The PulseCO pressure waveform algorithm is used for both LiDCO™plus and LiDCOrapid hemodynamic monitors to estimate cardiac output (CO). The accuracy of this CO estimate is achieved by comparing (calibrating) the PulseCO with a known CO, for example from the lithium dilution or LiDCO which scales the software algorithm and reduces the bias to the actual CO . This calibration/scaling factor (CF) is patient specific and should not significantly change over a period of a few days as it reflects the patient's aortic capacitance.
LiDCO™plus hemodynamic data files were obtained retrospectively from a university hospital medical/surgical ICU. Files were screened to obtain records with at least 24 hours of continuous hemodynamic data and at least one calibration at 24 hours following the initial set-up calibration. Data were analysed for change in CF from initial calibration over 24 and up to 48 hours. Relative change in CF was analyzed for correlation with changes in SVR and CO.
Twenty-one records contained at least 24 hours of data with a calibration at 24 hours. Eleven contained at least 48 hours of data and a further calibration at 48 hours. The 32 calibrations were reviewed for consistency of PulseCO and LiDCO results. The average variation in CF over 24 hours (7.4 ± 6.9% (mean ± SD)) and 48 hours (6.5 ± 6.2%) were similar (P = 0.362). Average changes in SVR were 24.7 ± 27.4% (range: 0 to 130%); average CO changed by 21.4 ± 17% (range: 0 to 57%) over 24 hours. There was no correlation between the variation in CF and changes in either CO or SVR.
It is known that the estimated coefficient of variation of a single measurement of LiDCO is 8%  and PulseCO is 2.4% . An orthogonally combined CV, reflecting the effects on CF, is estimated at 8.5%, or precision of 17%. Ninety-one per cent (29/32 ) of the observed CF changes were less than 17% of the initial CF. The PulseCO CF remains constant despite changes in SVR of up to 130% and CO of up to 57% over periods of 24 to 48 hours.