Respiratory Tract Pathogen Profiles of COVID-19 Pneumonia Patients and the Mortality Prediction

Background COVID-19 pneumonia is easily complicated with other respiratory pathogenic attacks, increasing the risk of death. Exploring the pathogen profiles of COVID-19 patients-related facilitated the clinical management and decisions to pursue better prognosis. Methods This study systematically compared the sputum culture results and death events of 170 non-COVID-19 and 197 COVID-19 patients. Statistical analysis was carried out to find the pathogen profile difference between the two populations. The death risk model was constructed for the infected COVID-19. Results It was found that co-infection with bacteria and fungi increased the mortality of COVID-19 pneumonia patients. The isolation rate of Acinetobacter baumannii in COVID-19 patients was significantly higher than that in non-COVID-19 patients and often showed multi-drug resistant phenotypes. The COVID-19 pneumonia patients showed a higher incidence of intensive care unit admission, ventilator-assisted ventilation and death with fungal infection. The serum levels of interleukin-1, interleukin-6, interleukin-8, TNF, lymphocytes, neutrophils and white blood cells in patients with COVID-19 pneumonia decreased. A death prediction model was constructed based on machine learning methods, achieving a prediction accuracy of 90.0%. The main factors affecting the survival rate of COVID-19 pneumonia patients co-infected with other pathogens were admission to the intensive care unit, days of hospital stay, ventilator-aided treatment, carbapenems administration, lymphocyte, serum aspartate aminotransferase level, Acinetobacter baumannii infection, and Candida infection. Conclusions This study provided necessary clinical indicators for timely and precise intervention of COVID-19 pneumonia patients when they were infected by other pathogens. The COVID-19-related secondary infection microorganisms were different compared with the pathogens isolated from non-COVID-19 patients.