Agreement in hemodynamic monitoring during orthotopic liver transplantation: a comparison of FloTrac/Vigileo at two monitoring sites with pulmonary artery catheter thermodilution.

Abstract

To study agreement in cardiac index (CI), systemic vascular resistance index (Systemic VRI) and stroke volume variation (SV variation) between the FloTrac/Vigileo at radial and femoral arterial cannulation sites, and pulmonary artery catheter (PAC) thermodilution, in patients undergoing orthotopic liver transplantation. A prospective observational study of 25 adult patients with liver failure. Radial and femoral arteries were cannulated with standardised FloTrac/Vigileo arterial transducer kits and a PAC was inserted. CI, SV variation and Systemic VRI were measured four times (30min after induction of anesthesia, 30min after portal vein clamping, 30min after graft reperfusion, 30min after commencement of bile duct anastomosis). The bias, precision, limits of agreement (LOA) and percentage errors were calculated using Bland-Altman statistics to compare measurements from radial and femoral arterial cannulation sites and PAC thermodilution. Neither radial nor femoral CI achieved acceptable agreement with PAC CI [radial to PAC bias (SD) 1.17 (1.49)L/min/m2, percentage error 64.40%], [femoral to PAC bias (SD) -0.71 (1.81)L/min/m2, percentage error 74.20%]. Agreement between radial and femoral sites for CI [mean difference (SD) -0.43 (1.51)L/min/m2, percentage error 70.40%] and Systemic VRI [mean difference (SD) 0.03 (4.17) LOA ±8.17mmHgminm2/L] were also unacceptable. Agreement in SV variation between radial and femoral measurement sites approached a clinically acceptable threshold [mean difference (SD) 0.68 (2.44)%), LOA ±4.78%]. FloTrac/Vigileo CI cannot substitute for PAC thermodilution CI, regardless of measurement site. SV variation measurements may be interchangeable between radial and femoral sites for determining fluid responsiveness.