End-tidal carbon dioxide's change to fluid challenge versus internal jugular vein dispensability index for predicting fluid responsiveness in septic patients: A prospective, observational study.

Abstract

The prediction of fluid responsiveness is crucial for the fluid management of septic shock patients. This prospective, observational study was conducted to compare end-tidal carbon dioxide (ETCO2) change due to fluid challenge (FC-induced ΔETCO2) versus internal jugular vein distensibility index (IJVDI) as predictors of fluid responsiveness in such patients. Septic hypoperfused mechanically ventilated patients were classified as fluid responders (Rs) and non-responders (NRs) according to the improvement of left ventricular outflow tract-velocity time integral (ΔLVOT-VTI) after fluid challenge (FC). The receiver operating characteristic (ROC) curves of FC-induced ΔETCO2, pre-(FC) IJVDI and their combination for prediction of fluid responsiveness were compared to that of ΔLVOT-VTI% as a gold standard. Of 140 patients who completed the study, 51 (36.4%) patients were classified as Rs and 89 (63.6%) patients as NRs. With regard to the prediction of fluid responsiveness, no significant difference (P. 0. 384) was found between the diagnostic accuracy of FC-induced ΔETCO2 >2 mmHg (area under the ROC curve [AUC] 0.908, P < 0.001) and that of pre-(FC) IJVDI >18% (AUC 0.938, P < 0.001), but a prediction model combining both markers, ΔETCO2 ≥3 mmHg and IJVDI ≥16%, achieved significantly higher accuracy (AUC 0.982, P < 0.001) than each independent one (P < 0.05). Under stable ventilatory and metabolic conditions, the predictivity of FC-induced ΔETCO2 >2 mmHg can be comparable to that of pre-(FC) IJVDI >18%. A predictive model combining both FC-induced ΔETCO2 ≥3 mmHg and IJVDI ≥16% can provide higher accuracy than that recorded for each one independently.