Decreased mortality with rollout of electronic pneumonia clinical decision support across 16 Utah hospital emergency departments

Introduction: ePNa open loop, electronic clinical decision support improved mortality and disposition for emergency department (ED) community-acquired pneumonia patients in 4 intervention hospitals vs 3 usual care hospitals (Annals EM 2015 66:511). ePNa provides ED clinicians with support for diagnosis and objective severity measurement, then disposition and antibiotic recommendations based on the DRIP score. We rolled out ePNa across 16 additional Intermountain Healthcare (Utah, USA) Hospitals beginning 2018 in a stepwise implementation trial. Objectives: Evaluate whether ePNa might improve clinical outcomes in a diverse group of non-teaching hospitals. Methods: 7293 ED patients >18 years 2017 to 2019 with complete data were identified electronically by ICD-10 codes; patients without confirmatory chest imaging (CheXED, Stanford AI model) were excluded. Results: Median age was 67 (IQR 52-79) years, 48% female. ePNa: use by clinicians averaged 47.4% after rollout but varied from 10% to 92% by month and hospital. Intention to treat analysis showed 30-day mortality of 5% (146/3101) vs 10% (417/4192) before rollout; outpatient disposition increased from 21% vs 48%. 7-day secondary hospitalization was 1% vs 3% after rollout. 30-day all-cause mortality was significantly lower after severity adjustment (electronicCURB, age, PaO2/FiO2, pleural effusion, HCAP, gender) by logistic regression (OR 0.57, 95% CI 0.46,0.69, p Conclusions: Rollout of ePNa clinical decision support into 16 hospital ED was associated with decreased mortality and hospital admission among pneumonia patients. We plan further interventions to increase ePNa use by ED clinicians.