Black History Month Focus and Health Disparities in the African American Community

Godstime

RacialDifferencesin....pdf

Home >Nursing homework help >Black History Month Focus and Health Disparities in the African American Community

Journal of Aging and Health 1 –26

sagepub.com/journalsPermissions.nav DOI: 10.1177/0898264315589576

jah.sagepub.com

Article

Racial Differences in Minnesota Nursing Home Residents’ Quality of Life: The Importance of Looking Beyond Individual Predictors

Tetyana P. Shippee, PhD1, Carrie Henning-Smith, MPH, MSW1, Taeho Greg Rhee, AM1, Robert N. Held, MBA2, and Robert L. Kane, MD1

Abstract Objectives: The aim of this study is to investigate racial differences in nursing home (NH) residents’ quality of life (QOL) at the resident and facility levels. Method: We used hierarchical linear modeling to identify significant resident- and facility-level predictors for racial differences in six resident-reported QOL domains. Data came from the following: (a) resident-reported QOL (n = 10,929), (b) the Minimum Data Set, and (c) facility-level characteristics from the Minnesota Department of Human Services (n = 376). Results: White residents reported higher QOL in five of six domains, but in full models, individual-level racial differences remained only for food enjoyment. On the facility level, higher percentage of White residents was associated with better scores in three domains, even after adjusting for all characteristics. Discussion: Racial differences in QOL exist on individual and aggregate levels. Individual differences are mainly explained

1University of Minnesota, Minneapolis, USA 2Minnesota Department of Human Services, Minneapolis, USA

Corresponding Author: Tetyana P. Shippee, Division of Health Policy and Management, School of Public Health, University of Minnesota, 420 Delaware Street SE, MMC 729, Minneapolis, MN 55455, USA. Email: [email protected]

589576 JAHXXX10.1177/0898264315589576Journal of Aging and HealthShippee et al. research-article2015

mailto:[email protected]

2 Journal of Aging and Health

by health status. The finding that facility racial composition predicts QOL more than individual race underscores the importance of examining NH structural characteristics and practices.

Keywords long-term care, quality of life, racial disparities, nursing home

Introduction

Members of racial/ethnic minority populations have been underrepresented in nursing homes (NHs) and other formal long-term care (LTC) services (Mui, Choi, & Monk, 1998). However, over the past decade, the proportion of older adults from minority groups residing in NHs has quickly grown (Agency for Healthcare Research and Quality [AHRQ], 2000). Racial/ethnic minority groups tend to receive poorer quality of care (QOC) in LTC settings (Smith, Feng, Fennell, Zinn, & Mor, 2007), although this may be due to patterns of use of a subset of facilities rather than outright discrimination in admission (Konetzka & Werner, 2009). Despite the increase in use of NH services by racial/ethnic minority groups, we know little about their quality of life (QOL) in these settings and how it compares to their White counterparts.

NH quality should not be solely defined as QOC, which is typically mea- sured by staff-reported clinical outcomes. Instead, NH quality should include resident-reported measures of QOL, which is multidimensional, incorporat- ing aspects of residents’ lives and experiences, physical environment and comfort, relationships with staff and other residents, food and meal enjoy- ment, social engagement, and mood (Kane, 2003; Shippee, Henning-Smith, Kane, & Lewis, 2013). Yet, most research on racial differences in NH quality has focused on QOC (Allsworth, Toppa, Palin, & Lapane, 2005; Christian, Lapane, & Toppa, 2003; Fennell, Miller, & Mor, 2000; Hughes & Lapane, 2004). Many of these studies have been limited to specific medical condi- tions and clinical procedures, such as diabetes (Allsworth et al., 2005) and medical management for stroke (Allsworth et al., 2005; Christian et al., 2003; Hughes & Lapane, 2004).

In contrast, this study uses a multidimensional, validated, resident-reported measure of QOL for the entire population of Medicaid-certified NHs in Minnesota, which is one of the few states in the nation to collect such mea- sures (Shippee et al., 2013). We use cumulative inequality (CI) theory (Ferraro & Shippee, 2009) to examine whether NH residents’ QOL differs by racial characteristics—and if so, whether these differences are explained by resident- or facility-level predictors. At the individual level, the outcome of

Shippee et al. 3

interest is whether individual QOL is related to race. At the structural level, the outcome of interest is whether facility-level QOL is associated with facil- ity racial composition and individual QOL. Ours is the first study to look at resident- and facility-level correlates of racial differences in NH QOL; our findings provide insights for policy and practice as the NH population contin- ues to diversify.

Racial Differences on Individual and Facility Levels

Few studies explicitly address racial differences in NH QOL. Resident-level differences in NHs’ QOC include more use of physical restraints among Black residents compared with White residents (Cassie & Cassie, 2013), less use of specialized dementia care among minority residents (Sengupta, Decker, Harris-Kojetin, & Jones, 2012), and lower rates of flu vaccination for Black residents versus White residents (Cai, Feng, Fennell, & Mor, 2011). A few, mainly qualitative, studies find generally lower QOL for minority older adults (Engle, Fox-Hill, & Graney, 1998; Ryvicker, 2011; Wu & Barker, 2008). For example, a study examining needs and preferences of Black and White NH residents close to death found that QOL was equally important for Black and White adults, but that Black residents reported more untreated pain (Engle et al., 1998). Another study found that older Asian NH residents reported lack of cultural understanding (e.g., culturally inappropriate food) and need for more social engagement (Wu & Barker, 2008). Another qualita- tive study found that residents in a NH serving a predominantly racial/ethnic minority population had lower quality of resident–staff interaction because the facility was less focused on promoting resident-centered care and because staff members were compared with those in a facility serving predominately White residents, had less resident individual information, and were less trained on how to communicate with residents (Ryvicker, 2011). Overall, minority NH residents, especially Black older adults, tend to have lower functional status (Jones, Sonnenfeld, & Harris-Kojetin, 2009), which affects QOL (DuBeau, Simon, & Morris, 2006; Shippee et al., 2013). Therefore, these residents may also report lower QOL ratings compared with their White counterparts.

Especially needed is better understanding of how facility-level predictors affect racial differences in QOL, particularly as racial segregation is an important reason for racial disparities in health care (Greene, Blustein, & Weitzman, 2006; Hayanga et al., 2009). A growing body of literature shows that minorities tend to be disproportionately served by hospitals with lower QOC (Jha, Fisher, Li, Orav, & Epstein, 2005). Research on disparities in LTC has demonstrated that long-standing policies of racial and socioeconomic

4 Journal of Aging and Health

segregation have resulted in segregated NH facilities (Smith et al., 2007, 2008). NHs that house higher proportions of Black residents often have seri- ous quality deficiencies, lower staffing ratios, and greater financial vulnera- bility (Stone, 2011). Also, minority neighborhoods have fewer NHs than predominately White neighborhoods (Smith et al., 2007), and the NHs that do operate in high-poverty neighborhoods, with a higher proportion of minority elderly residents, have higher numbers of closures and fare worse on quality performance measures (Feng et al., 2011b). Although these studies demon- strate troubling disparities in QOC for minority older adults, few studies have examined differences in QOL by race/ethnicity.

Existing literature focuses mostly on racial differences in admission to NHs and QOC. Black older adults have historically used NHs (and other forms of LTC) at a lower rate than their White counterparts (Feng, Fennell, Tyler, Clark, & Mor, 2011a), despite bearing higher disability burden. This difference may reflect personal choice, as well as difficulty accessing LTC. However, trends now indicate that Black older adults use NHs more than White older adults (Smith et al., 2008). Black older adults admitted to NHs are in worse health than their White counterparts (Cagney & Agree, 2005), in part because they are less likely to have a primary care provider and more likely to delay care (Ferraro & Shippee, 2008). Cultural preferences may explain differences in timing of admission; some research shows different norms around extended family providing care (Cagney & Agree, 2005) and historical mistrust of health care by minorities, especially Black older adults (Gamble, 1997). Payment source may be another obstacle; most Black older adults rely on Medicaid (Sengupta et al., 2012), and some NHs prefer to admit residents who are not on Medicaid.

