Fluid Boluses Possibly Harmful in Some Settings

Abstract

Source: Maitland K, Kiguli S, Poka RO, et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. Epub 2011 May 26; doi: 10.1056/NEJMoa1101549The Fluid Expansion as Supportive Therapy (FEAST) group of investigators from Africa and the United Kingdom evaluated the effects of rapid infusion normal saline or albumin compared with intravenous maintenance fluids in children with a severe febrile illness and impaired consciousness and/or respiratory distress accompanied by impaired perfusion. This clinical trial in Uganda, Kenya, and Tanzania enrolled more than 3,000 children aged 60 days to 12 years from January 2009 to January 2011. Stratum A included children without severe hypotension who had one or more of: capillary refill at least three seconds; lower extremity temperature gradient; weak radial pulse; or severe tachycardia. Stratum B included children with impaired perfusions and severe hypotension. Children with severe malnutrition, gastroenteritis, or noninfectious causes of shock were excluded. Stratum A children were randomized to receive 20 ml/kg normal saline, 20 ml/kg albumin, or no boluses (controls). Stratum B children were randomized to receive either normal saline or albumin bolus. Children received intravenous maintenance fluids, antibiotics, antimalarials, anticonvulsants, treatment for hypoglycemia, and transfusions for hemoglobin less than 5 g/dL. The primary end point was mortality at 48 hours. Secondary end points included hypotensive shock within 48 hours, neurologic sequelae at 4 or 24 weeks, and mortality at 4 weeks.Stratum A randomized 3,141 and Stratum B randomized 29 children. Baseline characteristics for all groups were similar: median age 24 months; 62% with prostration; 15% comatose; 83% in respiratory distress; 51% with moderate to severe acidosis; malaria in 57%; and 4% positive for HIV. Children in the two treatment arms received four times more fluid than did control children (40 ml/kg vs 10 ml/kg). Stratum A mortality within the first 48 hours was 10.5% receiving saline bolus, 10.6% receiving albumin, and 7.3% receiving maintenance fluids, with a relative risk of death of 1.45 for any bolus as compared to controls (95% CI, 1.13–1.86; P=.003). Mortality at four weeks was 12.0%, 12.2%, and 8.7% in the three groups, respectively (P=.004 for bolus vs control). Stratum B mortality was 56% in the saline group and 69% in the albumin group (P=.17). Prior to achieving the enrollment goal of 3,600, the study was stopped due to higher mortality among children receiving bolus fluids.The investigators conclude that fluid boluses significantly increased 48-hour mortality in critically ill children with impaired perfusion in these resource-limited settings.Drs Berg and Bratton have disclosed no financial relationship relevant to this commentary. This commentary does not contain a discussion of an unapproved/investigative use of a commercial product/device.This study calls into question the rapid administration of a fluid bolus as an important determinant of positive resuscitation outcome. 1 In resource-limited areas where intensive care facilities are lacking, WHO guidelines reserve fluid resuscitation for advanced shock. Hence, most children in sub-Saharan Africa receive no more than maintenance fluids along with blood transfusions for severe anemia.2There are unique factors in the study population and setting that deserve consideration before application to a wider population, including limited equipment, monitoring capability, and treatments as compared to a modern critical care unit. The value of the intensive care unit environment should not be underestimated.3 It allows the careful titration of fluid, vasopressors to optimize hemodynamics, and mechanical ventilation for pulmonary edema and respiratory failure. Fluid bolus administration in this study was done over one hour, much slower than recommended for sepsis in the critical care literature (up to 60 ml/kg in 15 minutes). Only 7% to 9% of subjects were dehydrated, an unusual finding in pediatric shock patients in resource-rich areas.One important potential confounder can be found in the Supplement Appendix Table 3a accompanying this article (nejmoa1101549_appendix.pdf). In the first hour, 20% of the no-bolus group received a blood transfusion, versus 4% in the saline and 2% in the albumin group. This may have skewed the results to show the beneficial effect of a blood transfusion in the control group, rather than a negative effect of bolus fluids in the intervention groups. The harmful effect of a saline bolus versus no-bolus was statistically significant in patients with a hemoglobin less than 5 gm/dL but not in those with a higher hemoglobin. Could this be because the severely anemic patients needed blood, not a fluid bolus? The conclusions, while perhaps applicable to patients in sub-Saharan Africa, should not be applied to wider populations or locations without further investigation. The results of a single study performed in a unique environment do not disprove the beneficial effects of bolus fluid resuscitation on mortality4 and hospital length of stay when advanced critical care resources are available.5,6