Abstract
Objective: Pericardial effusion is not an uncommon finding in hospitalized patients. Many pericardial effusions are found incidentally through computed tomography (CT) performed for other indications. Echocardiography is usually ordered when an incidentally discovered pericardial effusion is found on the CT to examine the effect of the effusion on hemodynamics and to detect early signs of tamponade. However, in clinical practice, the discrepancy between CT and echocardiography regarding the size of pericardial effusions is common. The accuracy of CT in the evaluation of the size of pericardial effusions is not well-studied. Our study aims to evaluate the accuracy of CT in assessing the size of a pericardial effusion compared with the gold standard echocardiography.
Methods: This is a retrospective study examining patients presenting to the University of Toledo Medical Center (UTMC) with pericardial effusions. One hundred and forty-one patient charts were reviewed and 45 subjects were excluded. Ninety-six patients in whom both CT and echocardiography were performed were enrolled in the final analysis. The time interval between both imaging modalities was limited to less than 14 days and no interventions on the effusion (e.g., pericardiocentesis) occurred in the time interval between the two imaging modalities.
Results: The size of the pericardial effusion was assessed similarly between CT and echocardiography in 50% of the cases (48/96). In the other half of the study population, the results were discrepant; CT was found to overestimate the size of pericardial effusion in 44% of the cases (42/96). The agreement rate between the two modalities is significantly low kappa = 0.111, P = 0.028.
The independent variables age, gender, body mass index (BMI), use of anticoagulants, and renal function had no effect on the agreement between CT and echocardiography.
Conclusion: Computerized tomography tends to overestimate the size of the pericardial effusion compared to echocardiography. Based on an incidental finding of pericardial effusion on CT scan, this discrepancy should be recognized prior to ordering an echocardiogram. Echocardiography can be considered in relevant clinical settings.