ABSTRACT
Background Multifaceted non-pharmaceutical interventions during the COVID-19 pandemic have not only reduced the transmission of SARS-CoV2, but have had an effect on the prevalence of other pathogens. This retrospective study aimed to compare and analyze the changes of respiratory pathogens in hospitalized children with community-acquired pneumonia.Methods From January 2019 to December 2020, children with community-acquired pneumonia were selected from the Department of Respiratory Medicine, Shanghai Children's Medical Center. On the first day of hospitalization, sputum, throat swabs, venous blood samples from them were collected for detection of pathogens.Results A total of 2596 children with community-acquired pneumonia were enrolled, including 1871 patients in 2019 and 725 in 2020. The detection rate in 2020 was lower than in 2019, whether single or multiple pathogens. Compared with 2019, the detection rate of virus, especially parainfluenza virus, influenza virus and respiratory syncytial virus, all decreased in 2020. On the contrary, the prevalence of human rhinovirus was much higher than that in 2019. In addition, the positivity rate for bacteria did not change much over the two years, which seemed to be less affected by COVID-19. And Mycoplasma pneumoniae which broke out in 2019 has been in low prevalence since March 2020 even following the reopening of school.Conclusions Strict public health interventions for COVID-19 in China have effectively suppressed the spread of not only SARS-CoV2 but parainfluenza virus, influenza virus and Mycoplasma pneumonia as well. However, it had a much more limited effect on bacteria and rhinovirus. Therefore, more epidemiological surveillance of respiratory pathogens will help improve early preventive measures.
Subject(s)
COVID-19 , Pneumonia , Pneumonia, MycoplasmaABSTRACT
There is an urgent need for automated methods to assist accurate and effective assessment of COVID-19. Radiology and nucleic acid test (NAT) are complementary COVID-19 diagnosis methods. In this paper, we present an end-to-end multitask learning (MTL) framework (COVID-MTL) that is capable of automated and simultaneous detection (against both radiology and NAT) and severity assessment of COVID-19. COVID-MTL learns different COVID-19 tasks in parallel through our novel random-weighted loss function, which assigns learning weights under Dirichlet distribution to prevent task dominance; our new 3D real-time augmentation algorithm (Shift3D) introduces space variances for 3D CNN components by shifting low-level feature representations of volumetric inputs in three dimensions; thereby, the MTL framework is able to accelerate convergence and improve joint learning performance compared to single-task models. By only using chest CT scans, COVID-MTL was trained on 930 CT scans and tested on separate 399 cases. COVID-MTL achieved AUCs of 0.939 and 0.846, and accuracies of 90.23% and 79.20% for detection of COVID-19 against radiology and NAT, respectively, which outperformed the state-of-the-art models. Meanwhile, COVID-MTL yielded AUC of 0.800 $\pm$ 0.020 and 0.813 $\pm$ 0.021 (with transfer learning) for classifying control/suspected, mild/regular, and severe/critically-ill cases. To decipher the recognition mechanism, we also identified high-throughput lung features that were significantly related (P < 0.001) to the positivity and severity of COVID-19.
Subject(s)
COVID-19ABSTRACT
The outbreak of novel coronavirus pneumonia (NCP) has become the most severe public health issue at the moment, threatening people′s lives. Pediatricians in Shanghai have recently launched a discussion on the focused questions of NCP, including the incidence situation, epidemiological features, essentials of early screening, treatment and nosocomial infection prevention of children′s novel coronavirus infection (2019-nCoV), and further put forward the experts proposal upon the patterns of disease occurrence, development, diagnosis and control, for the reference of frontline pediatricians.