Outcomes of Home Health Care for High-Risk Rural Medicare - PowerPoint PPT Presentation

Community Factors and Outcomes of Home Health Care for High-Risk Rural Medicare Beneficiaries TRACY MROZ, PHD – UNIVERSITY OF WASHINGTON LONG TERM SERVICES AND SUPPORTS INTEREST GROUP MEETING JUNE 24, 2017

Collaborators WWAMI Rural Health Research Center Department of Family Medicine University of Washington https://depts.washington.edu/fammed/rhrc/ • C. Holly A. Andrilla, MS • Susan M. Skillman, MS • Lisa A. Garberson, PhD • Davis G. Patterson, PhD • Eric H. Larson, PhD

Acknowledgement of Funding This research was supported by the Federal Office of Rural Health Policy (FORHP), Health Resources and Services Administration (HRSA), U.S. Department of Health and Human Services (HHS) under cooperative agreement #U1CRH03712. The information, conclusions, and opinions expressed in this presentation are those of the authors and no endorsement by FORHP, HRSA, or HHS is intended or should be inferred.

Background • Home health is playing an increasing role in post-acute care • Incentives exist to improve outcomes following acute care hospitalization and reward value over volume - Medicare Hospital Readmissions Reduction Program - Public quality reporting - Bundling demonstrations - Comprehensive care for joint replacement - Home health value-based purchasing model

Rural Home Health Care • Rural home health patients tend to be sicker and at higher risk for hospitalization (Probst & Bhavsar, 2014) • Delivering home health services in rural areas can be particularly challenging and access is sometimes limited despite current payment incentives to serve rural beneficiaries (CMS, 2014; Skillman et al., 2016; Probst et al., 2014) • Communities that are economically- disadvantaged may also have higher risk for readmissions (Kind, 2014)

Study Question Are community factors associated with outcomes of care for rural Medicare beneficiaries receiving home health for high-risk conditions? Hypothesis: Rurality, geographic region, available health resources, and local economy will be associated with hospital readmissions, emergency department use, and community discharge.

Design and Data Sources • Retrospective cohort analysis of rural Medicare beneficiaries who utilized home health care between 2011 and 2013 • Medicare administrative data: - Outcomes and Assessment Information Set (OASIS) - Home health claims - Enrollment file • Area Health Resource File (AHRF) • County Typology Codes from U.S. Department of Agriculture Economic Research Service (USDA ERS)

Eligibility • Medicare Fee-for-Service beneficiary • Aged 65 or over at time of home health admission • Rural- residing based on beneficiary’s ZIP code • Discharged from acute care hospital within 14 days prior to home health admission • Began home health episode on or after January 1, 2011 and ended on or before December 31, 2013 • Primary diagnosis of acute myocardial infarction, heart failure, pneumonia, or COPD

Outcomes • Community discharge following the initial episode of home health (Y/N) • Hospital readmission during the initial home health episode (Y/N) • Emergency department use during the initial home health episode (Y/N)

Community Factors • Rurality categorized as large rural, small rural, or isolated small rural based on 2010 Rural-Urban Commuting Area (RUCA) codes (Morrill et al., 1999) • Geographic location categorized into one of nine divisions defined by the U.S. Census Bureau • Available health resources including number of acute care hospital beds, skilled nursing facility beds, home health agencies, rural health clinics, and primary care physicians in each county standardized by county-level Medicare enrollment ages 65 and over • County-level economic indicators as dichotomous variables representing persistent poverty, low education, low employment, and population loss

Patient Characteristics • Age • Sex • Race • Dual eligibility status • Diagnosis • Clinical severity • Functional status at admission • Cognitive status at admission • Living situation • Caregiving needs

Analysis • Complete case analysis • Hierarchical multiple logistic regression models using general estimating equation methods to account for clustering of beneficiaries within counties

Results: Patient Characteristics 48,802 rural Medicare Fee-for-Service beneficiaries eligible • 58% female • 30% over age 85 • 8% non-white • 29% dually-eligible for Medicare and Medicaid • 31% live alone • 59% with cardiac dx versus 41% with pulmonary dx • Moderate (46%) to high (45%) clinical severity • Moderate (62%) to low (22%) functional status • Intact cognition (55%) to mild impairment (35%) • 78% required caregiver assistance with medication management and 61% required caregiver assistance for supervision and safety

