Fluid Resuscitation in Sepsis: “Get the Balance Right”*
In this issue of Critical Care Medicine, Sakr et al (6) present the very interesting results of their planned substudy of an observational multinational prospective audit, the so called “Intensive Care Over Nations database” (7). The authors concluded that a “higher cumulative fluid balance at day 3 but not in the first 24 hours following ICU admission was independently associated with an increase in the hazard of death.” At first sight, these findings support the current approach to stabilize the patient with “aggressive” fluid resuscitation initially and then be restrictive as soon as possible. However, this conclusion is put in perspective by a closer look at the data: Fluid input on day 1 with less than 3.5 L was relatively low suggesting that hemodynamic stabilization took already place before ICU admission. The authors attributed this issue to an increased awareness for sepsis, a circumstance that has also been discussed as a potential reason for the failure of “early goal-directed therapy” as proclaimed by Rivers et al (8) in the recent randomized, controlled trials (9–11). Nevertheless, the relevance of a negative fluid balance within the first three ICU days for the patients’ outcome is reinforced by the present study.
The second major finding is that the reduced fluid balance in survivors was exclusively caused by higher fluid outputs, whereas there was no difference in fluid input between survivors and nonsurvivors. This discovery raises (at least) two questions: what are the reasons for the reduced fluid output (summarizing diuresis, extracorporeal fluid elimination, and drainage fluid in the present study) and how does this information influence clinical practice? With regard to the first question, the authors tried to adjust for differences in renal function by including Sequential Organ Failure Assessment renal subscores in the multivariable analysis. Trusting this valid statistical approach, there must have been additional factors contributing to the reduced fluid output in nonsurvivors such as insufficient perfusion pressures and/or a lack of intravascular volume. Based on the observational design and the high number of participating centers worldwide, the applied strategies and goal variables for hemodynamic therapy probably differed substantially throughout the study. Unfortunately, the authors did not provide information about differences between survivors and nonsurvivors in respect to vasopressor support and hemodynamic parameters.