Accuracy and precision of echocardiography versus right heart catheterization for the assessment of pulmonary hypertension.

BACKGROUND Echocardiographic studies have contributed to progress in the understanding of the pathophysiology of the pulmonary circulation and have been shown to be useful for screening for and prognostication of pulmonary hypertension, but are considered unreliable for the diagnosis of pulmonary hypertension. We explored this apparent paradox with rigorous Bland and Altman analysis of the accuracy and the precision of measurements collected in a large patient population. METHODS A total of 161 patients referred for a suspicion of pulmonary hypertension were prospectively evaluated by a Doppler echocardiography performed by dedicated cardiologists within 1 h of an indicated right heart catheterization. RESULTS Nine of the patients (6%) were excluded due to an insufficient signal quality. Of the remaining 152 patients, 10 (7%) had no pulmonary hypertension and most others had either pulmonary arterial hypertension (36%) or pulmonary venous hypertension (40%) of variable severities. Mean pulmonary artery pressure, left atrial pressure and cardiac output were nearly identical at echocardiography and catheterization, with no bias and tight confidence intervals, respectively ± 3 mm Hg, ± 5 mm Hg and ± 0.3 L/min. However, the ± 2SD limits of agreement were respectively of + 19 and - 18 mm Hg for mean pulmonary artery pressure, + 8 and - 12 mm Hg for left atrial pressure and + 1.8 and - 1.7 L/min for cardiac output. CONCLUSIONS Doppler echocardiography allows for accurate measurements of the pulmonary circulation, but with moderate precision, which explains why the procedure is valid for population studies but cannot be used for the individual diagnosis of pulmonary hypertension.