Change in cardiac output during Trendelenburg maneuver is a reliable predictor of fluid responsiveness in patients with acute respiratory distress syndrome in the prone position under protective ventilation

Table 4 Diagnostic performance of five diagnostic tests to predict fluid responsiveness

Tests	Number of patients analyzed	AUC (CI_95%)	Optimal threshold	Gray zone of optimal threshold	Patients in gray zone, (number (%))	Sensitivity (CI_95%)	Specificity (CI_95%)	PLR	NLR
ΔCCI_TREND	33	0.90* (0.80–1.00)	8%	(5–12%)	10 (30%)	87% (67–100%)	89% (72–100%)	7.90	0.15
PPV_BASELINE-1	19	0.49 (0.21–0.77)	10%	(–Inf to Inf)	19 (100%)	33% (0–67%)	80% (50–100%)	1.65	0.84
PPV_VT8	19	0.52 (0.24–0.80)	9%	(–Inf to Inf)	19 (100%)	78% (44–100%)	40% (10–70%)	1.30	0.56
ΔPPV_6-8	19	0.59 (0.31–0.88)	29%	(17%–Inf)	16 (84%)	100% (100–100%)	40% (10–70%)	1.67	0
ΔCCI_EEO	33	0.65 (0.46–0.84)	10%	(−4% to 11%)	26 (79%)	33% (13–60%)	100 (100–100%)	Inf	0.67

AUC area under ROC curve. CI _95% 95% confidence interval, ΔCCI _EE change in continuous cardiac index during end-expiratory occlusion, ΔCCI _TREND change in continuous cardiac index during the Trendelenburg maneuver, ΔPPV _6-8 change in pulse pressure variation between ventilation with 6 and 8 ml.kg^-1 predicted body weight tidal volume, Inf infinity, NLR negative likelihood ratio, PLR positive likelihood ratio, PPV _BASELINE-1 pulse pressure variation at baseline-1, PPV _VT8 pulse pressure variation during ventilation with 8 ml.kg^-1 tidal volume
*p < 0.001 vs. an AUC of 0.5

ISSN: 1364-8535