[Advances in epidemiology, etiology, and treatment of community-acquired pneumonia].

Community-acquired pneumonia (CAP) is the third leading cause of death worldwide and one of the most commonly infectious diseases. Its epidemiological characteristics vary with host and immune status, and corresponding pathogen spectrums migrate over time and space distribution. Meanwhile, with the outbreak of COVID-19, some unconventional treatment strategies are on the rise. This article reviewed the epidemiological characteristics, pathogen spectrum and treatment direction of CAP in China over the years, and aimed to provide guidance for the diagnosis and treatment of CAP in clinical practice.

[Advances in SARS-CoV-2 diagnostic strategies during the later COVID-19 era].

Based on natural infection or vaccination, the protective barrier for population has been preliminarily established. However, with constant appearances of SARS-CoV-2 variants, breakthrough infection events cannot be completely avoided, and thus the diagnostic strategy is still the key to discovering epidemic sources and blocking the transmission chain. Currently, SARS-CoV-2 diagnosis technologies based on nucleic acid, antigen and antibody detections have developed and extended in diversity. Under the background of work resumption and epidemic-prevention normalization during the later COVID-19 era, it is necessary for us to choose appropriate detection methods to satisfy the need of epidemic prevention and control in various scenarios. We summarized the principles and applicable characteristics of existing SARS-CoV-2 detection technologies in this paper, aimed to provide guidance for clinical and public health personnel to make targeted decisions.

[Analysis of the clinical characteristics of COVID-19 reinfection cases].

Objective: To explore the epidemiological and clinical characteristics of COVID-19 reinfection cases. Methods: The published COVID-19 reinfection cases were reviewed and the relevant data were extracted, including the baseline characteristics of patients, the results of antibody tests, and the whole-genome sequencing results of the viral strains. Results: We reviewed 29 reinfection cases in 20 reports from 14 countries. The age of re-infected patients ranged from 21 to 90 years (median 53 years), and there was no significant difference in gender distribution. Among the 29 patients, 11 were health care workers, 6 received immunosuppressive drugs (including glucocorticoids), 17 presented more severe symptoms than their primary infections and 5 (all aged over 80 years) died. The interval of the two infections was usually less than 60 days when the patients were infected by the same viral strain,while the interval was much longer (median 78.5 days) when the patients were infected by different viral strains. Nine patients had negative antibody test results or low antibody titers when the reinfections were confirmed, and 5 of them had negative antibody test results even during the initial infection. Conclusions: Virus-specific antibodies had a protective effect against COVID-19 reinfection for the majority of the population, however, this effect may decrease over time. Occupational exposure, low levels of antibodies, or an inability to produce antibodies may be the main risk factors for reinfection. Advanced age was a major risk factor for a poor prognosis. Effective personal prevention and social distancing were still essential for the prevention of reinfection.

A Spatial-temporal Graph based Hybrid Infectious Disease Model with Application to COVID-19

As the COVID-19 pandemic evolves, reliable prediction plays an important role in policymaking. The classical infectious disease model SEIR (susceptible-exposed-infectious-recovered) is a compact yet simplistic temporal model. The data-driven machine learning models such as RNN (recurrent neural networks) can suffer in case of limited time series data such as COVID-19. In this paper, we combine SEIR and RNN on a graph structure to develop a hybrid spatio-temporal model to achieve both accuracy and efficiency in training and forecasting. We introduce two features on the graph structure: node feature (local temporal infection trend) and edge feature (geographic neighbor effect). For node feature, we derive a discrete recursion (called I-equation) from SEIR so that gradient descend method applies readily to its optimization. For edge feature, we design an RNN model to capture the neighboring effect and regularize the landscape of loss function so that local minima are effective and robust for prediction. The resulting hybrid model (called IeRNN) improves the prediction accuracy on state-level COVID-19 new case data from the US, out-performing standard temporal models (RNN, SEIR, and ARIMA) in 1-day and 7-day ahead forecasting. Our model accommodates various degrees of reopening and provides potential outcomes for policymakers.