Lithium dilution, pulse power analysis, and continuous thermodilution cardiac output measurements compared with bolus thermodilution in anaesthetized ponies.

BACKGROUND: This study compares cardiac output (CO) measurements obtained by lithium dilution (LiDCO), pulse power analysis (PulseCO), and continuous thermodilution (CTD) with bolus thermodilution (BTD) in ponies. METHODS: Eight isoflurane-anaesthetized Shetland ponies received xylazine, ketamine, and midazolam infusions (0.3, 1.2, and 0.018 mg kg(-1) h(-1), respectively). CO was measured with BTD, CTD, LiDCO, and PulseCO. Lithium was injected into the jugular vein and blood was sampled from the facial artery for lithium detection and this artery was also used for PulseCO. Measurements were obtained during four stable haemodynamic conditions in the following order: isoflurane 1% (end-tidal concentration), isoflurane 2%, isoflurane 1%, and isoflurane 1%+dobutamine 5 µg kg(-1) min(-1). RESULTS: The bias (2 sd) was 2.5 (2.1) and 0.5 (2.9) litre min(-1) for LiDCO-BTD and for CTD-BTD comparisons, respectively. The limits of agreement were wider than ±30%; therefore, interchangeability was rejected for both comparisons. A possible error in LiDCO might explain the bias observed because CTD-BTD comparison showed less bias. Changes in PulseCO did not correlate with those of BTD and a weak correlation (r(2)=0.23; P=0.018) and concordance (Pc=0.42) was found between CTD and BTD. CONCLUSIONS: This is the first study to show a large bias for LiDCO-BTD comparison in animals receiving xylazine, ketamine, and midazolam infusions. The trending abilities of neither PulseCO nor CTD were reliable. Further studies are needed to elucidate possible influences of drugs on the accuracy of the LiDCOplus system.