[Comparison of 2 modified Fick methods and thermodilution for determining the cardiac output in patients with mechanical ventilation].

Objective: To evaluate the effectiveness of two cardiac output measurement methods based on a modified Fick equation, that calculate the O2 consumption (VO2) dividing the CO2 production (VCO2) by a fixed respiratory quotient (RQ).

Design: Comparative study.

Setting: One 5 beds general intensive care unit in a 500 beds general hospital.

Patients: Ten mechanically ventilated critically ill patients requiring the right heart catheterization. The mean age was 65.5 years and the mean APACHE II score at admission was 24.

Measurements: The cardiac output was measured using two modified Fick methods. The VO2 was calculated dividing VCO2 by 0.9, while the arteriovenous O2 content difference was calculated using the mixed venous O2 content in the first method (COF), and the central venous O2 content in the second one (COFra). Moreover the cardiac output was measured with the thermodilution technique (COTD) as gold standard.

Results: The mean difference between the COTD and COF determinations was 0.571 L +/- 1.81 L, with limits of agreement ranging from -3.057 to +4.200 L. The mean difference between the COTD and COFra determinations was -0.322 L +/- 2.05 L, with limits of agreement ranging from -4.430 to +3.785 L. Both differences were nonsignificant. The correlation coefficients with COTD were: COF determinations 0.72, COFra determinations 0.70. In the group of COFra determinations less than 7 L the mean difference between COTD and COFra was 0.495 L with limits of agreement ranging from +2.208 L to -1.218 L.

Conclusions: The correlation coefficients of the two modified Fick methods with COTD are good, and the mean differences between their results and the gold standard are small, but the low precision of both tested methods demonstrated by the very large limits of agreement, severely reduce the clinical reliability of the measurements. Only for the less than 7 L cardiac outputs the COFra limits of agreement with COTD are narrow enough, and in this range the technique can be useful e.g. revealing a low cardiac output.