Results: Community Factors Rurality of Beneficiary Residence Isolated small rural 20% Large rural 52% Small rural 28%

Results: Community Factors Number of Beneficiaries by Census Division 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 New Middle East West South East West Mountain Pacific England Atlantic North North Atlantic South South Central Central Central Central

RESULTS: Community Factors • Among 1,797 rural counties: 37.5% were designated as “low employment” - 15.1% were designated as “low education” - 12.3% were designated as “persistent poverty” - 16.7% were designated as “population loss” - • On average rural counties had (per 1,000 Medicare beneficiaries ages 65+): 16.5 acute care hospital beds - 50.5 SNF beds - 2.9 primary care physicians - 0.6 rural health clinics - 0.3 home health agencies -

RESULTS: Adjusted Analyses • Rurality not significantly associated with any outcomes • Economic indicators - Counties with versus without low employment had:  9% higher odds of readmissions  10% higher odds of emergency department visits - Counties with versus without persistent poverty had:  12% lower odds of community discharge  10% higher odds of readmissions - Population loss and low education not significantly associated with any outcomes

RESULTS: Adjusted Analyses • Available health resources - Counties with lower numbers of skilled nursing facility beds had significantly:  Higher odds of community discharge (OR 1.12 lowest quartile versus highest quartile)  Lower odds of emergency department visits (OR .93 lowest quartile versus highest quartile)  Lower odds of readmissions (OR .94 lowest quartile versus highest quartile) - Availability of acute hospital beds, home health agencies, primary care physicians, and rural health clinics not significantly associated with any outcomes

RESULTS: Adjusted Analyses • Geographic location significantly associated with all outcomes Community Hospital ED Census Division Discharge Readmissions Visits New England (CT, ME, MA, NH, RI, VT) (reference) -- -- -- Middle Atlantic (NJ, NY, PA) .92 1.14 .93 East North Central (IL, IN, MI, OH, WI) .90 1.10 1.03 West North Central (IA, KS, MN, MO, NE, ND, SD) .98 .99 .93 South Atlantic (DE, DC, FL, GA, MD, NC, SC, VA, WV) .86 1.18 .94 East South Central (AL, KY, MS, TN) .64 1.52 1.05 West South Central (AR, LA, OK, TX) .45 1.66 1.19 Mountain (AZ, CO, ID, MT, NV, NM, UT, WY) .74 1.11 1.12 Pacific (AK, CA, HI, OR, WA) .92 0.93 1.01

Adjusted Rates of Outcomes of Home Health Care for High-Risk Rural Medicare Beneficiaries 80 70 60 50 40 30 20 10 0 New Middle East North West North South East South West South Mountain Pacific England Atlantic Central Central Atlantic Central Central Community Discharge Hospital Readmission Emergency Department Use

Outcomes of Home Health Care for High-Risk Rural Medicare Beneficiaries by Census Division

KEY FINDINGS • Region of U.S. highly related to outcomes of home health care for high-risk rural Medicare beneficiaries • Pacific, New England, and West North Central census divisions had the best overall outcomes while East South Central and West South Central had the worst overall outcomes • Most other community factors not associated with outcomes or magnitude of associations smaller compared to geographic region

STUDY LIMITATIONS • Selection bias into home health • No comparison with urban beneficiaries • Medicare Advantage not included • Lack of accounting of care processes in home health that may be related to outcomes • Could not examine outcomes after home health discharge or verify outcomes using hospital claims data

CONCLUSIONS & IMPLICATIONS • Regional variation generally consistent with previous research on acute and post-acute care outcomes • Region appears more important to outcomes than other community factors − Lessons to be learned from high performing areas − Room for improvement in low performing areas • Drivers of regional variation and targets for intervention • Impact of policies that incentivize value and efficiency on rural beneficiaries

Policy Brief Mroz TM, Andrilla CHA, Skillman SM, Garberson LA, Patterson DG. Community factors and outcomes of home health care for high-risk rural Medicare beneficiaries. Policy Brief #161. Seattle, WA: WWAMI Rural Health Research Center, University of Washington, Oct 2016. Available ◦ WWAMI Rural Health Research Center https://depts.washington.edu/fammed/rhrc/ ◦ Rural Health Research Gateway https://www.ruralhealthresearch.org/publications/1069