Does using two Doppler cardiac output monitors in tandem provide a reliable trend line of changes for validation studies?

The demise of the pulmonary artery catheter as a gold standard in cardiac output measurement has created the need for new standard. Doppler cardiac output can be measured suprasternally (USCOM) and via the oesophagus (CardioQ). Use in tandem they may provide a reliable trend line of cardiac output changes against which new technologies can be assessed. Data from three similar clinical studies was pooled. Simultaneous USCOM and CardioQ readings, 13 (7-27), were performed every 15-30 min intraoperatively. Within individual patient regression analysis was performed. Data was normalized, CardioQ against USCOM, to eliminate the systematic error component following calibration. Bland-Altman and trend, concordance and polar analysis, were performed on the grouped data. Cardiac output was indexed (CI) to BSA. Data from 53 patients, aged 59 (26-81) years, scheduled for major surgery were included. Within-individual mean (SD) CI was 3.4 (0.6) L min(-1) m(-2). Correlation was good to excellent in 83 % of cases, R(2) > 0.80, and reasonable in 96 %, R(2) > 0.60. Percentage error was 38 %, and decreased to 14 % with normalization. The estimated 95 % precision for a single Doppler reading was ±10 %. Concordance rate was 96.6 % (confidence intervals 94.7-99.5 %) and above the >92 % threshold for good trending ability. Polar analysis also confirmed good trending ability. The regression line between Doppler methods was offset with a slope of 0.9, thus CardioQ CI readings increased relative to USCOM. Both Doppler methods trended cardiac output reliably. Used in tandem they provide a new standard to assess cardiac output trending.