LOS ANGELES – Fluid administration of at least 1 L did not increase the incidence of acute respiratory or heart failure in severe sepsis, and actually seemed to decrease the need for dialysis in of 164 patients at Scott and White Memorial Hospital in Temple, Tex.
For every 1 mL of fluid administered per kilogram of body weight, the likelihood of dialysis decreased by 8.5% (odds ratio, 0.915; 95% confidence interval, 0.854-0.980; P = .0111), with no increase in heart or respiratory failure on univariate analysis. The 126 patients (77%) who received at least 1 L had a 68% reduction in the need for dialysis (OR, 0.32; CI, 0.117-0.890; P = .0288).
“The No. 1 reason we weren’t meeting benchmarks was fluid administration,” explained lead investigator Aruna Jahoor, MD, a pulmonary critical care and sleep medicine fellow at Texas Tech University Health Sciences Center.
Seventeen percent of patients received greater than or equal to 30 mL/kg of fluid resuscitation, while 28% received greater than or equal to 20 mL/kg of intravenous fluid resuscitation. It turned out that staff in the emergency department - where most of the patients were treated in the critical first 6 hours - were concerned about fluid overload and throwing patients into respiratory, heart, or renal failure, Dr. Jahoor said. The team didn’t find a difference in mortality when patients received 30 mL/kg - just over 2 L in a 70-kg patient - versus 20 mL/kg or 1 L. The patients’ in-hospital mortality rate and 28-day mortality rate were 27%, and 32%, respectively.
There also weren’t increased rates of heart failure, acute respiratory failure, or mechanical ventilation when patients received at least 1 L of fluid. “There were [also] lower rates of dialysis, which indicated that we weren’t overloading patients. Even when we looked at fluid as a continuous variable, we still didn’t see” complications, Dr. Jahoor said.
The findings should be reassuring to treating physicians. “When you have pushback against 30-mL/kg administration, you can say ‘well, at least let’s give a liter. You don’t have to worry as much about some of the complications you are citing,’ ” she said.
For very obese patients, “it can get a little uncomfortable to be given” enough fluid to meet the 30-mL/kg goal, “but you can give at least a liter” without having to worry too much, she said. The patients in the study were treated from 2010 to 2013; normal saline was the most common resuscitation fluid. The hospital has since added the 30-mL/kg fluid resuscitation to its sepsis admission orders, and compliance has increased significantly.
A multivariate analysis is in the works to control for confounders. “We will probably [still] see you are not having increased rates of congestive heart or respiratory failure, or needing dialysis,” Dr. Jahoor said. The protective effect against dialysis might drop out, “but I am hoping it doesn’t,” he said.
The investigators had no relevant financial disclosures.