Picture of Housing and Health: Medicare and Medicaid Use Among Older Adults in HUD-Assisted Housing

03/01/2014

Picture of Housing and Health: Medicare and Medicaid Use Among Older Adults in HUD-Assisted Housing

The Lewin Group

March 2014


This report was prepared under contract between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the Lewin Group. Additional funding was provided through Interagency Agreements with the HHS Administration on Aging and the U.S. Department of Housing and Urban Development. For additional information about this subject, you can visit the DALTCP home page at http://aspe.hhs.gov/office_specific/daltcp.cfm or contact the ASPE Project Officer, Emily Rosenoff, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W., Washington, D.C. 20201.  Her e-mail address is: Emily.Rosenoff@hhs.gov.

The opinions and views expressed in this report are those of the authors. They do not necessarily reflect the views of the Department of Health and Human Services, the Department of Housing and Urban Development, the contractor or any other funding organization.


TABLE OF CONTENTS

ACRONYMS
EXECUTIVE SUMMARY
I. INTRODUCTION
A. Study Overview and Task Objectives
B. Background
II. STUDY OBJECTIVES AND RESEARCH QUESTIONS
III. DATA
IV. RESULTS
A. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions
B. Calculate HUD and Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008
C. Estimate HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program
D. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions
V. DISCUSSION
A. Summary of Results
B. Limitations
C. Future Analysis
NOTES
APPENDICES
APPENDIX A: Data Sources and Variable Definitions
APPENDIX B: Study Samples and Analytic Approach
APPENDIX C: 2008 Detailed Results
APPENDIX D: Supplemental Results, All Years
LIST OF FIGURES
FIGURE 1: Demographic Characteristics of HUD-Assisted Individuals in 12 Study Jurisdictions, 2008
FIGURE 2: Distribution of HUD-Assisted Individuals by Property Type in 12 Study Jurisdictions, 2008, All Ages
FIGURE 3: Distribution of HUD-Assisted Individuals by Property Type in 12 Study Jurisdictions, 2008, Age 62+ Years
FIGURE 4: Proportion of HUD-Assisted Properties in 12 Study Jurisdictions Defined as "Elderly", 2008
FIGURE 5: Medicare and Medicaid SSN Match Rates for HUD-Assisted Individuals, by Age Group, 2008
FIGURE 6: Medicare and Medicaid Restrictive Match Rates for HUD-Assisted Individuals, by Age Group, 2008
FIGURE 7: Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid and Who Did Not Link, Age 65 or Older, 2008
FIGURE 8: HUD-Assisted Medicare Beneficiaries Part D LIS Status and Medicaid Participation
FIGURE 9: Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison in 12 Jurisdictions, All Ages, 2008
FIGURE 10: Medicare and Medicaid Enrollees with Chronic Conditions, 2008, 12 Study Jurisdictions
FIGURE 11: Comparison of Fee-for-Service HUD-Assisted MME and Unassisted MME Medicare Health Care Utilization
FIGURE 12: Comparison of Fee-for-Service HUD-Assisted MMEs and Unassisted MMEs Medicaid Health Care Utilization
FIGURE 13: HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries, 2008, by Gender
FIGURE 14: HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries, 2008, by Race/Ethnicity
FIGURE 15: Fee-for-Service HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries, 2008, by Part D LIS Status
FIGURE 16: HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Beneficiaries, Age 65+, 2008, by Chronic Condition Category Count and Prevalence
FIGURE 17: Comparison of Fee-for-Service HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries' Medicare Health Care Utilization
FIGURE B1: Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison, All Ages
FIGURE C1: Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison
LIST OF TABLES
TABLE A1: Property Type Variable Definition
TABLE A2: Assignment of Chronic Conditions into 9 Chronic Condition Groups
TABLE A3: Medicare Health Care Utilization Variable Definitions
TABLE A4: Medicare Health Care Spending Variable Definitions
TABLE B1: HUD-Assisted Individuals 2008 Unique Observations
TABLE C1: Demographic Characteristics of HUD-Assisted Individuals, 2008, 12 Study Jurisdictions
TABLE C2: Living Arrangement, All Ages and Age 62+ Years
TABLE C3: Living Arrangement, Household Level, All Ages
TABLE C4: Full and Partial HUD Housing Assistance, 2008, by Age Group
TABLE C5: Proportion of HUD-Assisted Properties in 12 Study Jurisdictions Defined as "Elderly"
TABLE C6: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals, 2008, by Age Group
TABLE C7: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI, 2008, by Age Group
TABLE C8: Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid, 2008, by Age Group
TABLE C9: HUD-Assisted Medicare Beneficiary Enrollment in Part D LIS, 2008
TABLE C10: HUD-Assisted Medicare Beneficiary Enrollment in Medicaid and Medicaid Coverage Characteristics, 2008
TABLE C11: Five Sub-groups for Comparison of Health and Health Care Utilization Among HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community
TABLE C12: HUD-Assisted and Unassisted MME Demographic Characteristics
TABLE C13: HUD-Assisted and Unassisted MME Medicare and Medicaid Enrollment Characteristics
TABLE C14: MMEs with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
TABLE C15: Comparison of HUD-Assisted MME and Unassisted MME Medicare Health Care Utilization
TABLE C16: Comparison of HUD-Assisted MME and Unassisted MME Medicaid Health Care Utilization per 1000 Member Months
TABLE C17: Comparison of HUD-Assisted MME and Unassisted MEE Medicaid Health Care Utilization, Proportion of MMEs with Any Use
TABLE C18: HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Beneficiaries, Age 65+, 2008, by Demographic and Coverage Characteristics
TABLE C19: HUD-Assisted Medicare/No Medicaid and Unassisted Beneficiaries with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
TABLE C20: Comparison of HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries' Medicare Health Care Utilization
TABLE D-A1: HUD-Assisted Individuals Full vs. Partial Year, by Year (All Ages)
TABLE D-A2: HUD-Assisted Individuals Full vs. Partial Year, by Year (Age 62+)
TABLE D-A3: HUD-Assisted Individuals Full vs. Partial Year, by Year (Age 65+)
TABLE D-A4: HUD-Assisted Individuals Across Consecutive Years Full vs. Partial
TABLE D-A5: Living Arrangement, HUD-Assisted Individuals, All Ages and Age 62+ Years
TABLE D-A6: Living Arrangement, HUD-Assisted Households, All Ages
TABLE D-A7a: HUD-Assisted Individual Demographics, 2008
TABLE D-A7b: HUD-Assisted Individuals Demographics, 2007-2009
TABLE D-A8: HUD-Assisted Individual Demographics by Geographic Area
TABLE D-A9: Demographics by Age Group
TABLE D-A10: Demographics by Property Type
TABLE D-A11: Demographics by Subside Type
TABLE D-A12: Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+
TABLE D-A13: Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+ by Geographic Location
TABLE D-A14: Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+
TABLE D-A15: Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+ or Designated Elderly by Geographic Location
TABLE D-B1: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals, 2008, by Age Group
TABLE D-B2: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI, 2008
TABLE D-B3: Restrictive Match Rates: All Ages
TABLE D-B4: Restrictive Match Rates: <62 Years of Age
TABLE D-B5: Restrictive Match Rates: Age 62 to 64
TABLE D-B6: Restrictive Match Rates: Age 65+
TABLE D-B7: Restrictive Match Rates: Heads of Household, All Ages
TABLE D-B8: Restrictive Match Rates: Heads of Household, <62 Years of Age
TABLE D-B9: Restrictive Match Rates: Heads of Household, Age 62 to 64
TABLE D-B10: Restrictive Match Rates: Heads of Household, Age 65+
TABLE D-B11: Restrictive Match Rates: Heads of Household, Receiving SSI
TABLE D-B12: Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid by Age Group, 2008
TABLE D-C1: Health Insurance Characteristics of Those that Link by Year by Age
TABLE D-C2: Health Insurance Characteristics of HUD Heads of Households Receiving SSI that Link in 2008 by Age
TABLE D-D1: Final 2008 Samples for Utilization and Cost Analysis
TABLE D-D2: Final 2008 Samples Limited to Individual with Medicaid Fee for Service -- Managed Care
TABLE D-D.A1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries, 2008
TABLE D-D.A2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.A3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.A4: Medicare and Beneficiary Health Care Costs
TABLE D-D.A1.1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.A1.2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.A1.3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.A1.4: Medicare and Beneficiary Health Care Costs
TABLE D-D.B1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.B2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.B3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.B4: Medicare and Beneficiary Health Care Costs
TABLE D-D.C1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.C2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.C3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.C4: Medicare and Beneficiary Health Care Costs
TABLE D-D.C1.1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.C1.2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.C1.3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.C1.4: Medicare and Beneficiary Health Care Costs
TABLE D1: Proportion of Beneficiaries with Part D Coverage, Enrolled in LIS

ACRONYMS

The following acronyms are mentioned in this report and/or appendices.

Acronyms
ADLActivity of Daily Living
AIDSAcquired Immune Deficiency Syndrome
AMIArea Median Income
ASCAmbulatory Surgery Center
 
BMIR  Below-Market Interest Rate
 
CBPPCenter on Budget and Policy Priorities
CBSACore Based Statistical Area
CCWChronic Condition Warehouse
CMSCenters for Medicare and Medicaid Services  
 
DIBDisability Insurance Benefits
DMEDurable Medical Equipment
 
E&MEvaluation and Management
EDEmergency Department
EREmergency Room
ESRDEnd Stage Renal Disease
 
FFSFee-For-Service
FIPSFederal Information Processing Standards
FPLFederal Poverty Level
 
GDITGeneral Dynamics Information Technology
 
HCBSHome and Community-Based Services
HHHome Health
HHSU.S. Department of Health and Human Services
HIVHuman Immunodeficiency Virus
HMOHealth Maintenance Organization
HUDU.S. Department of Housing and Urban Development
 
IADLInstrument Activity of Daily Living
IMGImaging Services
 
LISLow Income Subsidy
 
MAXMedicaid Analytic eXtract
MDSMinimum Data Set
MMEMedicare and Medicaid Enrollee
MR/DD  Mental Retardation and Other Developmental Disabilities  
MSPMedicare Savings Program
 
NFNursing Facility
 
PACProject Assistance Contract
PHAPublic Housing Authority
PHYSPhysician Office Services
PICPublic and Indian Housing Information Center
PIHPublic and Indian Housing
PMPMPer Member Per Month
PRACProject Rental Assistance Contract
 
QDWIQualified Disabled and Working Individual
QMBQualified Medicare Beneficiary
 
RAPRental Assistance Payment
RTIResearch Triangle Institute
 
SCHIP  State Children's Health Insurance Program
SNFSkilled Nursing Facility
SSASocial Security Administration
SSISupplemental Security Income
SSNSocial Security Number
 
TRACSTenant Rental Assistance Certification System  

EXECUTIVE SUMMARY

Overview

For the U.S. Departments of Health and Human Services (HHS) and Housing and Urban Development (HUD), the Lewin Group and its sub-contractors, Leading Age and the Moran Company, explored the potential for publicly-subsidized senior housing to serve as a platform for efficiently managing the population health of low-income older adults with various levels of physical and mental health risk. We developed a demonstration design and piloted creating a new dataset linking HUD tenant and Medicare/Medicaid claims data in 12 jurisdictions across the country.

Study Objective

This study task explored the feasibility of matching HUD administrative data to the HHS Centers for Medicare and Medicaid Services (CMS) administrative data in order to determine the extent to which this resource could track health and housing outcomes, and whether this approach could reliably support future research and policy analysis.

Data

We focused on 2008 individual-level administrative data provided for the 12 geographic areas from both HUD and CMS. The HUD multifamily and public housing data came from the Tenant Rental Assistance Certification System and the Public and Indian Housing Information Center. Medicare administrative data came from the Medicare Beneficiary Summary File and includes Medicare Parts A, B, and D enrollment, payments, and utilization, as well as information about chronic conditions. The Medicaid Analytic eXtract Person Summary file, based on state submission of Medicaid administrative data, provided enrollment, payment and utilization for Medicaid-covered services.

Study Objectives and Results

We provide a detailed description of the study results in the Summary Report and in Appendix C. The chart below provides a brief overview of the study objectives and corresponding results.

ObjectivesResults
Match Rate between HUD and CMS dataAmong HUD-assisted individuals age 65 or older in 2008, 93% matched to Medicare based on Social Security number (SSN) only; 85% matched to Medicare when requiring a match on SSN, gender, and date of birth.
Estimate HUD-assisted Medicare beneficiaries enrollment in Part D subsidy assistance programs or MedicaidHealth insurance assistance programs for Medicare beneficiaries reduce out-of-pocket health care expenses, which result in savings for HUD. Among HUD-assisted Medicare beneficiaries age 65 years or older:
  • Approximately 68% of HUD-assisted Medicare beneficiaries were dually enrolled in Medicare and Medicaid.
  • Approximately 90% of the Medicare beneficiaries had Part D (drug coverage) coverage. Among those with Part D coverage, 80% received premium and out-of-pocket assistance through the Low Income Subsidy Program.

While we would not expect the proportion to be 100%, as not all HUD-assisted Medicare beneficiaries meet the eligibility requirements for these programs because the income criteria for HUD-assisted housing can be higher than that for health insurance assistance, there may be room for improvement in the enrollment rates.

Compare Medicare and Medicaid payment and utilization among Medicare and Medicaid enrollees (MMEs)Most of our analyses to compare HUD-assisted individuals to unassisted individuals focused on Medicare beneficiaries dually enrolled in Medicaid (MMEs), age 65+ with fee-for-service (FFS) coverage compared to unassisted MME beneficiaries in the community in order to examine comparable groups in terms of economic status because this factor drives social determinants of health, use of services, and resulting payment.Health conditions:
  • HUD-assisted MMEs had more chronic conditions (55% of HUD-assisted MMEs had 5 or more compared to 43% of unassisted MME), which translated into higher health care utilization and costs than unassisted MMEs in the community.

Medicare payments and utilization:

  • HUD-assisted MME beneficiaries' average Medicare FFS per member per month (PMPM) payment was 16% higher than unassisted MMEs in the community ($1,222 compared to $1,054).
  • Higher utilization by HUD-assisted MMEs for home health visits (31% higher), ambulatory surgery center visits (45% higher), physician office visits (26%), and emergency department visits (13%) drove the higher payments.

Medicaid payments and utilization:

  • HUD-assisted MMEs Medicaid FFS PMPM payment was 32% higher than unassisted MMEs ($1,180 vs. $895).
  • HUD-assisted MMEs used over 100% more Personal Care services, 80% more "other HCBS", and over 67% more durable medical equipment services covered by Medicaid unassisted MMEs in the community.

