Abstract TMP52: Racial And Ethnic Disparities In Hospitalization Cost After Accounting For The Influence Of Hospital Length Of Stay

Abstract

Background: Hospitalization length of stay (LOS) and cost are used as quality metrics, but there are known racial and ethnic disparities. For individuals with ischemic stroke, the relationship between LOS, cost, and race/ethnicity has not been fully examined. Methods: We included adult discharges from the National Inpatient Sample 2016-2019 with a baseline NIHSS, primary discharge diagnosis of ischemic stroke (I63*), and LOS ≤14 days to limit outlier LOS admissions. The primary outcome was hospitalization cost, which the NIS defines as a cost-to-charge ratio times charges. The exposure was race/ethnicity, which was abstracted from the medical record. We fit linear regression models and compared models with and without LOS to determine if it changed the point estimates of our outcome. Results: The cohort contained 128,399 discharges (mean age 69.5±14.0, 49.0% Female, median NIHSS 4 (1-8), and median LOS of 3 (2-5) days). In the model without LOS, White individuals had the lowest cost. After LOS adjustment, Black individuals had significantly lower cost than other race/ethnicities (Figure 1). As shown in Table 1, after adjusting for LOS, Black stroke hospitalizations cost $256.8 less than White, while hospitalizations for other race/ethnicities cost anywhere from $300-867 more than White. A separate analysis of the same dataset yielded that a greater proportion of Black patients (16.4%) have Medicaid at discharge versus White patients (5.6%). Conclusions: Hospitalization cost for Black patients is confounded by their length of stay. More research into the factors influencing this mediation is necessary; considering for example, the influence of waiting time for public insurance.