Theoretical Framework

We draw on the CI theory (Ferraro & Shippee, 2009) to understand racial differences in QOL in NHs. CI theory is a middle-range theory that synthe- sizes elements from stress process (Pearlin, Schieman, Fazio, & Meersman, 2005), cumulative disadvantage (Dannefer, 2003), and life-course perspec- tives (Elder,1998) to focus on how inequality is generated in social systems and plays out over the life course. We view the link between race/ethnicity and QOL as resting on both individual and structural levels or resulting from the interplay “between institutional arrangements and individual life trajecto- ries” (O’Rand, 1996, p. 230). Despite this being a cross-sectional study, we believe CI theory is useful because it helps understand disparities in LTC as an outcome of lifetime processes of disadvantage/advantage. CI theory is useful to our study in at least two respects.

Shippee et al. 5

First, the theory emphasizes the role of accumulated risk and available resources for individual outcomes. Race/ethnicity is confounded with socio- economic status and is related to socially determined barriers and resultant health risks which might result in different patterns of LTC use. Black older adults (who constitute the majority of our non-White sample) have a greater likelihood of depending on Social Security benefits as their only source of income and of depending on Medicaid for health insurance compared with their White counterparts which might result in different options of facilities for LTC (Sengupta et al., 2012). Therefore, they may experience limited access to high quality NHs and worse health status when admitted. Race/ ethnicity is also related to one’s personal cultural attitudes and beliefs about health and the health care system. A history of racial discrimination and exploitation of vulnerable populations has led to greater distrust in health care providers and health care systems among minority populations than that of their White counterparts (Boulware, Cooper, Ratner, LaVeist, & Powe, 2003). Research shows systematic differences in the QOC received by White versus Black older adults in LTC (Cai et al., 2011).

Second, CI theory emphasizes that inequality has an institutional charac- ter, with attention to how structural arrangements, not simply individual char- acteristics or choices, maintain or exacerbate inequality on a variety of outcomes, including health care. Racial disparities in QOC on the facility level include greater numbers of deficiencies and lower staffing ratios in NHs with more minority residents (Smith et al., 2007). In addition, CI theory states that neighborhood context contributes to the development of inequality through access to care and available social support, among other factors (Ferraro, Shippee, & Schafer, 2009). Research shows that minority individu- als, because of their lower socioeconomic status, are more likely to live in impoverished, often racially segregated neighborhoods, characterized by resource-poor institutions and health care providers (Smith et al., 2007). Lack of these resources can result in a variety of health problems for the residents including health hazards (Cagney, Browning, & Wen, 2005), delayed access to primary care (Latkin & Curry, 2003), and lower QOC (Jha et al., 2005). NHs that are most likely to serve poor and non-White populations are char- acterized by limited resources, reliance on Medicaid, and poor QOC (Mor, Zinn, Zngelelli, Teno, & Miller, 2004) and are at greatest risk of closure (Feng et al., 2011b).

Based on CI theory and the existing literature, we hypothesized that on the individual level, minority residents in NHs would report lower individual QOL due to accumulated health and socioeconomic disadvantages over their life course, compared with their White counterparts (Hypothesis 1 [H1]). On the facility level, we expected that NHs that serve a higher proportion of

6 Journal of Aging and Health

White older adults would have higher average QOL scores compared with facilities that serve a higher proportion of non-White residents (Hypothesis 2 [H2]). However, there may be individual differences in QOL based on the facility-level racial composition. For example, non-White residents may have higher QOL in a facility with others “like them,” whereas White residents may do worse in a facility with a higher percentage of non-White residents. Thus, we also examine the effect of facility racial composition on individual QOL scores, building on the classic work on group composition and indi- vidual outcomes in education (e.g., Rosenberg & Simmons, 1971). We com- bine multidimensional resident-reported data on QOL with clinical data from the Minimum Data Set (MDS) and facility-level characteristics from the Minnesota Department of Human Services (DHS) to investigate these hypotheses.

Method

Sample and Data Sources

This study used data from three sources: (a) self-reported resident interviews using a multidimensional measure of QOL, (b) resident clinical data from the MDS version 2.0, and (c) facility-level characteristics from reports to the Minnesota DHS.

Self-reported resident QOL was compiled from the Resident Quality of Life and Satisfaction With Care Survey. The interview is conducted annually via two-stage random sampling, in which facilities provide a list of long-stay and short-stay residents. The interview sample includes separate random samples of long- and short-stay residents at each facility (Vital Research, 2010). Residents were eligible for either list if they were not in isolation due to communicable illness and if they or their guardian did not decline partici- pation. In 2010, 96% of residents (27,724) were eligible to participate, and 58% (16,187) were sampled to be approached. Of these, 15% had unsuccess- ful interview attempts with the most common reasons being inability to respond (5%), refusal (4%), and severe cognitive impairment (2%), leaving a survey response rate of 85% (n = 13,433). An average of 35 interviews per facility were completed (Vital Research, 2010).

Face-to-face interviews used a 52-item survey covering previously vali- dated QOL domains (Kane, 2003; Kane et al., 2003). The survey uses a sim- plified yes/no binary response structure to facilitate inclusion of respondents with mild to moderate cognitive impairment (except for mood items, which use a Likert-type scale from 1 to 4). The majority of respondents missed only one to five out of 52 items (58%), with 14% responding to all 52 items.

Shippee et al. 7

Patterns of missing data differed by resident characteristics, with older, lon- ger-stay, and more cognitively impaired residents being less likely to have a complete survey. We addressed missing values by using the multiple imputa- tion approach in Stata via the “mi” procedure (Rubin, 1996; StataCorp LP, 2011). Findings were robust to alternate strategies of handling missing data, including list-wide deletion.

Resident clinical data were drawn from the mandatory MDS 2.0 data, for all NH residents with a QOL report in 2010. Most independent variables had little missing data (less than 6%). Facility-level characteristics come from facility reports to the DHS. Specific facility characteristics are described below. We had no missing data on facility characteristics. For 2010, our full models used 10,969 resident surveys (385 non-White).

Measurement of QOL

Our measure of QOL consisted of six QOL domains: environment, personal attention, food enjoyment, engagement, negative mood, and positive mood, plus a summary score, which includes all domains. This work is based on the original conceptual work by Kane et al. (2003) with updated factor analyses to better match the Minnesota sample and revised survey instrument (Shippee et al., 2013). Prior work has found that mood is strongly associated with QOL (Kane et al., 2003) and positive and negative mood load in factor analyses as two independent domains (Shippee et al., 2013).

The environment domain includes four items addressing ease of navi- gating one’s room, arrangement of personal items, and the ability to take care of one’s possessions. Personal attention includes six items asking residents whether staff treat them politely and with respect, whether they are handled gently, whether they can get help when needed, and whether they feel listened to. Food enjoyment includes three items asking residents whether they enjoy the food and mealtimes and whether or not their favor- ite foods are served in the facility. Engagement includes nine items asking whether there are activities that the resident enjoys, whether staff members know what the resident likes, whether they feel known as a person by staff and other residents, and whether they would consider any staff or other residents to be friends. Negative mood includes six items asking residents how often in the past 2 weeks they have been bored, angry, worried, sad, afraid, or lonely. Negative mood scores are re-scaled so that higher scores indicate less negative mood. Finally, positive mood includes three items asking residents how often in the past 2 weeks they have felt peaceful, interested in things, and happy. In factor analyses, all domains are loaded with alpha scores >.60.

8 Journal of Aging and Health

Resident-Level Variables

Our selection of resident-level characteristics includes demographic and health characteristics that have been shown to affect NH residents’ QOL (Shippee et al., 2013).

Our main sociodemographic variable of interest is race/ethnicity, mea- sured as White versus non-White. Although it would be ideal to examine QOL for each racial/ethnic group separately, our data were limited by the number of non-White NH residents (only 3% of residents were non-White). Of those who were non-White, 55% were Black, 24% were Native American, 11% were Hispanic, and 10% were Asian. Other sociodemographic charac- teristics include age, gender, educational attainment (high school degree or more vs. less than high school), marital status (married vs. widowed/divorced/ never married), living arrangement prior to entering the facility (lived alone, lived with others, transferred from another facility), length of stay (in years), and payment source on admission (Medicaid, Medicare, and private insur- ance/self-family pay).

