Nursing Home Quality and Disparities of Care Alex Laberge, MBA, PT Department of Health Services Research, Management & Policy.

Nursing Home Quality and Disparities of Care

Alex Laberge, MBA, PT

Department of Health Services Research, Management & Policy

Project Team

University of Florida Robert Weech-Maldonado, Ph.D. Zhou Yang, Ph.D. Lloyd Dewald, MS

Texas A&M Christopher Johnson, Ph.D.

Collaboration with USF (Kathy Hyer, Ph.D.) and Florida State University (David McPherson, Ph.D.)

Acknowledgement: Supported in part by the Administration on Aging and the UCLA Center for Health Improvement

in Minority Elders (CHIME)/Resource Centers for Minority Aging Research, National Institute on Aging (AG-02-004)

Research Question Are there racial/ethnic and language

differences in the provision of nursing home stroke rehabilitation care after controlling for between-facility effects?

Stroke and Rehab Care Stroke

3rd leading cause of death Most common neurological reason for hospital

admission Leading cause of adult disability

Majority of stroke survivors need rehabilitation services that enhance their recovery and minimize their disability

Nursing Homes and Rehab Care Rehabilitation services offered through a variety

of acute and post-acute settings, such as hospitals, inpatient rehabilitation facilities, nursing homes, and home health agencies

Nursing homes increasingly expanding their role in the provision of rehabilitation care and post-acute care

Minorities make up 21% of those diagnosed with stroke in nursing homes in 2002

Medicare and Rehab Care Medicare the primary payer for post-acute

rehabilitation care in nursing homes Medicare provides 100% coverage of the first 20 days

and 80% of the next 80 days of eligible nursing home stays

Case-mix adjustment based on the Resource Utilization Group (RUG III) classification of a patient as reflected by the MDS

Rehabilitation RUG levels determined by the amount of therapy services. The incremental change in reimbursement between RUG levels is set so that a facility will benefit financially from providing more therapy

Racial/Ethnic Differences in Nursing Home Care

Prior research suggests the presence of racial/ethnic disparities in nursing home care Christian et al. (2003)- racial/ethnic minorities in

nursing homes less likely to receive medications for secondary prevention of stroke

Baumgarten et al. (2004)- Blacks had a higher incidence of nursing home acquired pressure sores

Racial/Ethnic Differences in Nursing Home Care

The observed racial/ethnic differences in nursing home quality of care may be a combination of Minorities receiving lower quality of care than Whites

within the same facility (within-facility differences) Minorities being clustered in facilities with lower

quality of care (between-facility differences). Prior research has found between-facilities disparities

in the nursing home industry E.g., Grabowski et al. (2004), Mor et al. (2004),

Smith et al. (2007)-segregation still exists in U.S. nursing homes with Blacks being much more likely to be placed in nursing homes with serious deficiencies, lower staffing ratios and greater financial vulnerability

Language Differences in Nursing Home Care ~ 47 million people in the U.S. speak a language

other than English at home, and over 21 million are limited English proficient (LEP) (US Census 2000)

Prior studies suggest that language barriers have a greater negative effect on patient experiences than race/ethnicity among Hispanics and Asians (Weech-Maldonado et al. 2001, 2003, and 2004)

3,279 (5.1%) of the nursing home stroke rehabilitation patients who had a MDS 14 day assessment had a language other than English as their first language.

Study Contributions To date there have been no studies examining

Within-facility differences Racial/ethnic differences in rehab care Language differences in nursing home care

Behavioral Model of Health Services Utilization (Andersen, 1998)

Predisposing

Enabling

Need

Utilization of Rehab Services

Data

2002 Nursing Home Minimum Data Set (MDS) 14-day Medicare MDS assessments

64,174 residents Sample limited to those with a stroke diagnosis,

whose care was paid by Medicare Part A Exclude hospital-based facilities Include only residents admitted from hospitals

Dependent Variables Therapy utilization for speech, occupational, and

physical therapy Number of minutes of therapy provided to the resident in

the 7-day observation period

Independent Variables Predisposing variables

Race/ethnicity and language White

English Non-English

Black Hispanic

English Spanish

Asian English Non-English

Age Gender

Enabling variables Support person Desire to be discharged Education BMI (> 30) Type of secondary insurance

