529 PERSPECTIVES ON CARDIAC OUTPUT EVALUATION IN PULMONARY HYPERTENSION PATIENTS UNDERGOING RIGHT HEART CATHETERIZATION

Abstract

Abstract Background and aim Right heart catheterization (RHC) is the gold standard for the diagnosing and classifying pulmonary hypertension (PH). Cardiac output (CO) and the derived cardiac index (CI) are cornerstones of the hemodynamic evalution allowing pulmonary vascular resistance (PVR) calculation and risk stratification. According to the current European Guidelines, CO should be assessed by the direct Fick method or thermodilution (TD). The indirect Fick (IF) method based on theorically estimated values for VO2, is deemed acceptable bul lacks reliability. Of note, data from the literature on a direct comparison among these three methods used for CO measurement are poor. The objective of our study is to compare CO evaluation by the DF method (DFCO) with CO assessed by IF (IFCO) and TD (TDCO). Secondly, we aimed at assessing the impact of CO estimation by different methods on PVR measurement and therefore on PH hemodynamic classification. This might translate into relevant diagnostic, prognostic and therapeutic implications. Methods 27 patients underwent RHC. Breath-by breath VO2 was measured through a Rudolph mask connected with a two-way respiratory valve to a metabolic cart and averaged over the course of 12 to 15 minutes. DFCO was calculated using the following formula: VO2 (ml/min)/(arteriovenous oxygen content difference×10). For IFCO, the same formula was used with VO2 computed by the Dehmer and the LaFarge-Miettinen formulas. TDCO was obtained through the injection of 10 ml of cold saline solution (0-4°C), in the right atrium and by measuring the temperature variation in the pulmonary artery. Results RHC revealed PH in 23 patients (85%) Average VO2 by Dehmer formula differed significantly from direct measurement (228.9 ± 27.9 ml/min and 204.4 ± 43.6 ml/min respectively; P = 0.017). By using LaFarge formula, mean estimated VO2 differed from direct measurement without reaching statistical significance (185.3 ± 31.9 ml/min and 204.4 ± 43.6 ml/min respectively, P= 0.072). Average CO values did not differ significantly from each other (DFCO 4.2 ± 1.5 L/min; IFCO 4.7 ± 1.3 L/min; TDCO 4.6 ± 1.2 L/min; P=NS for all pairwise comparisons) and showed significant Pearson's correlation. Out of 27 cases, TDCO differed for more (± 0.5 L/min) than 10% as compared to DFCO in 17 (63%). IFCO differed consistently from DFCO in 12 (44%) and IFCO from TDCO in 17 (63%). Despite discrepancies among different CO measurement methods, Cohen's kappa coefficient showed good agreement between the three methods in high PVR (≥3 WU) estimation (IF vs DF 0.79; TD vs DF 0.83 and IF vs TD 0.92 respectively). Conclusions Our results indicate consistent discrepancies among different approaches for VO2 and CO measurement, however this did not impact over PVR assessment and therefore in PH classification. Probably, this is due to the small sample size and the low number of patients (8 of 27) with values around the pathological cut-off for PVR (3 WU). Further studies are needed to better investigate the magnitude of CO measurement discrepancies among different methods on PVR calculation and hemodynamic classification.