Cardiac output determination from endotracheally measured impedance cardiography: clinical evaluation of endotracheal cardiac output monitor.

OBJECTIVES

To evaluate the accuracy, precision, and trending of a new endotracheally sourced impedance cardiography-based cardiac output (CO) monitor (ECOM; ConMed Corp, Irvine, CA).

SETTING

Two university hospitals.

PARTICIPANTS

Thirty patients scheduled for elective coronary artery bypass graft (CABG) surgery.

INTERVENTIONS

All patients received a pulmonary artery catheter (PAC), arterial catheter, endotracheal CO monitor (ECOM), endotracheal intubation, and transesophageal echocardiographic monitoring. ECOM CO was compared with CO measured with pulmonary artery thermodilution, and left ventricular CO measured with transesophageal echocardiography.

MEASUREMENTS

One hundred forty-five pairs of triplicate CO measurements using intermittent bolus pulmonary artery thermodilution (TD) and ECOM were compared at 5 distinct time points: postinduction, postinduction passive leg raise, poststernotomy, post-CABG completion, and post-chest closure. Eighty-seven pairs of triplicate CO measurements using transesophageal echocardiography were obtained at 3 time points: postinduction, post-CABG completion, and post-chest closure and compared with ECOM- and PA-derived CO measurements. The measurements at each time point were compared by using Bland-Altman and polar plot analyses.

RESULTS

The mean CO ranged from 2.16 to 9.41 L/min. ECOM CO, compared with TD CO, revealed a bias of 0.02 L/min, 95% limits of agreement of -2.26 to 2.30 L/min, and a percent error of 50%. ECOM CO showed trending with TD CO with 91% and 99% of values within 0.5L/min and 1 L/min limits of agreement, respectively. ECOM CO, compared with TEE CO, revealed a bias of -0.25 L/min, 95% limits of agreement of -2.41 to 1.92 L/min, and a percent error of 48%. ECOM CO showed trending with TEE CO with 83% and 95% of values within 0.5L/min and 1 L/min limits of agreement, respectively.

CONCLUSION

ECOM CO shows an acceptable bias with wide limits of agreement and a large percent error when compared with TD CO or TEE CO; however, it shows acceptable trending of CO to both modalities in patients undergoing cardiac surgery. Further studies are required to evaluate ECOM in other patient populations and clinical situations.