Preload dependence indices to titrate volume expansion during septic shock: a randomized controlled trial

Abstract

Introduction: In septic shock, pulse pressure or cardiac output variation during passive leg raising are preload dependence indices reliable at predicting fluid responsiveness. Therefore, they may help to identify those patients who need intravascular volume expansion, while avoiding unnecessary fluid administration in the other patients. However, whether their use improves septic shock prognosis remains unknown. The aim of this study was to assess the clinical benefits of using preload dependence indices to titrate intravascular fluids during septic shock.

Methods: In a single-center randomized controlled trial, 60 septic shock patients were allocated to preload dependence indices-guided (preload dependence group) or central venous pressure-guided (control group) intravascular volume expansion with 30 patients in each group. The primary end point was time to shock resolution, defined by vasopressor weaning.

Results: There was no significant difference in time to shock resolution between groups (median (interquartile range) 2.0 (1.2 to 3.1) versus 2.3 (1.4 to 5.6) days in control and preload dependence groups, respectively). The daily amount of fluids administered for intravascular volume expansion was higher in the control than in the preload dependence group (917 (639 to 1,511) versus 383 (211 to 604) mL, P = 0.01), and the same held true for red cell transfusions (178 (82 to 304) versus 103 (0 to 183) mL, P = 0.04). Physiologic variable values did not change over time between groups, except for plasma lactate (time over group interaction, P <0.01). Mortality was not significantly different between groups (23% in the preload dependence group versus 47% in the control group, P = 0.10). Intravascular volume expansion was lower in the preload dependence group for patients with lower simplified acute physiology score II (SAPS II), and the opposite was found for patients in the upper two SAPS II quartiles. The amount of intravascular volume expansion did not change across the quartiles of severity in the control group, but steadily increased with severity in the preload dependence group.

Conclusions: In patients with septic shock, titrating intravascular volume expansion with preload dependence indices did not change time to shock resolution, but resulted in less daily fluids intake, including red blood cells, without worsening patient outcome.

Trial registration: Clinicaltrials.gov NCT01972828. Registered 11 October 2013.

Figure 1

Treatment algorithm. ^*or hemoglobin <7 g.dL^-1; †Ht ≤30% or hemoglobin ≤10 g.dL^-1 in the first 6 hours following inclusion; ^‡ml.kg^-1 of predicted body weight. ACP, acute cor pulmonale; CI, cardiac index; CVP, central venous pressure; Ht, hematocrit; MAP, mean arterial pressure; MV, mechanical ventilation; PLR, passive leg raising test; PPV, pulse pressure variation; RBC, red blood cells; SR, sinus rhythm; SRM, spontaneous respiratory movements; ΔSV, stroke volume variation after fluid administration; V_T, tidal volume.

Figure 2

Study flow chart. ICU, intensive care unit.

Figure 3

Evolution of hemodynamic parameters over time. Symbols are mean parameter values over time (blue = control group, red = preload dependence group). Bars are standard deviation. CI, cardiac index; CVP, central venous pressure; MAP, mean arterial pressure; NS, non-statistically significant; ScvO₂, superior vena cava venous oxygen saturation.

Figure 4

Evolution of lactates over time (a) and lactate difference from inclusion (b) at each time point. Red symbols are mean parameter values over time. Black lines are individual parameter values over time.

Figure 5

Amount of fluids administered for intravascular volume expansion as a function of treatment arm and SAPS II. (a) Daily amount of fluids. (b) Amount of fluid administered from H0 to H12 after inclusion. In both groups, patients were classified into four categories of severity at inclusion according to quartiles of SAPS II score [20]. Bars are mean values and error bars standard deviation. NS, non-statistically significant; SAPS II, Simplified Acute Physiology Score II.

Figure 6

Kaplan-Meier plot of the probability of remaining under vasopressor therapy (a) and survival (b) from inclusion to day 28.