Antimicrobial resistance has evolved due to decades of inappropriate antibiotic prescribing by clinicians. In the emergency department (ED), nearly 50% of the 10 million antibiotic prescriptions written each year are inappropriate or unnecessary. However, antimicrobial stewardship programs (ASPs) for the ED have yet to be created. Failure to develop ED-based ASPs is a result of barriers unique to the ED, including lack of ED formatted guidelines, erratic workflow, rapid decision making, and diagnostic uncertainty. Electronic health record-based clinical decision support (EHR-CDS) can be customized to present ED- specific antibiotic recommendations at the point of care, in the usual clinical workflow. Therefore, EHR-CDS has great potential to overcome ED barriers and serve as platform for ED-based ASP. In this proposal, we build upon our preliminary single-center research to develop a multicenter EHR-CDS for antibiotic prescribing for two common pediatric ED infections, community acquired pneumonia (CAP) and urinary tract infection (UTI). National CAP and UTI antibiotic prescribing guidelines exist, but have not been adapted into ED context. We will use rigorous dissemination and implementation methods that this team has successfully used previously to create multicenter EHR-CDS. We also build upon our preliminary work to create a novel, multi- center EHR-CDS within the Pediatric Emergency Care Applied Research Network (PECARN). We will conduct workflow analyses and identify EHR triggers to determine the optimal timing for EHR-CDS activation, necessary to minimize alert fatigue. At each of the 3 participating sites, we will also develop an EHR-based mechanism to determine guideline adherent antibiotic prescribing for CAP and UTI. Through integration of the formatted ED antibiotic treatment guidelines for CAP and UTI, workflow analyses, and EHR activation triggers analyses, we will configure a prototype EHR-CDS for antibiotic prescribing. We will conduct heuristic review and subsequent scenario-based user testing at each site to produce a functional, scalable prototype EHR- CDS. Though we will initially build the EHR-CDS using a single her (with the largest market share), we will simultaneously develop a web-service version of the CDS for further scalability. Finally, as we have for prior studies, we will create automated clinician feedback reports using EHR data to provide antibiotic prescribing practices; these will complement the EHR-CDS as part of the overall ED-based ASP. The adaptable scalable EHR-CDS for CAP and UTI, and the provider feedback reports, will serve as the centerpieces for generalizable ED-based ASP. Upon completion of the work in this proposal, we will be poised to conduct a multicenter trial to test the effect of our EHR-CDS in pediatric and general ED settings.
Antimicrobial stewardship for the emergency department is necessary to reduce inappropriate antibiotic prescribing and development of resistant organisms. This proposal will result in an adaptable, scalable EHR- CDS for antibiotic prescribing ready for testing in a multicenter trial.
