Head-to-head comparison of cardiac magnetic resonance imaging and transthoracic echocardiography in the general population (MATCH)

BACKGROUND: Comparing transthoracic echocardiography (TTE) and cardiac magnetic resonance imaging (CMR) is crucial for cardiac assessment. This study aims to clarify their comparability in a large population sample. METHODS: CMR and two- (2D) and three-dimensional (3D) TTE were used to quantify left and right heart dimensions in participants of the Hamburg City Health Study. Intertechnique agreement was evaluated using Bland-Altman analyses, Pearson correlation coefficients, and Cohen's Kappa. RESULTS: Data from 2126 participants (median age 67 [IQR: 59-72] years, 897 (42.2%) female) were analyzed. Left ventricular (LV) diastolic volumes were similar (CMR: 117.0 [96.2, 138.0] ml, 2D-TTE: 111.8 [93.6, 134.3] ml, r = 0.7, p < 0.001), while systolic volumes were lower with CMR (CMR: 36.0 [26.9, 46.0] ml, 2D-TTE: 46.7 [37.9, 57.5] ml, r = 0.67, p < 0.001). CMR LV ejection fraction (LVEF) was 10% higher than 2D-TTE (CMR = 69.0 [64.0, 74.0]%, 2D-TTE = 58.3 [55.5, 61.7]%, p < 0.001; r = 0.40, p < 0.001). Left atrial volumes correlated moderately with low bias (CMR: 53.0 [40.0, 68.0] ml, 2D-TTE: 51.6 [41.5, 64.0] ml, r = 0.63, p < 0.001). LV mass showed good correlation but was higher using 2D-TTE (r = 0.74, p < 0.001). Right ventricular (RV) volumes showed the largest differences, with CMR demonstrating lower interobserver variability (ICC = 0.97 vs. 0.61 for 2D-TTE) and markedly larger volumes (RVEDV mean bias = 74.8 ml, r = 0.50, p < 0.001). CONCLUSION: In a large general population, CMR quantifies cardiac function and dimensions more reliably than TTE. Both modalities provide significantly different absolute values, limiting intertechnique transferability. TRIAL REGISTRATION: Retrospectively registered at ClinicalTrial.gov, registration number: NCT03934957, registration date: 04/01/2019, https://clinicaltrials.gov/study/NCT03934957 .