High-Minority Nursing Homes Disproportionately Affected by COVID-19 Deaths

Abstract

Racial/ethnic disparities in healthcare have been highlighted by the recent COVID-19 pandemic. Using the Centers for Medicare and Medicaid Services' Nursing Home COVID-19 Public File, this study examined the relationship between nursing home racial/ethnic mix and COVID-19 resident mortality. As of October 25, 2020, high minority nursing homes reported 6.5 COVID-19 deaths as compared to 2.6 deaths for nursing homes that had no racial/ethnic minorities. After controlling for interstate differences, facility-level resident characteristics, resource availability, and organizational characteristics, high-minority nursing homes had 61% more COVID-19 deaths [Incidence Rate Ratio (IRR) = 1.61; p < 0.001] as compared to nursing facilities with no minorities. From a policy perspective, nursing homes, that serve primarily minority populations, may need additional resources, such as, funding for staffing and personal protective equipment in the face of the pandemic. The COVID-19 pandemic has sharpened the focus on healthcare disparities and societal inequalities in the delivery of long-term care.

Keywords: COVID-19; disparities; mortality; nursing homes; race/ethnicity.

Figure 1

Nursing homes with COVID-19 death by resident racial/ethnic composition (n = 12,914). Source: Centers for medicare and medicaid services' nursing home COVID-19 public Blacks/Hispanics: Low proportion of minorities (0 < minority > 13.3); Medium proportion of minorities (13.3 < minority < 30.3); High proportion of minorities (30.3 < minority).

Figure 2

Incidence rate ratios of COVID-19 mortality by resident racial/ethnic mix (n = 12,914). CMS nursing home COVID-19 public file, brown University's LTCFocus, and Robert graham center's social deprivation index. No minorities is reference group; Low-proportion of minorities (0 < minority > 13.3 of Blacks/Hispanics); Medium-proportion of minorities (13.3 < minority < 30.3 of Blacks/Hispanics); High-proportion of minorities (30.3 < minority). Incidence rate ratios calculated from the negative binomial regression averaging over the remaining covariates. Model 1: Controlled for interstate differences and size. Model 2: Variables from model 1,and including facility-level resident characteristics. Model 3: Variables from model 2, and including resource availability. Model 4: Variables from model 3, and including organizational characteristics.