ABC | Volume 113, Nº2, August 2019

Original Article Test-Retest Reliability of Non-Invasive Cardiac Output Measurement during Exercise in Healthy Volunteers in Daily Clinical Routine Michael Thomas Coll Barroso, 1 Matthias Wilhelm Hoppe, 2, 3 P hilip Boehme, 1,4 Thomas Krahn, 5 Christian Kiefer, 1 Frank Kramer, 6 Thomas Mondritzki, 1,7 Phillipe Pirez, 6 Wilfried Dinh 1,6 Helios Clinics Wuppertal - Department of Cardiology, University Hospital Witten/Herdecke, 1 Wuppertal – Germany University Wuppertal - Department of Movement and Training Science, 2 Wuppertal – Germany Department of Orthopedic, Trauma, Hand and Neuro Surgery, Klinikum Osnabrück GmbH, 3 Osnabrück – Germany Bayer AG - Open Innovation & Digital Technologies, 4 Wuppertal – Germany Bayer AG - Research & Development, Pharmaceuticals, Biomarker Research, 5 Wuppertal – Germany Bayer AG - Translational Science, Clinical Sciences Experimental, 6 Wuppertal – Germany Bayer AG - Research & Development, Pharmaceutical, Preclinical Research, 7 Wuppertal – Germany Mailing Address: Wilfried Dinh • Bayer Pharma AG - Drug Discovery, Clinical Sciences, Experimental Medicine Aprather Weg 18a Wuppertal 46414 – Germany E-mail: wilfried.dinh@bayer.com Manuscript received July 19, 2018, revised manuscript October 31, 2018, accepted November 14, 2018 DOI: 10.5935/abc.20190116 Abstract Background: Thoracic bioreactance (TB), a noninvasive method for the measurement of cardiac output (CO), shows good test-retest reliability in healthy adults examined under research and resting conditions. Objective: In this study, we evaluate the test-retest reliability of CO and cardiac power (CPO) output assessment during exercise assessed by TB in healthy adults under routine clinical conditions. Methods: 25 test persons performed a symptom-limited graded cycling test in an outpatient office on two different days separated by one week. Cardiorespiratory (power output, VO 2peak ) and hemodynamic parameters (heart rate, stroke volume, CO, mean arterial pressure, CPO) were measured at rest and continuously under exercise using a spiroergometric system and bioreactance cardiograph (NICOM, Cheetah Medical). Results: After 8 participants were excluded due to measurement errors (outliers), there was no systematic bias in all parameters under all conditions (effect size: 0.2-0.6). We found that all noninvasively measured CO showed acceptable test-retest-reliability (intraclass correlation coefficient: 0.59-0.98; typical error: 0.3-1 . 8). Moreover, peak CPO showed better reliability (intraclass correlation coefficient: 0.80-0.85; effect size: 0.9-1.1) then the TB CO, thanks only to the superior reliability of MAP (intraclass correlation coefficient: 0.59-0.98; effect size: 0.3-1.8). Conclusion: Our findings preclude the clinical use of TB in healthy subject population when outliers are not identified. (Arq Bras Cardiol. 2019; 113(2):231-239) Keywords: Cardiac Output; Cardiography, Impedance/methods; Exercise; Exercise Test/methods; Echocardiography/ methods; Reproducibility of Results; Adult. Introduction Cardiac output (CO) is an important physiological surrogate parameter, reflecting the hemodynamic demands of the organism. CO measuring has a wide application spectrum 1 and can provide information on hemodynamic status in patients 2 as well as athletes. 3 In chronic heart failure, CO is decreased and patients suffer from exercise intolerance. 4,5 In contrast, the athlete´s heart shows structural and functional adaptations due to training 6 resulting in a higher CO. 7 Interestingly altered cardiac structure and function do not predict exercise intolerance 8,9 or CO response 3 in both cases. Thus, cardiopulmonary exercise testing is necessary and peak oxygen consumption (VO 2peak ) is measured to determine exercise capacity. 10,11 But, estimation of VO 2 is influenced by several non-cardiac factors, 4,12 and can, therefore, be misleading. 9,13,14 Furthermore, CO cannot be accurately predicted from cardiopulmonary exercise testing. 4,15 However, to evaluate hemodynamic status, catheter-based measuring (i.e., Fick method, thermodilution method) is considered as the clinical standard. 16,17 Since such invasive methods are associated with high risk, their applicability is restricted. 18,19 Therefore, noninvasive measuring methods (i.e., transoesophageal echocardiography, lithium dilution CO, pulse contour CO, partial CO 2 rebreathing, thoracic electrical bioimpedance) were developed. 17 Of the noninvasive measuring methods, especially the thoracic electrical bioimpedance was frequently used in clinical studies and evaluated for its reliability. 20 However, thoracic bioreactance (TB) is a further promising technology to noninvasively monitor CO. 21 TB is based on the measurement of blood flow-related phase shifts of transthoracic electric signals to monitor noninvasively and continuously CO. Therefore theoretically, TB is superior to other methods 22,23 and has been used in several clinical settings. 21,23-25 But, before TB can be adopted for clinical and performance decision making, test´s quality criteria, as the test-retest reliability, must be fulfilled. 231