Background: The Veterans Affairs (VA) hospital is a primary site for internal medicine (IM) residents at our program. It provides trainees with exposure to high-yield general medicine cases and an opportunity to care for a unique and vulnerable population. Unfortunately, the admitting structure presents a number of challenges. Factors such as the geographical separation between the emergency department (ED) and wards and increasing patient care requirements limit communication between the ED and IM teams, leading to a rising number of low-impact admissions and decreased physician satisfaction. We hypothesized that the addition of an IM triaging resident (TR) in the ED would facilitate communication practices and improve triaging and management processes, leading to a 10% reduction in medical admissions and a 20% improvement in provider satisfaction.
Methods: A third-year IM resident was stationed in the ED during peak times for two weeks in a triaging role. Medical patients were initially evaluated by ED physicians, but those destined for admission were then evaluated by the TR to assess for the necessity of admission and assist in the initial management. When probable low-impact admissions were identified, the TR presented a discharge plan to ED and/or IM attending physicians. These were instances in which admission was unlikely to add meaningful impact to patient care based on objective and subjective parameters. For all patients ultimately admitted after the triaging process, the TR helped assign the optimal level of care and oversaw the technical aspects of admission in conjunction with nursing and bed management. Physician satisfaction, patient safety, and hospital performance metrics were tracked.
Results: More than 60 patients were evaluated by the TR, and more than 40% of cases represented likely low-impact admissions. Nearly half of those patients were subsequently discharged from the ED, resulting in a 19% relative reduction in the rate of medical admission during this period. Admission decision times did not vary, and no adverse patient safety events occurred. Surveys administered to IM residents revealed that the addition of a TR led to statistically significant improvements in work flow and initial patient management and a trend towards more time for educational activities. ED staff also reported improved work flow and productivity.
Conclusions: The use of a TR at our local VA hospital led to a reduction in low-impact admissions and improved physician experiences by enhancing interdepartmental communication and supporting the activities of the ED and IM teams at peak times. Our intervention did not have a negative impact on patient safety or applicable ED performance metrics and also identified potential areas of improvement in transitions of care from outpatient clinics to the ED and wards. In the future, expansion of this model may present an opportunity to improve physician satisfaction and hospital utilization through continued collaboration between ED and IM providers.