Patterns in Duration of Emergency Department Boarding and Variation by Sociodemographic Factors.

introductionEmergency department (ED) boarding negatively affects patient outcomes, increasing length of stay, hallway care, and mortality. Prior research found disparities in capacity metrics like hallway care based on patient race and ethnicity. However, whether boarding differs by demographics is not well characterized. We examined boarding variation by sociodemographic factors in a hospital with a standardized bed-prioritization process. We hypothesized that a structured inpatient assignment method may be associated with reduced boarding inequity.

methodsThis single-center, retrospective, cohort study included adult patients boarding in the ED after admission to the non-intensive care inpatient medicine service between February 2020-February 2023 at an urban, academic, tertiary-care hospital with > 110,000 annual ED visits. Primary outcome was time from admission order to inpatient bed transport. Patient demographics (age, sex, race/ethnicity, language, insurance, and housing status), visit characteristics (Emergency Severity Index, time, and day), and bed request features (telemetry, sitter need, and isolation precaution) were obtained via the medical record. We assessed for bivariate relationships between boarding time and demographics with descriptive statistics and analysis of variance using adjusted and unadjusted regression analyses with generalized estimating equations to account for patient-level correlation.

resultsIn total, 22,291 encounters were included. Average age was 64 (SD ±19) years, and 47% were female. Approximately 12% identified as Hispanic, 70% as non-Hispanic White, and 10% as non-Hispanic Black. Most (97%) boarded ≥ 120 minutes. In adjusted analyses, patients with Medicaid waited an additional 85 minutes (95% CI 49-121), and patients with Medicare waited an additional 67 minutes (95% CI 32-103) compared to those with commercial coverage (both P < .001, respectively). Non-Hispanic Black patients boarded 14 minutes longer (95% CI 22-51), and non-English primary language speakers boarded 15 minutes longer (95% CI 17-47) than non-Hispanic White patients and English primary language speakers, respectively, although these two findings were not statistically significant.

conclusionAmong adult patients admitted to the inpatient medicine service, non-commercial insurance such as Medicaid and Medicare was significantly associated with longer ED boarding, whereas race/ethnicity and primary language were not. Further study should determine whether these findings are replicated elsewhere, how this impacts patients, and whether targeted intervention can reduce inequities.