Abstract: Objective To explore the risk factors that may be related to severe/critical COVID-19 patients，so as to early identify patients with severe/critical high-risk indicators and make appropriate interventions and adjustments in a timely manner. Methods A total of 474 patients were included in the study. The differences in general characteristics，underlying diseases，clinical symptoms，laboratory findings and outcomes between mild/normal and severe/critical patients were observed retrospectively. The binary logistics model was used for multi-factor analysis to explore the risk factors that may be related to the development of severe/critical type. Results Severe/critically patients had higher mortality，older age and longer hospital stays. There were 305 patients complicated with pre-existing comorbidities，and single factor study found that neutrophil count， the levels of blood urea nitrogen（BUN），fibrinogen，D-dimer，CRP，procalcitonin，creatine kinase isoenzyme and lactate dehydrogenase were higher in severe/critical patients. However，lymphocyte count，platelet count，macrophage count，albumin，globulin ratio and total protein were lower，and there was a significant difference between the two groups（P < 0. 01）. Multivariate binary logistic analysis showed that the patients complicated with pre-existing comorbidities OR = 1. 706，95%CI（1. 074，2. 709），P = 0. 024. Lymphocyte count OR = 0. 521，95%CI（0. 342，0. 792），P = 0. 002；BUN OR = 1. 054，95%CI（1. 001，1. 110），P = 0. 044；CRP OR = 1. 107，95%CI（1. 024，1. 197），P = 0. 011；LDH OR = 1. 004，95% CI（1. 001，1. 007），P = 0. 007. Conclusion There is a higher risk of developing into severe/critical cases for the COVID-19 patients complicated with pre-existing comorbidities and laboratory examination results suggest that the higher levels of BUN，CRP，LDH，and lower lymphocyte count. Therefore，monitoring and treatment regimens of patients in the hospital could be strengthened and timely adjusted.

Key words: COVID-19, SARS-CoV-2, clinical characteristics, risk factors