Fluid resuscitation in severe sepsis and septic shock: systematic description of fluids used in randomized trials.

INTRODUCTION  Fluid therapy is one of the cornerstones of initial management of sepsis, but the choice of fluids used for resuscitation is controversial.  OBJECTIVES  While trying to determine the effects of alternative fluids used in sepsis resuscitation randomized controlled trials (RCTs), we found that the precise description of those fluids was frequently not available. This report presents the result of our efforts to provide the characteristics of those fluids to both researchers and clinicians. METHODS  We searched the following electronic databases: CENTRAL, MEDLINE, EMBASE, CINAHL, and ACPJC, and examined the reference lists of recently published meta‑analyses of fluid therapies in critically ill patients. These databases were searched from inception until August 2013. The data abstraction stage included determination of fluid composition, pH, chloride concentration, and presence or absence of buffers. We relied on the original articles as well as on manufacturers' websites, contact with authors, and contact with experts in the field.  RESULTS  Our original search yielded 7002 articles. In consecutive stages, we reduced it to 20. The types of fluids varied widely, including chloride content (110-154 mmol/l) and presence or absence of buffering substances in colloid solutions. Those characteristics were frequently not presented and rarely emphasized in the original articles.  CONCLUSIONS  The basic characteristics of fluids used in fluid therapy trials are often not easily available, yet of increasingly recognized clinical importance. We provide the information concerning composition of fluids used in RCTs, which will be useful not only to future investigators and systematic reviewers but also to clinicians using those fluids in regular clinical practice.