Evaluating Mortality and Hospital Room Design after High-Risk Inpatient Surgery

Abstract

Introduction: Each year in the US we spend more than $48 billion on the construction of healthcare facilities and unfortunately rarely evaluate the performance of their design after occupation. Nonetheless, certain design features— including occupancy (single vs double), distance from nursing station, direct lines of site, and a window view— likely influence patient care and outcomes but have not been adequately evaluated. Methods: We linked two complementary datasets - Michigan Medicine’s 1.) Facilities of Operations Building Information and 2.) Research Database Warehouse- and identified a cohort of 13 different surgery procedures that have known rates of high complication (colectomy, pancreatectomy, kidney transplant, etc.). Admissions to the 5th or 8th floor of a University Hospital during 2016-2019 provided a sample of 3,964 inpatient surgeries. Patient rooms were coded based on their features and patient encounters were linked by room number to identify clinical outcomes (mortality, length of stay) related to room design. A logistical and Poisson regression were used to evaluate inpatient and 30-day mortality. Results: Inpatient mortality was significantly higher for rooms without a window view, when adjusting for patient demographics. 30-day mortality was higher for rooms without a window view or direct line of site. Mortality did not vary by room type when accounting for patient days. Conclusion: Mortality for high-risk procedures varies widely across room design features and room type. Risk adjustment methods can alter the estimate of mortality and more research should be dedicated to isolating additional room design features.