Discussion

This study demonstrates the feasibility and utility of linking HUD tenant data to CMS administrative data. The descriptive results summarized above highlight key areas for future analysis to better understand the health and health care utilization of HUD-assisted elderly individuals enrolled in Medicare. This includes supplementing current data sources with additional CMS data, refining matching algorithms and study samples to better determine HUD-assisted elderly individuals' eligibility and enrollment in Medicare programs, providing distributional analyses, and conducting multivariate regressions to determine if the differences observed in descriptive comparisons remain after adjusting for confounders.

I. INTRODUCTION

A. Study Overview and Task Objectives

The U.S. Department of Health and Human Services (HHS) Office of the Assistant Secretary for Planning and Evaluation's Office of Disability, Aging and Long-Term Care Policy, in partnership with U.S. Department of Housing and Urban Development (HUD) and the HHS Administration for Community Living, engaged the Lewin Group and its sub-contractors, Leading Age Center for Applied Research and the Moran Company to develop design options for a demonstration of targeted, coordinated housing, health and long-term care services and supports for low-income older adults. This project sought to: (1) identify and examine affordable housing with services models that enable low-income older adults to live in affordable, safe, and accessible housing with access to health and supportive services needed to "age in place"; and (2) propose a demonstration design to track and measure outcomes and costs associated with promising housing with services models. This report presents the results of Task 6: Data Analysis. It explores the feasibility of matching HUD administrative data to national health administrative data in order to determine whether health and housing outcomes can be tracked through existing administrative data sources and whether this approach can reliably support future research and policy analysis.

We specifically focused on the feasibility of linking data from the HHS Centers for Medicare and Medicaid Services (CMS) to the HUD individual tenant-level administrative data. CMS data included the Medicare Beneficiary Annual Summary File and the Medicaid Analytic eXtract (MAX) for the 12 geographic areas of interest in this study (see Section III). Given the study objective, we chose geographic areas that have unique public housing with services models. For example, Burlington, Vermont has the "Supports and Services at Home" program that incorporates an interdisciplinary team of community service providers to coordinate participating residents' health and long-term care needs.

This report lays the groundwork for federal efforts to use existing administrative data maintained separately by health and housing agencies to more effectively serve individuals (including elderly and non-elderly persons with disabilities) and communities thatcould benefit from a coordinated housing with services program. In addition, we conducted initial descriptive analyses of the linked datasets to describe the individuals present in both the HUD and CMS data sources and compared them to individuals covered by Medicare and/or Medicaid, but not receiving housing assistance. Understanding the characteristics of individuals and their use of health care services in different housing arrangements will ideally inform policy to promote rational and optimal care.

B. Background

A large and rapidly expanding pool of low-income and modest-income older adults face the dual challenges of finding affordable and safe housing that can also accommodate changing needs as they grow older. Millions of older adults who rent or own their own homes face excessive housing costs and/or live in housing with serious physical problems. In 2011, an estimated 3.9 million older renter households without children had very low incomes (50% or less of area median income (AMI)). Of these households, 37% faced severe housing cost burdens exceeding half their incomes without public housing assistance, and a similar proportion, 36%, did receive assistance.1

A report from the Summit on Aging in Place in Public Housing (2011) stated that 1.3 million older adult renters live in publicly-subsidized housing. These older renters may have increased difficulty as they age, since they may experience a decline in their physical, cognitive, and/or mental health conditions.2 While the majority of these older renters are relatively healthy, Assets and Health Dynamics Among the Oldest Old survey data (Wave 2) show that subsidized older residents report poorer health than unsubsidized renters, more chronic conditions, significantly higher numbers of limitations in their ability to carry out basic activities of daily living (ADLs) and instrumental activities of daily living (IADLs), and less likelihood to live in properties that offer services.3

The current system of multiple payers -- primarily Medicare and Medicaid -- provides few incentives for primary, acute and chronic care providers to collaborate with each other, let alone cooperate with low-income housing or aging and long-term services and supports providers.4 As a result, when many older adults most need integrated services, they experience a highly fragmented and poorly coordinated services system. The ability of older adults with chronic conditions and/or significant disability to continue living independently may be cut short, their health and safety compromised, and public and private health and long-term care costs may skyrocket as a result of premature transfers to more expensive nursing homes and residential care facilities, repeated trips by emergency medical technicians to an individual's home, repeated trips to hospital emergency departments, and frequent hospitalizations.

To address some of these issues, hundreds of publicly assisted largely not-for-profit housing providers and several states and private sector organizations have developed programs to bring enhanced services to residents. Innovative housing providers across the country, working with federal, state, and community partners have, largely at their own initiative, developed many prototypes of publicly assisted housing with enhanced services for older adults. Typically, these properties employ a service coordinator (available through HUD grants and, in some cases, incorporated into the properties operating budget), complemented by a wide array of community partnerships.

The following table provides a guide to HUD and CMS key concepts and terms used throughout this report.

Guide to Housing Assistance and Medicare and Medicaid Services
NOTE:
  1. Program statistics in this table come from HUD's Picture of Subsidized Households dataset for 2009, available from http://www.huduser.org/portal/datasets/picture/yearlydata.html.
  2. U.S. Department of Housing and Urban Development. HUD's Public Housing Program. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/topics/rental_assistance/phprog.
  3. U.S. Department of Housing and Urban Development. Housing Choice Vouchers Fact Sheet. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/topics/housing_choice_voucher_program_section_8.
  4. U.S. Department of Housing and Urban Development. Section 202 Supportive Housing for the Elderly Program. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/program_offices/housing/mfh/progdesc/eld202.
  5. Note that some tables refer to both "Section 202" and "Section 202/162." A small number of 202 properties contain units for non-elderly disabled that were funded with rental assistance created by Section 162 of the Housing and Community Development Act of 1987. Section 162 has been superseded by the Section 811 program, which supports housing for disabled individuals.
  6. U.S. Department of Housing and Urban Development. Section 811 Supportive Housing for Persons with Disabilities. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/program_offices/housing/mfh/progdesc/disab811.
  7. Lind, K.D. (2012). Setting the Record Straight about Medicare. Washington, DC: AARP Public Policy Institute. Accessed December 2013. http://www.aarp.org/health/medicare-insurance/info-02-2012/Setting-the-Record-Straight-about-Medicare-fact-sheet-AARP-ppi-health.html.
  8. Kaiser Family Foundation. (2012). The Medicare Program at a Glance. Accessed December 2013. http://kff.org/medicare/fact-sheet/medicare-at-a-glance-fact-sheet/.
  9. Barry, P. (2013). Do You Qualify for Medicare? Accessed December 2013. http://www.aarp.org/health/medicare-insurance/info-04-2011/medicare-eligibility.html.
  10. Kaiser Family Foundation. (2013). The Medicaid Program at a Glance. http://kff.org/medicaid/fact-sheet/the-medicaid-program-at-a-glance-update/.
  11. Kaiser Family Foundation (2011). Federal Core Requirements and State Options in Medicaid: Current Policies and Issues. Accessed December 2013. http://kff.org/health-reform/fact-sheet/federal-core-requirements-and-state-options-in/.
  12. "Low-income" for Medicaid and Medicare do not follow the same guidelines as HUD. They are subject to federal and state regulations.
U.S. Department of Housing and Urban Development (HUD) Programs
HUD provides housing assistance to about 4.6 million low-income households,1 including families with children, older adults, and people with disabilities. Tenant income eligibility is determined on the basis of AMI (generally calculated for metropolitan areas and non-metropolitan counties), adjusted for family size. Low-income families are defined as families whose incomes do not exceed 80% of the AMI and very low-income families are defined as families whose incomes do not exceed 50% of the AMI. Over three-fourths of assisted households in 2009 had extremely low incomes not exceeding 30% of AMI. HUD programs important for this report include:
  • Public Housing is housing administered by PHAs for eligible low-income families, older adults, and people with disabilities. Eligibility is based on annual gross income and United States citizenship or eligible immigration status.2 About 23% of HUD-assisted households live in public housing.
  • Housing Choice Vouchers (also known as Section 8 vouchers) provide rent subsidies used in private rental markets and, like public housing, are administered by PHAs. Eligibility is based on income and citizenship or eligible immigration status. A family or individual receiving a voucher must pay a specified percentage (often 30%) of their income toward rent and the PHA pays the balance of rent, subject to program limits.3 Voucher holders represent about 46% of HUD-assisted households.
  • Section 202 housing is the Supportive Housing for the Elderly Program. This is a program run by HUD's Office of Multifamily Housing Programs. It helps finance the development of affordable and accessible housing for low-income older adults. This capital is provided to private not-for-profit organizations and not-for-profit consumer cooperatives and need not be repaid as long as the buildings house low-income seniors.4 The program may also provide rent subsidies.5
  • Section 811 housing is supportive housing for persons with disabilities. It is also run by HUD's Office of Multifamily Housing Programs. The 811 program provides interest-free capital to eligible organizations to help them finance affordable housing with supportive services for low-income adults with disabilities. As with the 202 program, capital need not be repaid as long as the buildings house low-income disabled persons. The program may also provide rent subsidies.6
  • HUD's Office of Multifamily Programs has several programs termed here "other multifamily housing," in which rental assistance is provided in conjunction with programs that subsidize developments through below-market interest financing, mortgage insurance or other forms of assistance for the new construction or rehabilitation of housing for low-income individuals. Multifamily programs, including the Section 202 and Section 811 supportive housing programs represent about 30% of HUD-assisted renters. See Appendix D for the full list of multifamily housing properties included in the data analysis.
Centers for Medicare and Medicaid Services (CMS) Programs
Medicare7,8,9 is a federal program that provides health insurance for older adults and people with disabilities. This program was designed for people age 65 and older. This program is typically an entitlement for older adults who:
  • Are eligible for Social Security Payments, and
  • Have made payroll tax contributions for 10 or more years.

The spouse of someone who meets these guidelines is also eligible for Medicare. Those under 65 can qualify for a couple of reasons. One of the major reasons is being entitled to Social Security disability benefits for at least 2 years.There are multiple parts to Medicare, including:

  • Medicare Part A ("Hospital insurance") covers acute health care needs, including hospital care, SNF care, hospice, and HH. Part A is paid for by a portion of Social Security tax. The vast majority of people age 65 and over get Medicare Part A for free as long as the individual or his/her spouse paid Medicare taxes for a minimum of 10 years and did not enroll late for Medicare, meaning he/she enrolled when first eligible.
  • Medicare Part B ("Medical insurance") covers outpatient health care needs, including doctors' visits and other preventive care services. Individuals who elect Part B coverage must pay premiums.
  • Medicare Part C, or Medicare Advantage Plans, provide you with all your Part A and Part B benefits, and often also cover prescription drugs. Private companies that contract with Medicare offer various types of plans including HMOs, Preferred Provider Organizations, Private FFS Plans, Special Needs Plans, and Medicare Medical Savings Account Plans. Individuals who elect Part C coverage must pay premiums.
  • Medicare Part D provides prescription drug coverage. This is a voluntary program that is offered through private insurance companies that are under contract with Medicare, and include premiums.

Medicaid10,11 is a public health insurance program for low-income children and adults.12 States administer this program following federal rules, and it is jointly financed by both federal and state governments. The federal minimum standards for eligibility are:

  • State coverage of core groups, which include low-income elderly individuals, people with disabilities, pregnant women, children, and parents of children.
  • These core groups are subject to different minimum income levels. For older adults and people with disabilities, states typically provide coverage to those who receive SSI.

Beyond these federal minimums, states can set their own standards for eligibility within the allowed federal range and can opt to cover additional services. For example, all states provide HCBS to older adults through waivers and many states use the upper income limits of 300% of SSI for these services.Medicaid also covers the Medicare premiums and copayments for low-income Medicare beneficiaries through the MSP. Those with income 100% or less of the FPL receive premium and copayment coverage while those with income 101%-125% of the FPL receive premium coverage only. In addition, some individuals receive full Medicaid benefits for services not included in Medicare, such as assistance with personal care and long stay nursing facility care. Participants who are a part of both programs, MMEs, are often referred to as Dual Eligibles or Duals.

II. STUDY OBJECTIVES AND RESEARCH QUESTIONS

The specific objectives of Task 6: Data Analysis and the corresponding research questions are outlined below. Although our project focuses on older adults and housing with services, we assessed the feasibility and quality of the match to the Medicare and Medicaid administrative data under Objectives A and B based on all HUD-assisted individuals. For Objectives C and D, related to Medicare enrollment and health care spending and utilization, we restricted our sample to individuals age 65 and older. This resulted in the most representative comparison because the vast majority of individuals age 65 and older have Medicare coverage and the small subset of those age 60-64 with Medicare or Medicaid coverage likely have it due to a disability.

A. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions

The first component of the study was to understand the characteristics of HUD-assisted individuals--both elderly and non-elderly. We conducted basic descriptive statistics on the HUD individual tenant-level administrative data to answer the following research questions:

  1. What are the demographic characteristics of HUD-assisted individuals, including age, gender, subsidy type, property type, race, disability status, and living arrangement?
    • How do the characteristics vary across year? Across the 12 geographic locations? Across housing subsidy type? Across housing program type?
  2. What proportion of public and assisted housing properties are "elderly properties"?
    • How does this designation vary across the 12 geographic locations?

B. Calculate HUD and Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions

The second study component used both the HUD and CMS data sources. We linked individuals within the HUD data to the CMS data to answer the following research questions:

  1. What proportion of HUD-assisted individuals link to Medicare or Medicaid enrollment data ("match rate")?
    • How does the match rate vary across age groups? By geography? By housing program type?
  2. How do older individuals receiving HUD housing assistance and who link to Medicare or Medicaid enrollment data compare demographically to HUD-assisted elderly individuals who do not link to the Medicare or Medicaid data?

C. Estimate the Proportion of HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program (i.e., Medicaid)

HUD is interested in understanding what proportion of HUD-assisted Medicare beneficiaries are enrolled in programs that could reduce their out-of-pocket health care expenses and, therefore, reduce HUD's costs.

  1. What proportion of HUD-assisted Medicare beneficiaries are enrolled in Part D subsidy assistance programs or Medical Savings Program (i.e., Medicaid)?

D. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions

How do HUD-assisted Medicare beneficiaries compare to unassisted beneficiaries in the community? What are the differences in demographic characteristics, Medicare/Medicaid coverage characteristics, health and health care utilization and payments?This summary includes results for this research question for two sub-groups:

  1. MMEs, age 65+.
  2. Medicare/No Medicaid, age 65+.

III. DATA

We based the analysis on tenant-level administrative data from HUD and individual-level administrative data from CMS in the table below and described in detail in Appendix A. Note, that while we received HUD and CMS Medicare data for 2007-2009, the majority of our results are based on 2008 as 2009 data was not available for CMS Medicaid data.

 HUD Tenant-Level Administrative DataCMS Individual-Level Administrative Data
Years2007, 2008, & 2009
  • 2007-2009 (Medicare)
  • 2007 & 2008 (Medicaid)
Data Sources TRACSMedicare Administrative Data from the Medicare Beneficiary Summary File for 2007-2009
  • Parts A, B, & D
  • Chronic Conditions
  • Payment & Use
PIH/PICMedicaid Administrative Data from the MAX Person Summary file for 2007 & 2008

The data were limited to individuals in our 12 geographic areas of interest for this study. As mentioned in the Introduction, given the study objective, we chose geographic areas that have unique public housing with services models:

  1. New Haven-Milford, Connecticut
  2. Bridgeport-Stamford-Norwalk, Connecticut
  3. Milwaukee-Waukesha-West Allis, Wisconsin
  4. San Francisco-Oakland-Freemont, California
  5. Boston-Cambridge-Quincy, Massachusetts
  6. Durham-Chapel Hill, North Carolina
  7. Richmond, Virginia
  8. New York-Northern New Jersey-Long Island
  9. Columbus, Ohio
  10. Akron, Ohio
  11. Cleveland, Ohio
  12. The entire State of Vermont

IV. RESULTS

The results presented below are organized by study objective and based on 2008 -- the most recent year for which we had data on HUD, CMS Medicare, and CMS Medicaid enrollment. Additional results for 2008 are presented in Appendix C. We also produced results for Objectives A and B, and C by year (2007, 2008, and 2009), which are included in Appendix D, all years.

A. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions

Our sample of HUD-assisted individuals consisted of all individuals that appear in the Tenant Rental Assistance Certification System (TRACS) or Public and Indian Housing (PIH) Information Center (PIC) data at any point in time during 2008 in the 12 study jurisdictions. This included 2,057,893 unique individuals and 967,557 unique household heads that received housing assistance at some point during 2008. The Objective A results are stratified by housing assistance program in Appendix D, all years.

A1. Demographic Characteristics of HUD-Assisted Individuals

Figure 1 presents the demographic characteristics for the HUD-assisted individuals in the 12 study jurisdictions. Some highlights include:

  • Approximately 19% of the sample was older than 62 years of age with 37% younger than age 18.

  • The largest proportion of the sample, 38%, was Black non-Hispanic. Approximately 24% were Hispanic, 16% were White non-Hispanic, and 19% were unknown or declined to report.

  • Over half of the sample (59%) was located in the New York-Northern New Jersey-Long Island area. Nine geographies represent 5% or less of the sample. This distribution is due to our selection of certain geographies. Given the New York City area is the majority of our sample, the individuals living in this area heavily influence the results across all objectives.

FIGURE 1. Demographic Characteristics of HUD-Assisted Individuals in 12 Study Jurisdictions, 2008
 Unique Individuals(N=2,057,893)
MeanSD
Age
All Persons34.925.3
Household Heads54.218.8
 N%
Age group
<18764,54337.2%
18-44566,05927.5%
45-61327,06715.9%
62-6446,7002.3%
65+352,09117.1%
Unknown/declined to report1,4330.1%
Geographic area
Vermont22,6351.1%
New Haven- Milford50,2012.4%
Bridgeport- Stamford- Norwalk35,3261.7%
Milwaukee- Waukesha- West Allis51,5552.5%
San Francisco- Oakland- Fremont167,7958.2%
Boston- Cambridge- Quincy243,60111.8%
Durham- Chapel Hill17,9450.9%
Richmond43,1962.1%
New York- Northern New Jersey- Long Island1,213,20159.0%
Columbus72,3323.5%
Akron35,4091.7%
Cleveland104,6975.1%
Race/ethnicity
White non-Hispanic331,94416.1%
Hispanic489,33723.8%
Black non-Hispanic773,55037.6%
Asian56,9342.8%
American Indian/Alaskan3,6990.2%
Hawaiian/Pacific Islander2,7620.1%
Other non-Hispanic2590.0%
Mixed non-Hispanic2750.0%
Unknown/declined to report399,13319.4%

Figure 2 presents the proportion of HUD-assisted individuals by property type for all ages. Figure 3 presents the proportion of HUD-assisted individuals by property type for individuals age 62 years or older.5

FIGURE 2. Distribution of HUD-Assisted Individuals by Property Type in 12 Study Jurisdictions, 2008, All Ages

FIGURE 2, Pie Chart: Housing Choice Vouchers (46.7%); Public Housing (30.9%); Other Multi Family (18.5%); Section 202 (3.1%); Other (0.7%); Section 811 & 202/162 (0.2%).


  • The largest proportion of individuals received Housing Choice Vouchers (47%).

  • Approximately 31% resided in public housing, while 19% lived in other multifamily properties.

  • Among individuals age 62 years or older, public housing, other multifamily, and Housing Choice vouchers were the most common (31%, 29%, and 26%), with an additional 14% residing in Section 202 properties.

FIGURE 3. Distribution of HUD-Assisted Individuals by Property Type in 12 Study Jurisdictions, 2008, Age 62+ Years

FIGURE 3, Pie Chart: Public Housing (30.6%); Housing Choice Vouchers (26.1%); Other Multi Family (29.2%); Section 202 (13.5%); Section 811 & 202/162 (0.1%); Other (0.5%).


A2. HUD Elderly Properties

Figure 4 presents the proportion of properties that were defined as elderly6 among all public housing properties (from PIC) and private assisted multifamily properties (from TRACS) within the 12 study jurisdictions. Half of all multifamily properties had an individual age 62 years or older in over 50% of their units. Approximately 42% of all public housing properties were either designated as elderly by public housing authorities (PHAs) or had an individual age 62 years or older in over 50% of their units. Both of these proportions varied greatly by the 12 study jurisdictions, from 30% of multifamily properties in Durham-Chapel Hill, North Carolina to 68% of multifamily properties in Bridgeport-Stamford-Norwalk, Connecticut and only 19% of public housing properties in Cleveland compared to 90% of public housing properties in Vermont (data not shown, see Appendix D).

FIGURE 4. Proportion of HUD-Assisted Properties in 12 Study Jurisdictions Defined as "Elderly", 2008
Property TypesNumber of PropertiesPercent of Properties
Multifamily elderly properties (reported in TRACS)1,81150.0%
Public Housing elderly properties (reported in PIC)33542.0%

B. Calculate HUD and Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008

B1. Match Rate: HUD-Assisted Individuals to Medicare or Medicaid Enrollment Data

The main objective of this study task was to determine the ability to link HUD-assisted individuals to Medicare and Medicaid administrative data (i.e., enrolled in Medicare or Medicaid health care programs) in the 12 study jurisdictions. We accomplished this objective by calculating two measures that represent the ability to link the HUD and CMS data sources:

  1. Social Security Number (SSN) Match Rate. The proportion of HUD-assisted individuals enrolled in Medicare and/or Medicaid programs determined solely on a matched SSN between the datasets.

  2. Restrictive Match Rate. A fairly conservative method on which we established whether the individuals receiving housing assistance were definitively included in the Medicare and/or Medicaid data by requiring a direct match on SSN, gender, and date of birth.7

Figure 5 presents six separate SSN match rates. We calculated the match rate to Medicaid-only, Medicaid and Medicare, and Medicare only separately for HUD-assisted individuals less than age 65, and individuals age 65 or older. A match is defined as being present in both datasets at any point during 2008.

  • Approximately 73% of HUD-assisted individuals less than age 65 and 96% of recipients age 65 or older matched to Medicaid, Medicare, or both.

  • Among HUD-assisted individuals less than age 65, 66% linked to Medicaid-only, 6% link to both Medicare and Medicaid, and 1% link to Medicare only.

  • Among HUD-assisted individuals age 65 or older, 63% linked to both Medicare and Medicaid and 30% linked to Medicare only. Less than 3% linked to Medicaid-only. It is most likely that the 3% who linked to Medicaid-only are not eligible for Medicare because they or their spouse did not make payroll tax contributions for ten or more years.

FIGURE 5. Medicare and Medicaid SSN Match Rates for HUD-Assisted Individuals, by Age Group, 2008

FIGURE 5, Stacked Bar Chart: <65 years--Medicaid (66.4%), Medicaid & Medicare (5.8%), Medicare (1.2%), Total (73.4%); 65+ years--Medicaid (2.6%), Medicaid & Medicare (63.0%), Medicare (30.2%), Total (95.8%).


Figure 6 presents the same six match rates as Figure 5 using the restrictive match rate criterion. As expected, the match rate lowered given the more stringent criteria. The match rate for HUD-assisted individuals under age 65 decreased by four percentage points and by nine percentage points for recipients age 65 or older.

  • Approximately 69% of HUD-assisted individuals less than age 65 and 87% of recipients age 65 or older matched to Medicaid, Medicare, or both.

  • Among HUD-assisted individuals less than age 65, 63% linked to Medicaid only, 5% link to both Medicare and Medicaid, and 1% link to Medicare only.

  • Among HUD-assisted individuals age 65 or older, 58% linked to both Medicare and Medicaid and 27% linked to Medicare only. Approximately 2% linked to Medicaid only. It is most likely that the 2% who linked to Medicaid only are not eligible for Medicare because they or their spouse did not pay payroll tax contributions for ten or more years.

FIGURE 6. Medicare and Medicaid Restrictive Match Rate for HUD-Assisted Individuals, by Age Group, 2008

FIGURE 6, Stacked Bar Chart: <65 years--Medicaid (62.5%), Medicaid & Medicare (5.5%), Medicare (1.1%), Total (69.1%); 65+ years--Medicaid (2.3%), Medicaid & Medicare (57.8%), Medicare (27.3%), Total (27.3%).]


We also calculated the restrictive match rate for over age 65 by property type and geographic area (data not shown, see Appendix D).

  • While the range of restrictive match rate did not vary widely across property types, the match rate was highest for Section 202 (elderly) properties at 90%. Individuals residing in public housing had the lowest match rate at 84%.

  • These percentages are also high across geographic areas (86.2% in New York-Northern New Jersey-Long Island to 94.2% in Vermont) with the exception of Richmond, where only 31.2% linked.

The analyses in the remainder of the report rely on the sample that met the restricted match rate criteria. The more conservative restrictive match rate creates a sample for which we can be very confident the HUD-assisted individual was in fact enrolled in Medicare and/or Medicaid.

B2. HUD-Assisted Individuals Matched to CMS Administrative Data Compared to Recipients Who Did Not Match

Next we wanted to better understand the characteristics of HUD-assisted individuals that matched to the CMS administrative data relative to the recipients that did not match using the restrictive approach.

FIGURE 7. Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid and Who Did Not Link, Age 65 or Older, 2008
 Linked to 2008 Medicare or Medicaid Enrollment Files(N=305,153)Did Not Link to 2008 Medicare or Medicaid Enrollment Files(N=44,505)
N%N%
  1. As defined in the HUD data source.
Gender
Male91,24129.9%13,41630.1%
Female213,91270.1%30,96569.6%
Unknown/declined to report00.0%1240.3%
Race/ethnicity
White non-Hispanic65,94221.6%6,46214.5%
Hispanic46,76015.3%8,52819.2%
Black non-Hispanic48,78516.0%10,72924.1%
Asian11,3733.7%1,2502.8%
Other8850.3%1030.2%
Unknown/declined to report131,40843.1%17,43339.2%
Property type
Public housing89,01329.2%16,70937.5%
Housing Choice Vouchers74,79624.5%9,00520.2%
Other multifamily91,46930.0%12,83628.8%
Section 20248,42415.9%5,75012.9%
Other multifamily2980.1%370.1%
Other1,1530.4%1680.4%
Elderly designated property1
Yes160,39852.6%20,10245.2%
No69,37722.7%15,32834.4%
Unknown75,37824.7%9,07520.4%
SSI
Receives SSI134,48444.1%16,94438.1%

Figure 7 presents a comparison of HUD-assisted individuals age 65 or older8 who were linked to Medicaid or Medicare enrollment files to those that were not linked based on the restrictive match rate. In other words, we compared the 87% of HUD-assisted individuals age 65 or older who linked to the 13% who did not link to any CMS data source. Individuals who linked to Medicare or Medicaid were more likely to be White (22% vs. 15% of individuals who did not link);9 less likely to live in public housing (29% vs. 38%); and more likely to receive Supplemental Security Income (SSI) (44% vs. 38%).

C. Estimate HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program (i.e., Medicaid)

As detailed in an October 2012 Center on Budget and Policy Priorities (CBPP) memo,10 HUD-assisted individuals may have unreimbursed medical expenses deducted from their income for the purpose of determining the level of assistance provided by HUD. According to the CBPP, approximately 700,000 assisted households claim such expenses, which increase HUD rental subsidy costs by $400-$500 million per year. Therefore, HUD is interested in understanding what proportion of HUD-assisted Medicare beneficiaries (defined as HUD-assisted individuals who linked to Medicare based on the restrictive match criteria) are enrolled in programs that could reduce their out-of-pocket health care expenses and therefore reduce HUD's costs. These programs include:

  1. Medicare Part D (drug coverage) Low Income Subsidy (LIS). Medicare Part D provides coverage for prescription medications and the LIS program assists beneficiaries with paying the premium and out-of-pocket expenses for their medication.

  2. Medicare Savings Program (MSP). This is more commonly referred to as being a Medicare and Medicaid enrollee (MME) or dually eligible. Medicare beneficiaries may be eligible for Medicaid to assist with out-of-pocket Medicare expenses (premiums and copayments) and to receive Medicaid-covered services that are not covered by Medicare, including and importantly long-term services and supports.

Figure 8 presents the proportion of HUD-assisted Medicare beneficiariesage 65 years or older enrolled in Part D LIS and Medicaid.

  • Approximately 90% of the Medicare beneficiaries had Part D coverage (data not shown, see Appendix D). Among those with Part D coverage, 80% received premium and out-of-pocket assistance through the LIS Program.

  • Approximately 68% of Medicare beneficiaries were dually enrolled in Medicare and Medicaid. Almost 89% of those dually enrolled in Medicare and Medicaid were receiving Medicaid assistance with Medicare expenses and fully eligible for Medicaid-covered services.