Health characteristics include difficulties with activities of daily living (scored from 0 to 28; high score indicates more impairment; Morris, Fries, & Morris, 1999), Alzheimer’s disease, mood/anxiety disorder, count of chronic conditions (scored from 0 to 4 or more, includes cancer, Parkinson’s disease, multiple sclerosis, stroke, arthritis, diabetes mellitus, and hip fracture), and cognitive status (1 = better cognitive performance, corresponding to score of 0-3 on the original measure vs. 0 = score of 4-6 corresponding to higher cog- nitive impairment; Morris et al., 1994).

Facility-Level Variables

In addition to the percentage of minority residents, facility-level characteris- tics included structural features that may affect QOL (Shippee, Hwanhee, Henning-Smith, & Kane, 2014). We included physical environment (size and structure) and organizational factors of the service system (care delivery and staffing) in our models (Lucas et al., 2007). Physical environment characteris- tics included location (rural, metropolitan, and micropolitan), resident acuity level (an aggregate measure of resident clinical severity, derived from the Minnesota Case Mix Classification Index, a score based on Resource Utilization Groups, which are determined by MDS items and have been shown to predict utilization among NH residents), percentage of private rooms, size (fewer than 75 beds vs. 75 beds or more), and ownership (for-profit, non- profit, and government). Organizational characteristics consisted of care delivery and staffing measures. These include staff retention (percentage of

Shippee et al. 9

direct care staff not leaving each year) and hours of care per resident day by different staff specialties (e.g., activity staff, licensed social workers, Certified Nursing Assistants (CNAs), Registered Nureses (RNs), and Licensed Practical Nurses (LPNs)). To see the impact of QOC on QOL, we include a quality improvement score (a measure of performance by the Centers for Medicare and Medicaid (CMS) based on health inspections, QOC rating, and staffing combined into an overall rating on a 1 to 5 scale). Finally, we included a mea- sure of participation in Minnesota’s Performance-Based Incentive Payment Program (PIPP), which allows facilities to apply for funding for self-initiated quality improvement projects. Participation in PIPP may indicate a facility’s motivation to improve QOL for residents, as well as their organizational abil- ity to do so.

Analytic Plan

The analysis was divided into three main stages. First, we examined racial differences for each QOL domain (on the bivariate level). We estimated racial differences using Pearson’s chi-square test of significance. Second, to evalu- ate the relationship between individual race/ethnicity and each QOL domain on the resident level, we used hierarchical linear models (HLMs), adjusted for relevant covariates. Because residents are nested within NHs, resident characteristics may be correlated with NH characteristics. HLM was suitable because it accounts for “within-group” correlation and provides better infer- ences (Raudenbush & Bryk, 2002). Third, to investigate racial differences in QOL on the facility level, we examined the role of racial composition (per- centage of non-White residents in each NH) for aggregate facility-level QOL using ordinary least squares (OLS) regression, with QOL scores and other characteristics aggregated at the mean for each facility. Finally, we used HLM to estimate the relationship between percentage of White (on the facil- ity level) and individual resident-reported QOL, controlling for resident- and facility-level covariates. We used “xt” family of commands in Stata v.12 for multivariate HLM analyses.

Results

Table 1 presents descriptive statistics and bivariate tests of racial differences. At the bivariate level, non-White NH residents rated QOL lower in all domains but environment and the summary score. A number of significant differences between White and non-White NH residents emerged. The non- White NH residents were approximately 14 years younger than their White counterparts. Non-White residents were also less likely to be female, tended

10 Journal of Aging and Health

Table 1. Descriptive Statistics for Dependent and Independent Variables.

Range White Non-White

Quality of life domains Environment 0-4 3.13 3.28 Personal attention 0-6 5.27 5.01*** Food 0-3 2.32 2.07*** Engagement 0-9 6.37 6.01* Negative mood 0-6 4.27 4.16* Positive mood 0-3 2.30 2.20* Summary score 0-31 23.68 22.73 Resident-level characteristics Race/ethnicity Non-Hispanic White 0-1 1.00 0.00*** Non-Hispanic Black/African American 0-1 0.00 0.55*** Native American 0-1 0.00 0.24*** Asian 0-1 0.00 0.10*** Hispanic 0-1 0.00 0.11*** Age 21-111 84.38 70.82*** Female 0-1 0.70 0.53*** Married 0-1 0.21 0.12*** High school education 0-1 0.66 0.64 Prior living arrangement Lived alone 0-1 0.39 0.28*** Transferred from another facility 0-1 0.26 0.24*** Activities of daily living impairments 0-28 14.08 12.23*** Alzheimer’s disease 0-1 0.12 0.06*** Anxiety/mood disorders 0-3 0.65 0.62 Count of chronic conditions 0-4 1.10 1.19* Good cognitive performance (vs. impaired) 0-1 0.88 0.88 Length of stay (years) 0-46 3.06 3.81*** Insurance coverage Private only 0-1 0.14 0.05*** Medicaid 0-1 0.43 0.85*** Medicare 0-1 0.38 0.08*** Dual 0-1 0.05 0.02*** Facility-level characteristics Metropolitan status Rural 0-1 0.28 0.07*** Metro 0-1 0.52 0.82*** Micro 0-1 0.20 0.12***

(continued)

Shippee et al. 11

to have longer lengths of stay, and were much more likely to be covered by Medicaid, compared with White NH residents. White NH residents had more functional dependency and higher rates of Alzheimer’s than their non-White counterparts (opposite of our expected findings based on CI theory). Non- White NH residents lived in larger facilities, which had fewer private rooms compared with those occupied by White NH residents. Non-White NH resi- dents were more likely to reside in for-profit NHs than White NH residents, and most non-White NH residents were in facilities located in urban areas, compared with White NH residents.

To evaluate the association between individual race and individual QOL, we estimated HLMs predicting each individual QOL domain, plus the sum- mary score, controlling for resident and facility characteristics, with residents nested within facilities (Table 2).

We found that individual resident characteristics explained most of the variability in QOL, and thus showed the effect of resident-related covariates only (facility-related predictors accounted for only 3% of the variability in

Range White Non-White

Acuity level (case-mix) 0.63-1.41 1.05 1.03*** Percent private rooms 0-100 39.09 29.42*** Active bed count 15-397 93.41 116.41*** Ownership For-profit 0-1 0.26 0.43*** Non-profit 0-1 0.64 0.54*** Government 0-1 0.10 0.03*** Staff retention 0.28-0.97 0.74 0.78*** Staff direct care hours (per resident day) Activities staff 0-0.62 0.24 0.20*** CNAs 0-4.23 2.37 2.10*** Licensed mental health/social workers 0-1.22 0.11 0.14*** LPNs 0.13-1.63 0.72 0.75*** RNs 0-1.53 0.43 0.49*** Quality improvement score 1-5 3.03 2.92* Participation in PIPP 0-1 0.19 0.18 n 10,538 385

Note. Mean for continuous variables and proportion for categorical variables. PIPP = Performance-Based Incentive Payment Program. Differences between White and non-White significant at *p < .05. **p < .01. ***p < .001.

Table 1. (continued)

T a b

le 2

. R

eg re

ss io

n C

o ef

fic ie

nt s

Fr o m

H ie

ra rc

hi ca

l L in

ea r

M o de

ls P

re di

ct in

g Q

ua lit

y o f Li

fe D

o m

ai ns

o n

th e

R es

id en

t Le

ve l.

R es

id en

t le

ve l

En vi

ro nm

en t

Pe rs

o na

l at

te nt

io n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

b b

W hi

te a

− 0.

04 0.

07 0.

12 *

0. 21

− 0.

08 0.

02 0.

24 A

ge −

0. 00

3* *

0. 00

4* **

− 0.

01 **

* −

0. 02

** *

− 0.

01 **

* −

0. 04

** *

Fe m

al e

− 0.

05 0.

07 **

0. 02

0. 19

** *

0. 00

0. 03

0. 26

* M

ar ri

ed −

0. 02

0. 08

* 0.

00 0.

19 **

0. 13

* 0.

08 **

0. 45

** H

ig h

sc ho

o l e

du ca

ti o nb

0. 02

0. 07

* −

0. 03

− 0.

