Concordance of Pericardial Effusion Size Between Computed Tomography and Echocardiography

Discrepancy between computed tomography (CT) and transthoracic echocardiography (TTE) regarding pericardial effusion (PEff) size is common, but there is limited data regarding the correlation between these 2 imaging methods. The aim of this study is to examine the real-world concordance of observed PEff size between CT and TTE. We performed a retrospective analysis of all imaging reports available from 2013 to 2019 and identified patients with a PEff who underwent both a chest CT and TTE within a 24-hour period. We evaluated the agreement between CT and TTE in assessing PEff size. Of 1,118 patients included in the study, mean age was 66 (±17 years) and 54% were female. The median time interval between the 2 studies was 9.4 hours (interquartile range 3.5 to 16.6). Patients within a half-grade or full-grade of agreement were 71.9% and 97.2%, respectively. The mean difference in grade of agreement (TTE minus CT) between the 2 imaging methods was -0.1 (±0.6, p <0.0001). CT was more likely to report a higher grade (i.e. larger PEff size) when compared with TTE (261 patients vs 157 patients, p <0.001). The weighted kappa was 0.73 (95% confidence interval 0.69 to 0.76). After excluding patients with trace/no effusion, 42.3% and 94.1% of patients' studies were within a half-grade or full-grade of agreement, respectively. Of the 18 patients who had large discrepancies, 9 patients had loculated effusions, 2 patients had large pleural effusions, and 6 patients had suboptimal TTEs images. In conclusion, TTE and CT showed relatively strong agreement in estimation of PEff size, with CT sizes larger than TTE, on average. Large discrepancies in size may be related to reduced image quality, large pleural effusions, and loculated PEff.