FIGURE 8. HUD-Assisted Medicare Beneficiaries (Age 65+): Part D LIS Status and Medicaid Participation

FIGURE 8, Bar Chart: With Part D LIS (79.7%), With Medicaid (67.9%).


It is important to note that the results presented in Figure 8 above are the proportion of all HUD-assisted Medicare beneficiaries age 65+ who match based on the restrictive match criteria. It is not the proportion of HUD-assisted Medicare beneficiaries who are eligible for these programs based on national criteria for Part D LIS and criteria established by each state for Medicaid benefits. In other words, we would not expect the proportion to be 100% as not all HUD-assisted Medicare beneficiaries are eligible for these programs because the income criteria for HUD-assisted housing can be higher than that for health insurance assistance. In fact, among HUD-assisted individuals, an examination of the proportion of Medicare beneficiaries dually enrolled in Medicaid, where enrollment should be automatic for the Medicare Part D prescription drug LIS, found 99% received LIS, while only about one-third of those with Medicare and no Medicaid coverage had this coverage.

D. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions

After we identified the HUD-assisted individuals who were Medicare and/or Medicaid beneficiaries in 2008, the next objective was to compare HUD-assisted Medicare beneficiaries to unassisted beneficiaries in the community with regards to enrollment, chronic conditions, health care payment and utilization. Although our overall study focused on individuals age 62 and older receiving housing assistance, we narrowed this analysis to individuals age 65 and older because the vast majority of individuals age 65 and older have Medicare coverage. Before comparing groups of beneficiaries, we first refined our study sample of Medicare beneficiaries (both HUD-assisted and unassisted beneficiaries in the 12 study jurisdictions).

Figure 9 presents a graphical depiction of the inclusion criteria, regardless of age. Medicare beneficiaries had to be enrolled in Part A Hospital and Part B Physician Services for all 12 months of 2008 or up until death and have no Medicare managed care (i.e., Medicare Part C or Medicare Advantage)11 enrollment in 2008. This reduced the sample of Medicare beneficiaries by 38%, down to 3.6 million. Approximately 7% of the Medicare beneficiaries were HUD-assisted individuals and 93% were unassisted in the community.

FIGURE 9. Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison in 12 Jurisdictions, All Ages, 2008

FIGURE 9, Inverted Pyramid: Top layer, All Medicare beneficiaries (N=5,752,662); Middle layer, Enrolled in A&B for entire 12 month period or up until death (N=4,799,074 or 83%); Lower layer, Medicaid Fee-for-service only (N=3,562,032 or 62%).


We expected HUD-assisted beneficiaries to differ dramatically from unassisted beneficiaries in the community with regards to health status, utilization, and payment given the differences in demographic factors (e.g., race, income) and Medicare and Medicaid program eligibility due to low income (e.g., Part D LIS and Medicaid participation defined above). This was supported by preliminary comparisons of HUD-assisted Medicare beneficiaries age 65+ and unassisted Medicare beneficiaries age 65+ in the community. We found:

  • After limiting our sample to HUD-assisted Medicare beneficiaries with continuous enrollment in Medicare Parts A and B with no Medicare managed care during 2008, approximately 70% of HUD-assisted Medicare fee-for-service (FFS) beneficiaries age 65+ were dually enrolled in Medicaid compared to only 13% of unassisted Medicare FFS beneficiaries in the community.

  • The average total medical and prescription drug per member per month (PMPM) Medicare payment was $1,479 for HUD-assisted Medicare FFS beneficiaries age 65+ compared to only $937 for unassisted Medicare FFS beneficiaries age 65+ in the community.

The higher expenditures for HUD-assisted Medicare beneficiaries likely reflects the much higher proportion of MMEs in the HUD-assisted sample. According to a recent study by the Kaiser Family Foundation, spending for MMEs was 1.8 times higher than for Medicare/No Medicaid counterparts.12 Based on this knowledge, we stratified the Medicare beneficiaries into five sub-groups in an attempt to group similar beneficiaries into distinct samples to reduce variability between the HUD-assisted beneficiaries and the unassisted beneficiaries when comparing their Medicare enrollment characteristics, payments, and utilization. See Appendix C for detail on stratification and definitions for all five sub-groups. In this summary report, we present the results for the following two sub-groups:

  • Sub-group A1 (MMEs): MMEs, Age 65+, with no Medicare skilled nursing facility (SNF) or Medicaid nursing facility13 use in 2008.

  • Sub-group C1 (Medicare only): Medicare beneficiaries with no Medicaid enrollment, age 65+, no SNF use in 2008.

D1. Comparison of HUD-Assisted MMEs and Unassisted MMEs in the Community, Age 65+ in 12 Study Jurisdictions, 2008 (Sub-group A1)

In 2008 in the 12 study jurisdictions, there were 112,045 HUD-assisted MMEs and 249,490 unassisted FFS MMEs in the community age 65+. Figure 10, Figure 11, and Figure 12 present the key findings when comparing HUD-assisted MMEs and unassisted MMEs in the community.

Figure 10 compares the count of chronic condition groupings among MMEs. HUD-assisted MMEs, on average, experience greater chronic illness than unassisted MMEs in the community.

  • HUD-assisted MMEs are more likely to have multiple chronic diseases; 55% have five or more compared to 43% of unassisted MMEs in the community.

  • The prevalence of eight of the nine chronic condition categories is also higher among HUD-assisted Medicare beneficiaries (data not shown, see Appendix D). The chronic condition groupings are: Cardiovascular; Cancer; Endocrine and Renal; Alzheimer's-related; Depression; Musculoskeletal; Pulmonary; Opthalmic; and, Other (Anemia, Hyperlipidemia, Hyperplasia, Hypertension).

FIGURE 10. Medicare and Medicaid Enrollees (MMEs) with Chronic Conditions (Age 65+), 2008, 12 Study Jurisdictions

Medicare and Medicaid Enrollees (MMEs) with Chronic Conditions (Age 65+), 2008, 12 Study Jurisdictions


Medicare and Medicaid Enrollees (MMEs) with Chronic Conditions (Age 65+), 2008, 12 Study Jurisdictions

HUD assisted MMEsUnassisted MMEs in the community
0 Conditions5.49.8
1 Conditions5.98.3
2 Conditions8.711.1
3 Conditions12.013.8
4 Conditions13.614.0
5+ Conditions54.543.1

Figure 11 provides a comparison of health care service utilization between HUD-assisted FFS MMEs and FFS unassisted MMEs in the community. HUD-assisted MMEs were higher utilizers of health care services. With the exception of dialysis, the difference in utilization was statistically significant for all health care services.

  • The higher utilization for HUD-assisted MMEs was most notable for home health visits (31% higher), ambulatory surgery center (ASC) visits (45% higher), other procedures (78% higher), durable medical equipment (DME) (22% higher), and Part B and D drugs (22% and 24% higher, respectively).

  • HUD-assisted MME beneficiaries also had a 26% higher rate of physician office visits and 13% higher rate of emergency department visits than unassisted MME beneficiaries.

  • As expected, this higher utilization resulted in 16% higher average medical and pharmacy PMPM Medicare payments for HUD-assisted MME beneficiaries than unassisted MME beneficiaries ($1,222 vs. $1,054) (data not shown, see Appendix C).

FIGURE 11. Comparison of Fee-for-Service HUD-Assisted MME and Unassisted MME Medicare Health Care Utilization
Medicare Services Annual Utilization per 1000 Member Months1HUD-Assisted MME(N=112,045)Non-HUD-Assisted MME(N=249,490)Ratio of HUD-Assistedvs. Unassisted MMEs
MeanMean
NOTES:
  1. Utilization per 1000 member months calculated as number of total services across all MMEs divided by number of months enrolled for all MMEs in 2008 multiplied by 1000.
  2. Non-acute inpatient settings include long-term care settings, like inpatient rehabilitation facilities.
  3. This variable is the count of unique revenue center dates (as a proxy for visits) in the hospital outpatient setting for a given year.
  4. Limited to Medicare beneficiaries with Part D coverage.

** p<0.01; * p<0.05

Total admissions32.830.91.06**
   Acute stay admissions31.429.41.07**
   Other inpatient admissions21.41.60.90**
Hospital readmissions5.24.91.06*
Medicare HH visits581.5445.51.30*
Medicare hospice days36.9208.10.18*
Hospital outpatient visits3686.5590.51.16**
Total ED visits58.451.61.13**
   ED visits without an admission36.630.61.19**
   ED visits resulting in an admission21.921.01.04**
Physician office visits1,652.31,307.91.26**
ASC visits14.510.01.45**
Dialysis events17.017.30.98
Anesthesia events27.823.71.17**
Imaging events510.5420.11.21**
Test events1,762.21,428.51.23**
Other procedures1,464.6821.61.78**
DME369.2301.41.22**
Part B drugs296.1241.81.22**
Part D4 drugs5,080.14,094.31.24**

Figure 12 compares the utilization of Medicaid-covered services14 by FFS HUD-assisted MMEs and unassisted MMEs in the community. HUD-assisted MMEs use a much greater amount of Medicaid-covered services than unassisted MMEs in the community. It should be noted that one driver of the higher utilization may be that a higher proportion of HUD-assisted MMEs were eligible for full Medicaid benefits than unassisted MMEs in the community (91% vs. 85%, respectively):

  • HUD-assisted MMEs used over 100% more Personal Care services,15 80% more "other HCBS", and over 67% more DME services covered by Medicaid than unassisted MMEs in the community. This suggests that the poorer health status of HUD-assisted MMEs increases their likelihood of needing HCBS and/or relationships between the property management and community supports make these services more available to HUD-assisted individuals.

  • On the other hand, unassisted MMEs in the community used 60% more Residential Care settings than HUD-assisted MMEs. Residential care settings, including assisted living facilities, have traditionally provided assistance and oversight to persons with physical and mental impairments who cannot or choose not to live at home alone.16

  • Due to the higher utilization of services, HUD-assisted MMEs Medicaid PMPM of $1,180 was 32% higher than unassisted MMEs PMPM of $895 (data not shown, see Appendix D).

FIGURE 12. Comparison of Fee-for-Service HUD-Assisted MMEs and Unassisted MMEs Medicaid Health Care Utilization1
Medicare Service Utilization per 1000 Member MonthsHUD-Assisted MMEswith Medicaid FFS(N=106,764)Unassisted MMEs withMedicaid FFS(N=227,186) Ratio of HUD-Assistedvs. Unassisted MMEs
MeanMean
NOTES:
  1. Limited to individuals with no Medicaid managed care enrollment since our data do not include managed care utilization and cost.
  2. Private duty nursing, adult day care, HH, rehab, targeted case management, transportation, and hospice.

** p<0.01

Personal Care services4,512.42,149.12.09**
Residential care38.963.70.61**
DME380.0227.71.67**
Other HCBS23,309.81,840.61.79**

D2. Comparison of HUD-Assisted Medicare/No Medicaid17 Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries in the Community, Age 65+ in 12 Study Jurisdictions, 2008 (Sub-group C1)

FIGURE 13. HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries, 2008, by Gender

FIGURE 13, Bar Chart: Male--HUD assisted beneficiaries (28.7%), Unassisted beneficiaries in the community (43.0%); Female--HUD assisted beneficiaries (71.3%), Unassisted beneficiaries in the community (57.0%).


During 2008 in the 12 study jurisdictions, there were 46,806 FFS HUD-assisted Medicare/No Medicaid beneficiaries and 2,318,394 unassisted beneficiaries in the community age 65+. Figure 13, Figure 14 and Figure 15 present the key findings when comparing HUD-assisted Medicare beneficiaries and unassisted beneficiaries in the community. Unlike the previous comparison of MMEs, these estimates do not control for income. We would expect that the HUD-assisted Medicare/No Medicaid beneficiaries would generally have lower income and experience greater social and economic constraints to good health than their community counterparts.

FIGURE 14. HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Beneficiaries, 2008, by Race/Ethnicity

FIGURE 14, Bar Chart: Non-Hispanic White--HUD assisted beneficiaries (64.3%), Unassisted beneficiaries in the community (87.3%); Hispanic--HUD assisted beneficiaries (9.5%), Unassisted beneficiaries in the community (3.2%); Black--HUD assisted beneficiaries (24.2%), Unassisted beneficiaries in the community (6.6%); Other--HUD assisted beneficiaries (2.0%), Unassisted beneficiaries in the community (2.9%).

* Other category includes categories of Asian/Pacific Islander, American Indian/Alaska Native, Other, Unknown. Race/ethnicity is based on CMS enrollment data.


Figure 13, Figure 14, and Figure 15 summarize the distribution of HUD-assisted Medicare beneficiaries and unassisted Medicare beneficiaries in the community by various demographic characteristics. As expected, the profile of the HUD-assisted beneficiaries does differ from the unassisted beneficiaries:

  • Medicare/No Medicaid HUD-assisted beneficiaries are much more likely to be female (71% vs. 57%), Black (24% vs. 7%), and over age 85 years (23% vs. 16%) than unassisted beneficiaries (age data not shown, see Appendix D).

  • Over 87% of unassisted beneficiaries in the community were non-Hispanic White compared to only 64% of HUD-assisted beneficiaries.

  • HUD-assisted Medicare/No Medicaid beneficiaries are also more likely to have received Part D LIS (32% vs. 6%) than unassisted beneficiaries in the community.

FIGURE 15. Fee-for-Service HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries, 2008, by Part D LIS Status

FIGURE 15, Bar Chart: HUD-assisted beneficiaries (32.4%); Unassisted beneficiaries in the community (5.9%).


Figure 16 shows the proportion of HUD-assisted and unassisted Medicare beneficiaries with chronic conditions. Unlike the HUD-assisted MMEs age 65 or older (Sub-group A1 above), the HUD-assisted Medicare/No Medicaid beneficiaries of the same age group are more similar to the unassisted Medicare/No Medicaid beneficiaries with regards to health status.

  • Approximately 14% of HUD-assisted Medicare/No Medicaid beneficiaries had zero chronic conditions compared to 13% of unassisted Medicare/No Medicaid beneficiaries. However, 38% of HUD-assisted Medicare/No Medicaid beneficiaries had five or more conditions compared to only 33% of unassisted Medicare/No Medicaid beneficiaries in the community.

  • The prevalence of chronic conditions was not consistently higher for the HUD-assisted Medicare/No Medicaid beneficiaries. In fact, the unassisted Medicare/No Medicaid beneficiaries had a higher proportion of cancer (10% vs. 9%) and ophthalmic conditions (37% vs. 32%).

