Abstract

INTRODUCTION: The reliability of pulse pressure analysis to estimate cardiac output is known to be affected by arterial load changes. However, the contribution of each aspect of arterial load could be substantially different. In this study, we evaluated the agreement of eight non-commercial algorithms of pulse pressure analysis for estimating cardiac output (PPCO) with esophageal Doppler cardiac output (EDCO) during acute changes of arterial load. In addition, we aimed to determine the optimal arterial load parameter that could detect a clinically significant difference between PPCO and the EDCO. METHODS: We included mechanically ventilated patients monitored with a prototype esophageal Doppler (CardioQ-Combi™, Deltex Medical, Chichester, UK) and an indwelling arterial catheter who received a fluid challenge or in whom the vasoactive medication was introduced or modified. Initial calibration of PPCO was made with the baseline value of EDCO. We evaluated several aspects of arterial load: total systemic vascular resistance (TSVR=mean arterial pressure [MAP]/EDCO*80), net arterial compliance (C=EDCO-derived stroke volume/pulse pressure), and effective arterial elastance (Ea=0.9*systolic blood pressure/EDCO-derived stroke volume). We compared CO values with Bland-Altman analysis, four-quadrant plot and a modified polar plot (with least significant change analysis). RESULTS: A total of 16,964-paired measurements in 53 patients were performed (median 271; interquartile range: 180-415). Agreement of all PPCO algorithms with EDCO was significantly affected by changes in arterial load, although the impact was more pronounced during changes in vasopressor therapy. When looking at different parameters of arterial load, the predictive abilities of Ea and C were superior to TSVR and MAP changes to detect a PPCO-EDCO discrepancy≥10% in all PPCO algorithms. An absolute Ea change>8.9±1.7% was associated with a PPCO-EDCO discrepancy≥10% in most algorithms. CONCLUSIONS: Changes in arterial load profoundly affected the agreement of PPCO and EDCO, although the contribution of each aspect of arterial load to the PPCO-EDCO discrepancies was significantly different. Changes in Ea and C mainly determined PPCO-EDCO discrepancy.