Comparison of clinical characteristics in patients with coronavirus disease and influenza A in Guangzhou, China.

Jiang, Wen-Qiang; Li, Xu-Sheng; Zhong, Wen-Hong; Huang, Lin-Qiang; Lin, Xiao-Jun; Wen, Miao-Yun; Deng, Yu-Jun; Li, Xin; Zeng, Hong-Ke

BACKGROUND This study aims to compare the epidemiological, clinical and laboratory characteristics between patients with coronavirus disease (COVID-19) and influenza A (H1N1), and to develop a differentiating model and a simple scoring system. METHODS We retrospectively analyzed the data from patients with COVID-19 and H1N1. The logistic regression model based on clinical and laboratory characteristics was constructed to distinguish COVID-19 from H1N1. Scores were assigned to each of independent discrimination factors based on their odds ratios. The performance of the prediction model and scoring system was assessed. RESULTS A total of 236 patients were recruited, including 20 COVID-19 patients and 216 H1N1 patients. Logistic regression revealed that age >34 years, temperature ≤37.5 °C, no sputum or myalgia, lymphocyte ratio ≥20% and creatine kinase-myocardial band isoenzyme (CK-MB) >9.7 U/L were independent differentiating factors for COVID-19. The area under curves (AUCs) of the prediction model and scoring system in differentiating COVID-19 from H1N1 were 0.988 and 0.962, respectively. CONCLUSIONS There are certain differences in clinical and laboratory features between patients with COVID-19 and H1N1. The simple scoring system may be a useful tool for the early identification of COVID-19 patients from H1N1 patients.