Computed tomography scan in COVID-19: a systematic review and meta-analysis

Introduction
Computed tomography (CT) scan is a commonly used tool for the diagnosis of the novel coronavirus disease 2019 (COVID-19), similarly to reverse transcription-polymerase chain reaction (RT-PCR). Because of the limitations of RT-PCR, there is growing interest in the usability of the CT scan. The present systematic review and meta-analysis aims to summarize the available data on the CT scan features of COVID-19.

Material and methods
We conducted a systematic search in electronic databases to find eligible studies published between 1 December 2019 and 4 April 2020, which investigated the computed tomographic features of patients with COVID-19. All preprint and peer-reviewed articles were included. No language limitation was applied. For proportional data, pooled prevalence was calculated using a Freeman-Tukey double arcsine transformation, with a 95% confidence interval (CI).

Results
Eighty-six studies were eligible to be included in the meta-analysis. For 7956 patients, the most common CT findings were bilateral pattern of involvement (78%; 95% CI: 0.73-0.82; p < 0.001), involvement of more than 1 lobe (75%; 95% CI: 0.68-0.82; p < 0.001), ground-glass opacities (GGO) (73%; 95% CI: 0.67-0.78; p < 0.001), and periphe­ral distribution of signs (69%; 95% CI: 0.61-0.76; p < 0.001). Only 5% of patients had a normal CT scan (95% CI: 0.03-0.07; p < 0.001). The proportion of paediatric patients (age < 18 years) with unremarkable CT findings was higher (40%; 95% CI: 0.27-0.55; p < 0.001). Subgroup analysis showed that patients with the severe or critical type of COVID-19 were more likely to have pleural effusion (RR 7.77; 95% CI: 3.97-15.18; p < 0.001) and consolidation (RR 3.13; 95% CI: 1.57-6.23; p < 0.001). CT results in patients with COVID-19 were comparable with those of people having pneumonia from other causes, except for the lower incidence of consolidation (RR 0.81; 95% CI: 0.71-0.91; p < 0.001) and higher risk of showing GGO (RR 1.45; 95% CI: 1.13-1.86; p < 0.001). The mortality rate was slightly higher in patients with bilateral involvement (RR 3.19; 95% CI: 1.07-9.49; p = 0.04).

Conclusions
Our study results show that COVID-19 shares some features with other viral types of pneumonia, despite some differences. They commonly present as GGO along with vascular thickening, air bronchogram and consolidations. Normal CT images, lymphadenopathies, and pleural effusions are not common. Consolidations and pleural effusions correlate with more severe disease. CT features are different between COVID-19 and non-COVID-19 pneumonia. Also, they differ by age, disease severity, and outcomes within COVID-19 patients.