10 *

0. 05

0. 04

0. 06

Li ve

d al

o ne

( re

f: no

Y es

0. 02

0. 06

* 0.

01 0.

10 0.

11 *

0. 07

* 0.

36 *

th er

f ac

ili ty

0. 02

0. 00

− 0.

03 −

0. 03

0. 14

** 0.

09 **

0. 19

A ct

iv it ie

s o f da

ily li

vi ng

c −

0. 06

** *

− 0.

01 **

* −

0. 00

3* −

0. 01

** −

0. 01

** *

− 0.

01 **

* −

0. 10

** *

A lz

he im

er ’s

d is

ea se

0. 26

** *

− 0.

02 0.

11 **

* −

0. 29

** *

− 0.

22 **

* −

0. 14

** *

− 0.

30 A

nx ie

ty /m

o o d

di so

rd er

− 0.

06 **

* −

0. 07

** *

− 0.

06 **

* −

0. 16

** *

− 0.

05 *

− 0.

02 −

0. 43

** *

C o un

t o f co

nd it io

ns d

− 0.

05 **

* −

0. 01

− 0.

01 0.

06 *

0. 02

0. 03

Le ss

c o gn

it iv

el y

im pa

ir ed

− 0.

06 0.

20 **

* −

0. 16

** *

0. 12

0. 23

** *

0. 16

** *

0. 51

** Le

ng th

o f st

ay (

ye ar

s) 0.

01 **

* 0.

00 0.

01 **

0. 03

** *

− 0.

07 **

* −

0. 04

** *

− 0.

05 **

Pa yo

r (r

ef : p

ri va

te o

nl y)

ed ic

ai d

0. 02

− 0.

04 −

0. 04

− 0.

05 0.

04 0.

01 −

0. 06

ed ic

ar e

0. 07

0. 00

0. 07

0. 09

0. 05

0. 35

ua l

0. 07

− 0.

05 0.

01 −

0. 07

0. 02

0. 00

− 0.

(c on

tin ue

R es

id en

t le

ve l

En vi

ro nm

en t

Pe rs

o na

l at

te nt

io n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

b b

n 10

,7 06

10 ,7

06 10

,7 06

10 ,7

06 10

,7 06

10 ,7

06 10

,7 06

F- st

at is

ti c

51 .7

2* **

8. 12

** *

6. 93

** *

9. 18

** *

12 .0

0* **

12 .6

9* **

14 .1

6* **

ψ 1

0. 01

0. 03

0. 02

0. 09

0. 08

0. 02

0. 93

ψ 2

1. 23

1. 36

0. 84

4. 85

3. 79

1. 15

31 .5

6 In

te rc

ep t

3. 98

** *

4. 32

** *

2. 27

** *

7. 16

** *

2. 02

** *

1. 07

** *

20 .9

3* **

IC C

( R

es id

en t

le ve

l) .9

9* **

.9 8*

** .9

8* **

.9 8*

** .9

8* **

.9 8*

** .9

7* **

N ot

e. M

o de

ls a

re a

dj us

te d

fo r

th e

fo llo

w in

g fa

ci lit

y ch

ar ac

te ri

st ic

s: lo

ca ti o n,

a cu

it y,

p er

ce nt

ag e

o f pr

iv at

e ro

o m

s, s

iz e,

o w

ne rs

hi p,

p ay

er s

o ur

ce , s

ta ff

re te

nt io

n, s

ta ff

di re

ct c

ar e

ho ur

s, q

ua lit

y im

pr o ve

m en

t sc

o re

, a nd

p ar

ti ci

pa ti o n

in P

IP P.

P IP

P =

P er

fo rm

an ce

-B as

ed I nc

en ti ve

P ay

m en

t Pr

o gr

am ; ψ

1 =

e in

te rc

ep t

va ri

an ce

; ψ 2

= t

he r

es id

ua l v

ar ia

nc e;

I C

C =

in tr

ac la

ss c

o rr

el at

io n

co ef

fic ie

nt .

a N o n-

W hi

te in

cl ud

ed N

at iv

e A

m er

ic an

, A si

an /P

ac ifi

c Is

la nd

er , B

la ck

, a nd

H is

pa ni

c/ La

ti no

. b C

o m

pa re

d w

it h

le ss

t ha

n a

hi gh

s ch

o o l e

du ca

ti o n.

c M ea

su re

d o n

a sc

al e

o f 0-

28 ; h

ig he

r in

di ca

te s

gr ea

te r

im pa

ir m

en t.

d M ea

su re

d o n

a sc

al e

o f 0-

4; h

ig he

r in

di ca

te s

m o re

c hr

o ni

c co

nd it io

ns .

*p <

.0 5.

* *p

< .0

1. *

** p

< .0

01 .

T a b

le 2

. (c

o n

ti n

u e d

)

14 Journal of Aging and Health

individual QOL; analyses available upon request). After accounting for resi- dent status and health characteristics, racial differences in QOL became non- significant except for food enjoyment, with White residents reporting higher food enjoyment scores than their non-White counterparts. Age, Activities of Daily Living (ADL) limitations, Alzheimer’s disease, diagnosis of anxiety/ mood disorders, cognitive limitations, and length of stay were all significant predictors of QOL (and many of these were significantly different for White vs. non-White residents as shown in bivariate analyses).

In Table 3, we examine the role of facility racial make-up on facility- average QOL scores, controlling for resident case-mix and facility-level covariates. Having a higher percentage of White residents on the facility level was associated with better average scores for personal attention, food enjoy- ment, social engagement, and the summary score. There was no association between percentage of White and the environment or mood domains on the facility level. No other resident-related characteristic (aggregated to the facil- ity level) was significant for the aggregate QOL scores. Of facility character- istics, direct care hours per resident day, especially activity staff hours, were most consistently associated with QOL.

Finally, to understand how facility racial composition may impact indi- vidual ratings of QOL, we estimated HLM models with percentage of White as the main predictor of resident QOL. We adjusted for resident and facility characteristics as shown in Table 4 (with percentage of White operationalized as a facility-level predictor). Percentage of White was significantly associ- ated with all but the environment domain. Even after accounting for resident and facility characteristics, having a higher percentage of White residents on the facility level was associated with five out of six domains: better QOL for personal attention, food, engagement, mood (less negative and more posi- tive), and the summary score. Among facility-level covariates, direct care hours per resident day, especially among activities staff, were consistently associated with QOL.

In sensitivity analyses, we investigated selected characteristics of the facilities with the highest proportion of non-White NH residents (n = 9). The proportion of non-White residents in these facilities ranged from 96% to 23%, compared with an overall mean of 3% across all facilities in our sample. All residents in all nine of these facilities had Medicaid as their payment source on admission (compared with 47% of the total sample).. In this small sample of facilities, residents tended to be younger (69 vs. 84 years), had lower ADL impairment scores (10.5 vs. 14), and had longer lengths of stay (4.21 vs. 3.08 years), on average. We also examined the characteristics of the communities in which these facilities were located, using U.S. Census data at the zip-code level. Compared with Minnesota as a whole, the median income

T a b

le 3

. R

eg re

ss io

n C

o ef

fic ie

nt s

Fr o m

O rd

in ar

y Le

as t

Sq ua

re s

R eg

re ss

io n

Pr ed

ic ti ng

Q ua

lit y

o f Li

fe o

n th

e Fa

ci lit

y Le

ve l b

y Pe

rc en

t W

hi te

En vi

ro nm

en t

Pe rs

o na

l A

tt en

ti o n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

b b

Pe rc

en t

W hi

te −

0. 34

0. 71

** 0.

33 **

1. 14

** 1.

44 0.

66 1.

07 **

* Fa

ci lit

y- re

la te

d co

va ri

at es

ca ti o n

(r ef

: r ur

al )

M et

ro 0.

02 −

0. 01

− 0.

03 −

0. 12

− 0.

32 −

0. 04

− 0.

ic ro

− 0.

07 0.

02 −

0. 04

0. 09

− 0.

35 −

0. 03

− 0.

A cu

it y

− 0.

44 *

− 0.

11 −

0. 11

− 0.

38 −

2. 24

* −

0. 77

− 0.

