Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares.

BACKGROUND Central venous pressure (CVP) is used almost universally to guide fluid therapy in hospitalized patients. Both historical and recent data suggest that this approach may be flawed. OBJECTIVE A systematic review of the literature to determine the following: (1) the relationship between CVP and blood volume, (2) the ability of CVP to predict fluid responsiveness, and (3) the ability of the change in CVP (DeltaCVP) to predict fluid responsiveness. DATA SOURCES MEDLINE, Embase, Cochrane Register of Controlled Trials, and citation review of relevant primary and review articles. STUDY SELECTION Reported clinical trials that evaluated either the relationship between CVP and blood volume or reported the associated between CVP/DeltaCVP and the change in stroke volume/cardiac index following a fluid challenge. From 213 articles screened, 24 studies met our inclusion criteria and were included for data extraction. The studies included human adult subjects, healthy control subjects, and ICU and operating room patients. DATA EXTRACTION Data were abstracted on study design, study size, study setting, patient population, correlation coefficient between CVP and blood volume, correlation coefficient (or receive operator characteristic [ROC]) between CVP/DeltaCVP and change in stroke index/cardiac index, percentage of patients who responded to a fluid challenge, and baseline CVP of the fluid responders and nonresponders. Metaanalytic techniques were used to pool data. DATA SYNTHESIS The 24 studies included 803 patients; 5 studies compared CVP with measured circulating blood volume, while 19 studies determined the relationship between CVP/DeltaCVP and change in cardiac performance following a fluid challenge. The pooled correlation coefficient between CVP and measured blood volume was 0.16 (95% confidence interval [CI], 0.03 to 0.28). Overall, 56+/-16% of the patients included in this review responded to a fluid challenge. The pooled correlation coefficient between baseline CVP and change in stroke index/cardiac index was 0.18 (95% CI, 0.08 to 0.28). The pooled area under the ROC curve was 0.56 (95% CI, 0.51 to 0.61). The pooled correlation between DeltaCVP and change in stroke index/cardiac index was 0.11 (95% CI, 0.015 to 0.21). Baseline CVP was 8.7+/-2.32 mm Hg [mean+/-SD] in the responders as compared to 9.7+/-2.2 mm Hg in nonresponders (not significant). CONCLUSIONS This systematic review demonstrated a very poor relationship between CVP and blood volume as well as the inability of CVP/DeltaCVP to predict the hemodynamic response to a fluid challenge. CVP should not be used to make clinical decisions regarding fluid management.