Need variables (Stroke severity) Cognitive Performance

Scale ADL Function Scale

Analysis Two-part model of health services utilization of

rehabilitation services First part: logistic regression to estimate the probability of

any use of services within the population State fixed effects

Second part: multivariate regression analysis to predict utilization conditional on whether the enrollee used any rehab therapy services Facility fixed effects

Huber/White correction to account for potential correlation among observations from the same facility

Descriptive Statistics

Dependent Variables White

English(n=51713)

White Non-

English(n=987)

Hispanic

English(n=669)

Hispanic Spanish(n=1714)

Asian English

(n=255)

Asian

Non- English

(n=578)

Black(n=8640)

Speech 58.0 48.9 48.1 43.4 52.4 48.2 52.7F=14(0.00)

Physical Therapy 194.4 181.2 180.7 161.1 186.4 165.2 161.9

F=141 (0.00)

Occupational Therapy 170.6 154.7 158.1 133.5 155.4 138.5 145.9

F=104 (0.00)

Number of Minutes

Logistic Regression Results Odd Ratios (Confidence Intervals)

Compared to English speaking Whites; *p< 0.10 **p<0.05 ***p<0.01

Race/Ethnicity Physical Therapy OccupationalTherapy

Speech Therapy

White Non English

0.94(0.76- 1.16)

0.94(0.78-1.13)

1.05(0.90-1.23)

Hispanic English

0.94(0.74- 1.19)

0.94(0.76- 1.16)

0.86(0.72-1.04)

Hispanic Spanish

0.67***(0.53 – 0.84)

0.69***(0.56-0.86)

0.62***(0.50-0.76)

Asian English 0.92(0.62- 1.38)

0.87(0.61-1.24)

0.88(0.65- 1.20)

Asian Non English

0.75*(0.54- 1.04)

0.87(0.65-1.17)

0.73**(0.56- 0.96)

Black 0.77 ***(0.71- 0.84)

0.87***(0.80-0.93)

0.81***(0.75- 0.86)

Results

White English speakers have greater odds of receiving therapy services when compared to Black, Hispanic Spanish and Asian Non-English nursing home residents with stroke

30% greater odds for PT, 15% greater odds for OT, and 23% greater odds for ST than Blacks

49% greater odds for PT, 45% greater odds for OT, and 61% for ST than Hispanic Spanish

33% greater odds for PT and 37% greater odds for ST than Asians

Predicted Therapy Utilization (Minutes)

Dependent Variables

White

English

White

Non-

English

Hispanic

English

Hispanic

Spanish

Asian

English

Asian

Non-

English Black

Physical Therapy 199.2 197.1** 197.1** 193.4** 197.3* 191.7** 192.4**

Occupational Therapy 175.8 174.7* 172.0** 169.5** 178.8** 172.1** 169.0**

SpeechTherapy 67.4 68.2* 59.2** 50.2** 63.3** 53.9** 58.6**

*p<0.01**p<0.001

Results Racial/ethnic minorities with stroke generally received

less therapy minutes than White English speakers in nursing homes across all therapy types

Among Hispanics, Asians, and Whites, non-English speakers generally received less therapy minutes than their English counterparts

Examples Blacks received 7 minutes less PT, 7 minutes less OT, and 9

minutes less ST Hispanic Spanish speakers received 6 minutes less PT, 6

minutes less OT, and 17 minutes less ST Asian non-English speakers received 8 minutes less PT, 4

minutes less OT, and 14 minutes less ST

Conclusions Racial/ethnic and language minorities are less likely to

receive any rehabilitation service among Medicare nursing home residents with stroke

Of those who actually receive some rehabilitation service, minorities tend to receive less therapy minutes even after controlling for between-facility effects, as well as predisposing, enabling and need factors

Nursing homes should address the observed racial/ethnic and language differences in processes of care as part of their quality improvement efforts

Future Research Causes for the observed racial/ethnic and language

differences in the use of rehabilitation services Lack of racial/ethnic and language concordance

between residents and therapists Lack of access to interpreter services Differences in health beliefs or cultural preferences Systemic bias

Impact of the observed lower utilization of nursing home rehabilitation therapies on outcomes of care among racial/ethnic and language minorities with stroke Walk improvement ADL improvement