provides a tool that can rapidly and accurately distinguish between COVID-19 and influenza cases. This finding would be particularly useful in regions having massive COVID-19 and influenza cases while limited resources for laboratory test of specific pathogens. Funding: National Natural Science Foundation of China (81900097) and the Emergency Response Project of Hubei Science and Technology Department (2020FCA002, 2020FCA023). Declaration of Interests: None reported. Ethics Approval Statement: This study was approved by the Medical Ethics Committee, Zhongnan Hospital of Wuhan University (Clinical Ethical Approval No. 2020020).