FIGURE 16. HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Beneficiaries, Age 65+, 2008, by Chronic Condition Category County and Prevalence
CCW ConditionHUD-Assisted Medicare BeneficiariesUnassisted Medicare Beneficiaries
N%N%
Prevalence of Chronic Conditions46,806100.0%2,318,394100.0%
0 Conditions6,62414.2%299,93612.9%
1 Condition3,9728.5%249,96210.8%
2 Conditions5,24511.2%312,46213.5%
3 Conditions6,47213.8%358,21915.5%
4 Conditions6,66914.2%339,58614.6%
5 or More Conditions17,82438.1%758,22932.7%
Prevalence of Chronic Conditions by Category
Cardiovascular20,63844.1%952,06141.1%
Cancer4,1058.8%238,41610.3%
Endocrine and Renal19,31341.3%800,92434.5%
Alzheimer's-related3,5437.6%179,3047.7%
Depression3,9308.4%167,1197.2%
Musculoskeletal15,49433.1%719,06531.0%
Pulmonary6,77714.5%246,61910.6%
Ophthalmic15,09832.3%847,51236.6%
Other (Anemia, Hyperlipidemia, Hyperplasia, Hypertension)  35,03074.8%1,696,31973.2%

Figure 17 provides a comparison of Medicare health care utilization between the Medicare/No Medicaid HUD-assisted beneficiaries and unassisted beneficiaries in the community. The utilization by HUD-assisted beneficiaries was higher for some services and lower for others when compared to unassisted Medicare/No Medicaid beneficiaries. In particular:

  • HUD-assisted Medicare/No Medicaid beneficiaries used 42% more emergency room visits per 1000 member months than unassisted beneficiaries.

  • The admission rate was 19% higher and the readmission rate was 17% higher for HUD-assisted Medicare/No Medicaid beneficiaries.

  • On the other hand, HUD-assisted Medicare/No Medicaid beneficiaries were more likely to not have any office visits. Approximately 13% of HUD-assisted beneficiaries did not have any office visits during 2008 compared to 9% of unassisted Medicare/No Medicaid beneficiaries (data not shown, see Appendix D).

  • Overall, the Medicare PMPM was 8% higher for HUD-assisted Medicare/No Medicaid beneficiaries than unassisted Medicare/No Medicaid beneficiaries ($617 vs. $574).

FIGURE 17. Comparison of Fee-for-Service HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries' Medicare Health Care Utilization
Medicare Services Annual Utilization per 1000 Member MonthsHUD-Assisted Medicare Beneficiaries(N=46,806)Unassisted Medicare Beneficiaries(N=2,318,394)Ratio of HUD-Assisteds. Unassisted
MeanMean
  1. Non-acute inpatient settings include long-term care settings, like inpatient rehabilitation facilities.

** p<0.01; * p<0.05

Total Admissions23.019.31.19**
   Acute stay admissions21.718.21.19**
   Other inpatient admissions11.31.11.20**
Hospital Readmissions2.72.31.17**
Medicare HH visits220.5136.01.62**
Medicare hospice days36.577.60.47**
Total outpatient visits444.8409.51.09**
Total ED visits43.030.41.42**
   ED visits without an admission27.218.81.45**
   ED visits resulting in an admission15.811.61.37**
Physician office visits1,145.21,129.11.01*
ASC visits12.716.00.79**
Dialysis events6.54.81.36**
Anesthesia events25.929.20.89**
Imaging events378.2382.20.99*
Test events1,347.01,416.90.95**
Other procedures527.2635.40.83**
DME180.7139.01.30**
Part B Drugs205.7242.20.85**
Part D Drugs2983.92496.01.20**

V. DISCUSSION

To our knowledge, this study was the first attempt to link the HUD individual tenant-level and CMS beneficiary level administrative data (enrollment and payment and utilization claims data). This study demonstrates the feasibility of linking these rich data sources to conduct numerous informative analyses that can shed light on the advisability of investing in programs that might improve the health and well-being of individuals with HUD-assisted housing. This study also provides descriptive comparisons of HUD-assisted Medicare beneficiaries to unassisted Medicare beneficiaries, highlighting areas of potential future analyses.

A. Summary of Results

Match Rate

We found that among HUD-assisted individuals age 65 or older in 2008, 93% matched to Medicare based on SSN only; 85% matched to Medicare when requiring a match on SSN, gender, and date of birth.

HUD-Assisted Medicare Beneficiaries Enrollment in Part D Subsidy Assistance Programs or Medicaid

Some HUD-assisted Medicare beneficiaries age 65 and older not enrolled in programs that could reduce their out-of-pocket health care expenses, such as the Medicare Part D (drug coverage) LIS and MSP (being an MME, may be costing HUD). Approximately 68% of HUD-assisted Medicare beneficiaries were dually enrolled in Medicare and Medicaid. Among HUD-assisted Medicare beneficiaries with Part D coverage, 80% received premium and out-of-pocket assistance. We would not expect the proportion to be 100% as not all HUD-assisted Medicare beneficiaries are eligible for these programs because the income criteria for HUD-assisted housing can be higher than that for health insurance assistance. We plan to examine the issue of participation among eligible individuals in future analyses.

Compare Medicare and Medicaid Payment and Utilization

The average total FFS medical and prescription drug PMPM Medicare paid amount was $1,479 for HUD-assisted Medicare beneficiaries age 65 and older compared to only $937 for unassisted Medicare beneficiaries in the community. The higher expenditures for HUD-assisted Medicare beneficiaries likely reflects the much higher proportion of MMEs in the HUD-assisted sample than in the unassisted beneficiaries in the community (70% vs. 13%), as spending for MMEs is almost two times higher than for Medicare/No Medicaid counterparts.

Given the difference in demographic characteristics, health conditions, and MME enrollment among HUD-assisted beneficiaries and unassisted beneficiaries in the community, we created similar sub-groups for comparison, including MMEs age 65+ with no SNF or nursing facility use and Medicare/No Medicaid age 65+ with no SNF use.

Fee-for-Service MME, Age 65+, No SNF/Nursing Facility Use

HUD-assisted MMEs age 65+ had more chronic conditions which translated into higher health care utilization and payments than unassisted MMEs in the community. HUD-assisted MME beneficiaries' average Medicare FFS PMPM was $1,222 compared to $1,054 PMPM for unassisted beneficiaries. This was driven by higher utilization for all services, with the exception of hospice. Most notably, HUD-assisted MMEs had more home health visits (31% higher), ASC visits (45% higher), physician office visits (26%), and emergency department visits (13%). These results are strictly descriptive. In other words, they are not adjusted for demographic characteristics or health care conditions beyond MME status that are known to be associated with health care utilization.

We found similar results when comparing Medicaid-covered FFS health care service utilization and payment by MMEs. HUD-assisted MMEs Medicaid FFS average PMPM was 32% higher than unassisted MMEs ($1,180 vs. $895). HUD-assisted MMEs used over 200% more Personal Care services, 80% more "other HCBS", and over 67% more DME services covered by Medicaid than unassisted MMEs in the community. This suggests that the poorer health status of HUD-assisted MMEs increases their likelihood of needing HCBS and/or relationships between the property management and community supports make these services more available to HUD-assisted individuals. On the other hand, unassisted MMEs in the community used 60% more Residential Care services, which includes assisted living, than HUD-assisted MMEs.

Fee-for-Service Medicare/No Medicaid Beneficiaries, Age 65+

While the comparison among MMEs found that HUD-assisted beneficiaries had more chronic conditions and higher health care utilization and Medicare payment almost across all health care services, the results of the Medicare/No Medicaid beneficiary comparison was mixed. Ultimately, HUD-assisted Medicare/No Medicaid beneficiaries had 8% higher Medicare PMPM than unassisted beneficiaries. The HUD-assisted FFS Medicare/No Medicaid beneficiaries had substantially higher utilization of costly services. HUD-assisted beneficiaries used 42% more emergency room visits, 19% more admissions, and had a 17% higher readmission rate. On the other hand, 87% of HUD-assisted beneficiaries had a physician office visit compared to 91% of unassisted beneficiaries. Unlike the previous comparison of MMEs, these estimates do not control for income. We would expect that the HUD-assisted Medicare/No Medicaid beneficiaries would generally have lower income and experience greater social and economic constraints to good health than their community counterparts. Similar to above, these results are descriptive only and, therefore, additional analyses would need to be conducted to determine if the results remain when controlling for differences among the two samples.

B. Limitations

While this study provides insight into the feasibility of linking HUD and CMS administrative data and a preliminary analysis of health and health care utilization among elderly HUD-assisted individuals who are Medicare beneficiaries, this study is not without limitations. First, because New York represents over half the individuals studied, the differences in New York's assisted population could account for a number of the observed differences. Thus, the findings of this report are only suggestive of future paths for research. Second, the Medicare plan paid amounts are not standardized for geographic differences in payment rates. Therefore, care must be used when comparing average payment for a particular service across the geographies. However, the distribution by the 12 geographies of the HUD-assisted beneficiaries and the unassisted beneficiaries was similar in our datasets reducing concerns about payments being driven by differences in geography. Third, we were unable to identify all nursing facility stays, regardless of payer, with our current data sources. Therefore, we had to eliminate any Medicare beneficiaries with any Medicaid nursing facility or Medicare SNF stay as a proxy for nursing facility use. While this only reduced our Medicare/No Medicaid sample by 5%, it reduced our sample of MMEs by 19%.

Finally, our analyses are descriptive and no policy recommendations should be derived based on these results alone. The results are, however, a starting point. This research demonstrates that it is feasible to match health and housing agency administrative data and suggests that while we have much to learn about the relationship between housing and health, the methods demonstrated here can be used to build that knowledge cost-effectively and in ways that may ultimately improve both housing and health program effectiveness. See future analysis for recommendations related to addressing these limitations.

C. Future Analysis

This study demonstrates that it is feasible to link HUD tenant data to CMS administrative data. The descriptive results summarized above highlight key areas for future analysis to better understand the health and health care utilization of elderly HUD-assisted individuals enrolled in Medicare. This includes supplementing current data sources with additional CMS data, refining matching algorithms and study samples to better determine HUD-assisted individuals' eligibility and enrollment in Medicare programs, and conducting multivariate regressions to determine if the differences observed in descriptive comparisons remain after adjusting for confounders.

  1. Refine match algorithm. Currently, we matched the datasets based on SSN and then SSN, gender, and date of birth without any discrepancies. We can refine the algorithms to require more than just a match on SSN, but not be as conservative as the "restrictive match." For instance, we may want to allow some differences in date of birth (i.e., consider it a match if mm/yy aligns).

  2. Look more closely at the eligibility/enrollment issue to estimate missed opportunities for HUD-assisted elderly individuals to enroll in CMS programs. As highlighted above, we calculated the proportion of all HUD-assisted Medicare beneficiaries, age 65 or older, receiving Part D LIS and dually enrolled in Medicaid. Future analyses could limit the denominator to HUD-assisted beneficiaries eligible for the Part D LIS program and Medicaid program separately to determine what proportion of those eligible were enrolled (i.e., look at income in HUD data vs. enrollment in Medical Savings Program, for example). This analysis could assist HUD in identifying if under-enrollment among those eligible is a missed opportunity for HUD-assisted individuals and better estimate the potential cost savings to HUD of increasing enrollment into these programs.

  3. Conduct multivariate regressions to determine if the differences observed in descriptive analyses above remain after adjusting for potential confounders available in the HUD and CMS data sources (e.g., race/ethnicity, age, geography, presence of chronic conditions, market characteristics).

  4. Update analysis with more recent data. The current analysis is based on 2008 data. While we had 2009 and 2010 Medicare data, 2008 was the most recent year of Medicaid data. We could update the analysis, with the exception of Medicaid-covered service comparison, on 2010 data.

  5. Add to existing data. We recommend appending the Minimum Data Set (MDS) detailing the utilization of Nursing Facility services, regardless of payer (Medicare, Medicaid, private payer) and the Timeline file. MDS would allow us to determine which HUD-assisted Medicare beneficiaries used nursing facility services compared to unassisted Medicare beneficiaries. The Timeline file creates a flag for each day of the year for each Medicare beneficiary. Each day is assigned as: community, community with home health care, nursing facility, SNF, inpatient, and death. This data would allow us to look at key events and paths following the event (e.g., hospital stay) and the associated payments. Finally, we suggest expanding this analysis to the entire nation. As mentioned previously, one limitation of this analysis is it is heavily influenced by the New York City metropolitan area and it is unknown if any findings are representative of HUD-assisted elderly housing recipients nationally.