Pe rc

en t

pr iv

at e

ro o m

s 0.

00 0.

00 4*

* 0.

00 2*

rg e

si ze

( >

75 b

ed s)

− 0.

05 −

0. 04

− 0.

02 −

0. 17

* 0.

11 −

0. 08

− 0.

O w

ne rs

hi p

(r ef

: f o r-

pr o fit

)

o n-

pr o fit

0. 05

0. 02

0. 00

0. 04

0. 25

− 0.

17 0.

o ve

rn m

en t

0. 05

0. 04

0. 01

0. 08

0. 15

− 0.

18 0.

Pa ye

r so

ur ce

M ed

ic ai

d 0.

05 −

0. 04

− 0.

04 −

0. 03

− 0.

28 −

0. 12

− 0.

ed ic

ar e

0. 08

* 0.

03 0.

00 −

0. 03

− 0.

06 −

0. 09

0. 00

af f re

te nt

io n

0. 01

− 0.

04 0.

04 −

0. 31

− 0.

90 0.

13 0.

D ir

ec t

ca re

h o ur

s pe

r re

si de

nt d

ct iv

it ie

s st

af f

0. 28

0. 74

** *

0. 28

** 1.

67 **

* 2.

16 *

0. 56

1. 22

** *

C N

A s

− 0.

13 **

0. 06

0. 00

0. 03

0. 08

− 0.

ce ns

ed m

en ta

l h ea

lt h/

so ci

w o rk

er s

0. 04

0. 33

0. 06

1. 00

* −

0. 59

0. 55

0. 24

LP N

s −

0. 08

− 0.

01 −

0. 09

− 0.

19 −

0. 03

0. 01

R N

s 0.

10 0.

11 0.

04 −

0. 13

− 0.

25 0.

43 *

0. 18

(c on

tin ue

En vi

ro nm

en t

Pe rs

o na

l A

tt en

ti o n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

b b

ua lit

y im

pr o ve

m en

t sc

o re

0. 03

0. 05

** 0.

01 0.

07 *

0. 09

0. 08

* 0.

05 *

rt ic

ip at

io n

in P

IP P

0. 03

0. 04

0. 00

0. 02

0. 05

0. 02

0. 05

R es

id en

t- re

la te

d co

va ri

at es

ge 0.

00 0.

01 0.

00 0.

Fe m

al e

− 0.

01 0.

04 0.

02 −

0. 05

− 0.

08 0.

13 0.

M ar

ri ed

0. 02

0. 09

0. 05

0. 01

0. 05

0. 09

0. 07

ig h

sc ho

o l e

du ca

ti o n

o r

hi gh

er 0.

03 −

0. 03

− 0.

03 −

0. 05

0. 02

0. 06

0. 00

io r

liv in

g ar

ra ng

em en

t (r

ef : W

it h

o th

er s)

Li ve

d al

o ne

− 0.

02 −

0. 02

0. 00

0. 02

− 0.

04 0.

03 −

0. 01

T ra

ns fe

rr ed

f ro

m a

no th

er f ac

ili ty

− 0.

04 −

0. 04

0. 01

− 0.

15 −

0. 09

− 0.

06 −

0. 05

ea n

ac ti vi

ti es

o f da

ily li

vi ng

0. 00

− 0.

01 0.

00 0.

A lz

he im

er ’s

d is

ea se

0. 04

0. 00

− 0.

01 0.

07 −

0. 06

0. 01

0. 02

nx ie

ty d

is o rd

er s

0. 00

− 0.

01 −

0. 01

0. 02

− 0.

06 −

0. 03

− 0.

C o un

t o f ch

ro ni

c co

nd it io

ns −

0. 01

− 0.

02 0.

01 −

0. 02

− 0.

02 0.

00 0.

Po o r

co gn

it iv

e pe

rf o rm

an ce

0. 04

0. 00

0. 07

− 0.

06 0.

Le ng

th o

f st

ay (

ye ar

s) 0.

01 0.

00 0.

01 0.

00 0.

00 n

35 2

F- st

at is

ti c

2. 41

** *

4. 20

** *

2. 67

** *

4. 45

** *

2. 41

** *

1. 98

** 3.

94 **

* In

te rc

ep t

4. 03

** *

3. 82

** *

1. 32

** *

5. 43

** *

12 .4

1* **

5. 48

** *

6. 62

** *

N ot

e. R

es ul

ts a

re p

re se

nt ed

a s

co ef

fic ie

nt s

fr o m

o rd

in ar

y le

as t

sq ua

re s

re gr

es si

o n

an al

ys es

. *

p <

.0 5.

* *p

< .0

1. *

** p

< .0

01 .

T a b

le 3

. (c

o n

ti n

u e d

)

T a b

le 4

. R

eg re

ss io

n C

o ef

fic ie

nt s

Fr o m

H ie

ra rc

hi ca

l L in

ea r

M o de

l P re

di ct

in g

N ur

si ng

H o m

e R

es id

en ts

’ Q ua

lit y

o f Li

fe b

y Pe

rc en

ta ge

o f W

hi te

En vi

ro nm

en t

Pe rs

o na

l at

te nt

io n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

Pe rc

en t

W hi

te 0.

07 0.

94 **

* 0.

41 *

1. 75

** *

0. 72

* 0.

48 *

4. 29

** *

Fa ci

lit y-

re la

te d

co va

ri at

Lo ca

ti o n

(r ef

: R ur

al )

M et

ro 0.

03 −

0. 03

− 0.

04 −

0. 15

* −

0. 08

− 0.

03 −

0. 30

M ic

ro −

0. 05

0. 01

− 0.

04 0.

06 −

0. 08

− 0.

04 −

0. 13

cu it y

0. 21

0. 09

− 0.

22 −

0. 20

0. 14

0. 02

0. 03

Pe rc

en t

pr iv

at e

ro o m

s 0.

00 0.

00 2*

0. 00

1 0.

01 *

rg e

si ze

( >

75 b

ed s)

− 0.

01 −

0. 04

− 0.

03 −

0. 16

** −

0. 05

− 0.

03 −

0. 34

O w

ne rs

hi p

(r ef

: f o r-

pr o fit

)

o n-

pr o fit

0. 08

0. 03

− 0.

01 0.

06 0.

13 0.

06 0.

o ve

rn m

en t

0. 06

0. 05

0. 02

0. 13

0. 16

0. 07

0. 48

ye r

so ur

ed ic

ai d

− 0.

07 −

0. 43

* 0.

20 −

0. 53

− 0.

04 −

0. 01

− 0.

ed ic

ar e

0. 15

− 0.

08 0.

26 *

− 0.

18 0.

16 0.

09 0.

St af

f re

te nt

io n

− 0.

10 −

0. 05

− 0.

03 −

0. 27

− 0.

08 −

0. 06

− 0.

D ir

ec t

ca re

h o ur

s pe

r re

si de

nt d

ct iv

it ie

s st

af f

0. 46

** 0.

69 **

* 0.

62 **

* 1.

61 **

* 1.

19 **

* 0.

63 **

* 5.

20 **

N A

s −

0. 05

0. 04

− 0.

01 0.

01 −

0. 08

− 0.

04 −

0. 11

Li ce

ns ed

m en

ta l h

ea lt h/

so ci

al w

o rk

er s

− 0.

13 0.

54 **

0. 30

0. 82

* −

0. 20

− 0.

06 1.

N s

− 0.

15 *

0. 04

− 0.

09 −

0. 19

− 0.

28 *

− 0.

14 *

− 0.

79 *

R N

s 0.

01 0.

01 −

0. 08

− 0.

22 −

0. 01

0. 02

− 0.

Q ua

lit y

im pr

o ve

m en

t sc

o re

0. 02

0. 05

** 0.

02 0.

04 0.

06 *

0. 03

* 0.

23 **

rt ic

ip at

io n

in P

IP P

0. 07

* 0.

03 0.

01 0.

04 0.

02 0.

(c on

tin ue

En vi

ro nm

en t

Pe rs

o na

l at

te nt

io n

Fo o d

En ga

ge m

en t

N eg

at iv

e m

o o d

Po si

ti ve

o o d

Su m

m ar

y sc

o re

R es

id en

t- re

la te

d co

va ri

at es

ge −

0. 00

4* *

0. 00

4* **

− 0.

