Do Highly Rated Nursing Homes Provide Better Post-Acute Care? - - PowerPoint PPT Presentation

Do Highly Rated Nursing Homes Provide Better Post-Acute Care?

Portia Y. Cornell (Providence VAMC & Brown University) David C. Grabowski (Harvard Medical School) Edward C. Norton (University of Michigan & NBER) Momotazur Rahman (Brown University) AcademyHealth Annual Research Meeting June 25-27, 2017 New Orleans, LA

Background

• In 2008, CMS adopted a five-star quality rating system for

SNFs in an effort to improve quality of care by informing consumer choices.

• A number of concerns about the 5-star rating system remain

 awareness and mistrust of the data limiting consumers use of the rating system  whether the “one size fits all” approach of the star rating system gives relevant in fomration to the many different types of patients seeking nursing home care  whether choosing a highly rated SNF improves patient

• utcomes is unknown

2

Patient and nursing-home data

• Medicare: inpatient and post-acute claims, dates
• f service, diagnoses
• Minimum Data Set (MDS): clinical assessments
• Online Survey Certification and Reporting

(OSCAR)

• American Hospital Association (AHA)
• Monthly Star rating from Nursing Home Compare
• 1,303,484 FFS patients from 3,173 hospitals

discharged to 14,750 SNFs

3

Outcome measure

• Follow patients for 180 days after admission to

SNF

• Assigned each day to one of 6 settings

Mean number of days

• Std. Dev.

Death 25.04 52.88 Hospice 3.69 17.26 Hospital 5.48 10.96 Skilled nursing facility 48.24 49.22 Home with home health 28.33 38.08 Home without home health 70.21 64.14

4

Patient-level controls

• Demographics: age, sex, language, race,

ethnicity, marital status, dual-eligible

• Utilization: intensive care, coronary care,

length of stay

• Health status: functional/cognitive scales,

diagnoses

5

• Outcome: Number of days in a setting out of

180

• Overall Star rating, integer 1 - 5
• Patient characteristics
• Time fixed effects
• ZIP code fixed effects

Reduced-form equation

• Yizhtn
• Qint
• Xi
• qt
• z z

6

Selection bias

• Causal effect of star rating on patient outcomes

difficult to parse

7

Hospitals send highest-risk patients to best nursing homes Patients who use quality ratings may be systematically different

Case Mix

Distance as instrumental variable

• Patients tend to choose a nursing home close to

home

• Within a zip code, quality of closest nursing

home can change month to month, or with market exit/entry

• Zip code fixed effects control for geographic

characteristics correlated with star rating

8

Strength of first stage

Log of distance to nearest SNF rating Effect on Q, star rating of admitting nursing home 1-star 0.0683*** 2-star 0.0633*** 3-star 0.0101*** 4-star

• 0.0804***

5-star

• 0.121***

R-squared 0.262 F-statistic 2612.3 9 The further away the nearest 1-star, the higher the quality rating The further away the nearest 5-star, the lower the quality rating Monotonic effect from 1-star to 5-star

Mean Percent change from mean

• 1.440***
• 6.0%

[-3.711]

• 0.0951
• 0.0%

[-0.691]

• 0.182**
• 3.6%

[-2.195]

• 1.139***
• 2.3%

[-3.001] 1.124*** +3.9% [3.731] 1.732*** +2.4% [4.027] (t-statistics in parentheses)

Main IV Results

Outcome IV estimate with ZIP fixed effects Death days

• 1.440***

[-3.711] Hospice days

• 0.0951

[-0.691] Hospital inpatient days

• 0.182**

[-2.195] Nursing home days

• 1.139***

[-3.001] Home health days 1.124*** [3.731] Home days 1.732*** [4.027] (t-statistics in parentheses)

10

Instrumented vs. OLS

11

Full IV Results

• Without FE and only hospital FE closer to OLS
• ZIP-FE similar with and without hospital FE

12

Outcome No fixed effects Hospital fixed effects Zip code fixed effects Hospital and zip code fixed effects Death days

• 1.704***
• 1.060***
• 1.440***
• 1.476***

[-17.23] [-5.172] [-3.711] [-3.785] Hospice days

• 0.00812

0.0216

• 0.0951
• 0.0811

[-0.234] [0.296] [-0.691] [-0.586] Inpatient days

• 0.262***
• 0.229***
• 0.182**
• 0.189**

[-12.20] [-5.032] [-2.195] [-2.260] Nursing home days

• 1.661***

0.617***

• 1.139***
• 1.070***

[-16.69] [3.032] [-3.001] [-2.824] Home health days

• 1.262***
• 0.621***

1.124*** 1.143*** [-16.07] [-3.915] [3.731] [3.785] Home days 4.898*** 1.273*** 1.732*** 1.672*** [43.41] [5.517] [4.027] [3.888]

Results by day

• Estimate probability of being in a particular

setting of each of 180 days

• Effects persist over time, not just in initial

days after admission

13

Results by day

14

Summary

• Patients who go to higher-rated SNFs spent

more time in better outcomes and less time in worse outcomes.

• OLS underestimates the mortality and hospital

in effect and overestimates the effect on days at home

15

Discussion

• Controlling for endogeneity of selection

matters

• Validates star rating as meaningful for patients

looking for care

16