Appendices


 

TABLE OF CONTENTS

APPENDIX A: DATA SOURCES AND VARIABLE DEFINITIONS
Introduction
A. Data Sources
B. Variable Definitions
APPENDIX B: STUDY SAMPLES AND ANALYTIC APPROACH
A. Study Samples
B. Analytic Approach
APPENDIX C: 2008 DETAILED RESULTS
Introduction
A. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions
B. Calculate HUD and Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008
C. Estimate Proportion of HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program
D. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions
APPENDIX D: SUPPLEMENTAL RESULTS, ALL YEARS
Objective A: Understand Demographic Characteristics of HUD-Assisted Individuals/Households in the 12 Study Jurisdictions
Objective B: Calculate Number of HUD-Assisted Individual Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008
Objective C: Estimate HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs of Medical Savings Program
Objective D: Compare Medicare and Medicaid Cost and Utilization for HUD-Assisted Medicare Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions
Part D LIS Enrollment
NOTES
LIST OF FIGURES
FIGURE B1: Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison, All Ages
FIGURE C1: Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison
LIST OF TABLES
TABLE A1: Property Type Variable Definition
TABLE A2: Assignment of Chronic Conditions into 9 Chronic Condition Groups
TABLE A3: Medicare Health Care Utilization Variable Definitions
TABLE A4: Medicare Health Care Spending Variable Definitions
TABLE B1: HUD-Assisted Individuals 2008 Unique Observations
TABLE C1: Demographic Characteristics of HUD-Assisted Individuals, 2008, 12 Study Jurisdictions
TABLE C2: Living Arrangement, All Ages and Age 62+ Years
TABLE C3: Living Arrangement, Household Level, All Ages
TABLE C4: Full and Partial HUD Housing Assistance, 2008, by Age Group
TABLE C5: Proportion of HUD-Assisted Properties in 12 Study Jurisdictions Defined as "Elderly"
TABLE C6: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals, 2008, by Age Group
TABLE C7: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI, 2008, by Age Group
TABLE C8: Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid, 2008, by Age Group
TABLE C9: HUD-Assisted Medicare Beneficiary Enrollment in Part D LIS, 2008
TABLE C10: HUD-Assisted Medicare Beneficiary Enrollment in Medicaid and Medicaid Coverage Characteristics, 2008
TABLE C11: Five Sub-groups for Comparison of Health and Health Care Utilization Among HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community
TABLE C12: HUD-Assisted and Unassisted MME Demographic Characteristics
TABLE C13: HUD-Assisted and Unassisted MME Medicare and Medicaid Enrollment Characteristics
TABLE C14: MMEs with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
TABLE C15: Comparison of HUD-Assisted MME and Unassisted MME Medicare Health Care Utilization
TABLE C16: Comparison of HUD-Assisted MME and Unassisted MME Medicaid Health Care Utilization per 1000 Member Months
TABLE C17: Comparison of HUD-Assisted MME and Unassisted MEE Medicaid Health Care Utilization, Proportion of MMEs with Any Use
TABLE C18: HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Beneficiaries, Age 65+, 2008, by Demographic and Coverage Characteristics
TABLE C19: HUD-Assisted Medicare/No Medicaid and Unassisted Beneficiaries with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
TABLE C20: Comparison of HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries' Medicare Health Care Utilization
TABLE D-A1: HUD-Assisted Individuals Full vs. Partial Year, by Year (All Ages)
TABLE D-A2: HUD-Assisted Individuals Full vs. Partial Year, by Year (Age 62+)
TABLE D-A3: HUD-Assisted Individuals Full vs. Partial Year, by Year (Age 65+)
TABLE D-A4: HUD-Assisted Individuals Across Consecutive Years Full vs. Partial
TABLE D-A5: Living Arrangement, HUD-Assisted Individuals, All Ages and Age 62+ Years
TABLE D-A6: Living Arrangement, HUD-Assisted Households, All Ages
TABLE D-A7a: HUD-Assisted Individual Demographics, 2008
TABLE D-A7b: HUD-Assisted Individuals Demographics, 2007-2009
TABLE D-A8: HUD-Assisted Individual Demographics by Geographic Area
TABLE D-A9: Demographics by Age Group
TABLE D-A10: Demographics by Property Type
TABLE D-A11: Demographics by Subside Type
TABLE D-A12: Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+
TABLE D-A13: Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+ by Geographic Location
TABLE D-A14: Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+
TABLE D-A15: Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+ or Designated Elderly by Geographic Location
TABLE D-B1: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals, 2008, by Age Group
TABLE D-B2: Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI, 2008
TABLE D-B3: Restrictive Match Rates: All Ages
TABLE D-B4: Restrictive Match Rates: <62 Years of Age
TABLE D-B5: Restrictive Match Rates: Age 62 to 64
TABLE D-B6: Restrictive Match Rates: Age 65+
TABLE D-B7: Restrictive Match Rates: Heads of Household, All Ages
TABLE D-B8: Restrictive Match Rates: Heads of Household, <62 Years of Age
TABLE D-B9: Restrictive Match Rates: Heads of Household, Age 62 to 64
TABLE D-B10: Restrictive Match Rates: Heads of Household, Age 65+
TABLE D-B11: Restrictive Match Rates: Heads of Household, Receiving SSI
TABLE D-B12: Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid by Age Group, 2008
TABLE D-C1: Health Insurance Characteristics of Those that Link by Year by Age
TABLE D-C2: Health Insurance Characteristics of HUD Heads of Households Receiving SSI that Link in 2008 by Age
TABLE D-D1: Final 2008 Samples for Utilization and Cost Analysis
TABLE D-D2: Final 2008 Samples Limited to Individual with Medicaid Fee for Service -- Managed Care
TABLE D-D.A1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries, 2008
TABLE D-D.A2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.A3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.A4: Medicare and Beneficiary Health Care Costs
TABLE D-D.A1.1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.A1.2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.A1.3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.A1.4: Medicare and Beneficiary Health Care Costs
TABLE D-D.B1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.B2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.B3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.B4: Medicare and Beneficiary Health Care Costs
TABLE D-D.C1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.C2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.C3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.C4: Medicare and Beneficiary Health Care Costs
TABLE D-D.C1.1: Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
TABLE D-D.C1.2: Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
TABLE D-D.C1.3: Medicare Health Care Utilization per 1000 Member Months
TABLE D-D.C1.4: Medicare and Beneficiary Health Care Costs
TABLE D1: Proportion of Beneficiaries with Part D Coverage, Enrolled in LIS

 

APPENDIX A. DATA SOURCES AND VARIABLE DEFINITIONS

Introduction

The following table provides a guide to HUD and CMS key concepts and terms.

Guide to Housing Assistance and Medicare and Medicaid Services
NOTES:
  1. Program statistics in this table come from HUD's Picture of Subsidized Households dataset for 2009, available from http://www.huduser.org/portal/datasets/picture/yearlydata.html.
  2. U.S. Department of Housing and Urban Development. HUD's Public Housing Program. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/topics/rental_assistance/phprog.
  3. U.S. Department of Housing and Urban Development. Housing Choice Vouchers Fact Sheet. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/topics/housing_choice_voucher_program_section_8.
  4. U.S. Department of Housing and Urban Development. Section 202 Supportive Housing for the Elderly Program. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/program_offices/housing/mfh/progdesc/eld202.
  5. Note that some tables refer to both "Section 202" and "Section 202/162." A small number of 202 properties contain units for non-elderly disabled that were funded with rental assistance created by Section 162 of the Housing and Community Development Act of 1987. Section 162 has been superseded by the Section 811 program, which supports housing for disabled individuals.
  6. U.S. Department of Housing and Urban Development. Section 811 Supportive Housing for Persons with Disabilities. Accessed December 2013. http://portal.hud.gov/hudportal/HUD?src=/program_offices/housing/mfh/progdesc/disab811.
  7. Lind, K.D. (2012). Setting the Record Straight about Medicare. Washington, D.C.: AARP Public Policy Institute. Accessed December 2013. http://www.aarp.org/health/medicare-insurance/info-02-2012/Setting-the-Record-Straight-about-Medicare-fact-sheet-AARP-ppi-health.html.
  8. Kaiser Family Foundation. (2012). The Medicare Program at a Glance. Accessed December 2013. http://kff.org/medicare/fact-sheet/medicare-at-a-glance-fact-sheet/.
  9. Barry, P. (2013). Do You Qualify for Medicare? Accessed December 2013. http://www.aarp.org/health/medicare-insurance/info-04-2011/medicare-eligibility.html.
  10. Kaiser Family Foundation. (2013). The Medicaid Program at a Glance. http://kff.org/medicaid/fact-sheet/the-medicaid-program-at-a-glance-update/.
  11. Kaiser Family Foundation (2011). Federal Core Requirements and State Options in Medicaid: Current Policies and Issues. Accessed December 2013. http://kff.org/health-reform/fact-sheet/federal-core-requirements-and-state-options-in/.
  12. "Low-income" for Medicaid and Medicare do not follow the same guidelines as HUD. They are subject to federal and state regulations.
U.S. Department of Housing and Urban Development (HUD) Programs
HUD provides housing assistance to about 4.6 million low-income households,1 including families with children, older adults, and people with disabilities. Tenant income eligibility is determined on the basis of AMI (generally calculated for metropolitan areas and non-metropolitan counties), adjusted for family size. Low-income families are defined as families whose incomes do not exceed 80% of the AMI and very low-income families are defined as families whose incomes do not exceed 50% of the AMI. Over three-fourths of assisted households in 2009 had extremely low incomes not exceeding 30% of AMI. HUD programs important for this report include:
  • Public Housing is housing administered by PHAs for eligible low-income families, older adults, and people with disabilities. Eligibility is based on annual gross income and United States citizenship or eligible immigration status.2 About 23% of HUD-assisted households live in public housing.
  • Housing Choice Vouchers (also known as Section 8 vouchers) provide rent subsidies used in private rental markets and, like public housing, are administered by PHAs. Eligibility is based on income and citizenship or eligible immigration status. A family or individual receiving a voucher must pay a specified percentage (often 30%) of their income toward rent and the PHA pays the balance of rent, subject to program limits.3 Voucher holders represent about 46% of HUD-assisted households.
  • Section 202 housing is the Supportive Housing for the Elderly Program. This is a program run by HUD's Office of Multifamily Housing Programs. It helps finance the development of affordable and accessible housing for low-income older adults. This capital is provided to private not-for-profit organizations and not-for-profit consumer cooperatives and need not be repaid as long as the buildings house low-income seniors.4 The program may also provide rent subsidies.5
  • Section 811 housing is supportive housing for persons with disabilities. It is also run by HUD's Office of Multifamily Housing Programs. The 811 program provides interest-free capital to eligible organizations to help them finance affordable housing with supportive services for low-income adults with disabilities. As with the 202 program, capital need not be repaid as long as the buildings house low-income disabled persons. The program may also provide rent subsidies.6
  • HUD's Office of Multifamily Programs has several programs termed here "other multifamily housing," in which rental assistance is provided in conjunction with programs that subsidized developments through below-market interest financing, mortgage insurance or other forms of assistance for the new construction or rehabilitation of housing for low-income individuals. Multifamily programs, including the Section 202 and Section 811 supportive housing programs represent about 30% of HUD-assisted renters. See Appendix D for the full list of multifamily housing properties included in the data analysis.
Centers for Medicare and Medicaid Services (CMS) Programs
Medicare7,8,9 is a federal program that provides health insurance for older adults and people with disabilities. This program was designed for people age 65 and older. This program is typically an entitlement for older adults who:
  • Are eligible for Social Security Payments.
  • Have made payroll tax contributions for 10 or more years.

The spouse of someone who meets these guidelines is also eligible for Medicare. Those under 65 can qualify for a couple of reasons. One of the major reasons is being entitled to Social Security disability benefits for at least two years.There are multiple parts to Medicare, including:

  • Medicare Part A ("Hospital insurance") covers acute health care needs, including hospital care, SNF care, hospice, and HH. Part A is paid for by a portion of Social Security tax. The vast majority of people age 65 and over get Medicare Part A for free as long as the individual or his/her spouse paid Medicare taxes for a minimum of 10 years and did not enroll late for Medicare, meaning he/she enrolled when you were first eligible.
  • Medicare Part B ("Medical insurance") covers outpatient health care needs, including doctors' visits and other preventive care services. Individuals who elect Part B coverage must pay premiums.
  • Medicare Part C, or Medicare Advantage Plans, provide you with all your Part A and Part B benefits, and often also cover prescription drugs. Private companies that contract with Medicare offer various types of plans including HMOs, Preferred Provider Organizations, Private FFS Plans, Special Needs Plans, and Medicare Medical Savings Account Plans. Individuals who elect Part C coverage must pay premiums.
  • Medicare Part D provides prescription drug coverage. This is a voluntary program that is offered through private insurance companies that are under contract with Medicare, and include premiums.

Medicaid10,11 is a public health insurance program for low-income children and adults.12 States administer this program following federal rules, and it is jointly financed by both federal and state governments. The federal minimum standards for eligibility are:

  • State coverage of core groups, which include low-income elderly individuals, people with disabilities, pregnant women, children, and parents of children.
  • These core groups are subject to different minimum income levels. For older adults and people with disabilities, states typically provide coverage to those who receive SSI.

Beyond these federal minimums, states can set their own standards for eligibility within the allowed federal range and can opt to cover additional services. For example, all states provide HCBS to older adults through waivers and many states use the upper income limits of 300% of SSI for these services.Medicaid also covers the Medicare premiums and copayments for low-income Medicare beneficiaries through the MSP. Those with income 100% or less of the FPL receive premium and copayment coverage while those with income 101-125% of the FPL receive premium coverage only.Participants who are a part of both programs, MMEs, are often referred to as Dual Eligibles or Duals. In addition, some individuals receive full Medicaid benefits for services not included in Medicare, such as assistance with personal care and long stay nursing facility care.

A. Data Sources

We based the analysis on individual-level administrative data from both HUD and CMS described in detail below.

1. HUD Individual-Level Administrative Data

We received individual-level administrative data from HUD for 2007, 2008, and 2009. The data originated from two HUD data systems. Given the study objective, we chose geographic areas that have unique public housing with services models.

The data were limited to individuals in our 12 geographic areas of interest:

  1. New Haven-Milford, Connecticut
  2. Bridgeport-Stamford-Norwalk, Connecticut
  3. Milwaukee-Waukesha-West Allis, Wisconsin
  4. San Francisco-Oakland-Fremont, California
  5. Boston-Cambridge-Quincy, Massachusetts
  6. Durham-Chapel Hill, North Carolina
  7. Richmond, Virginia
  8. New York-Northern New Jersey-Long Island
  9. Columbus, Ohio
  10. Akron, Ohio
  11. Cleveland, Ohio
  12. The entire State of Vermont

Tenant Rental Assistance Certification System (TRACS)

The TRACS system collects certified tenant data from owners and management agents of privately owned multifamily housing projects, as well as from local PHAs, and state housing agencies acting as subsidy contract administrators for HUD. The programs covered in TRACS include: Section 8, Rent supplement, Rental Assistance Payment (RAP), Section 236 (Interest Reduction and RAPs), Section 211 Below-Market Interest Rate (BMIR) mortgage insurance, Section 202 Project Rental Assistance Contract (PRAC), Section 811 PRAC, and Section 202/162 Project Assistance Contract (PAC). Note that there are two Section 202 categories. A small number of 202 properties contain units for non-elderly disabled that were funded with rental assistance created by Section 162 of the Housing and Community Development Act of 1987. Section 162 has been superseded by the Section 811 program, which supports housing for disabled individuals. Data contain information on both an individual level (for each individual who resides in the household) and a household level.

Public and Indian Housing (PIH) Information System (PIC)

PHAs regularly collect and electronically submit information to HUD about the households they assist, and the housing assistance that is provided. This study limited its review of housing assistance to the following programs administered by HUD: Section 8 Certificates, Mod Rehab, Public Housing, and Section 8 Vouchers, Multifamily. Data were available for the head of household and each individual that resides in the household.

As described in Appendix B, we combined the TRACS and PIC data into one dataset for our sample. For individuals that appear in both datasets (across all three years this was 38,326 individuals, or 1.54%), we created only one observation per individual based on the PIC data. Individuals were excluded if an individual's move out date in a given year occurred prior to the current data year (N=194,674).

