Merits and limitations of fluid balance

Nobody will argue with the observation that a positive fluid balance is associated with increased mortality rates (1-3). Indeed, this finding is not surprising, because patients who are more severely ill are more likely to develop edema whether or not they have renal failure. Therefore, studies investigating the impact of fluid balance on outcomes must use multivariable or propensity-matched analyses that include many variables in order to determine whether fluid balance is independently associated with outcome. This has been done in several recent studies. In a study conducted in our Department of Intensive Care in Brussels, fluid balance was more positive in non-surviving than in surviving patients with sepsis, and these factors remained related even after adjustment for many variables (2). The monocentric nature of the study may be seen as a weakness, but it is also a strength because it means that variability in patient management was limited. Larger observational studies conducted in Europe (3) and worldwide (1) have shown similar results. These two studies included many variables, enabling quite extensive multivariable analyses to be conducted. In the SOAP study (3), a positive fluid balance had an impressively high prognostic value, just below that of the SAPS II score. In the recent analysis of the ICON database, the effects of a positive fluid balance, again significantly associated with outcome, were particularly important after the first 24 hours (1).