Uncalibrated arterial pressure waveform analysis for cardiac output monitoring is biased by low peripheral resistance in patients with intracranial haemorrhage.

BACKGROUND: Cardiac output (CO) monitoring by uncalibrated arterial pressure waveform analysis (APCO) using the FloTrac/Vigileo™ is feasible in patients with intracranial haemorrhage, but the results of validation studies are contradictory. The aim of the present study was to analyse the clinical agreement between the intermittent bolus thermodilution technique (TDCO) and APCO in patients with non-traumatic intracranial haemorrhage.
METHODS: This was a prospective observational clinical study in a university level intensive care unit. We studied patients who underwent CO monitoring according to clinical indications using TDCO. Simultaneously, APCO was applied using the radial arterial pressure curve. The difference in CO values measured by APCO with a mid-chest calibration level was compared with a calibration level at the angle of the eye.
RESULTS: A total of 407 data pairs from 16 patients were obtained. The mean CO(TDCO) was 7.6 litre min(-1) and CO(APCO) was 6.0 litre min(-1), with a bias corrected for repeated measures of 1.5 litre min(-1) and 95% limits of agreement of -2.4 to 5.4 litre min(-1). The percentage error was 58%. The increasing bias correlated with low peripheral resistance (ρ=-0.53, P=0.036). The calibration level at the patient's eye angle did not affect CO values (median bias 0 litre min(-1) with 25th-75th percentile -0.1 to 0.2 litre min(-1)).
CONCLUSIONS: The second generation of FloTrac(®)/Vigileo(®) monitoring system underestimates the TDCO in patients with non-traumatic intracranial haemorrhage. The bias correlates with measured systemic vascular resistance. The upper calibration level does not affect the results.