2. CMS Individual-Level Administrative Data

We received individual-level administrative data from CMS for both Medicare and Medicaid. This included information on enrollment, eligibility, presence of chronic conditions, health care payments, and health care use for individuals covered by Medicare and/or Medicaid who resided in any of our 12 geographic areas of interest. We received the most recent data available for the data sources: 2007-2009 for Medicare data and 2007 and 2008 for Medicaid data. Note, that while we received HUD and CMS Medicare data for 2007-2009, given 2009 data was not available for CMS Medicaid data, the majority of our results are based on 2008.

a. Medicare administrative data

We received three segments of the Medicare Beneficiary Summary File for each year of 2007-2009. Data contain information on the individual level for any Medicare beneficiary that had coverage at any point in time during the four year period. The three segments included:

  • Parts A, B, and D. This file contains information on the beneficiary's demographic, enrollment and eligibility for Medicare coverage. For example, variables include gender, race, first date of Medicare coverage, date of death if applicable, length of Medicare enrollment, indicatorfor any Medicare managed care coverage, indicator for Part D coverage, and number of months eligible for both Medicare and Medicaid.

  • Chronic Conditions. This file contains flags for 28 chronic conditions based on validated criteria, as defined and identified by CMS. Example conditions include acute myocardial infarction, hypertension, cancers, and Alzheimer's. For each of the 28 conditions, the file includes a year-end flag, mid-year flag, and the first date of occurrence for the condition.

  • Cost and Use. This file includes the total utilization, Medicare payments, and beneficiary payments for the given year by type of health care service for each beneficiary. Examples of health care services include utilization of outpatient emergency room, acute inpatient hospital, SNF days covered by Medicare, imaging, and DME.

b. Medicaid Administrative Data

We received the MAX Person Summary file for 2007 and 2008. The Person Summary file contains demographic characteristics, Medicaid enrollment, payment, and utilization variables for each beneficiary that had Medicaid coverage at any point in time during the calendar year.

B. Variable Definitions

The variables are organized by study objective and data source. We created the variables for each year of data.

1. Demographic Characteristics of HUD-Assisted Housing Residents

The variables below were based on information in the HUD TRACs and PIC datasets. Each variable was created by year for all HUD-assisted housing residents in the 12 geographic areas of interest. The level of observation was beneficiary year. Variables were created for each year of HUD data -- 2007, 2008, and 2009.

  • Race/Ethnicity. Created using the fields from TRACS (RACE_WHITE, RACE_AMRIND_ALKNTV, RACE_ASIAN, RACE_BLCK_AFR_AMR, RACE_HAWAIIAN_PI, RACE_OTHER, and RACE_DECLND_TO_RPT) and PIC (MBR_ETHNICITY, WHITE_IND, BLACK_IND, A_INDIAN_ALASKAN_IND, ASIAN_IND, and HAWAIIAN_PACIFIC_IND). Categories are: White/non-Hispanic, Black, Asian, Hispanic, North American Native, Other and Unknown/did not report.

  • Living Arrangement (individual level). Original values in the data included: Head of Household, spouse, child, other adult, foster child, adult co-head, live-in caretaker, other person >18 years old or full-time student, full-time student 18+, other youth under 18, did not report. We combined values to create sub-groups of interest. The categories were assigned as follows: If there was one person in a household, then they were marked as living alone. If there were two people in a household and one of them was a spouse, then each person in the household was marked as living with a spouse only. If there was a live-in aide in the household then everyone in the household was marked as living with a live-in aide. If there was more than one person in the household and one of the persons was a youth, foster youth, or student, and there was no live-in aide, then everyone in the household was marked as living with children. If there was more than one person in the household and there were no youths and no live-in aides and the people in the household were adults, co-heads, or foster adults, then everyone in the household was marked as living with other adults. If the people in the household had responses missing or other responses to their relationship status question (e.g., not spouse, child, student, live-in aide, etc.) then everyone in the household was marked as other living arrangement.

  • Living Arrangement (household level). Households are determined to be households according to the SSN of the head of household. Each person in a household is grouped together by the SSN of the household head. After grouping all of the households, the living arrangement by household was determined. Living arrangement at the household level was assigned the same way as the individual level living arrangement variable (see above). If a household appeared in both the TRACS and PIC data in the same year and they had differing living arrangements, then the household's living arrangement was marked as unknown.

  • Household Heads. Heads of household are noted in both the TRACS and PIC data in relationship fields. If a person selected head of household ("H"), then they were considered a household head.

  • Disability or Handicapped Status. Disability was set to Y/N/Not reported. The PIC field of "disability_ind" was a direct map to this variable. The TRACS data was mapped using the "special status" field. If a person's SPECIAL_STATUS_CD contained an "H" (handicapped) response, they were considered disabled, regardless of other SPECIAL_STATUS_CD responses. For example, a person could have a special status of "ESH", meaning they were an elderly (E) student (S) who was handicapped (H). Whenever handicapped (H) was one of the responses, they were considered disabled.

  • Gender. We utilized the gender variable in TRACS/PIC dataset. Values include: male, female, unknown/did not report.

  • Age. Age was calculated for each year. The age was calculated as the year of data (i.e., 2007, 2008) minus the birth year. Age in 2009 was used as the default when looking at individuals in all three years. If a person's age was not available in the 2009 data, their age in 2008 or 2007 was used and the appropriate amount of years were added to determine what their age would be in 2009. Two age group variables were created based on this variable: 0-18, 18-44, 45-61, 62-64,18 65+, unknown/not reported; and <62, 62-64, 65+, unknown/not reported.

  • Age 62+ with Children in the Home. This variable took all persons age 62 or older from the individual living arrangement variable and noted whether that person lived with children in the household or had a different living arrangement.

  • Elderly Properties. Each property in the TRACS data is identified using a Property ID. Similarly, each property in PIC is identified using a Development Code. For TRACs properties, a property was defined as elderly if at least 50% of the households within a Property ID had an individual age 62 or older. PHAs can explicitly designate properties as elderly. Therefore, a property in the PIC data was defined as elderly if it had an elderly designation or at least 50% of the households within a Development Code had an individual age 62 or older.

  • Subsidy Type.19 The PIC variable used for subsidy type was "program type" and for TRACS the variable used was "subsidy type". The variable categories created were: Public housing, Housing Choice Vouchers, Project-Based Section 8, Section 202, Section 202/162, Section 811, and Other. For PIC, program type was mapped as follows: public housing to "Public housing"; Section 8 certificates and Section 8 vouchers to "Housing Choice Vouchers"; and Mod Rehab converted to "Other". For TRACS, subsidy type was mapped as follows: Section 8 to "Project-Based Section 8", Section 202 PRAC to "Section 202", Section 202/162 PAC to "Section 202/162", Section 811 PRAC to "Section 811", and Rent supplement, RAP, Section 236, and BMIR to "Other".

  • Property Type. We mapped the property type variable from the HUD property level data files to our individual-level files based on Development Code for PIC individuals and Property ID for TRACS individual. The PIC variable used for property type was ‘program type name.' And for TRACS the variable used was "subsidy_type". The variable categories created were: "public housing", "housing choice vouchers", "other multifamily", "Section 202", "Section 811 and 202/162" and "other". "Housing choice vouchers' include all types of buildings in which housing choice voucher holders live, from a single-family home to a large apartment complex. The assignment of PIC and TRACS property types to these categories is displayed in Table A1. For the small proportion of TRACS individuals whom do not have a property type assigned to their property, we assigned their property type value based on their subsidy type. The purpose of the property type variable was to assign individuals to the actual type of property in which they live whereas the subsidy type variable assigned individuals to a category based on their subsidy type regardless of the type of property in which they live. While there is much overlap between the two variables, the main difference is for individuals receiving a project-based Section 8 subsidy. The property type variable distributes individuals receiving a project-based Section 8 subsidy to the actual property in which they live -- Other multifamily, Section 202, Section 811 and 202/162, or Other. The majority of individuals with project-based Section 8 subsidies were mapped to "Other multifamily". However, a larger portion of individuals age 65+ with a project-based Section 8 subsidy were mapped to "Section 202" property.20

TABLE A1. Property Type Variable Definition
 Public HousingHousing Choice VouchersOther MultifamilySection 202Section 811 & 202/162Other
PIC property types (based on development code' variable)Public housingSection 8 vouchers   Other
TRACS property types (based on "property_id" variable)  
  • Sec. 8
  • 515/8 new construction
  • Sec 8 substantial rehab
  • Sec 8 new construction
  • HFDA 8/new construction
  • PD/8 Existing
  • Loan Management Set-Aside
  • Preservation
  • Pension Fund
  • Rent Supplement
  • RAP
  • Section 202 with PACs
  • 202/8 new construction
  • 202/8 substantial rehab
  • Section 811 with PRACs
  • 202/162 new construction
 
  • Geographic Area. Individuals were assigned to one of the 12 geographic areas of interest of our study based on Core Based Statistical Area (CBSA) or State Code: Vermont, New Haven-Milford, Bridgeport-Stamford-Norwalk, Milwaukee-Waukesha-West Allis, San Francisco-Oakland-Fremont, Boston-Cambridge-Quincy, Durham-Chapel Hill, Richmond, New York-Northern New Jersey-Long Island, Columbus, Akron, and Cleveland.

Metropolitan Statistical Area(CBSA Code)CountyCounty Census CodeSSA State CodeSSA County Code
State of Vermont 5000047000
New Haven- Milford (35300)Connecticut--New Haven County--state Federal Information Processing Standards (FIPS) code: 090900907040
Bridgeport- Stamford- Norwalk (14860)Connecticut--Fairfield County--state FIPS code: 090900107000
Milwaukee- Waukesha- West Allis (33340)Wisconsin--state FIPS code: 55 52 
Milwaukee County5507952390
Ozaukee County5508952440
Washington County5513152650
Waukesha County5513352660
San Francisco- Oakland- Fremont (41860)California--state FIPS code: 06 05 
Alameda County0600105000
Contra Costa County0601305060
Marin County0604105310
San Francisco County0607505480
San Mateo County0608105510
Boston- Cambridge- Quincy (14460)Massachusetts--state FIPS code: 25   
Norfolk County2502122130
Plymouth County2502322150
Suffolk County2502522160
Middlesex County2501722090
Essex County2500922040
New Hampshire--state FIPS code: 33   
Rockingham County3301530070
Strafford County3301730080
Durham- Chapel Hill (20500)North Carolina--state FIPS code: 37   
Chatham County3703734180
Durham County3706334310
Orange County3713534670
Person County3714534720
Richmond (40060)Virginia--state FIPS code: 51   
Amelia County5100749030
Caroline County5103349160
Charles City County5103649180
Chesterfield County5104149200
Cumberland County5104949240
Dinwiddie County5105349260
Goochland County5107549370
Hanover County5108549420
Henrico County5108749430
King and Queen County5109749480
King William County5110149500
Louisa County5110949540
New Kent County5112749621
Powhatan County5114549720
Prince George County5114949740
Sussex County5118349910
New York- Northern New Jersey- Long Island (35620)New Jersey--state FIPS code: 34   
Middlesex County3402331270
Monmouth County3402531290
Ocean County3402931310
Somerset County3403531350
Hunterdon County3401931250
Morris County3402731300
Sussex County3403731360
Union County3403931370
Bergen County3400331100
Hudson County3401731230
Passaic County3403131320
Essex County3401331200
New York--state FIPS code: 36   
Nassau County3605933400
Suffolk County3610333700
Bronx County3600533020
Kings County3604733331
New York County3606133420
Putnam County3607933580
Queens County3608133590
Richmond County3608533610
Rockland County3608733620
Westchester County3611933800
Pennsylvania--Pike County4210339630
Columbus (18140)Ohio   
Delaware County3904136210
Fairfield County3904536230
Franklin County3904936250
Licking County3908936460
Madison County3909736500
Morrow County3911736600
Pickaway County3912936660
Union County3915936810
Akron (10420)Portage County3913336680
Summit County3915336780
Cleveland (17460)Cuyahoga County3903536170
Geauga County3905536280
Lake County3908536440
Lorain County3909336480
Medina County3910336530
  • Fullyear2007. This field indicates if an individual received housing assistance for the entire year or only a portion of the year. The variable was created differently for individuals in TRACS and PIC data. For TRACS, the following three variables were used to determine full year assistance: move in date, certification type, and effective date. If an individual's "move in date" is before the first of the year, and he/she did not have a "certification type" of Termination or Move Out with an "effective date" during the year or prior to the year, fullyear2007=1. If the individual's "move in date" is after the first of the year or if the individual had a "certification type" of Termination or Move Out with an "effective date" during the year, fullyear2007=0. For PIC, the three variables used were: admission date, type of action, and effective date. If the individual's "admission date" is before the first of the year and he/she did not have a "type of action" of Termination with an "effective date" during the year or prior to the year, fullyear=1. If the individual's "admission date" is after the first of the year or the individual had a "type of action" of Termination with an "effective date" during the year, fullyear2007=0. For both TRACS and PIC, some individuals had a move out or termination dated back to a previous year (2006 or earlier). In this instance, the individual was deleted from the 2007 dataset.

  • Fullyear2008. Created similarly to fullyear2007. Some individuals have a move out or termination dated back to a previous year (2007 or earlier). For example, an individual may show up as present in the 2007 and 2008 data. In the 2008 data, it shows that they moved out in 2007. This individual's fullyear2007 variable was then set to 0 ("partial-year") and they were deleted from the 2008 dataset.

  • Fullyear2009. Created similarly to fullyear2007. Some individuals have a move out or termination dated back to a previous year (2008 or earlier). For example, an individual may show up as present in the 2008 and 2009 data. In the 2009 data, it shows that they moved out in 2008. This individual's fullyear2008 variable was then set to 0 ("partial-year") and they were deleted from the 2009 dataset.

  • Fullyear0708. Using fullyear2007 and fullyear2008 variables, this variable categorizes individuals as receiving housing assistance for all of 2007 and 2008 or for only a partial portion of those two years. The variable was defined as followings: fullyear0708=1 if fullyear07=1 and fullyear08=1. Fullyear0708=0 if either fullyear07 or fullyear08 were equal to zero. Similar variable was created for fullyear0809.

  • Fullyearallyears. Using fullyear2007, fullyear2008, and fullyear2009 variables, this variable categorizes individuals as receiving housing assistance for all three years or for only a partial portion of the three years. The variable was defined as followings: fullyearallyears=1 if fullyear07=1, fullyear08=1, and fullyear09=1. Fullyearallyears=0 if either fullyear07 or fullyear08 or fullyear09 were equal to zero.

