Comparison of three haemodynamic monitoring methods in comatose post cardiac arrest patients.

Objectives. Haemodynamic monitoring during post arrest care is important to optimise treatment. We compared stroke volume measured by minimally-invasive monitoring devices with or without thermodilution calibration, and transthoracic echocardiography (TTE), and hypothesised that thermodilution calibration would give stroke volume index (SVI) more in agreement with TTE during targeted temperature management (TTM). Design. Comatose out-of-hospital cardiac arrest survivors receiving TTM (33 degrees C for 24 hrs) underwent haemodynamic monitoring with arterial pulse contour analyses with (PiCCO2 (R)) and without (FloTrac (R)/Vigileo (R) monitor (R)) transpulmonary thermodilution calibration. Haemodynamic parameters were collected simultaneously every fourth hour during TTM (hypothermia) and (normothermia). SVI was measured with TTE during hypothermia and normothermia. Bland-Altman analyses were used for determination of SVI bias (+/- 1SD). Results. Twenty-six patients were included, of whom 77% had initial shockable rhythm and 52% discharged with good outcome. SVI (bias +/- 2SD) between PiCCO (after thermodilution calibration) vs FloTrac/Vigileo, TTE vs FloTrac/Vigileo and TTE vs PiCCO were 1.4 (+/- 25.8), -1.9 (+/- 19.8), 0.06 (+/- 18.5) ml/m2 during hypothermia and 9.7 (+/- 23.9), 1.0 (+/- 17.4), -7.2 (+/- 12.8) ml/m2 during normothermia. Continuous SVI measurements between PiCCO and FloTrac/Vigileo during hypothermia at reduced SVI (<35 ml/m2) revealed low bias and relatively narrow limits of agreement (0.5 +/- 10.2 ml/m2). Conclusion. We found low bias, but relatively wide limits of agreement in SV with PiCCO, FloTrac/Vigileo and TTE during TTM treatment. The methods are not interchangeable. Precision was not improved by transpulmonary thermodilution calibration during hypothermia.