RESUMO
Objective@#This study evaluated the deep learning (DL) algorithm performance to detect lesions that suggest pneumonia in chest X-rays (CXR) of suspected coronavirus disease 2019 (COVID-19) patients. @*Methods@#This retrospective study included consecutive patients who visited a screening clinic in Daegu, and were suspected to be afflicted with the COVID-19 during the COVID-19 epidemic. CXR were analyzed using the commercial artificial intelligence product that provides free online DL algorithms to the public for COVID-19. Computerized tomography was used as the standard reference. Performance of the DL algorithm was evaluated by the sensitivity and specificity, and results were compared to the CXR records of emergency physicians (EP) in charge of the actual screening triage clinic during the COVID-19 epidemic. @*Results@#Totally, 114 patients were evaluated, of which 38 patients were positive for COVID-19. In 85 CXRs examined (36 COVID-19 and 49 non-COVID-19) with findings of pneumonia in computerized tomography, the DL algorithm showed significantly higher sensitivity as compared to the EP (DL, 98.8% [93.6%-99.9%] vs. EP, 85.9% [76.6%-92.5%]; P<0.01). Moreover, the DL algorithm showed significantly higher sensitivity for detecting CXRs with COVID-19 pneumonia, as compared to the EP (DL, 100.0% [90.3%-100%] vs. EP, 91.7% [77.5%-98.3%]; P=0.08). @*Conclusion@#We conclude that for examining the CXR of patients with suspected COVID-19, sensitivity of the DL algorithm is superior than the EP for detecting lesions suggesting pneumonia. Thus, the application of the DL algorithm is potentially useful in screening triage clinics to detect COVID-19 pneumonia.
RESUMO
Objective@#This study evaluated the deep learning (DL) algorithm performance to detect lesions that suggest pneumonia in chest X-rays (CXR) of suspected coronavirus disease 2019 (COVID-19) patients. @*Methods@#This retrospective study included consecutive patients who visited a screening clinic in Daegu, and were suspected to be afflicted with the COVID-19 during the COVID-19 epidemic. CXR were analyzed using the commercial artificial intelligence product that provides free online DL algorithms to the public for COVID-19. Computerized tomography was used as the standard reference. Performance of the DL algorithm was evaluated by the sensitivity and specificity, and results were compared to the CXR records of emergency physicians (EP) in charge of the actual screening triage clinic during the COVID-19 epidemic. @*Results@#Totally, 114 patients were evaluated, of which 38 patients were positive for COVID-19. In 85 CXRs examined (36 COVID-19 and 49 non-COVID-19) with findings of pneumonia in computerized tomography, the DL algorithm showed significantly higher sensitivity as compared to the EP (DL, 98.8% [93.6%-99.9%] vs. EP, 85.9% [76.6%-92.5%]; P<0.01). Moreover, the DL algorithm showed significantly higher sensitivity for detecting CXRs with COVID-19 pneumonia, as compared to the EP (DL, 100.0% [90.3%-100%] vs. EP, 91.7% [77.5%-98.3%]; P=0.08). @*Conclusion@#We conclude that for examining the CXR of patients with suspected COVID-19, sensitivity of the DL algorithm is superior than the EP for detecting lesions suggesting pneumonia. Thus, the application of the DL algorithm is potentially useful in screening triage clinics to detect COVID-19 pneumonia.
RESUMO
Background@#When an emergency-care patient is diagnosed with an emerging infectious disease, hospitals in Korea may temporarily close their emergency departments (EDs) to prevent nosocomial transmission. Since February 2020, multiple, consecutive ED closures have occurred due to the coronavirus disease 2019 (COVID-19) crisis in Daegu. However, sudden ED closures are in contravention of laws for the provision of emergency medical care that enable the public to avail prompt, appropriate, and 24-hour emergency medical care. Therefore, this study ascertained the vulnerability of the ED at tertiary hospitals in Daegu with regard to the current standards. A revised triage and surveillance protocol has been proposed to tackle the current crisis. @*Methods@#This study was retrospectively conducted at 6 level 1 or 2 EDs in a metropolitan city where ED closure due to COVID-19 occurred from February 18 to March 26, 2020. The present status of ED closure and patient characteristics and findings from chest radiography and laboratory investigations were assessed. Based on the experience from repeated ED closures and the modified systems that are currently used in EDs, revised triage and surveillance protocols have been developed and proposed. @*Results@#During the study period, 6 level 1 or 2 emergency rooms included in the study were shut down 27 times for 769 hours. Thirty-one confirmed COVID-19 cases, of whom 7 died, were associated with the incidence of ED closure. Typical patient presentation with respiratory symptoms of COVID-19 was seen in less than 50% of patients, whereas abnormal findings on chest imaging investigations were detected in 93.5% of the study population. The chest radiography facility, resuscitation rooms, and triage area were moved to locations outside the ED, and a new surveillance protocol was applied to determine the factors warranting quarantine, including symptoms, chest radiographic findings, and exposure to a source of infection. The incidence of ED closures decreased after the implementation of the revised triage and surveillance protocols. @*Conclusion@#Triage screening by emergency physicians and surveillance protocols with an externally located chest imaging facility were effective in the early isolation of COVID-19 patients. In future outbreaks of emerging infectious diseases, efforts should be focused toward the provision of continued ED treatment with the implementation of revised triage and surveillance protocols.