Changes in central venous-to-arterial PCO2 difference and central venous oxygen saturation as markers to define fluid responsiveness in critically ill patients: a pot-hoc analysis of a multi-center prospective study

Abstract

The main aim of the study whether changes in central venous-to-arterial CO2 difference (ΔP(v–a)CO2) and central venous oxygen saturation (ΔScvO2) induced by volume expansion (VE) are reliable parameters to define fluid responsiveness (FR) in sedated and mechanically ventilated septic patients. We also sought to determine whether the degree of FR was related to baseline ScvO2 and P(v–a)CO2 levels. This was a post-hoc analysis of a multicenter prospective study. We included 205 mechanically ventilated patients with acute circulatory failure. Cardiac index (CI), P(v–a)CO2, ScvO2, and other hemodynamic variables were measured before and after VE. A VE-induced increase in CI > 15% defined fluid responders. Areas under the receiver operating characteristic curves (AUCs) and the gray zones were determined for ΔP(v–a)CO2 and ΔScvO2. One hundred fifteen patients (56.1%) were classified as fluid responders. The AUCs for ΔP(v–a)CO2 and ΔScvO2 to define FR were 0.831 (95% CI 0.772–0.880) (p < 0.001) and 0.801 (95% CI 0.739–0.853) (p < 0.001), respectively. ΔP(v–a)CO2 ≤ 2.1 mmHg and ΔScvO2 ≥ 3.4% after VE allowed the categorization between responders and non-responders with positive predictive values of 90% and 86% and negative predictive values of 58% and 64%, respectively. The gray zones for ΔP(v–a)CO2 (− 2 to 0 mmHg) and ΔScvO2 (− 1 to 5%) included 22% and 40.5% of patients, respectively. ΔP(v–a)CO2 and ΔScvO2 were independently associated with FR in multivariable analysis. No significant relationships were found between pre-infusion ScvO2 and P(v–a)CO2 levels and FR. In mechanically critically ill patients, ΔP(v–a)CO2 and ΔScvO2 are reliable parameters to define FR and can be used in the absence of CI measurement. The response to VE was independent of baseline ScvO2 and P(v–a)CO2 levels. Clinical trial registration The study was registered in the ClinicalTrials.gov registry: NCT03225378, date: July 20, 2017.