02 **

* −

0. 02

** *

− 0.

01 **

* −

0. 04

** *

m al

e −

0. 05

0. 07

* 0.

02 0.

18 **

* −

0. 01

0. 02

0. 23

ar ri

ed −

0. 02

0. 07

* 0.

00 0.

18 **

0. 12

* 0.

07 *

0. 43

H ig

h sc

ho o l e

du ca

ti o n

o r

hi gh

er 0.

01 0.

06 *

− 0.

03 −

0. 10

* 0.

05 0.

04 0.

Pr io

r liv

in g

ar ra

ng em

en t

(r ef

: W it h

o th

er s)

Li ve

d al

o ne

0. 01

0. 06

* 0.

01 0.

10 0.

11 *

0. 07

* 0.

36 *

T ra

ns fe

rr ed

f ro

m a

no th

er f ac

ili ty

0. 02

0. 00

− 0.

03 −

0. 03

0. 14

** 0.

09 **

0. 19

6 M

ea n

ac ti vi

ti es

o f da

ily li

vi ng

− 0.

06 **

* −

0. 01

** *

− 0.

00 3*

− 0.

01 **

− 0.

01 **

* −

0. 01

** *

− 0.

10 **

* A

lz he

im er

’s d

is ea

se 0.

26 **

* −

0. 02

0. 11

** *

− 0.

29 **

* −

0. 22

** *

− 0.

14 **

* −

0. 30

A nx

ie ty

d is

o rd

er s

− 0.

06 **

* −

0. 07

** *

− 0.

06 **

* −

0. 15

** *

− 0.

05 *

− 0.

02 −

0. 43

** *

C o un

t o f ch

ro ni

c co

nd it io

ns −

0. 05

** *

− 0.

01 −

0. 02

0. 06

* 0.

02 0.

02 Po

o r

co gn

it iv

e pe

rf o rm

an ce

− 0.

06 0.

20 **

* −

0. 16

** *

0. 13

0. 22

** *

0. 15

** *

0. 50

** Le

ng th

o f st

ay (

ye ar

s) 0.

01 **

* 0.

00 0.

01 **

0. 03

** *

− 0.

06 **

* −

0. 04

** *

− 0.

05 **

n 10

,6 96

10 ,6

96 10

,6 96

10 ,6

96 10

,6 96

10 ,6

96 10

,6 96

ψ 1

0. 01

0. 03

0. 02

0. 08

0. 07

0. 02

0. 82

ψ 2

1. 24

1. 36

0. 84

4. 56

3. 79

1. 15

31 .5

9 F-

st at

is ti c

53 .2

0* **

9. 64

** *

7. 31

** *

10 .5

2* **

12 .6

1* **

13 .3

9* **

15 .4

3* **

IC C

.0 1*

* .0

2* **

.0 2*

** .0

2* **

.0 2*

** .0

2* **

.0 3*

** In

te rc

ep t

4. 01

** *

3. 79

** *

1. 79

** *

6. 12

** *

1. 62

* 0.

81 *

18 .2

1* **

N ot

e. R

es ul

ts a

re p

re se

nt ed

a s

co ef

fic ie

nt s.

ψ 1

= t

he in

te rc

ep t

va ri

an ce

; ψ 2

= t

he r

es id

ua l v

ar ia

nc e;

I C

C =

in tr

ac la

ss c

o rr

el at

io n

co ef

fic ie

nt .

*p <

.0 5.

* *p

< .0

1. *

** p

< .0

01 .

T a b

le 4

. (c

o n

ti n

u e d

)

Shippee et al. 19

for the communities of these facilities was an average of US$47,017 (vs. US$59,126 statewide), the poverty rate was 24% (vs. 11%), and 45% of the zip-code residents were non-White (vs. 15% statewide). We carried out sen- sitivity analyses to examine QOL by race in these nine facilities and found that often non-White residents had higher QOL scores than their White coun- terparts but the findings varied by domain and by facility. However, with the sample of nine, we cannot draw any statistically relevant conclusions. (Full results available upon request.)

Discussion

We sought to elucidate the role of race/ethnicity for NH QOL by examining not only the individual-level disparities but also structural factors. The analy- sis revealed two main findings. First, on the individual level, most of these differences between non-White and White NH residents’ QOL became non- significant after we accounted for other resident and facility characteristics, indicating that the difference was not due to race but due to resident case-mix (failing to support H1). Second, NHs with a higher mix of non-White resi- dents had lower average QOL scores compared with NHs with a higher mix of White residents (full support of H2). This finding suggests that most of the racial/ethnic disparity in QOL reflects how care is delivered in facilities that house a higher proportion of non-White residents.

Our finding that most of the individual-level racial differences in QOL were explained by resident characteristics (except food) points to the impor- tance of the economic and health resources and risks over the life course as stated by CI theory (Ferraro & Shippee, 2009). Previous work on racial dif- ferences in QOC finds racial disparities in numerous outcomes, including diabetes (Allsworth et al., 2005), drug management to prevent secondary stroke (Allsworth et al., 2005; Christian et al., 2003; Hughes & Lapane, 2004), and overall QOC, even when controlling for severity of conditions and socioeconomic status (Kang, 2011). However, our study shows that much of the racial disparity in QOL on the individual level can be explained by resi- dent case-mix. Our non-White sample differed from White residents in a number of ways; they were younger, had fewer ADL limitations, had longer stays, and had Medicaid as primary form of payment. Others have shown that Black older adults rely more on Medicaid as their main source of payment compared with White older adults, as well as a higher likelihood of receiving lower QOC (Konetzka & Werner, 2009). Yet, unlike other studies, our sample of non-White NH residents was not more disabled than their White counter- parts at the time of admission (Wallace, Levy-Storms, Kington, & Andersen, 1998).

20 Journal of Aging and Health

When we examined the impact of facility racial composition on facility- average QOL scores, we found that NHs with a higher mix of non-White resi- dents had lower average QOL scores for three out of six domains (personal attention, food, and engagement) and lower summary scores, controlling for resident case-mix and other resident and facility characteristics. These find- ings provide partial support for previous literature on racial disparities in NH QOC showing that many of the disparities are reflected on the facility, not individual levels (Fennell et al., 2000). The findings also provide partial sup- port for CI theory which calls attention to the role of institutions as a feature of structural inequality (Ferraro & Shippee, 2009).

Finally, we also tested the relationship between facility racial composition and resident-reported QOL. Previous literature on QOC has linked racial composition to lower NH care (Mor et al., 2004) and poorer hospital out- comes (Jha et al., 2005). However, prior studies have not addressed the role of racial composition on individual QOL. Building from the literature on school performance and individual outcomes (e.g., Rosenberg & Simmons, 1971), we found that having a higher percentage of White residents on the facility level was associated with better QOL for five out of six domains (except environment), and the summary score. These findings show further support to the role of institutional structures for individual outcomes as stated by CI theory (Ferraro & Shippee, 2009).

Our findings reveal a complex relationship between race/ethnicity and NH resident QOL. Confounding factors are tricky to tease apart, clearly, such as socioeconomic status, health burden, and forms of payment, all of which are known to influence QOL. Our findings may simply suggest racial/ethnic dif- ferences in QOL. Yet, alternative explanations exist for the association with facility characteristics. The residents using the predominantly non-White facilities are distinctly different (e.g., younger, longer Length of Stay (LOS)). Hence, facilities that serve a higher proportion of non-White older adults may provide a different package of services (or provide the typical package that is less relevant) to meet these different needs; then, percentage of non-White may be a proxy for other factors. Others have found that facilities serving a higher proportion of non-White residents are more likely to deliver inferior care (Feng et al., 2011b), but the nine NHs in our sample with the highest proportion of non-White residents had a range of QOC scores that suggest a more complex story (i.e., these nine NHs were not uniformly low-quality providers but other attributes affected QOL). The most consistent finding about these particular NHs was that all the residents relied on Medicaid (compared with 47% of the total sample). In most instances, we might expect that Medicaid payment would be a major factor because it pays below pri- vate-pay rates (Grabowski, Angelelli, & Mor, 2004), but Minnesota uses rate

Shippee et al. 21

equalization, which prohibits NHs from charging private-pay patients more than Medicaid patients for similar services. However, NHs can charge pri- vate-pay patients a higher rate for a private room and other additional ser- vices, still allowing for some differences. Thus, NHs with less private-pay or Medicare patients may have fewer resources to provide adequate staffing or high quality care compared with those who primarily rely on Medicaid. In addition, these nine NHs tended to be located in poorer communities within metro areas, where non-White individuals are more likely to live. This find- ing provides some support to the role of geographical segregation as the rea- son for racial differences, claimed by some as the driver of racial inequality in NH care (Smith et al., 2008), although it may also reflect institutional and interpersonal bias not captured by our data (Mullings & Schulz, 2006).

