The diagnostic accuracy of estimated continuous cardiac output compared with transthoracic echocardiography

Estimated continuous cardiac output (esCCO) is a new and noninvasive cardiac output (CO) monitoring device using pulse wave transit time. The aim of this study was to assess rapid changes in CO using esCCO (ΔCOesCCO) without invasive calibration and to compare the results with those using transthoracic Doppler echocardiography (ΔCOTTE).

Fifty-four consecutive patients were enrolled in this study following elective cardiac surgery. The COesCCO and COTTE were collected during four consecutive steps: 1) at baseline, 2) during passive leg raising (PLR), 3) at return to baseline, and 4) after a fluid challenge. The relationship between ΔCOesCCO and ΔCOTTE induced by PLR and a fluid challenge was assessed and a polar plot analysis was performed. Relationship, Bland-Altman analysis, and percentage error for absolute values of COesCCO and COTTE were also performed.

Twenty-four patients were excluded from the analysis. No correlation was found between ΔCOesCCO and ΔCOTTE during PLR (r = 0.07; P = 0.732; n = 30) and after a fluid challenge (r = 0.24; P = 0.394; n = 14). The polar plot analysis showed that 21 data points (87%) of significant changes in CO were above the 30° radial sector lines and confirmed that esCCO was unable to track changes in CO. A weak positive relationship was found between absolute values of COesCCO and COTTE (r = 0.28; P = 0.004). Bias, precision, and limits of agreement were 0.25 L·min(-1), 2.4 L·min(-1), and -4.4 to 4.9 L·min(-1), respectively. The percentage error was 80%.

Estimated continuous cardiac output without external calibration seems unable to assess rapid changes in CO following cardiac surgery and was not interchangeable with transthoracic Doppler echocardiography.