Meta-analysis of the effects of normal saline on mortality in intensive care.

INTRODUCTION AND OBJECTIVE: To evaluate, by means of a meta-analysis, the effect of normal saline on mortality in intensive care patients, when compared with the use of balanced crystalloids.

MATERIAL AND METHOD: Published controlled clinical trials, randomised and sequential prospective studies in time, evaluating the mortality when physiological saline was used in patients admitted to intensive care units. Electronic search was performed in Medline, Embase, Cochrane Library, ISI Proceedings, and Web of Science, as well as a manual search of selected references. An independent evaluation was performed by 2 investigators. Discrepancies were resolved by consensus in the working group. Contingency tables were performed, and the OR with confidence intervals of each study were obtained. Heterogeneity was assessed by I2 . Publication bias was assessed using funnel plot and Egger test.

RESULTS: A total of 8 articles were selected for the meta-analysis of mortality, which included a total of 20,684 patients. A significant association was observed between the use of saline and mortality in intensive care patients (OR 1.0972; 95% CI 1.0049-1.1979), when compared to the use of balanced crystalloids. No statistical evidence of publication bias (Egger, P=.5349) was found. In the sensitivity analysis, none of the studies substantially modified the overall outcome if it was eliminated from the meta-analysis.

CONCLUSIONS: There may be an increase in mortality associated with the use of saline in patients admitted to intensive care when comparing with the use of balanced crystalloids.