Previous Abstract | Next Abstract
Printable Version
Fick Partial Rebreathing CO2 Cardiac Output: Is There Clinical Agreement with Intraoperative Transesophageal Echocardiography?
Wilfred R. Lewis, M.D.; Pamela Gray, M.D.; Albert C. Perrino, M.D.
Anesthesiology, VA Connecticut Health Care System & Yale University School of Medicine, West Haven, Connecticut, United States
Introduction: By using partial rebreathing of carbon dioxide, the NICO monitor (Novametrix Medical Systems, Wallingford, CT) provides a fast, simple and non-invasive method of Fick cardiac output (CO) determination. Version 4.0 of this device uses a rebreathing period of 50 seconds and from measurements of changes in both CO2 elimination and end-tidal CO2, provides CO updates every 3 minutes. Non-invasive measurements of CO using a partial CO2 rebreathing technique is particularly promising for intraoperative application. To assess the intraoperative performance of the NICO, we designed a clinical evaluation study to compare CO measurements obtained by NICO to those obtained by transesophageal echocardiography(TEE).

Methods: After IRB approval and written informed consent, matched sets of CO measurements from the NICO and TEE (SONOS 5500, Hewlett Packard, Andover, MA) were collected in intubated, mechanically ventilated surgical patients. Matched sets of CO measurements from the two techniques were collected at specific time periods per protocol. Preselected periods included post-incision, aortic cross-clamp, reaming, inotropic infusion, etc. Fick CO determination was performed with the NICO "FAST" mode which avoids algorhythmic averaging of COs from sequential rebreathing cycles. TEE CO was determined using multiplane continuous wave Doppler measurements of ascending aortic flow (transgastric longitudinal view) and planimetry of 2-D aortic valve area (midesophageal shortaxis view). Doppler spectral data was recorded on SVHS tape and analyzed off-line by experienced echocardiographers who were blinded to the NICO data. TEE determinations were compared to the NICO CO using Bland-Altman analysis. Statistical comparison of means was performed using Student's T-test with a p<0.05 considered significant.

Results: A total of 113 matched CO measurements were obtained from 17 of the 21 enrolled patients. Data analysis was not performed on 2 patients due to alterations in ventilation, in 1 patient due to sub-optimal TEE imaging and in 1 patient due to corrupted CO2 signal. TEE cardiac outputs ranged from 2.8 to 10.5 (mean=5.6) L/m and NICO cardiac outputs ranged from 3.4 to 9.6 (mean=6.0) L/m. There was no statistical difference between the mean CO of the two techniques. Bland-Altman analysis resulted in a mean difference of 0.39 L/m (p=ns) and a standard deviation of the differences of 1.01 L/m (Figure).Discussion: These data show the NICO device produces little systematic error in CO determinations. The standard deviation of the differences of 1.0 L/m demonstrates that the two techniques are not always in agreement. However, clinical CO techniques have inherent variability with a 15% variation currently regarded as clinically acceptable.(Ref 1) Therefore, we conclude that the degree of agreement between NICO and TEE demonstrated in this study is within clinically acceptable limits.


1. Critchley LAH, Critchley JAJH. A meta-analysis of bias and precision statistics to compare cardiac output measurement techniques. J of Clin Monit 1999; 15:85-91.

Anesthesiology 2001; 95:A574
Figure 1