Cardiac output measurement by thermodilution.

The accuracy of thermodilution for measuring cardiac output was studied by comparing this method with measurements obtained by an electromagnetic flowmeter in the dog. Thermal curves were computed by a cardiac output computer and absolute flows were obtained by pump calibration of the flowmeter. Regression analysis showed an excellent correlation (P less than 0.001) between thermodilution measurements with injectates at 0 degrees C and flowmeter measurements using either cannulating probes (n = 105, r = 0.98) or periaortic probes (n = 100, r = 0.96). With the use of normal room temperature injectates, a good correlation was also found with absolute flows (n = 75, r = 0.92, P less than 0.001). Thermodilution overestimated cardiac output by 3% to 9%. In 32 pairs of successive cardiac output determinations by thermodilution, there was no significant difference between the two measurements (P greater than 0.05). In eight patients cardiac output was measured postoperatively by thermodilution with injectates at 0 degrees C and 24 degrees C administered in rapid succession. Regression analysis of 50 pairs of measurements at the two temperatures showed a strong correlation between the two techniques (r = 0.96) and the two results were not significantly different (P = 0.00001).