This study was a retrospective, observational study conducted in Shenzhen Children’s Hospital, a 1300-bed tertiary care facility in Shenzhen, China. The study population consisted of all consecutive patients with acute respiratory symptoms hospitalized between May 2019 and Aug 2019. Children with positive adenovirus test by immunofluorescence assay or PCR in respiratory tract specimens and radiographic findings of pneumonia were included. Children with any of the following factors were excluded: newborns; infection of HIV; leukemia; known or suspected active tuberculosis; receiving immunosuppressive agents; immunodeficiency; chemotherapy; and chronic conditions (malnutrition, congenital heart disease; chronic lung disease).
Classification of pneumonia severity
Classification of pneumonia severity was performed by the criteria of the community-acquired pneumonia guideline in China . Based on the clinical symptoms and chest imaging findings, patients were divided into severe pneumonia group and mild pneumonia group. Severe cases were identified in the presence of at least one of the following signs: disturbance of consciousness, significant tachypnea (respiratory rate > 70 breaths per minute in infants and > 50 breaths per minute in older children), cyanosis, dyspnea, oxygen saturation < 92%, extrapulmonary complication, dehydration, refusal to eat, and severe chest imaging findings (pneumothorax, pleural effusion, pulmonary atelectasis, or multilobe infiltrates).
Data collection and management
The clinical variables were measured every day during hospitalization. Blood draws were done during hospitalization as required for guiding management decisions. Demographic information (age and sex), signs and symptoms (temperature, blood pressure, pulse and respiratory rate, cough, tachypnea, cyanosis, etc.), laboratory results (hematology, organ function, pathogen tests, etc.), chest image results (chest x-ray and/or computed tomography), bronchoscopy results, treatment (oxygen supply, endotracheal intubation, antimicrobial, etc.) and outcome (survival, death, recovery or discharged against medical advice) were recorded. Co-infection of mycoplasma pneumoniae (MP) was defined as positive PCR test of mycoplasma pneumoniae DNA in respiratory tract specimens (oropharyngeal swab or bronchoalveolar lavage fluid) during hospitalization. Co-infection of influenza virus was defined as positive antigen or PCR test of influenza virus in respiratory tract specimens (nasopharyngeal swab or bronchoalveolar lavage fluid) during hospitalization.
Sample management and virus detection
All samples were transported to the laboratory within 4 hours. Respiratory tract samples were tested for adenovirus by D3® UltraTM DFA Respiratory Virus Screening and ID Kit (Diagnostic Hybrids, Inc. USA) or Adenovirus DNA Detection Kit (Shenzhen Puruikang Biotech Co.,Ltd). Serum samples were stored at -80℃ until adenovirus PCR analysis. Quantification of adenovirus in serum samples was performed on a commercial fluorescence quantitative PCR kit (Daan Gene, Cat. Guangzhou, China) following the protocol of the manufacturer. The limit of detection (LOD) was 500 copies/mL.
We did a univariate correlation analysis of demographic and laboratory variables to determine the statistical significance of the pairwise associations between the severe and mild pneumonia group. Mann-Whitney test and chi-square test were used for quantitative and qualitative variables, respectively. We further did binary logistic regression analysis to identify independent demographic and laboratory risk factors for severe pneumonia.
Log10-transformed concentrations of serum viral load were used as independent variables in analysis. Serum viral load below the LOD were assigned a viral load of 1 copy/mL (0 log10 copies/mL). Continuous variables were summarized as mean (standard deviation, SD) when they were normally distributed and as median (interquartile range, IQR) if they had a skewed distribution. Sex, age, highest white blood cell (WBC) count during hospitalization, mycoplasma pneumoniae co-infection, influenza virus co-infection, and highest serum viral load in the disease course were applied as independent variables. Data analysis was performed by SPSS 26.0 software. All P-values were two-tailed, and P < 0.05 was considered to indicate statistical significance.