Validation data extraction and analysis
To verify the results of the meta-analysis, all laboratory-confirmed COVID-19 cases in two hospitals in Wenzhou (the First and Second affiliated Hospitals of Wenzhou Medical University) from January 22 to February 22, 2020 were included. Follow-up was performed until discharge, death, or March 16, 2020. The basic characteristics of the patients were extracted and compared between severe and non-severe cases.
We extracted the following information: (1) Demographic characteristics, including age, gender, height and weight. According to Chinese adult obesity standards, patients were classified as obese (BMI≥28.0), overweight (24.0≤BMI<28.0), normal (18.5≤BMI<24.0), and lean (BMI<18.5) 15. (2) Wuhan exposure history: the patients were divided into four groups: direct exposure to Wuhan (the patient recently lived or visited Wuhan); indirect contact with Wuhan (close contact with people who had been to Wuhan during the epidemic); contact with other confirmed diagnoses patients; no clear history of contact. (3) Clinical symptoms, onset to hospitalization time.