Background: Central venous pressure (CVP) and intrathoracic blood volume index (ITBVI) were used to assess the fluid status. It has previously been shown that CVP is not as accurate as ITBVI for all the shock patients. We therefore hypothesized that the change of CVP has the ability to predict fluid responsiveness in some clinical cases of shock.
Methods: From September 1st 2009 to September 1st 2011, sixty-three patients with shock from different Intensive Care Unit (ICU) were collected into this retrospective study. All the patients received fluid challenge strategy (infusing 300 ml hydroxyethyl starch in 20 minutes), were monitored with CVP and pulse-indicated continuous cardiac output (PICCO). The correlation between changes in cardiac index (ΔCI), CVP (ΔCVP) and ITBVI (ΔITBVI) were analyzed. Fluid responsiveness was defined as an increase in CI ≥ 10%. Receiver operating characteristic (ROC) curves were generated for ΔCVP and ΔITBVI.
Results: For all the patients, there was no correlation between ΔCI and ΔCVP (P = 0.073), but in the subgroup analysis, the correlation between ΔCI and ΔCVP was significant in those younger than 60 years old (P = 0.018) and those with hypovolemic shock (P = 0.001). The difference of areas under the ROC curves of ΔCVP and ΔITBVI were not statistically significant in the group younger than 60 years old or hypovolemic shock group (P > 0.05, respectively). However, no similar results can be found in the group older than 60 years old and the other two shock type groups from ROC curves of ΔCVP and ΔITBVI.
Conclusions: ΔCVP is not suitable for evaluating the volume status of the shock patients with fluid resuscitation regardless of their condition. However, in some ways, ΔCVP have the ability to predict fluid responsiveness in the younger shock patients or in the hypovolemic shock patients.