Determination of cardiac output by thermodilution during hypothermia.

The thermodilution method for estimating cardiac output was compared with the electromagnetic flowmeter technique in 10 mongrel dogs at normothermia and during surface-induced deep hypothermia. Thermodilution curves obtained during cooling or rewarming must be corrected for the baseline drift caused by changing core temperature. At normothermia, the correlation coefficient between the two methods was 0.96 and the reproducibility of the thermodilution technique was 5 percent. Comparable correlation was present during hypothermia. Curves corrected for baseline drift resulted in significantly different output values from those derived from uncorrected curves (p < 0.05). The thermodilution method is valid at low body temperatures. Clinical confirmation of these results, particularly during open heart surgery in infants, is warranted.