1674: TARGETING VASOACTIVE THERAPIES TO WARM VERSUS COLD PEDIATRIC SEPTIC SHOCK

Abstract

Learning Objectives: Pediatric sepsis guidelines recommend vasopressors for “warm” shock and inotropes for “cold” shock. We hypothesized that clinicians either follow these guidelines in choosing vasoactive based on overall determination of shock type or use another single clinical sign to choose vasoactive, and that discordance between shock type and vasoactive selection is associated with poor outcomes. Methods: We performed a retrospective observational study of patients ≤18 years who had vasoactive (vasopressor/inotrope) medication initiated for septic shock in a single academic PICU between 2012 and 2016. Clinicians explicitly documented shock type and five clinical signs: extremity temperature, pulse strength, capillary refill, diastolic blood pressure, and pulse pressure. Each sign was categorized as indicating “warm” or “cold” shock (e.g. bounding pulse=warm; weak pulse=cold). Alignment between specified shock type or clinical sign and first vasoactive was analyzed using Cohen’s κ. Association of shock type-vasoactive discordance (e.g. patients with cold shock treated with norepinephrine) and complicated course (death within 28 days or ≥2 organ system dysfunctions at 7 days) was analyzed using multivariable logistic regression. Results: 117 patients met inclusion criteria. Median age was 5 (IQR 1-10) years, PIM-3 mortality risk was 4.0% (IQR 0.7-11.3%), and 35 (29.9%) patients had a complicated course. Overall, alignment of warm shock/clinical sign with vasopressor and cold shock/clinical sign with inotrope was poor (к 0-0.24). Patients with documented warm shock were more likely to be treated with discordant vasoactive (i.e. inotrope) than cold shock patients with vasopressors (64.3% vs 17.0%, p<0.001; overall discordance 54.7%). After adjusting for age, PIM-3, sepsis source, shock type, and time cohort, discordance between shock type and vasoactive selection was not associated with complicated course (aOR 0.87, 95% CI 0.28-2.72). Conclusions: Clinicians often did not target vasoactive selection to shock type or individual clinical sign, especially for warm shock. However, discordance between shock type and first vasoactive was not associated with worse outcomes. Additional studies are needed to update pediatric septic shock guidelines with the best way to target vasoactive support.