Abstract
OBJECTIVE: To evaluate whether changes in central venous pressure during fluid expansion and baseline cyclic respiratory variation in the central venous pressure amplitude (RespCVP) curve could be used to discriminate between fluid responders and nonresponders. METHODS: This prospective observational study included critically ill adult patients who underwent fluid expansion in the form of a fluid bolus or fluid challenge with crystalloids. All patients were under mechanical ventilation and adequately sedated. We determined the central venous pressure at baseline (CVPT0) and the changes at 5 (ΔCVPT5), 10 (ΔCVPT10) and 15 (ΔCVPT15) minutes during fluid infusion. We also measured the RespCVP at baseline. Fluid responsiveness was defined as a cardiac index increase of ≥ 15%. RESULTS: The study included 30 patients (11 responders and 19 nonresponders). The CVPT0 and the changes after a fluid challenge at all three time points did not adequately predict fluid responsiveness, as determined by their area under the curve values (CVPT0: 0.70, (95%CI: 0.49 - 0.90; ΔCVPT5: 0.78, (95%CI: 0.57 - 0.99; ΔCVPT10: 0.63, (95%CI: 0.39 - 0.88; ΔCVPT15: 0.68, ((95%CI: 0.45 - 0.92). The RespCVP at baseline also had a poor performance (area under the curve: 0.70; 95%CI: 0.50 - 0.91). CONCLUSION: Changes in central venous pressure have limited value in predicting fluid responsiveness.