Samir H Shah
University of Tennessee Health Sciences Center, USA
Title: Innovation in earlier identification of pediatric severe sepsis: Application of an electronic pediatric severe sepsis screening tool (E-assist study)
Biography
Biography: Samir H Shah
Abstract
Background: Earlier recognition of pediatric Severe Sepsis (SS) is critical for timely goal-directed therapy and improving patient outcomes. Early indicators of SS and therapy administered in response to an Electronic Medical Record (EMR)-integrated, pediatric SS screening algorithm ('sniffer') have not been described.
Methods: All adolescents (13-18 years) admitted to Le Bonheur Children's Hospital from over a 1-year period (n=3138) were screened using an EMR SS screening algorithm. Demographics, diagnoses, and interventions on all patients were collected for potential CRS cases.
Results: Using mixed logistic regression on 15 alert indicators, pulse (p=0.004), temperature (p=0.008), ALT (p=0.015) and BUN/Creatinine ratio (p=0.013) were found to be key independent determinants of pediatric SS. Alert-mediated, early goal-directed therapies (administered within 6 hours of a positive alert) were fluid bolus (33%, n=12), antibiotic use (61%, n=44), and oxygen support (96%, n=55). The median time (hours) from the first positive alert to therapy was 9.4 - fluid bolus (IQR 3.1-68.9), 5.3 - antibiotic use (IQR 1.8-9.6), and 0.8 - oxygen support (IQR 0.4-1.6). True positive cases (n=66) had increases in hospital length of stay (LOS; in days) - 5.4 (p=0.004); ICU LOS - 8.5 (p<0.001); mechanical ventilator use - 3.3 (p=0.003); pressor support - 0.8 (p=0.17). SS mortality was 2.5% (n=5).
Conclusions: An EMR-integrated SS real-time screening algorithm offers the potential to facilitate early goal-directed therapy and decrease sepsis-related mortality.