Accuracy, Precision, and Trending of 4 Pulse Wave Analysis Techniques in the Postoperative Period.

ABSTRACT

OBJECTIVE: The aim of this study was to analyze the accuracy, precision, and trending ability of the following 4 pulse wave analysis devices to measure continuous cardiac output PiCCO2 ([PCCO]; Pulsion Medical System, Munich, Germany); LiDCORapid ([LCCO]; LiDCO Ltd, London, UK); FloTrac/Vigileo ([FCCO]; Edwards Lifesciences, Irvine, CA); and Nexfin ([NCCO]; BMEYE, Amsterdam, The Netherlands). DESIGN: Prospective, observational clinical study. SETTING: Intensive care unit of a single-center, teaching hospital. PARTICIPANTS: The study comprised 22 adult patients after elective coronary artery bypass surgery. INTERVENTIONS: Three measurement cycles were performed in all patient durings their immediate postoperative intensive care stay before and after fluid loading. Hemodynamic measurements were performed 5 minutes before and immediately after the administration of 500 mL colloidal fluid over 20 minutes. MEASUREMENTS AND MAIN RESULTS: PCCO, LCCO, FCCO, and NCCO were assessed and compared with cardiac output derived from intermittent transpulmonary thermodilution (ICO). One hundred thirty-two matched sets of data were available for analysis. Bland-Altman analysis using linear mixed effects models with random effects for patient and trial revealed a mean bias ±2 standard deviation (%error) of -0.86 ± 1.41 L/min (34.9%) for PCCO-ICO, -0.26 ± 2.81 L/min (46.3%) for LCCO-ICO, -0.28 ± 2.39 L/min (43.7%) for FCCO-ICO, and -0.93 ± 2.25 L/min (34.6%) for NCCO-ICO. Bland-Altman plots without adjustment for repeated measurements and replicates yielded considerably larger limits of agreement. Trend analysis for all techniques did not meet criteria for acceptable performance. CONCLUSIONS: All 4 tested devices using pulse wave analysis for measuring cardiac output failed to meet current criteria for meaningful and adequate accuracy, precision, and trending ability in cardiac output monitoring.