Patients
We analyzed the laboratory test results of 3057 COVID-19 patients, including 1455 mild or moderate, 1417 severe, 150 critical, and 35 unclassified cases, admitted to Wuhan Huoshenshan Hospital from February 4 to March 30, 2020. A total of 3051/3057 (99.8%) patients were older than 18 years old. The severity degree of each patient was determined according to the clinical classification criterion in Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia released by the National Health Commission (trial version 7; http://en.nhc.gov.cn/2020-03/29/c_78469.htm). Patients who met any of the following criteria were diagnosed as severe cases: (1) shortness of breath defined by respiration rate ≥ 30 breaths/min, (2) oxygen saturation ≤ 93 at rest, and (3) alveolar oxygen partial pressure/fraction of inspiration O2 (PaO2/FiO2) ≤ 300 mmHg (1 mmHg = 0.133 kPa). Patients whose pulmonary imaging showed significant progression of lesions > 50% within 24–48 h were also treated as severe cases. Patients who met any of the following conditions were diagnosed as critical cases: (1) respiratory failure requiring mechanical ventilation, (2) shock, and (3) organ failure needing intensive care unit (ICU) monitoring and treatment. Also, the severity degree of each patient in this study was defined as the most serious disease state during hospitalization. We obtained the clinical characteristics and laboratory findings of all patients from the electronic medical records of the hospital. This study was approved by the Medical Ethical Committee of Wuhan Huoshenshan Hospital. Written informed consent was obtained from each patient. The summary of necessary information (Supplementary Data Sheet 1), biochemical indicators (Supplementary Data Sheet 2), immune phenotype (Supplementary Data Sheet 3) and antibody level (Supplementary Data Sheet 4) were provided. High-dose steroids and tocilizumab were not used in this cohort, while the information about the use of low-dose steroids in patients (Supplementary Data Sheet 5) was provided.
The lymphocyte subgroup assay
The lymphocyte subgroups were measured by Flow cytometry (CytoFLEX flow cytometry system, Beckman coulter, Inc.) using commercially available kits (Beckman coulter, Inc.) according to the manufacture’s protocol. Briefly, the reagents of the BD six-color lymphocyte subgroup (FITC-CD3, PE-CD16/PE-CD56, PerCP-Cy5.5-CD45, PE-Cy7-CD4, APC-CD19, and APC-Cy7–CD8) were mixed with the whole blood and incubated at room temperature for 20 min, followed by adding 1 mL of a lysis solution with 30 min incubating. The proportion of CD3+, CD3+/CD4+, CD3+/CD8+, CD3−/CD19+, CD3−/CD56+/CD16+ cells in lymphocytes was analyzed with the software.
Serum anti-SARS-CoV-2 antibodies assay
Total SARS-CoV-2 IgM or IgG in the serum was measured by chemiluminescence using commercially available kits (Shenzhen YHLO Biotech Co., Ltd.), which was coated with N and S proteins, in 1850 patients at different time points. In addition, 416 of these patients were tested for S-specific, RBD-specific, and N-specific IgM and IgG levels at different time points by chemiluminescence using commercially available kits (Nanjing RealMind Biotech Co., Ltd.). Briefly, the blood samples were centrifuged at room temperature, the supernatant was taken and incubated with antigen-coated magnetic beads. The antigen-antibody complex is then captured, incubated, and reacted with hydrogen peroxide in an excitatory buffer. Relative luminescence intensity was recorded in the ACL2800 chemrenaliluminescence system (Nanjing RealMind Biotech Co., Ltd.). The relative luminescence intensity was converted to AU/ML antibody levels. Relative antibody levels were presented as the measured chemiluminescence values divided by the constant derived from the linear correlation, which was signal-to-cutoff (S/CO). S/CO > 1 was defined as positive and S/CO ≤1 as negative (Nanjing RealMind Biotech Co., Ltd.). Similarly, S/CO > 10 was defined as positive and S/CO ≤10 as negative (Shenzhen YHLO Biotech Co., Ltd.). Also, we validated the performance of commercial kits for antibody detection. None of nine healthy controls, five patients infected with hepatitis B virus, or five patients with syphilis tested positive for S-IgM, S-IgG, RBD-IgM, RBD-IgG, N-IgM, or N-IgG. All of nine COVID-19 patients tested positive for S-IgM, S-IgG, RBD-IgM, RBD-IgG, N-IgM, or N-IgG (Supplementary Figure S1, Supplementary Data Sheet 6).
Definition of physiological function abnormalities
Patients whose B-type natriuretic peptide (BNP) level was not within the normal range (0–100 pg/ml) (Supplementary Table S1) were defined as patients with abnormal cardiac function. Patients whose creatinine (CRE) level was not within the normal range (57–111 umol/L vs. 41–81 umol/L in males and females, respectively) were defined as patients with renal function abnormality. If any of the indicators, which were total bile acid (TBA, 0–10 umol/L), total bilirubin (TBIL, 0–26 umol vs. 0–21 umol in males and females, respectively), direct bilirubin (DBIL, 0–8 umol/L), indirect bilirubin (IBIL, 0–14 umol/L), glutamic-pyruvic transaminase (GPT, 9–60 IU/L), glutamic oxalacetic transaminase (GOT, 7–45 IU/L), and alkaline phosphatase (ALP, 45–125 IU/L vs. 35–135 IU/L in males and females, respectively) was not within the normal range, these patients were defined as patients with abnormal liver function.
Statistical analysis
Statistical analysis was performed in R version 3.6.0. We used the Wilcoxon rank-sum test or Fisher’s exact test to compare the difference between groups where appropriate. Continuous and categorical variables were presented as median (IQR) and n (%), respectively.
Survival was estimated according to the Kaplan–Meier method by R package “survival”. The log-rank test was used to assess statistical significance.
To recognize the risk factors for death from COVID-19 patients, those variables associated with survival of COVID-19 patients (age, sex, pre-existing diseases, days from symptom onset to admission, and days from admission to discharge) were evaluated using univariable Cox regression models by “coxph” function in R package “survival”. P-value < 0.05 was considered statistically significant. Those variables with not significant P-value from the Wald test were removed one at a time, while the significant variables of univariable analysis were entered into the multivariable Cox proportional hazards model and analyzed by “coxph” function in R package “survival”. Also, P-value < 0.05 was considered statistically significant in multivariable analysis.