Characterization of Children with Septic Shock Cared for by Emergency Medical Services.

Objective

To inform the future development of a pediatric prehospital sepsis tool, we sought to 1) describe the characteristics, emergent care, and outcomes for children with septic shock who are transported by emergency medicine services (EMS) and compare them to those self-transported; and 2) determine the EMS capture rate of common sepsis screening parameters and the concordance between the parameters documented in the EMS record and in the emergency department (ED) record.

Methods

This is a retrospective cohort study of children ages 0 through 21 years who presented to a pediatric ED with septic shock between 11/2013 and 06/2016. Data, collected by electronic and manual chart review of EMS and ED records, included demographics, initial vital signs in both EMS and ED records, ED triage level, site of initial ED care, ED disposition, ED therapeutic interventions, outcomes, and times associated with processes. Potential screening parameters were dichotomized as normal vs. abnormal based on age-dependent normative data.

Results

Of the children with septic shock treated in our ED, 19.3% arrived via EMS. These children as compared to those self-transported were more likely (i.e., p < 0.05) to be male, have public insurance, receive initial care in the ED resuscitation suite, be hypotensive on arrival, receive their first ED fluid bolus sooner (33 vs. 58 minutes), receive vasoactive agents, be mechanically ventilated in the first 24 hours, and have slightly longer length of hospital stays. Both groups had similar times to antibiotics. While poor outcomes were rare, the 3- and 30-day mortalities were similar for both groups. EMS capture rates were highest for heart rate and respiratory rate and lowest for temperature, glucose, and blood pressure. Interrater reliability was highest for heart rate.

Conclusions

Children presenting to the ED with septic shock transported by EMS represent a critically ill subset of modest proportions. Realization of a sepsis screening tool for this vulnerable population will require both creation of a tool containing a limited subset of objective parameters along with processes to ensure capture.