Introduction: Vasoactives like epinephrine (Epi) and norepinephrine (NE) are a mainstay of supportive therapy for children with septic shock, however there is no consensus about which of these should be used as first-line. The objective of this study was to compare the outcomes associated with the choice of first-line Epi vs. NE in a multicenter observational dataset of children with septic shock using a target trail emulation design. Methods: We analyzed encounters of children with septic shock based on the Phoenix criteria who were admitted to one of 13 U.S. pediatric ICUs between 2012-2018 and who were treated with either Epi or NE as first-line vasoactive within 24 hours of admission. Using a target trial emulation, we compared Epi vs. NE on all-comers with septic shock and in secondary analysis, we compare these treatments in patients with acute hypoxemic respiratory failure (AHRF), defined as an oxygenation index ≥4 or an oxygen saturation index ≥5. Outcomes included in-hospital mortality, multiple organ dysfunction syndrome (MODS) at 1 week, need for a second vasoactive, and 6-hour vasoactive infusion score (VIS). We used propensity score matching (PSM) and inverse probability treatment weighting (IPTW) to adjust for confounders (PRISM III, malignancy, age, admission epoch, and site). Results: There were 1,755 children with septic shock treated with either Epi or NE as first-line who met inclusion criteria. In both the PSM and IPTW analyses, NE as first-line was associated with a lower mortality, lower MODS at 1 week, and lower VIS score at 6 hours compared to Epi (all p< 0.001). In the 431 patients with AHRF, Epi as first-line was associated with lower rates of needing a second vasoactive when compared to NE in IPTW only (p=0.02), but there were no differences in the other outcomes. Conclusions: In this target trial emulation, we found that NE as first-line vasoactive was associated with more favorable outcomes in all comers with septic shock when compared to Epi. On the other hand, in the subset of patients with AHRF, there were no differences in outcomes except that Epi was associated with lower need for a second vasoactive. Our findings warrant prospective validation as they may lead to more targeted treatment in pediatric septic shock.
Grage et al. (Sun,) studied this question.