Evaluation of the use of the fourth version FloTrac system in cardiac output measurement before and after cardiopulmonary bypass

• Published in: Journal of clinical monitoring and computing Vol. 32; no. 5; pp. 807 - 815
• Main Authors: Lin, Sheng-Yi, Chou, An-Hsun, Tsai, Yung-Fong, Chang, Su-Wei, Yang, Min-Wen, Ting, Pei-Chi, Chen, Chun-Yu
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.10.2018; Springer Nature B.V
• Subjects: Aged; Anesthesiology; Bias; Cardiac Output; Cardiopulmonary Bypass; Catheterization, Peripheral; Catheters; Critical Care Medicine; Female; Health Sciences; Heart surgery; Hemodynamic Monitoring - instrumentation; Hemodynamic Monitoring - methods; Hemodynamic Monitoring - statistics & numerical data; Humans; Intensive; Male; Medicine; Medicine & Public Health; Middle Aged; Monitoring, Intraoperative; Original Research; Patients; Pulmonary arteries; Pulmonary Artery; Radial Artery; Reproducibility of Results; Statistics for Life Sciences; Surgery; Thermodilution; Cardiopulmonary bypass; FloTrac; Cardiac output; Cohen κ coefficient; Arterial waveform analysis
• ISSN: 1387-1307; 1573-2614
• DOI: 10.1007/s10877-017-0071-6

The FloTrac system is a system for cardiac output (CO) measurement that is less invasive than the pulmonary artery catheter (PAC). The purposes of this study were to (1) compare the level of agreement and trending abilities of CO values measured using the fourth version of the FloTrac system (CCO-FloTrac) and PAC-originated continuous thermodilution (CCO-PAC) and (2) analyze the inadequate CO-discriminating ability of the FloTrac system before and after cardiopulmonary bypass (CPB). Fifty patients were included. After exclusion, 32 patients undergoing cardiac surgery with CPB were analyzed. All patients were monitored with a PAC and radial artery catheter connected to the FloTrac system. CO was assessed at 10 timing points during the surgery. In the Bland–Altman analysis, the percentage errors (bias, the limits of agreement) of the CCO-FloTrac were 61.82% (0.16, − 2.15 to 2.47 L min) and 51.80% (0.48, − 1.97 to 2.94 L min) before and after CPB, respectively, compared with CCO-PAC. The concordance rates in the four-quadrant plot were 64.10 and 62.16% and the angular concordance rates (angular mean bias, the radial limits of agreement) in the polar-plot analysis were 30.00% (17.62°, − 70.69° to 105.93°) and 38.63% (− 10.04°, − 96.73° to 76.30°) before and after CPB, respectively. The area under the receiver operating characteristic curve for CCO-FloTrac was 0.56, 0.52, 0.52, and 0.72 for all, ≥ ± 5, ≥ ± 10, and ≥ ± 15% CO changes (ΔCO) of CCO-PAC before CPB, respectively, and 0.59, 0.55, 0.49, and 0.46 for all, ≥ ± 5, ≥ ± 10, and ≥ ± 15% ΔCO of CCO-PAC after CPB, respectively. When CO < 4 L/min was considered inadequate, the Cohen κ coefficient was 0.355 and 0.373 before and after CPB, respectively. The accuracy, trending ability, and inadequate CO-discriminating ability of the fourth version of the FloTrac system in CO monitoring are not statistically acceptable in cardiac surgery.