Background: Cardiac output (CO) is an important cardiac parameter, however its determination is difficult in clinical routine. Non-invasive inert gas rebreathing (IGR) measurements yielded promising results in recent studies. It directly measures pulmonary blood flow (PBF) which equals CO in absence of significant pulmonary shunt flow (Q(S)). A reliable shunt correction requiring the haemoglobin concentration (c(Hb)) as only value to be entered manually has been implemented. Therefore, the aim of the study was to evaluate the effect of various approaches to Q(S) correction on the accuracy of IGR.
Methods: Cardiac output determined by cardiac magnetic resonance imaging (CMR) served as reference values. The data was analysed in four groups: PBF without correcting for Q(S) (group A), shunt correction using the patients' individual c(Hb) values (group B), a fixed standard c(Hb) of 14.0 g dl(-1) (group C) and a gender-adapted standard c(Hb) for male (15.0 g dl(-1)) and female (13.5 g dl(-1)) probands each (group D).
Results: 147 patients were analysed. Mean CO(CMR) was 5.2 +/- 1.4 l min(-1), mean CO(IGR) was 4.8 +/- 1.3 l min(-1) in group A, 5.1 +/- 1.3 in group B, 5.1 +/- 1.3 l min(-1) in group C and 5.1 +/- 1.4 l min(-1) in group D. The accuracy in group A (mean bias 0.5 +/- 1.1 l min(-1)) was significantly lower as compared to groups B, C and D (0.1 +/- 1.1 l min(-1); P<0.01).
Conclusion: IGR allows a reliable non-invasive determination of CO. Since PBF significantly increased the measurement bias, shunt correction should always be applied. A fixed c(Hb) of 14.0 g dl(-1) can be used for both genders if the exact c(Hb) value is not known. Nevertheless, the individual value should be used if any possible.