We performed a prospective observational study in post-operative cardiac surgery patients. All patients had a PAC with CCI monitoring capability. We connected the LiDCO apparatus to a radial artery line and performed a one-time calibration with a lithium dilution indicator. In order to test the least invasive method possible, we used a peripheral intravenous (IV) line for indicator delivery rather than the conventional central line technique. We recorded paired PAC/LiDCO-PulseCO CCI measurements every minute for 3 h. We blinded investigators and clinicians to minimally invasive data with an opaque shield over the monitor. We assessed agreement with Bland-Altman analysis.
We obtained 1485 paired measurements in 8 subjects. The mean CI was 2.9 L/min/m2. By Bland-Altman plot, PAC and LiDCO measurements showed minimal bias (2212;0.01), but the 95 % limits of agreement (±2SD) of ± 1.3 L/min/m2 were relatively wide with respect to the mean.
This calibrated minimally invasive (i.e. radial arterial line and peripheral IV) technique demonstrated low bias compared with CCI measured by PAC. However, the relatively wide confidence limits indicate that differences in the two measurements could still be clinically significant.