Objective To understand epidemiological and clinical characteristics of coronavirus disease 2019 (COVID-19) confirmed cases and excluded cases, and provide recommendations for clinical diagnosis and treatment. Methods Patients with suspected COVID-19 staying in a hospital were studied, those who were with positive 2019-nCoV nucleic acid test were as confirmed case group, those who were with negative 2019-nCoV nucleic acid test for at least two times, and excluded after consultation by Jing'an District expert group of Shanghai were as excluded case group, differences in epidemiology, clinical manifestation, laboratory examination and imaging data between two groups were compared. Results Of 14 confirmed cases, male-female ratio was 1.33:1, median age was 51.5 years (38.3, 63.0), 10 cases (71.43%) had a clear epidemiological exposure history; of 84 excluded cases, male-female ratio was 0.95:1, median age was 37 years (29.0, 59.75), 14 cases (16.67%) had a clear epidemiological exposure history; there were both significant differences in age distribution and epidemiological exposure history between two groups (both P<0.05), but no significant differences in gender distribution, residence distribution and clinical manifestations (all P>0.05). The percentage of patients with normal or decreased leukocytes in confirmed case group was higher than that in excluded case group (92.86% vs 72.62%, P<0.001), percentage of patients with decreased lymphocyte was lower than that in excluded case group (P=0.001). Comparison of chest CT of two groups showed that the percentage of ground glass exudation shadow and spot exudation shadow in confirmed case group was higher than that in excluded case group, difference was statistically significant (P<0.05); the median time from scattered or multiple pneumonia to ground glass pneumonia in confirmed case group was 3 (2, 5) days. Conclusion Characteristic factors for COVID-19 are age, clear epidemiological history, leukocyte count, percentage of lymphocyte, and CT manifestations of ground glass-like lung, clinicians should discriminate in diagnosis and treatment, further improve diagnosis accuracy of COVID-19.