Meta-analysis of chest CT features of patients with COVID-19 pneumonia

The objective of this paper is to perform a meta-analysis regarding the chest computed tomography (CT) manifestations of coronavirus disease-2019 (COVID-19) pneumonia patients. PubMed, Embase, and Cochrane Library databases were searched from 1 December 2019 to 1 May 2020 using the keywords of "COVID-19 virus," "the 2019 novel coronavirus," "novel coronavirus," and "COVID-19." Studies that evaluated the CT manifestations of common and severe COVID-19 pneumonia were included. Among the 9736 searched results, 15 articles describing 1453 common patients and 697 severe patients met the inclusion criteria. Based on the CT images, the common patients were less frequent to exhibit consolidation (odds ratio [OR] = 0.31), pleural effusion (OR = 0.19), lymphadenopathy (OR = 0.17), crazy-paving pattern (OR = 0.22), interlobular septal thickening (OR = 0.27), reticulation (OR = 0.20), traction bronchiectasis (OR = 0.40) with over two lobes involved (OR = 0.07) and central distribution (OR = 0.18) while more frequent to bear unilateral pneumonia (OR = 4.65) involving one lobe (OR = 13.84) or two lobes (OR = 6.95) when compared with severe patients. Other CT features including ground-glass opacities (P = .404), air bronchogram (P = .070), nodule (P = .093), bronchial wall thickening (P = .15), subpleural band (P = .983), vascular enlargement (P = .207), and peripheral distribution (P = .668) did not have a significant association with the severity of the disease. No publication bias among the selected studies was suggested (Harbord's tests, P > .05 for all.) We obtained reliable estimates of the chest CT manifestations of COVID-19 pneumonia patients, which might provide an important clue for the diagnosis and classification of COVID-19 pneumonia.