Continuous thermodilution cardiac output: Clinical validation against a reference technique of known accuracy

Abstract

Objective: To evaluate the accuracy and precision of continuous thermodilution (CCO) by using a validated bolus thermodilution (BCO) reference technique as criterion standard. Design: Under circulatory steady state conditions, a CCO system (Vigilance, software versions 4.35 and 4.39) was validated with regard to CCO as well as iced and room temperature BCO. Setting: Intensive care unit at a university hospital. Patients: Method comparison was conducted in 56 cardiac surgical patients, 28 patients being allocated to one of the two software versions, and 14 within each group to either iced or room temperature BCO. Measurements and results: CCO readings were registered in duplicate before and after three to five bolus injections conducted with both the Vigilance and reference systems. Iced BCO showed excellent agreement between the Vigilance and reference systems, yielding SDs of bias of 0.41 and 0.37 l/min and linear correlation coefficients (r) of 0.97 and 0.96. With room temperature BCO, agreement was significantly less. CCO, irrespective of software version, showed higher SDs of bias (0.90 and 0.84 l/min)and lower r values (0.84 and 0.81) than iced BCO (p < 0.0001). CCO measurements with software version 4.39 yielded a similar SD of bias to that with room temperature BCO. Conclusion: Decreased precision-of CCO as compared to iced BCO may, in clinical settings, be outweighed by the advantages of automated and continuous monitoring. Under research conditions, however, iced BCO remains the method of choice.