Chest CT features of coronavirus disease-19 (COVID-19) pneumonia: which findings on initial CT can predict an adverse short-term outcome?

objectiveTo study the spectrum of chest CT features in coronavirus disease-19 (COVID-19) pneumonia and to identify the initial CT findings that may have the potential to predict a poor short-term outcome.

methodsThis was a retrospective study comprising 211 reverse transcriptase-polymerase chain reaction (RT-PCR) positive patients who had undergone non-contrast chest CT. Prevalence, extent, pattern, distribution and type of abnormal lung findings were recorded. Patients with positive CT findings were divided into two groups; clinically stable (requiring in-ward hospitalization) and clinically unstable [requiring intensive care unit (ICU) admission or demised] based on short-term follow-up.

resultsLung parenchymal abnormalities were present in 42.2% (89/211) whereas 57.8% (122/211) cases had a normal chest CT. The mean age of clinically unstable patients (63.6 ± 8.3 years) was significantly different from the clinically stable group (44.6 ± 13.2 years) (

conclusionCertain imaging findings on initial CT have the potential to predict short-term outcome in COVID-19 pneumonia. Extensive pulmonary abnormalities, evaluated by combined anteroposterior, central-peripheral and a higher percentage of the total lung involvement, indicate a poor short-term outcome. Similarly, the presence of crazy paving pattern, consolidation with air bronchogram and segmental vascular changes are also indicators of poor short-term outcome. ADVANCES IN KNOWLEDGE: Certain findings on initial CT can predict an adverse short-term prognosis in COVID-19 pneumonia.