Association Between Patient Diversity in Hospitals and Racial/Ethnic Differences in Patient Length of Stay.

TitleAssociation Between Patient Diversity in Hospitals and Racial/Ethnic Differences in Patient Length of Stay.
Publication TypeJournal Article
Year of Publication2022
AuthorsGhosh AK, Unruh MA, Ibrahim S, Shapiro MF
JournalJ Gen Intern Med
Volume37
Issue4
Pagination723-729
Date Published2022 Mar
ISSN1525-1497
KeywordsCross-Sectional Studies, Healthcare Disparities, Hospitals, Humans, Length of Stay, Retrospective Studies, United States
Abstract

BACKGROUND: Hospitals serving a disproportionate share of racial/ethnic minorities have been shown to have poorer quality outcomes. It is unknown whether efficiencies in inpatient care, measured by length of stay (LOS), differ based on the proportion patients served by a hospital who are minorities.

OBJECTIVE: To examine the association between the racial/ethnic diversity of a hospital's patients and disparities in LOS.

DESIGN: Retrospective cross-sectional study.

PARTICIPANTS: One million five hundred forty-six thousand nine hundred fifty-five admissions using the 2017 New York State Inpatient Database from the Healthcare Cost and Utilization Project.

MAIN MEASURE: Differences in mean adjusted LOS (ALOS) between White and Black, Hispanic, and Other (Asian, Pacific Islander, Native American, and Other) admissions by Racial/Ethnic Diversity Index (proportion of non-White patients admitted to total patients admitted to that same hospital) in quintiles (Q1 to Q5), stratified by discharge destination. Mean LOS was adjusted for patient demographic, clinical, and admission characteristics and for individual intercepts for each hospital.

KEY RESULTS: In both unadjusted and adjusted analysis, Black-White and Other-White mean LOS differences were smallest in the most diverse hospitals (Black-White: unadjusted, -0.07 days [-0.1 to -0.04], and adjusted, 0.16 days [95% CI: 0.16 to 0.16]; Other-White: unadjusted, -0.74 days [95% CI: -0.77 to -0.71], and adjusted, 0.01 days [95% CI: 0.01 to 0.02]). For Hispanic patients, in unadjusted analysis, the mean LOS difference was greatest in the most diverse hospitals (-0.92 days, 95% CI: -0.95 to -0.89) but after adjustment, this was no longer the case. Similar patterns across all racial/ethnic groups were observed after analyses were stratified by discharge destination.

CONCLUSION: Mean adjusted LOS differences between White and Black patients, and White and patients of Other race was smallest in most diverse hospitals, but not differences between Hispanic and White patients. These findings may reflect specific structural factors which affect racial/ethnic differences in patient LOS.

DOI10.1007/s11606-021-07239-w
Alternate JournalJ Gen Intern Med
PubMed ID34981364
PubMed Central IDPMC8904308
Grant ListKL2 TR002385 / TR / NCATS NIH HHS / United States
K24 AR055259 / AR / NIAMS NIH HHS / United States