Development of SMART on FHIR interoperable clinical decision support for emergency department patients with pneumonia and pilot deployment into novel Epic electronic health record environments

Project Summary In the United States, more than 1.5 million adults are hospitalized each year with community-acquired pneumonia. Despite availability of high-quality guidelines for the treatment of patients with pneumonia, guideline adherence remains poor and patient care is variable. Lack of objective risk assessment leads to unnecessary hospital admissions and misallocation of scarce intensive care resources. Thus, there is a critical need for well-designed clinical decision support (CDS) tools to assist clinicians with accurate diagnosis and appropriate treatment for patients with pneumonia. Intermountain Healthcare developed and deployed real-time electronic CDS (called ePNa) embedded in the Cerner electronic health record (EHR) across 20 adult hospital emergency departments. ePNa gathers real- time and historical data from the EHR combined with artificial intelligence processing of chest images (ChexED) to guide pneumonia diagnosis, risk stratification, site of care, risk of antibiotic resistance, and treatment. ePNa smooths transitions of care between clinicians. Our work incorporates the Five Rights of CDS to optimize the technology’s strengths in the clinical environment. Two controlled studies demonstrate that ePNa increases guideline-recommended, best practice care and decreases 30-day all-cause mortality in adults with pneumonia. ePNa is ready to be deployed to additional environments with distinct patient populations and EHRs. However, siloed EHRs limit technology developed within one vendor from being shared with others. The objective of this proposal is to enable broad dissemination of ePNa by optimizing its interoperability, scalability, and usability. Accordingly, we aim to (1) Develop a SMART on FHIR ePNa application capable of interfacing with both the Cerner and Epic EMRs and conduct a formative evaluation through an iterative design process. We will rapidly refine and validate the new ePNa application. (2) Engage patients and clinicians to inform the patient-centeredness of ePNa’s clinical interface. The project will incorporate feedback from stakeholder clinicians, administrators, and patients to understand contextual factors affecting implementation success in a novel geographic and demographic environment. (3) Evaluate the feasibility and acceptability of ePNa through a pilot implementation trial guided by the RE-AIM QuEST framework. Our proposal is likely to facilitate guideline adherent care for patients with pneumonia broadly across different healthcare systems.