Prognostic value of chest computed tomography in adult inpatients with coronavirus disease 2019 pneumonia in Goma, in the Democratic Republic of the Congo

Abstract

Objectives: Chest computed tomography (CT) plays a crucial role in evaluating patients with coronavirus disease 2019 (COVID-19) pneumonia. This study aimed to assess the association between initial chest CT findings and mortality in adult inpatients with COVID-19 in the city of Goma, North Kivu province, the Democratic Republic of the Congo. Materials and Methods: This was a multicenter retrospective study of patients hospitalized in Goma with COVID-19 pneumonia who underwent chest CT from January 1, 2021, to December 31, 2022. For each patient included, two experienced radiologists independently reviewed the initial chest CT. A multivariate logistic regression was performed to identify chest CT findings associated with mortality at the 5% significance level. Results: Of 76 patients included in the study, 26 (34.2%) died. The degree of lung parenchymal involvement >25% (adjusted odds ratio [aOR] = 16.27 [3.30–80.16]) and the presence of consolidations (aOR = 3.33 [1.00–11.47]) were predictive of mortality with an area under the receiver operating characteristic curve of 0.8392. Sensitivity was 73.08%, specificity was 82%, positive predictive value was 67.86%, and negative predictive value was 85.42%. Conclusion: Lung involvement >25% and consolidations on chest CT appear to predict death in adult inpatients with COVID-19 pneumonia.