A comparison of noninvasive cardiac output measurement by partial carbon dioxide rebreathing with the lithium dilution technique in anesthetized foals

Newer techniques for cardiac output (Q) determinations that are minimally invasive remain to be validated in neonatal foals against other accepted techniques such as the lithium technique (LiDCO). This study compares Q determinations using the partial CO₂ rebreathing technique (NICO) with LiDCO in anesthetized neonatal foals. Ten foals were instrumented for NICO and LiDCO determinations. For each foal low, intermediate and high levels of cardiac output were achieved in that order using an end‐tidal isoflurane (ETI) concentration of 1.3 – 2.1% for the lowest rate; an ETI of 0.85–1.4% and a constant‐rate infusion of dobutamine (1–3 ?g/kg/min) for the intermediate rate; and an ETI of 0.83–1% and dobutamine (2–6 ?g/kg/min) for the highest rate. Four foals also received IV intermittent doses (total cumulative dose of 1.1–1.7 mg) of phenylephrine at the highest rate of Q. The measurements were obtained in duplicate or triplicate for each Q technique after achieving a stable hemodynamic plane for at least 15 minutes at each rate of Q. For the lithium technique, all foals received 1.1–1.9 mL (0.16–0.28 mmol) of lithium. A Bland‐Altman analysis was used to compare the bias and precision of the two techniques. Eighty seven comparisons were determined between the two techniques. Eight were excluded due to more than 20% variation between the LiDCO determinations or technical errors at the time of determination. The correlation coefficient between the two methods was 0.67 for all Q determinations. Mean LiDCO and NICO values from 79 measurements were 130 ± 40 mL^–1 kg minute^–1 (range, 68– 237) and 152 ± 31 mL^–1 kg minute^–1 (89 – 209), respectively. The mean ( mL^–1 kg minute^–1) of the differences of LiDCO – NICO was = –0.7248 + 0.8602 NICO. The precision (1.96 SD) of the differences between LiDCO and NICO was 58.9 mL^–1 kg minute^–1 (–80.9–+36.9) with a mean difference of –22 mL^–1 kg minute^–1 (bias; 95% CI – 15.2 to ‐28.7). In conclusion, given the small bias compared to the limits of agreement, the NICO technique for determining Q deserves further consideration for adoption into clinical practice in neonatal foals.