Continuous versus intermittent thermodilution for cardiac output measurement during alveolar recruitment manoeuvres in sheep

Abstract

ObjectiveTo evaluate interchangeability of a thermodilution based STAT mode continuous cardiac output (CCO) measurement method with bolus thermodilution (BTD). Study designRandomized crossover study. AnimalsTen 9 month old healthy male sheep. MethodsEach sheep was anaesthetized twice for laparoscopy. On one occasion mechanical ventilation was used immediately after anaesthetic induction (IPPV treatment) and on the other occasion the start of IPPV was delayed and two periods of alveolar recruitment manoeuvres were also performed (RM treatment). Cardiac output (CO) was measured simultaneously with both CCO and BTD at 6 time points. Data were analysed using difference versus mean plots. A priori limits of acceptance were set at ±30% of the mean of every paired measurement. If <5% of the data fell outside of these limits (Chi-square test, p < 0.05) the interchangeability of methods was accepted. Proportions of data outside of these limits were also compared between treatments (Fisher's test, p < 0.05). Cardiac output data from each treatment and measurement method were also analyzed separately with one-factorial anova and Bonferroni test (p < 0.05). ResultsA total of 119 measurements were obtained. Cardiac output ranged from 1.9 to 10.4 L minute−1 (CCO) and from 1.1 to 9.8 L minute−1 (BTD). The bias and limits of agreement were 0.5 ± 1.9 L minute−1. More than 5% of all data fell outside of the limits of acceptance (24/119), and a larger proportion fell outside of these limits in the RM (20/59) compared to the IPPV treatment (4/60). The Bonferroni test detected significant decreases of CO over time in both treatments when measured with BTD but not with CCO. Conclusions and clinical relevanceThe STAT mode CCO method is not interchangeable with BTD during acute haemodynamic changes caused by recruitment manoeuvres, thus the results of STAT mode CCO should be interpreted with caution because decreases in CO may not be detected.