Precision of echocardiographic estimates of right atrial pressure in patients with acute decompensated heart failure.

BACKGROUND: Several methods that estimate right atrial pressure (RAP) from echocardiographic parameters have been proposed. However, their precision (i.e., how much they decrease RAP estimation uncertainty) is unknown. The aim of this prospective study was to evaluate and compare the precision of previously proposed RAP estimates in patients with acute decompensated heart failure.

METHODS: Echocardiographic and invasive hemodynamic data were acquired in 75 patients with acute decompensated heart failure. Measurements were made at the start and 48 to 72 hours after the beginning of treatment. RAP was estimated by method 1, using the cutoffs defined by inferior vena cava diameter (IVCd) and IVCd percentage change (IVCd%change) during inspiration, and by method 2, using IVCd%change and systolic to diastolic hepatic flow ratio (S/Dhep). Method 3 was used in patients with sinus rhythm, using the ratio of early tricuspid inflow and early diastolic tissue Doppler tricuspid annular velocities (E/E'ta). RAP was also estimated by resting IVCd, IVCd during inspiration, IVCd%change, right ventricular regional isovolumetric relaxation time, E/E'ta, right atrial volume index, S/Dhep, right ventricular Tei index, right ventricular E/A, and right atrial emptying fraction. Precision gain was measured as the difference between the standard deviation of RAP and the standard error of the estimate of RAP.

RESULTS: Method 1 (r = 0.48, P < .05), IVCd during inspiration (r = 0.49, P < .0001), IVCd%change (r = 0.41, P < .0001) and IVCd (r = 0.40, P < .0001) had the highest correlation with RAP. The highest gain in precision was also observed with the above methods (9%, 13%, 9%, and 8%, respectively). All other parameters had poor correlation with RAP.

CONCLUSION: In patients with advanced heart failure, echocardiographic RAP prediction methods showed only modest precision. Furthermore, none of the tested methods resulted in clinically relevant improvements of RAP estimates. Estimating RAP from a single IVCd measurement is at least as precise as using complex prediction methods.