Prevention And Control Of COVID-19 In Emergency Department And Establishment Of Early Warning System For Viral Pneumonia

BackgroundCoronavirus disease 2019(COVID-19)is threatening people’ s life throughout the world.Early warning and prevention system needs to be improved.ObjectiveTo summarize the composition of respiratory virus pathogens in fever clinic and analyze the clinical characteristics of different respiratory virus infection.To understand the effect of fever clinic in COVID-19 epidemic.To establish an artificial intelligence early warning system for viral pneumonia.MethodsRetrospectively collected patients with definite etiological results using nasal and pharyngeal swabs in fever clinic.Make a questionnaire survey to collect and analyze the basic information and response of fever clinic to the epidemic situation of COVID-19.Establish an artificial intelligence early warning system for viral pneumonia based on imaging and other clinical characteristics of pneumonia patients.ResultsPart 1Totally,1860 patients were screened and 136 patients were enrolled.72(52.94%)of them were diagnosed as influenza(Flu)A virus infection.32(23.53%)of them were diagnosed as Flu B virus infection.18(13.24%)and 14(10.29%)of them were diagnosed as COVID-19 and respiratory syncytial virus(RSV)infection respectively.COVID-19 group had a higher rate of contact with epidemic area within 14 days and clustering onset than other groups.The most common symptom was fever.The ratio of fever and the highest temperature were higher in Flu A virus infection patients than in COVID-19 patients.COVID-19 patients had lower white blood cell and neutrophil count than Flu A and RSV infection group,but higher lymphocyte count than Flu A and B infection groups.COVID-19 group(83.33%)had higher rate of pneumonia in chest CT scan than Flu A and B virus infection groups.Part 289.29%hospitals had fever clinic for screening of potential infectious diseases before the outbreak of COVID-19 and the percentage increased to 100%in response to the epidemic of COVID-19.Most of the fever clinics were managed by the infection department,and the proportion had no significant change before and after the COVID-19 epidemic.After the outbreak of COVID-19,the number of patients,the percentage of fever clinics setting observation rooms and rescue rooms,the number of observation beds in fever clinics increased compared with those before the outbreak.Both the proportion of rescue equipment including Emergency Rescue Vehicles,monitors,invasive and non-invasive ventilator and the proportion of setting radiation examinations in fever clinics also increased.This study showed that no hospital infection occurred among medical staff or other patients and their families except 2 hospitals in Wuhan.Part 3In this study,190 cases(6018 pictures)of pneumonia were collected,including 104 cases(3041 pictures)of bacterial pneumonia and 86 cases(3067 pictures)of viral pneumonia.Based on the above data and lung imaging,an artificial intelligence diagnosis model of viral pneumonia was established,and the model was verified by chest CT and clinical data of 10 cases of bacterial and viral pneumonia respectively.The accuracy rate of single chest CT was 86%.Taking the average value of all the chest CT prediction results of a patient and combining with the medical history and laboratory data as the basis for judging bacterial or viral pneumonia,the accuracy rate was 85%.ConclusionsPart 1Influenza viruses accounted for a large proportion of respiratory virus infection even though during the epidemic of COVID-19 in Beijing.No single symptom or laboratory finding was suggestive of specific respiratory virus,however,epidemic history was important for screening of COVID-19.Part 2Fever clinic played an important role in control of COVID-19 in China.But the early warning system was not perfect enough.The big data of fever clinic might be helpful for improving the early warning system for infectious diseases.The unified management mode of emergency department and fever clinic and ensuring enough emergency space in the setting might improve the response ability of fever clinic to epidemic.Part 3This study proposes an artificial intelligence diagnostic tool that can quickly identify viral pneumonia.It may play an active role in early warning,prevention and control of sudden major respiratory infectious diseases.