A New HRCT Score for Diagnosing SARS-CoV-2 Pneumonia: A Single-Center Study with 1153 Suspected COVID-19 Patients in the Emergency Department

Abstract

The 2019 coronavirus disease (COVID-19) pandemic is affecting millions of people worldwide. Chest high-resolution computed tomography (HRCT) is commonly used as a diagnostic test for suspected COVID-19; however, despite numerous attempts, there is no single scoring system that is widely accepted and used in clinical practice to estimate the probability of SARS-CoV-2 pneumonia. The aim of this single-center retrospective study is to develop a radiological score to predict the probability of COVID-19 with HRCT. Patients admitted to the emergency department with symptoms suggestive of COVID-19 who underwent both HRCT and RT-PCR on nasopharyngeal swab to detect SARS-CoV-2 infection between 1 March and 30 April 2020 were included. A multivariable regression analysis was conducted to identify all HRCT signs independently associated with a positive RT-PCR assay for SARS-CoV-2 and build the HRCT score. A total of 1153 patients were enrolled in this study. The number of segments with ground glass opacities (OR 1.18, 95% CI 1.11–1.26), number of segments with linear opacities (OR 1.21, 95% CI 1.05–1.42), crazy paving patterns (OR 6, 95% CI 3.79–9.76), and vascular ectasia in each segment (OR 2.46, 95% CI 1.1.5–5.8) were included in the score. The HRCT score showed high discriminatory power (area under the ROC curve of 0.8267 [95% CI 0.8–0.85]) with 72.2% sensitivity, 86.6% specificity, 78% PPV, and 83% NPV for its best cut-off. In summary, the HRCT score has good diagnostic and discriminatory accuracy for COVID-19 and is easy and quick to perform.