RESUMEN
(1) Background: Previous studies reported limited performance of arterial pressure waveform-based cardiac output (CO) estimation (FloTrac/Vigileo system; CO-FloTrac) compared with the intermittent thermodilution technique (COint). However, errors due to bolus maneuver and intermittent measurements of COint could limit its use as a reference. The continuous thermodilution technique (COcont) may relieve such limitations. (2) Methods: The performance of CO-FloTrac was retrospectively assessed using continuous recordings of intraoperative physiological data acquired from patients who underwent off-pump coronary artery bypass graft (OPCAB) surgery with CO monitoring using both CO-FloTrac and COcont. Optimal time adjustments between the two measurements were determined based on R-squared values. (3) Results: A total of 134.2 h of data from 30 patients was included in the final analysis. The mean bias was -0.94 (95% CI, -1.35 to -0.52) L/min and the limits of agreements were -3.64 (95% CI, -4.44 to -3.08) L/min and 1.77 (95% CI, 1.21 to 2.57) L/min. The percentage error was 66.1% (95% CI, 52.4 to 85.8%). Depending on the time scale and the size of the exclusion zone, concordance rates ranged from 61.0% to 75.0%. (4) Conclusion: Despite the time adjustments, CO-FloTrac showed non-negligible overestimation, clinically unacceptable precision, and poor trending ability during OPCAB surgery.
