Risk of Prolonged Intensive Care Length of Stay Relative to Extent of Fluid Resuscitation and Time to Vasopressor Initiation in Pediatric Sepsis

Abstract

Purpose: Septic shock continues to be a leading causes of pediatric mortality worldwide. Currently there is only limited literature detailing what constellation of signs and symptoms may predict children with septic shock. In particular, there is scant literature on whether excessive fluid resuscitation (> 60 mL/kg) following a diagnosis of hypotension or if delays in vasopressor initiation following fluid-refractory septic shock lead to worse outcomes. I hypothesize that in children in the ED with septic shock, greater amounts of fluids and the time to vasopressor initiation after last fluid bolus is associated with worse outcomes. Methods: This project is part of a larger retrospective, time series study describing baseline data for quality of care metrics in the diagnosis and management of pediatric septic shock. Children 0-18 years of age were eligible for the study if they presented to the ED from November 2012 to December 2014 with signs and symptoms concerning for sepsis and had hypotension for age during their ED course. Predictors of outcomes were total volume of fluid in milliliters per kilogram received in the ED following hypotension and time to vasopressor initiation following determination of time of fluid-refractory shock. Outcomes of mortality, total hospital length of stay, intensive care length of stay, and duration of vasopressors were assessed. Outcomes were controlled for potential confounders of age and severity of illness using the Pediatric Risk of Mortality score III. Results: 408 patients underwent analysis. 220 (53.9%) were males (95% Confidence Interval: 49.0%-58.7%). Fluid administration of more than 60 mL/kg in the ED was not associated with worsened mortality, intensive care or hospital length of stay, or longer vasopressor requirement. After controlling for severity of illness and age, the intensive care length of stay and duration of vasopressor support was statistically longer (p value = .045 and 0.008, respectively). Specifically, for every minute of vasopressor delay, intensive care length of stay and duration of time spent on vasopressor support increased by 19 and 13 minutes, respectively (p<.05). Conclusions: Delays in vasopressor initiation from fluid refractory shock is associated with an increased intensive care length of stay and longer time on vasopressors. These results call for the exploration of interventions that can improve timeliness of vasopressor initiation.