Arterial pressure-based cardiac output in septic patients: different accuracy of pulse contour and uncalibrated pressure waveform devices

Table 3 Summary of the comparisons between the different techniques used for measuring cardiac index

	Changes induced by volume expansion (Group 1)	Changes induced by introduction/increase in norepinephrine (Group 2)
CIpc versus CItd
Bland-Altman analysis for the changes in absolute value	Bias, -0.20 ± 0.63 L/min/m²	Bias, -0.05 ± 0.74 L/min/m²
Linear regression for the changes in percentage	r = 0.72 (P < 0.05)	r = 0.78 (P < 0.05)
Ability of CIpc to detect an increase in CItd ≥ 15%	Cut-off, CIpc increase ≥ 12%	Cut-off, CIpc increase ≥ 15%
	Specificity, 74 (49 to 91)%	Specificity, 93 (68 to 99)%
	Specificity, 95 (76 to 99)%	Specificity, 88 (69 to 97)%
	Area under the ROC curve, 0.878 (0.736 to 0.960)	Area under the ROC curve, 0.924 (0.795 to 0.983)
CIpw versus CItd
Bland-Altman analysis for the changes in absolute value	Bias, -0.23 ± 0.95 L/min/m²	Bias, -0.01 ± 1.75 L/min/m²
Linear regression for the changes in percentage	r = 0.33 (P < 0.05)	r = -0.03 (P = 0.65)
Ability of CIpw to detect an increase in CItd ≥ 15%	Cut-off, CIpw increase ≥8%	Cut-off, CIpw increase ≥34%
	Sensitivity, 56 (33 to 80)%	Sensitivity, 27 (8 to 55)%
	Specificity, 71 (48 to 89)%	Specificity, 96 (80 to 99)%
	Area under the ROC curve, 0.564 (0.398 to 0.720)^a	Area under the ROC curve, 0.541 (0.377 to 0.700)^a

n = 40 in Group 1, and n = 40 in Group 2.
^aP < 0.05 vs. CIpc.
CItd, cardiac index measured by transpulmonary thermodilution; CIpc, pulse contour-based cardiac index measured with the PiCCO device; CIpw, arterial pressure waveform-based cardiac index measured with the Vigileo device; AUC, area under the curve; ROC, receiving operating characteristics.

ISSN: 1364-8535