352 - Association between Hospital Occupancy and Mortality, 2016-2024

Publication Number: 4345.352

Gabrielle D'Ambrosi, Boston Children's Hospital, Somerville, MA, United States; Eva Robinson, boston childrens hospital, Boston, MA, United States; Hannah J. Lev, Boston Children's Hospital, Boston, MA, United States; Carly E.. Milliren, Boston Children's Hospital, Boston, MA, United States; Carter Petty, Boston Children's Hospital, Boston, MA, United States; Rachel K. D'Anna, Boston Children's Hospital, Boston, MA, United States

Background: Hospital occupancy strain, described as an increase in patient census relative to hospital resources, may have negative impacts on patients including an increase in mortality. Efforts to improve care during times of severe strain, such as during public health emergencies, are imperative to improve patient outcomes.

Objective: This study examines the relationship between hospital occupancy and overall mortality utilizing data from the Pediatric Health Information System (PHIS). Additionally, this study considers the effect of hospital characteristics and temporal trends on mortality.

Design/Methods: A retrospective analysis was performed using daily hospital census and mortality data from 2016-2024 across 50 U.S. hospitals in PHIS. Daily hospital occupancy was quantified as the number of daily census counts over each hospital's maximum annual census (i.e., capacity) and expressed as a percentage. Occupancy was then divided into quartiles. A mixed-effects linear regression model, accounting for clustering of daily observations within hospitals, assessed the association between daily mortality rates and hospital occupancy, adjusting for year, region, season, weekday, and COVID period.

Results: From 2016-2024, 159,272 hospital-days were examined. Daily percent mortality decreased as hospital occupancy increased, with lower mortality on days in the higher three occupancy quartiles compared to the lowest (all p < 0.001) in the unadjusted mixed-effects model (Table 1). Mortality rates remained significantly lower on highest occupancy days after adjusting for year, region, season, and COVID period (-0.005%, 95% CI: (-0.008%, -0.002%), p = 0.002). Mortality declined over time but was moderately higher in Southern hospitals, over weekends and summer months, and during the COVID pandemic period.

Conclusion(s): While effect sizes were small, our findings indicate that mortality rates declined as hospitals approached higher occupancy from 2016-2024. In addition, we observed distinct temporal and regional variations that can inform future resource planning and patient care strategies.

Table 1: Mixed-effects models examining association between hospital factors and mortality at 50 PHIS hospitals (N=159,272 hospital days)
*Estimates presented are direct percentages
**Pre-onset defined as 2016-2019, during pandemic defined as 2020-2022, post-pandemic defined as 2023-2024