Impact of an Overnight Internal Medicine Academic Hospitalist Program on Patient Outcomes

Background: Many academic hospitals have implemented overnight hospitalists to supervise house staff and improve outcomes, but few studies have described the impact of this role. Objective: To investigate the effect of an overnight academic hospitalist program on patient-level outcomes. Secondary objectives were to describe the program’s revenue generation and work tasks. Design: Retrospective interrupted time-series analysis of patients admitted to the medicine service before and after implementation of the program. Participants: All patients aged 18 and older admitted to the acute or intermediate care units between 7:00 p.m. and 6:59 a.m. during the period before (April 2011–August 2012) and after (September 2012–April 2014) program implementation. Intervention: An on-site attending-level physician directly supervising medicine house staff overnight, providing clinical care during high-volume periods, and ensuring safe handoffs to daytime providers. Main Measures: Primary outcomes included in-hospital mortality, 30-day hospital readmissions, length of stay, and upgrades in care on the night of admission and during hospitalization. Multivariable models estimated the effect on outcomes after adjusting for secular trends. Revenue generation and work tasks are reported descriptively. Key Results: During the study period, 6484 patients were admitted to the medicine service: 2722 (42 %) before and 3762 (58 %) after implementation. No differences were found in mortality (1.1 % vs. 0.9 %, p=0.38), 30-day readmissions (14.8 % vs. 15.6 %, p=0.39), mean length of stay (3.09 vs. 3.08 days, p=0.86), or upgrades to intensive care on the night of admission (0.4 % vs. 0.7 %, p=0.11) or during hospitalization (3.5 % vs. 4.2 %, p=0.20). During the first year, hospitalists billed 1209 patient encounters (3.3/shift) and 63 procedures (0.2/shift), and supervised 1939 patient admissions (6.12/shift) while supervising house staff 3-h/shifts. Conclusions: Implementation of an overnight academic hospitalist program showed no impact on several important clinical outcomes, and revenue generation was modest. As overnight hospitalist programs develop, investigations are needed to delineate the return on investment and focus on other outcomes that may be more sensitive to change, such as errors and provider/patient satisfaction.