Background: The field of hospital medicine has grown significantly, warranting investigation of staffing models to maximize outcomes and promote patient safety. Handoffs are a source of increased risk for patients and result in longer lengths of stay, increases in error, and missed items for follow up. Our current hospital medicine model uses specialized care teams including Oncology, Kidney and Liver Transplant. These teams see general medicine patients, but when specialized service is needed, general medicine patients are moved to other services. A Jeopardy service is utilized when the number of patients exceeds safe capacity of the standard teams and primarily sees general medicine patients. It is also used for coverage for providers unable to work due to illness or emergency. We sought to analyze the impact of our system on handoffs for general medicine patients.
Methods: Our study is a retrospective cohort reviewing all patients who had at least one note written by a hospital medicine provider from April 2018 to June 2019. Data on filed notes were collected from the electronic health record (EHR) reporting database including patient demographic data, encounter admission and discharge date and time, note type, author, specialty service and file time. First contact providers (FCP) were matched with the note level data by encounter number. Additional data was collected from hospital medicine scheduling software and was matched with the note author to confirm service information. Encounters were analyzed as to whether a patient had a jeopardy provider, advanced practice provider (APP) and numbers of unique note authors and FCPs.
Results: Patients with General Medicine needs had a significantly higher number of patients who identified as Black (N=1494, 81%) compared to the Liver (N=234, 48%), Transplant (N=293, 65%), and Oncology Teams (N= 834, 59%). General medicine patients were more likely to end up being seen by a jeopardy provider (N=148), 8% for General Medicine vs Liver (N=30, 6%), Transplant (N=18, 4%) and Oncology Teams (N= 56, 4%) respectively. When compared to standard teams, patients seen on jeopardy were more likely to have an increased number of unique note authors (2.48 vs 2.14, p < 0.001) despite having a shorter length of stay (6.95 vs 8.46, p < 0.001) . General medicine patients on Jeopardy services compared to standard care, trended towards a longer length of stay (8.68 vs 7.44, p=0.07), had more unique note authors (3.03 vs 2.18, p < 0.001) and more FCPs involved in their care (3.29 vs 2.37, p < 0.001).
Conclusions: With hospital medicine becoming a foundational aspect of hospital structure, it is critical to optimize the structure to best serve patients. Our hospital medicine model at UCMC results in increased handoffs and fragmentation of care for our general medicine population, which has a higher proportion of minorities compared to other specialty services. Multiple studies have demonstrated the health inequities experienced by racial minorities in this country, both historically and more recently with the COVID-19 pandemic. In addition, there is a great deal of literature highlighting the risks associated with discontinuity in the hospital, perhaps most concerning is an increased risk of mortality. This indicates our current hospital medicine structure may result in increased vulnerability amongst our general medicine and minority patient populations, and risks adding to the inequality experienced by racial minorities, specifically African Americans, in this country.
