Effects of the time response of the temperature sensor on thermodilution measurements

Thermodilution is widely used to measure cardiac output, ejection fraction and end diastolic volume. Even though the method is based on dynamic temperature measurements, little attention has been paid to the characterization of the dynamic behavior of the temperature sensor and to its influence on the accuracy of the method. This paper presents several theoretical and empirical results related to the thermodilution method. The results show that, at flow velocities above 0.2 m s(-1), the response of temperature sensors embedded in Swan-Ganz catheters can be accurately described by a convolution operation between the true temperature of the blood and the impulse response of the sensor. The model developed is used to assess the influence of the probe response on the measurement of cardiac output, and this study leads us to the conclusion that the probe response can cause errors in the cardiac output measurement, but this error is usually small (2% in cases with a high degree of arrhythmia). The results show that these small errors appear during arrhythmias that affect the R-R interval and when the real temperature distribution at the pulmonary artery does not possess a shape with perfect temperature plateaux.