The objective of the Connected Emergency Care (CEC) Patient Safety Learning Lab is to reduce health and financial harm for patients with lower respiratory tract infection caused by a fragmented emergency care system. Within this system, lack of longitudinal relationships between clinicians and patients drives care with- out context and practice-based learning and outcomes-based decision making are limited by absence of post- encounter feedback. These disconnects are compounded in the emergency department (ED) by a hazardous decision making environment fraught with time-pressure and excessive cognitive loading. Electronic health rec- ords (EHR) and health information technology (HIT) infrastructures hold potential for improvement of these conditions, but are currently optimized toward billing and documentation and often exacerbate disconnect be- tween clinicians and patients. We intend to reduce health and financial harms caused by suboptimal diagnosis, treatment, and disposition decisions using advanced data science methods and EHR-integrated clinical deci- sion support (CDS). Using this approach, we will establish a connected (closed-loop) emergency care sys- tem through which the following specific aims are achieved: 1. Optimize diagnostic performance for patients with suspected LRTI by (A) minimizing overuse of avoidable imaging and laboratory testing, and (B) expediting detection of patients with severe LRTI. 2. Increase the specificity of antibiotic treatment for patients with an LRTI by (A) reducing inappropriate prescribing, and (B) enabling targeted antibiotic therapy; indication, spectrum, and dose. 3. Improve transition of care outcomes after the ED encounter is complete by reducing (A) unneces- sary hospitalizations and sudden care-level changes for those admitted, and (B) 30-day post-encounter acute care utilization for those discharged. Clinical decision making for patients with LRTI is complicated by nonspecific and overlapping disease presen- tations and diagnosis and treatment and disposition decisions are highly variable. Our aims target these three decision-points in the LRTI care continuum to maximize impact on health and cost outcomes. These aims will be executed in parallel by taking a systems engineering approach to CDS development that connects ED clinicians to the patient’s pre-encounter context, post-encounter outcomes, and intra-encoun- ter current and projected state. Advanced data-science, user-centered design, human factors engineering, de- cision sciences, and principles of health informatics will be employed during EHR-embedded CDS develop- ment and in the products that result. Central to their design will be the ability to adapt to the uniqueness of di- verse ED settings while having the capacity to scale leveraging newly launched EHR dissemination platforms (e.g., Epic App Orchard) and AHRQ CDS Connect.


Funding Source


Project Period