Correlation between thorax computed tomography findings and clinical and laboratory data on patients with coronavirus disease 2019

Abstract

OBJECTIVE: We investigated the correlation between computed tomography (CT) scores, laboratory findings, and clinical symptoms in patients with coronavirus disease 2019 (COVID-19). MATERIALS AND METHODS: Clinical, laboratory, and thorax CT findings on the admission of 121 patients with COVID-19 were retrospectively evaluated. CT scores based on lobe involvement and CT patterns (i.e., ground-glass abnormalities, consolidation, and crazy-paving patterns) were estimated, and the relationship between CT score and symptomatic (e.g. fever, cough) versus asymptomatic (e.g., inflammation, coagulation, liver and kidney function) clinical laboratory findings were statistically analyzed. RESULTS: Sixty-eight of 121 patients (56%) were symptomatic; 53 (44%) were asymptomatic. The CT scores of symptomatic patients, especially those with coughing and dyspnea, were statistically higher (2 [0–9] vs. 0 [0–1]; P < 0.001). Erythrocyte sedimentation rate, C-reactive protein, ferritin, D-dimer, fibrinogen, aspartate aminotransferase, lactate dehydrogenase, glucose, prothrombin time and alanine amino transferase values were correlated with CT scores (ρ = 0.638, P < 0.001, ρ = 0.512, P < 0.001; ρ = 0.325, P = 0.001; ρ = 0.452, P < 0.001; ρ = 0.525, P < 0.001; ρ = 0.379, P < 0.001; ρ = 0.445, P < 0.001; ρ = 0.332, P < 0.001, ρ = 0.296, P = 0.003; ρ = 0.222, P = 0.015, respectively). Albumin values were negatively correlated with CT scores (ρ = −0.398, P < 0.001). CONCLUSION: CT scores may help clinicians evaluate the severity of COVID-19 pneumonia and thus help in managing the disease.