Influence of xylazine on the function of the LiDCO sensor in isoflurane anaesthetized horses.

Objective: Previous studies showed an influence of xylazine on the LiDCO sensor in vitro and in standing horses, but did not prove that this interaction caused error in LiDCO measurements. Therefore, agreement of cardiac output (CO) measurements by LiDCO and bolus-thermodilution (BTD) was determined in horses receiving xylazine infusions.

Study Design: Prospective, experimental study.

Animals: Eight Warmblood horses.

Methods: All horses were premedicated with xylazine. Anaesthesia was induced with midazolam and ketamine and was maintained with isoflurane in oxygen. During six hours of anaesthesia CO measurements and blood samples were taken before, during and after a 60 minute period of xylazine infusion. Pairs of LiDCO and bolus thermo-dilution (BTD) measurements of CO were performed. Sensor voltages exposed to blood and saline were measured before, during and after xylazine infusion and compared using Bland-Altman method of agreement with corrections for repeated measures.

Results: The CO values (mean ± SD) before xylazine were 34.8 ± 7.3 and 36.4 ± 8.1 L minute(-1) for BTD and LiDCO, respectively. After starting the xylazine infusion, the CO values for BTD decreased to 27.5 ± 6.1 L minute(-1) whereas CO values measured by LiDCO increased to 54.7 ± 18.4 L minute(-1) . One hour after discontinuing xylazine infusion, CO values were 33 ± 6.7 and 36.5 ±11.9 L minute(-1) for BTD and LiDCO, respectively. The difference between saline and blood exposed sensor voltages decreased during xylazine infusion and these differences were positive numbers before but negative during the infusion. There were correlations between xylazine plasma concentrations, CO differences and sensor voltage differences (saline - blood).

Conclusions And Clinical Relevance: This study proved that xylazine infusion caused concentration dependent bias in LiDCO measurements leading to an overestimation of readings. Sensor voltage differences (saline - blood) may become valuable clinical tool to predict drug-sensor interactions.