Racial differences in QOL were less likely to be associated with individual characteristics and more likely to reflect overall facility quality. This indi- cates a need to focus on structural facility-level characteristics in addressing QOL and reducing disparities for racial/ethnic minorities. Our results high- light several possible areas for intervention. Specifically, programs like Minnesota’s PIPP program and quality improvement measurement appear to have positive impacts on QOL (Shippee et al., 2014). Other states might con- sider adopting similar programs to reduce disparities between facilities, but such programs should be tailored to the needs of minority populations when setting incentives (Weissman et al., 2012). Our findings about staff hours have particular policy relevance. Activity staff hours per resident day (which varied significantly by race), in particular, have a strong positive association with QOL (Shippee et al., 2013). Increasing funding for activity staff and programming may improve the QOL of all residents.

This study has limitations. This study was conducted in only one state. Minnesota is atypical in several regards. (a) It is one of only a few states to implement resident-reported QOL measures on all Medicaid-certified NHs in the state (over 95% of all Minnesota NHs); thus, national data using these mea- sures are not available. (b) The racial composition is different from many other states as Minnesota has a low non-White population of older adults (although this trend is starting to change). (c) Minnesota has adopted rate equalization, which prohibits NHs from charging private-pay patients more than Medicaid patients for similar services (Punelli, 2013). Minnesota has one of the highest rates of private rooms in NHs in the country; it is also rated highly in the state scorecard on LTC services and supports, holding the first place among all states for overall LTC and QOC and QOL (AARP, 2014). Thus, findings from Minnesota may not be generalizable to other, more diverse states with less pro- gressive policies. Our estimates are likely conservative as the levels of racial disparities in NH care are likely to be much higher in other states.

22 Journal of Aging and Health

We examine data from only 2010, the most recent year of data available. Although it would be useful to longitudinally examine racial differences in QOL over time, our data are collected from a random sample of NH residents each year, thus permitting longitudinal analyses only on a facility level. We are also limited by the racial composition of our sample. With 3.5% of our sample (approximately 385 residents) identifying as non-White, we lacked power to analyze differences by racial group. However, while our White ver- sus non-White dichotomy is a blunt measure of race, it is an important step toward understanding racial disparities in NH QOL, and sample sizes of vari- ous racial groups will increase as NHs become increasingly diverse. Furthermore, while we control for several observable characteristics that are correlated with both race and QOL (e.g., age, marital status, Alzheimer’s dis- ease, health status, and length of stay), there are likely unobservable charac- teristics that are correlated with both race and QOL. For instance, we do not have a measure of racial discrimination or residential segregation, which may be associated with QOL for non-White NH residents. Results from our exam- ination of the nine NHs with the highest proportion of non-White residents indicated that they were located within segregated neighborhoods. However, we do not have information on where residents move from. Prior research indicates that residents often choose NHs based on proximity to home or family (Bell, 1996), which would mean that individuals who select into more segregated NHs likely come from more segregated neighborhoods. Future research should attempt to investigate the role of life-course exposure to dis- crimination and segregation on NH resident QOL.

In spite of these limitations, our study makes an important first step toward understanding QOL for racial/ethnic minority groups living in NHs. We make novel use of this data to identify how race is associated with QOL on the individual and facility level and are also able to identify resident- and facility-level predictors of QOL. The facility-level character- istics, in particular, offer insight into policy and programmatic interven- tions that might lessen racial disparities in NH QOL. In particular, attention should be paid to structural characteristics that limit NH options for racial/ ethnic minority populations, such as geographic location, staffing, and payment structure. Future research should examine how QOL differs by specific racial/ethnic group, especially as NH settings continue to become more racially diverse.

Authors’ Note

The content is solely the responsibility of the authors and does not necessarily repre- sent the official views of the National Institutes of Health.

Shippee et al. 23

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Support for this research was provided to the first author by the Fesler-Lampert Chair on Aging, University of Minnesota Center on Aging, and a grant from the National Center for Advancing Translational Sciences of the National Institutes of Health (Award Number 1KL2RR033182-03). Carrie Henning-Smith was supported by the Hearst Fellowship in Public Health and Aging.

References

American Association of Retired Persons (AARP). (2014). Minnesota: 2013 state long-term services and supports scorecard results. Retrieved from http:// www.longtermscorecard.org/~/media/Microsite/State%20Fact%20Sheets/ Minnesota%20Fact%20Sheet.pdf

Agency for Healthcare Research and Quality. (2000). Has the gender or racial com- position of nursing homes changed? Retrieved from http://archive.ahrq.gov/ news/press/pr2000/nurshmpr.htm

Allsworth, J. E., Toppa, R., Palin, N. C., & Lapane, K. L. (2005). Racial and ethnic disparities in the pharmacologic management of diabetes mellitus among long- term care facility residents. Ethnicity & Disease, 15(2), 205-212.

Bell, J. (1996). Decision making in nursing home placement. Journal of Women & Aging, 8(1), 45-60.

Boulware, L. E., Cooper, L. A., Ratner, L. E., LaVeist, T. A., & Powe, N. R. (2003). Race and trust in the health care system. Public Health Reports, 118, 358-365.

Cagney, K. A., & Agree, E. M. (2005). Racial differences in formal LTC: Does the timing of parenthood play a role? The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 60, 137-145.

Cagney, K. A., Browning, C. R., & Wen, M. (2005). Racial disparities in self- rated health at older ages: What difference does the neighborhood make? The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 60, S181-S190.

Cai, S., Feng, Z., Fennell, M. L., & Mor, V. (2011). Despite small improvement, black nursing home residents remain less likely than whites to receive flu vac- cine. Health Affairs, 30, 1939-1946.

Cassie, K. M., & Cassie, W. (2013). Racial disparities in the use of physical restraints in U.S. nursing homes. Health & Social Work, 38, 207-213.

Christian, J. B., Lapane, K. L., & Toppa, R. S. (2003). Racial disparities in receipt of secondary stroke prevention agents among US nursing home residents. Stroke, 34, 2693-2697.

http://www.longtermscorecard.org/~/media/Microsite/State%20Fact%20Sheets/Minnesota%20Fact%20Sheet.pdf

http://archive.ahrq.gov/news/press/pr2000/nurshmpr.htm

24 Journal of Aging and Health

Dannefer, D. (2003). Cumulative advantage/disadvantage and the life course: Cross- fertilizing age and social science theory. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 58, S327-S337.

DuBeau, C. E., Simon, S. E., & Morris, J. N. (2006). The effect of urinary inconti- nence on quality of life in older nursing home residents. Journal of the American Geriatric Society, 54, 1325-1333.

Elder, G. H., Jr. (1998). The life course and human development. In W. Damon & R. M. Lerner (Eds.), Theoretical Models of Human Development: Vol. 1. Handbook of child psychiatry (5th ed., pp. 939-991). New York, NY: Wiley.

Engle, V. F., Fox-Hill, E., & Graney, M. J. (1998). The experience of living-dying in a nursing home: Self-reports of black and white older adults. Journal of the American Geriatrics Society, 46, 1091-1096.

Feng, Z., Fennell, M. L., Tyler, D. A., Clark, M., & Mor, V. (2011a). The care span: Growth of racial and ethnic minorities in US nursing homes driven by demo- graphics and possible disparities in options. Health Affairs, 30, 1358-1365.

Feng, Z., Lepore, M., Clark, M. A., Tyler, D., Smith, D. B., Mor, V., & Fennell, M. L. (2011b). Geographic concentration and correlates of nursing home closures: 1999-2008. Archives of Internal Medicine, 171, 806-813.

