Cardiac output estimation by pulse wave analysis using the pressure recording analytical method and intermittent pulmonary artery thermodilution: A method comparison study in patients after off-pump coronary artery bypass surgery

BACKGROUND: Invasive pulse wave analysis is used in peri-operative settings to estimate cardiac output (CO). The 'pressure recording analytical method' (PRAM) implemented in the MostCareUp CO monitor is an invasive pulse wave analysis method using high-frequency sampling and analysis of the pulse wave to directly estimate the arterial impedance as a key variable of the proprietary CO estimation algorithm.

OBJECTIVE: To compare CO estimated by PRAM (PRAM-CO; test method) with CO measured by pulmonary artery thermodilution (PATD-CO; reference method).

DESIGN: Prospective observational method comparison study. PRAM-CO and PATD-CO were assessed simultaneously at five time points with at least 20 min between measurements. Arterial pressure waveforms were carefully checked for damping artefacts and a proprietary electronic filter of the MostCareUp CO monitor was used to optimise waveform quality.

SETTING: ICU of a German university hospital from August 2018 until April 2019.

PATIENTS: We included adult patients admitted to the ICU after elective off-pump coronary artery bypass surgery who were monitored with a radial arterial catheter and a pulmonary artery catheter. Patients with severe heart valve insufficiency or persistent arrhythmia were excluded.

MAIN OUTCOME MEASURES AND ANALYSIS: PATD-CO and PRAM-CO were compared using Bland-Altman analysis accounting for repeated measurements, the percentage error and trending analysis (four-quadrant plot, concordance rate).

RESULTS: We analysed 195 paired CO values of 41 patients. Mean PATD-CO and PRAM-CO were 4.99 ± 1.02 and 4.92 ± 1.05 l min, respectively. PATD-CO and PRAM-CO ranged from 3.04 to 8.74 and 2.79 to 8.01 l min, respectively. The mean of the differences between PATD-CO and PRAM-CO was -0.08 ± 0.74 l min with 95% limits of agreement of -1.55 to +1.40 l min. The percentage error was 29.8%. The concordance rate in four-quadrant plot analysis was 92%.

CONCLUSION: Using the system's electronic waveform filter PRAM-CO shows good agreement and trending ability compared with PATD-CO in adults after off-pump coronary artery bypass surgery.