Two simple echo-Doppler measurements can accurately identify pulmonary hypertension in the large majority of patients with chronic heart failure.

BACKGROUND: The assessment of pulmonary hypertension in patients with heart failure is of great clinical importance not only for diagnostic purposes but also for prognostication. The present study was undertaken on a consecutive basis with a group of patients with chronic heart failure. Patients were evaluated for their suitability for heart transplantation: (1) to explore the diagnostic accuracy of several echo Doppler parameters of pulmonary hemodynamics in predicting the presence of elevated pulmonary artery pressure (defined as pulmonary artery systolic pressure > or =35 mmHg and mean pulmonary artery pressure >20 mmHg); (2) to assess the diagnostic ability of the same parameters to identify patients with elevated pulmonary vascular resistance; and (3) to evaluate the influence of right ventricular function and degree of tricuspid regurgitation in modifying diagnostic accuracy.
METHODS: Echo Doppler examination and right heart catheterization were performed consecutively within 24 hours in 86 patients. The optimal cut-off value for a series of echo Doppler parameters capable of identifying patients with pulmonary hypertension was obtained by dividing the entire sample into 2 groups; the optimal threshold (highest sensitivity and specificity) of the echo and Doppler parameters used to classify patients with and without pulmonary hypertension was determined in 67% of cases by means of the receiver operating characteristic (ROC) curve: this was the testing sample. The proportion of cases classified correctly according to the selected cut-off was computed. The remaining 33% of cases represented the validation sample: sensitivity, specificity and predictive values (and their 95% confidence intervals [CIs]) for identifying pulmonary hypertension were calculated from the proposed cut-offs in this second sample. Finally, the overall performance of the echo Doppler parameters was assessed over the whole sample by considering the extent of the area under the ROC curve (A-ROC) and its 95% CI, for the dichotomic measurement.
RESULTS: On right heart catheterization, a pulmonary artery systolic (PAPs) pressure > or =35 mmHg plus a mean pressure (mean PAP) >20 mmHg was documented in 49 of 86 cases (57%), for whom mean values were 56 +/- 17 and 38 +/- 11 mmHg, respectively. The proportion of cases identified correctly as having pulmonary hypertension was highest for PAPs (88%) and mean PAP (85%) in addition to acceleration time of pulmonary artery systolic flow (ACT) (79%) and pulmonary artery diastolic pressure obtained utilizing the early phase of the tricuspid regurgitation spectral flow (PAPd/TR) (75%). PAPd/TR performed better in the validating sample in terms of diagnostic ability, with high sensitivity and specificity (100% and 60%) and positive and negative predictive values (PPV 80%, NPV 100%). PAPs, mean PAP, ACT and PAPd/TR confirmed their prevailing diagnostic ability (A-ROC from 0.74 to 0.86) in identifying pulmonary hypertension with fair to high feasibility (67% to 91%) and an odds ratio (OR) indicative of strong association. ACT and PAPd/TR, the 2 parameters with the highest feasibility, allowed us to identify 46 of 49 (94%) hypertensive cases. The same parameters did not perform well in identifying patients with increased vascular resistance, with A-ROC ranging from 0.55 to 0.69. Heterogeneity of effect, due to right ventricular function or tricuspid regurgitation degree, could not be demonstrated in the ability of the echo Doppler measurements to identify pulmonary hypertension.
CONCLUSIONS: ACT, PAPd/TR, PAPs and mean PAP have been shown to accurately classify patients with chronic heart failure with or without pulmonary hypertension. In particular, ACT and PAPd/TR alone allowed reliable and accurate definition of pulmonary hypertension in 94% of patients, regardless of right ventricular function or degree of tricuspid regurgitation. Non-invasive pulmonary pressure assessment by the referred method might be useful in the evaluation of heart transplant candidates.