Fennell, M. L., Miller, S. C., & Mor, V. (2000). Facility effects on racial differences in nursing home quality of care. American Journal of Medical Quality, 15, 174-181.

Ferraro, K. F., & Shippee, T. P. (2008). Black and White chains of risk for hospitaliza- tion over 20 years. Journal of Health and Social Behavior, 49, 193-207.

Ferraro, K. F., & Shippee, T. P. (2009). Aging and cumulative inequality: How does inequality get under the skin? The Gerontologist, 49, 333-343.

Ferraro, K. F., Shippee, T. P., & Schafer, M. H. (2009). Cumulative inequality theory for research on aging and the life course. In V. L. Bengtson, M. Silverstein, N. M. Putney, & D. Gans (Eds.), Handbook of theories of aging (pp. 413-434). New York, NY: Springer.

Gamble, F. N. (1997). Under the shadow of Tuskegee: African Americans and health care. American Journal of Public Health, 87, 1773-1778.

Grabowski, D. C., Angelelli, J. J., & Mor, V. (2004). Medicaid payment and risk- adjusted nursing home quality measures. Health Affairs, 23, 243-249.

Greene, J., Blustein, J., & Weitzman, B. C. (2006). Race, segregation, and physicians’ participation in Medicaid. Milbank Quarterly, 84, 239-272.

Hayanga, A. J., Kaiser, H. E., Sinha, R., Berenholtz, S. M., Makary, M., & Chang, D. (2009). Residential segregation and access to surgical care by minority pop- ulations in US counties. Journal of the American College of Surgeons, 208, 1017-1022.

Hughes, C. M., & Lapane, K. L. (2004). Factors associated with the initiation and discontinuation of secondary stroke prevention agents in nursing homes. Journal of Stroke & Cerebrovascular Diseases, 13, 164-170.

Jha, A. K., Fisher, E. S., Li, Z., Orav, E. J., & Epstein, A. M. (2005). Racial trends in the use of major procedures in the elderly. New England Journal of Medicine, 353, 683-691.

Shippee et al. 25

Jones, A., Sonnenfeld, N. L., & Harris-Kojetin, L. D. (2009). Racial differences in functioning among elderly nursing home residents, 2004. Hyattsville, MD: U.S. Department of Health and Human Service, Centers for Disease Control and Prevention, National Center for Health Statistics.

Kane, R. A. (2003). Definition, measurement, and correlates of quality of life in nursing homes: Toward a reasonable practice, research, and policy agenda. The Gerontologist, 43(Suppl. 2), 28-36.

Kane, R. A., Kling, K. C., Bershadsky, B., Kane, R. L., Giles, K., Degenholtz, H. B., . . . Cutler, L. J. (2003). Quality of life measures for nursing home residents. The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 58, 240-248.

Kang, Y. (2011). Racial disparities in nursing home quality of care: A multilevel analysis using 2004 National Nursing Home Survey Data. Dissertation Abstracts International, Section B: The Sciences and Engineering, 71(7-B), 4209.

Konetzka, R. T., & Werner, R. M. (2009). Disparities in long-term care: Building equity into market-based reforms. Medical Care Research and Review, 66, 491-521.

Latkin, C. A., & Curry, A. D. (2003). Stressful neighborhoods and depression: A prospective study of the neighborhood disorder. Journal of Health and Social Behavior, 44, 34-44.

Lucas, J. A., Levin, C. A., Lowe, T. J., Robertson, B., Akincigil, A., Sambamoorthi, U., . . . Crystal, S. (2007). The relationship between organizational factors and resident satisfaction with nursing home care and life. Journal of Aging & Social Policy, 19(2), 125-151.

Mor, V., Zinn, J., Zngelelli, J., Teno, J. M., & Miller, S. C. (2004). Driven to tiers: Socioeconomic and racial disparities in the quality of nursing home care. Milbank Quarterly, 82, 227-256.

Morris, J. N., Fries, B. E., Mehr, D. R., Hawes, C., Phillips, C., Mor, V., & Lipsitz, L. A. (1994). MDS Cognitive Performance Scale©. Journal of Gerontology, 49, M174-M182.

Morris, J. N., Fries, B. E., & Morris, S. A. (1999). Scaling ADLs within the MDS. The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 54, M546-M553.

Mui, A. C., Choi, N. G., & Monk, A. (1998). Long-term care and ethnicity. Westport, CT: Auburn House.

Mullings, L., & Schulz, A. J. (2006). Intersectionality and health: An introduction. In A. Schulz & L. Mullings (Eds.), Gender, race, class, & health (pp. 3-17). San Francisco, CA: SAGE.

O’Rand, A. M. (1996). The precious and the precocious: Understanding cumulative disadvantage and cumulative advantage over the life course. The Gerontologist, 36, 230-238.

Pearlin, L. I., Schieman, S., Fazio, E. M., & Meersman, S. C. (2005). Stress, health, and the life course: Some conceptual perspectives. Journal of Health and Social Behavior, 46, 205-219.

26 Journal of Aging and Health

Punelli, D. (2013). Nursing facility reimbursement and regulation (Information Brief). Retrieved from http://www.house.leg.state.mn.us/hrd/pubs/nfreimb.pdf

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models (2nd ed.). Thousand Oaks, CA: SAGE.

Rosenberg, M., & Simmons, S. R. (1971). Black and white self-esteem: The urban school child. Washington, DC: American Sociological Association.

Rubin, D. B. (1996). Multiple imputation after 18+ years. Journal of the American Statistical Association, 91, 473-489.

Ryvicker, M. (2011). Staff-resident interaction in the nursing home: An ethnographic study of socio-economic disparities and community contexts. Journal of Aging Studies, 25, 295-304.

Sengupta, M., Decker, S. L., Harris-Kojetin, L., & Jones, A. (2012). Racial differ- ences in dementia care among nursing home residents. Journal of Aging and Health, 24, 711-731.

Shippee, T. P., Henning-Smith, C. E., Kane, R. L., & Lewis, T. (2013). Facility and resident level predictors of quality of life in long-term care. The Gerontologist. Advance online publication. doi:10.1093/geront/gnt148

Shippee, T. P., Hwanhee, H., Henning-Smith, C., & Kane, R. L. (2014). Longitudinal changes in nursing home-resident reported quality of life: The role of facility characteristics. Research on Aging, 1-26.doi:10.1177/0164027514545975

Smith, D. B., Feng, Z., Fennell, M. L., Zinn, J., & Mor, V. (2007). Separate and unequal: Racial segregation and disparities in quality across U.S. nursing homes. Health Affairs, 26, 1448-1458.

Smith, D. B., Feng, Z., Fennell, M. L., Zinn, J., & Mor, V. (2008). Racial disparities in access to long-term care: The illusive pursuit of equity. Journal of Health Politics, Policy and Law, 33, 861-881.

StataCorp LP. (2011). Stata Statistical Software: Release 12 (12th ed.). College Station, TX: Author.

Stone, R. (2011). Long-term care for the elderly. Washington, DC: The Urban Institute Press.

Vital Research. (2010). Implementation of the 2010 Consumer Satisfaction and Quality of Life Survey in Minnesota nursing homes and satisfaction survey of fam- ily members of residents in Minnesota nursing homes: Final report (Prepared for the Minnesota Department of Human Services). Los Angeles, CA: Vital Research.

Wallace, S. P., Levy-Storms, L., Kington, R. S., & Andersen, R. (1998). The persistence of race and ethnicity in the use of long-term care. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 53, S104-S112.

Weissman, J. S., Hasnain-Wynia, R., Weinick, R. M., Kang, R., Vogeli, C., Iezzoni, L., & Landrum, M. B. (2012). Pay-for-performance programs to reduce racial/ ethnic disparities: What might different designs achieve? Journal of Healthcare for the Poor and Underserved, 23, 144-160.

Wu, S., & Barker, J. C. (2008). Hot tea and juk: The institutional meaning of food for Chinese elders in an American nursing home. Journal of Gerontological Nursing, 34(11), 46-54.

http://www.house.leg.state.mn.us/hrd/pubs/nfreimb.pdf