Accuracy of a Novel Approach to Measuring Arterial Thermodilution Cardiac Output During Intra-Aortic Counterpulsation

Objective: To assess the agreement between a novel approach of arterial and the pulmonary artery bolus thermodilution for measuring cardiac output in critically ill patients during aortic counterpulsation. Methods: Eighteen male patients aged 37-80years, undergoing preoperative insertion of an intra-aortic balloon pump (IABP) and elective coronary artery bypass grafting. A thin 1.3FG thermistor was introduced through the pressure lumen to the tip of an 8FG IABP catheter, and the pump rate was set at 1:1. After arrival in the intensive care unit cardiac output (CO) was measured under haemodynamic steady-state conditions hourly for 8-11h, and arterial bolus thermodilution (BCOiabp) and pulmonary artery bolus thermodilution (BCOpulm) were determined after the patients' admission to the intensive care unit. Results: A total of 198 data pairs were obtained: 177 with aortic counterpulsation and 21 without. During aortic counterpulsation, median CO was 6.8l/min for BCOiabp and 6.1l/min for BCOpulm, without aortic counterpulsation; corresponding values were 7.1l/min for BCOiabp and 6.5l/min for BCOpulm with aortic counterpulsation. Mean bias was +0.77l/min, limits of agreement ( ± 2SD) were -1.27/+2.81l/min, and mean error (2SD/[(BCOiabp+BCOpulm)/2] was 31.4%. Without aortic counterpulsation, corresponding values were +0.43l/min, -1.03/+1.87l/min, and 22.4%. Conclusions: Agreement between BCOiabp and BCOpulm was satisfactory for CO values between 2.0 and 10l/min only without aortic counterpulsation. BCOiabp CO measurements during aortic counterpulsation after coronary artery bypass grafting cannot be recommended at the present tim