Objective: The determination of basal cardiac output (CO) and of its variations during different therapeutic interventions liable to increase or decrease it in mechanically ventilated patients using transesophageal echocardiography (TEE).
Design: To compare CO measurements simultaneously obtained by transmitral single-plane TEE and thermodilution.
Setting: Medical intensive care unit.
Patients: Twenty-two consecutive mechanically ventilated patients hospitalized for various medical conditions were included.
Interventions: The comparisons between transmitral single-plane TEE and thermodilution measurements were made at baseline and after different therapeutic interventions affecting CO (fluids or dobutamine infusion or positive end-expiratory pressure titration).
Measurements: Seventy-four measurements were obtained. Cardiac output using TEE was the product of the mitral valve area, the time-velocity integral of flow at the same site and the heart rate.
Results: A significant correlation was observed between thermodilution and TEE measurements of CO (n = 74, r = 0.78, p < 0.001) despite wide limits of agreement (mean +/- 2SD = -0.3 +/- 3.1 l/min). Thermodilution and TEE CO determinations both had significant inverse correlation with the arterial-venous oxygen content difference in ten consecutive patients (r = 0.77, p < 0.01 and r = 0.71, p < 0.01, respectively). The correlation between variations of CO greater than 20% obtained by thermodilution and TEE was significant (r = 0.89, p < 0.001). The operative characteristics implied the ability of TEE to predict significant variations of thermodilution CO (sensitivity 85% and negative predictive values 86%). Moreover, arterial-venous oxygen content difference changes of 5% or more were better detected using TEE than thermodilution.
Conclusions: These results suggest that although transesophageal CO measurements cannot replace thermodilution ones, the determination of CO variations obtained using TEE may be useful in the management of critically ill mechanically ventilated patients. This technique may make it possible to monitor hemodynamics during initial therapeutic interventions in those patients in whom right heart catheterization cannot be performed immediately.