Accuracy and sensitivity of three-dimensional echocardiography to detect changes in right ventricular volumes: comparison study with cardiac magnetic resonance.

We aimed to investigate the ability of three-dimensional transthoracic echocardiography (3DE) to detect changes in RV volumes compared to cardiac magnetic resonance (CMR). Eighty-five subjects including 45 with no known cardiac disease and 40 patients with a variety of cardiac diseases were included. Two- and three-dimensional echocardiography as well as CMR of the RV was performed before and after infusion of on average two litres of saline. Examinations were analysed with estimation of RV dimensions, volumes and ejection fraction (RVEF). Intra- and inter-examiner variability was evaluated in 25 patients randomly selected from the cohort. Three-dimensional echocardiography underestimated volumes and RVEF compared to CMR with mean differences and 95% limits of agreement of 110.3 ± 59 mL for RV end-diastolic volume (RVEDV), 43.3 ± 32 mL for RV end-systolic volume (RVESV) and 3.5 ± 10.7% for RVEF. CMR was more reproducible than 3DE, with intra-observer coefficient of variation (CV) of 4% vs. 14.2% for RVEDV, 9.7% vs. 16.7% for RVESV and 6.3% vs. 8.6% for RVEF. The RVEDV, RVESV and RV stroke volume (RVSV) by CMR significantly increased after saline infusion by 15.3 ± 16.2 mL, 3.5 ± 14.2 mL and 11.8 ± 12.6 mL, respectively, as well as RVEF by 1.5 ± 4.6% (p < 0.05). However, 3DE was not able to detect any of these changes in RV volumes (p ≥ 0.05). Compared to CMR imaging of the RV, three-dimensional echocardiography appears unable and unreliable in detecting RV volume changes of less than 15%, highlighting the need for cautious utility of 3DE in these circumstances.