Predictive Accuracy of Hemodynamic Surrogate Indices in Patients with Advanced Heart Failure

Abstract

<h3>Purpose</h3> Right heart catheterization (RHC) is an essential tool in the evaluation advanced heart failure (AHF) patients. It is unclear which of the established hemodynamic surrogate indices are most predictive of poor outcomes. We aimed to compare the predictive accuracy of published hemodynamic surrogates in AHF patients. <h3>Methods</h3> In a pooled registry of patients undergoing RHC for evaluation of decompensated heart failure, we analyzed several hemodynamic indices with cut-off values taken from previously published studies, including: systolic blood pressure / pulmonary capillary wedge pressure (SBP/PCWP) ratio < 4; cardiac power index (CPI=MAP x CI/451) <0.44; cardiac power efficiency (CPE= CPI/PCWP) <0.018; aortic pulsatility index (API=SBP-DBP/PCWP) <1.45; pulmonary artery pulsatility index (PAPi=PASP-PADP/CVP) <2.0; CVP + PCWP >30; CVP/PCWP >0.6. The primary composite outcome was defined as freedom from advanced heart failure therapies, inotropes, or death. <h3>Results</h3> A total of 182 patients (median age 59, 71% male) were included and were on oral medical therapy at the time of RHC. The median (mmHg) SBP was 113 (101, 122), MAP 86 (76.3, 94), RA 15 (10, 18), mean PA 39.5 (32, 47.6), PCWP 26 (20, 32), cardiac index 1.8 (L/m/m2) (1.5, 2.1). At 30 days, 122 patients met the primary composite outcome including 5 deaths. At end of study follow up (up to 365 days) 15 additional patients (n=137) met the primary outcome. All hemodynamic surrogates at their designated cut-off values were associated with an increased risk of the primary outcome at 30 days (<i>P</i><0.05), and at follow-up up to one year with the exception of API (Figure). The receiver operator characteristic analyses for all indices at their pre-designated cut-off values had a comparable range of risk discrimination of 0.58 to 0.64 (Figure) at one-year. <h3>Conclusion</h3> Novel hemodynamic surrogates may be additive in risk prognostication of AHF patients, but can vary in diagnostic accuracy at their pre-specified cut-off values when applied to external clinical cohorts.