  • Reason for Partial-Year. For individuals where fullyear=0 (partial-year), this variable specifies why they were not receiving housing assistance for the entire year. The values include: moved in after the first of the year, moved out during the year, terminated21 during the year.

2. Feasibility of linking HUD and CMS data sources

The main objective of this study task was to determine the ability to link HUD-assisted housing recipients to Medicare and Medicaid administrative data (i.e., enrolled in Medicare or Medicaid health care programs) in the 12 study jurisdictions. We accomplished this objective by calculating two measures that represent the ability to link the HUD and CMS data sources:

  1. SSN Match Rate. The proportion of HUD-assisted housing recipients enrolled in Medicare and/or Medicaid programs (determined solely on a matched SSN between the datasets).

  2. Restrictive Match Rate. A fairly conservative method on which we established whether the individuals receiving housing assistance were definitively included in the Medicare and/or Medicaid data by requiring a direct match on SSN, gender, and date of birth.22

The restrictive match rate was calculated by year and across years for both Medicare and Medicaid programs using various combinations of the below indicators.

  • Linked to Medicare Eligibility File by Year (2007, 2008, 2009). For each individual in the TRACS/PIC datasets, we created a series of binary variables (Y/N) indicating whether the individual could be linked to the Medicare eligibility by year. If the individual was linked based on SSN, date of birth, and gender, this individual was assigned a Y for the given year. If the individual was not located on the Medicare eligibility file, this individual was assigned a N.

  • Linked to Medicaid eligibility file by year (2007 and 2008). For each individual in the TRACS/PIC datasets, we created two binary variables (Y/N) indicating whether the person could be linked to the Medicaid eligibility file by year. If the individual was linked based on SSN, date of birth, and gender, this individual was assigned a Y for the given year. If the individual was not located on the Medicaid eligibility file, this individual was assigned a N.

3. Medicare Beneficiary Demographic and Coverage Characteristics

The variables below were based on information in the Medicare Beneficiary Summary File. Each variable was created by year for all Medicare beneficiaries in the 12 geographic areas of interest. The level of observation was beneficiary year.

  • Age. Age at end of year. Age groups were also created as followed: under age 65 (<62, 62-64); 65+ (65-74, 75-84, 85+).

  • Gender. Values included unknown, female, and male.

  • Race. Research Triangle Institute (RTI) Race Code which is based on enhanced race/ethnicity designation based on first and last name algorithms. Categories include: unknown, non-Hispanic White, Black, other, Asian/Pacific Islander, Hispanic, American Indian/Alaska Native.

  • Died. Each beneficiary was assigned a Y/N flag based on the beneficiary date of death variable. If date of death=mmddyyyy, then Died=Y(1). If date of death is missing, then Died=N(0).

  • Part A Coverage Indicator. Part A coverage was assigned as a Y/N variable using the Beneficiary HI Coverage Months variable. If the beneficiary had a minimum of one month of Part A coverage, Part A Coverage indicator=Y(1), else=N(0). Medicare Part A includes inpatient services coverage, such as hospital care, SNF care, nursing home care (with exceptions), hospice, and home health services. The costs covered for these services depend on a number of factors.

  • Part B Coverage Indicator. Part B coverage was assigned as a Y/N variable using the Beneficiary Severe Mental Illness Coverage Months variable. If the beneficiary had a minimum of one month of Part B coverage, Part B Coverage indicator=Y(1), else=N(0). Medicare Part B covers outpatient services, such as doctor's visits. The services covered by Part B are medically necessary services (i.e., those needed to diagnosis or treat a medical condition) or preventive services (e.g., a flu shot).23

  • Month Count with Part A Coverage. The count of months when the beneficiary had Medicare Part A coverage. Values ranged from 0 to 12.

  • Month Count with Part B Coverage. The count of months when the beneficiary had Medicare Part B coverage. Values ranged from 0 to 12.

  • Annual Part A and B Enrollment. Beneficiaries were identified as being enrolled in both Part and B for the entire year if the month count with Part A coverage=12 and month count with Part B coverage=12.

  • Medicare Health Maintenance Organization (HMO)/Managed Care Indicator. The Medicare HMO/Managed Care indicator was assigned as a Y/N variable using the HMO coverage total months variable. If the beneficiary had a minimum of one month of HMO coverage, HMO/Managed Care Coverage indicator=Y(1), else=N(0). Medicare HMOs, or Managed Care, are plans that incorporate Medicare Parts A and B; these are called Medicare Part C, or Medicare Advantage plans. In these plans, a private Medicare-approved company provides oversight of a beneficiary's health care utilization.

  • Month Count with HMO/Managed Care Coverage. The count of months when the beneficiary had Medicare HMO/Managed Care coverage. Values ranged from 0 to 12.

  • Original Reason for Medicare Entitlement. Reason why the beneficiary was initially entitled to Medicare coverage. Values include: 0=Old age and survivors insurance; 1=Disability insurance benefits (DIB); 2=End Stage Renal Disease (ESRD); and 3=Both DIB and ESRD.

  • Current Reason for Medicare Entitlement. Reason why the beneficiary was currently entitled to Medicare coverage during reference year. Values include: 0=Old age and survivors insurance; 1=DIB; 2=ESRD; and 3=Both DIB and ESRD.

  • Medicare Coverage Status. Medicare coverage status was assigned based on the monthly buy-in indicators which specify if the beneficiary had Part A and/or Part B coverage by month. Medicare coverage status values were based on the following mapping:

    • If no buy-in monthly flags=(1,2,3,A,B,C) then beneficiary was assigned to 1=Not Entitled.
    • If all buy-in monthly flags=(1 or A) then beneficiary was assigned to 2=Part A coverage only category.
    • If all buy-in monthly flags=(2 or B) then beneficiary was assigned to 3=Part B coverage only category.
    • If all buy-in monthly flags=(3 or AB) then beneficiary was assigned to 4=Part A and B coverage only category.
    • If enrollee has any other combination of buy-in monthly flags they were assigned to the 5=Combination coverage category.
  • Part D Coverage Indicator. Part D coverage was assigned as a Y/N variable using the Plan D coverage monthsvariable. If the beneficiary had a minimum of one month of Part D coverage, Part D Coverage indicator=Y(1), else=N(0). Medicare Part D is prescription drug coverage. This is optional coverage that Medicare beneficiaries can purchase through private plans.24 Having Part D coverage adds more drug coverage to Medicare Part A and B plans, in addition to some other Medicare plans.

  • Month Count with Part D Coverage. The count of months when the beneficiary had Medicare Part D coverage. Values ranged from 0 to 12.

  • Part D Coverage by Cost Share Group Code. The Part D benefit includes cost sharing provisions. State Medicaid and other government-sponsored subsidized premiums and/or copayments/coinsurance for low-income individuals are allowed. Additionally, unlike Medicare A and B, the Part D benefit allows for means-testing. The LIS provides assistance to certain low-income individuals to supplement the premium and cost sharing (including deductibles and cost sharing during the coverage gap) associated with the Part D benefit. Subsidies may also be provided to employers to cover eligible beneficiaries. All of these cost sharing provisions are indicated within this variable. Each beneficiary was assigned to one cost share group code based on the 12 monthly cost share indicators.25 Months with a value of 10, 11, 12, 13, 00, or XX (not enrolled in Part D or missing) were not considered when counting the most often that occurs. If all months were equal to 09 ("No premium-subsidy nor cost sharing"), then PartD_costshrgrp=9. Else, partD_costshrgrp is set equal to the value that occurs most often across the 12 monthly indicators (01=Bene is deemed with 100% premium-subsidy and no copayment; 02=Bene is deemed with 100% premium-subsidy and low copayment; 03=Bene is deemed with 100% premium-subsidy and high copayment; 04=Bene with LIS, 100% premium-subsidy and high copayment; 05=Bene with LIS, 100% premium-subsidy and 15% copayment; 06=Bene with LIS, 75% premium-subsidy and 15% copayment; 07=Bene with LIS, 50% premium-subsidy and 15% copayment; 08=Bene with LIS, 25% premium-subsidy and 15% copayment).

  • MME Status. MME Status was assigned to one of the following categories based on the variable "EL_MDCR_DUAL_ANN". Categories included: 00 (or NA), 99 (or unknown), 01 (qualified Medicare beneficiary (QMB) only), 02 (QMB Plus), 03 (specified low-income Medicare beneficiary (SLMB) only), 04 (SLMB plus), or 05 (Other MME status). These categories include the array of categories for MMEs, meaning that someone eligible for both Medicare and Medicaid may fall under some of these categories.26

  • Qualified Medicare Beneficiary (QMB). For an individual to be a QMB,they must be entitled to Medicare Part A, have an income at or below 100% of the Federal Poverty Level (FPL) and resources that do not exceed twice the SSI eligibility limit. These individuals also cannot be otherwise eligible for full Medicaid. This program pays for Part A and B premiums, as well as deductibles, coinsurance, and copayments.

    • For QMBs, Medicaid pays for their Medicare Part A and B premiums, if any, and Medicare deductibles and coinsurance for Medicare services provided by Medicare providers (as is consistent with the Medicaid State Plan).
  • A QMB Plus is someone who meets all of the eligibility criteria of a QMB; however, they are eligible for full Medicaid benefits.

    • For QMB Plus individuals, Medicaid also pays for these benefits and provides full Medicaid benefits.
  • SLMB. These individuals are entitled to Medicare Part A, have an income above 100% FPL, but below 120% FPL and resources that do not exceed twice the SSI eligibility limit. These individuals are not otherwise eligible for Medicaid.

    • For SLMB only individuals, Medicaid pays for their Medicare Part B premiums only.
  • The SLMB Plus population meets all of the SLMB requirements; however, they are eligible for full Medicaid benefits.

    • For SLMB plus individuals, Medicaid pays for their Medicare Part B premiums and provides full Medicaid benefits.
  • Other MME Status. This category captures those individuals who are eligible for Medicare and Medicaid, but are not QMB and SLMB. These can include qualified disabled and working individuals (QDWI) and Medicaid-only Dual Eligibles.

  • Any MME Status. A binary Y/N variable based on DUAL_MO count. If DUAL_MO count not equal to zero, MME Status=Y; else N.

  • Number of months with dual (Medicare and Medicaid) coverage. Equal to the number of months within the DUAL_MO variable.

4. Medicaid Coverage Characteristics

The variables below were based on information in the Medicaid person summary file. Each variable was created by year for all Medicaid beneficiaries in the 12 geographic areas of interest. The level of observation was beneficiary year.

  • Month Count of Medicaid Coverage. The count of months when the beneficiary has Medicaid coverage (managed care or FFS). Values ranged from 0 to 12.

  • Annual Medicaid Enrollment. Beneficiaries were identified as being enrolled in Medicaid for the entire year if the month count ("EL_ELGBLTY_MO_CNT")=12.

  • Medicaid HMO/Managed Care Indicator. The Medicaid HMO/Managed Care indicator was assigned as a Y/N variable using the private insurance month count variable. If the beneficiary had a minimum of one month of private insurance, Medicaid HMO/Managed Care Coverage indicator=Y(1), else=N(0).

  • Month Count with Medicaid HMO/Managed Care Coverage. The count of months when the beneficiary had Medicaid HMO/Managed Care coverage. Values ranged from 0 to 12.

  • Medicaid Basis for Eligibility. Reason why the beneficiary was initially eligible for Medicaid coverage. Values include: 00=00 (not eligible); 99=99 (unknown); 11, 21, 31, 41, or 51=01 (aged); 12, 22, 32, 42, or 52=02 (blind/disabled); 14, 24, 34, 44, or 54=03 (child); 15, 25, 35, 45, or 55=04 (adult); 16=05 (child of unemployed adult); 17=06 (unemployed adult); 48=07 (foster care child); or 3A=08 (covered under breast and cervical cancer prevention act).

  • Medicaid Maintenance of Assistance. The maintenance of assistance for the Medicaid beneficiary. Values include: 0=not eligible; 1=Cash or Section 1931 of the Act; 2=MN; 3=Poverty or State Children's Health Insurance Program (SCHIP); 4=Other; 5=foster care child; 6=Section 1115 Demo Expansion. Medicaid maintenance of assistance refers to how someone qualifies for Medicaid:27, 28

    • Value 0 means someone is not eligible for Medicaid.
    • Value 1 refers to Section 1931 of the Social Security Act, establishes rules for Medicaid coverage for qualified low-income families.29
    • Value 2 refers to those that are Medically Needy.
    • Value 3 means someone qualifies for Medicaid under poverty-related eligibility criteria.
    • Value 4 captures, but does not specify, other ways someone is eligible for Medicaid.
    • Value 5 means someone qualifies for Medicaid because they are a foster care or adoptive child.
    • Value 6 means someone is eligible under a Section 1115 demonstration.
  • Type of 1915c Waiver Status Code. 1915c waivers are HCBS waivers. Individuals can qualify for a 1915c waiver based on the codes listed above. For example, someone can qualify for a state Medicaid 1915c waiver specifically for people who are aged and disabled. Most recent type of 1915c waiver, if applicable. Values include Blank=unknown, missing eligibility; G=aged and disabled; H=aged; I=disabled; J=brain injured; K=HIV/AIDS; L=MR/DD; M=mentally ill/severely emotionally disturbed; N=technology-dependent/medically fragile; O=Other or unknown; P=Autism/Autism spectrum disorder; 0=never enrolled in a 1915c waiver during the year.

  • Any 1915c Waiver Status Code. Indicates if individual had any 1915c waiver status code during 2008. If MAX_1915C_WAIVER_TYPE_LTST is not equal to "blank" or 0, then indicator=1, else indicator=0.

5. Chronic Conditions

The Chronic Condition Warehouse (CCW) flags, as defined by CMS, were used to identify individuals with Medicare coverage who had any of the 27 chronic conditions. A binary variable (Y/N) was created for each of the 28 CCW conditions for each year of Medicare data available (2007, 2008, 2009, 2010) using the "end-of-year" flags. Individuals with an "end-of-year" value=1 ("claims met") or 3 ("claims and coverage met") for a condition were assigned a "Y" for the given CCW condition. A summary variable was created for each of the four years which summed up the flags across 26 of the 27 conditions (excluding Alzheimer's chronic condition to not double count with Alzheimer's and related conditions flag) for a given individual (e.g., an individual with three CCW condition "end-of-year" flags would be assigned a three for the given calendar year).

We grouped the 27 chronic conditions into nine categories, shown in Table A2.

TABLE A2. Assignment of Chronic Conditions into 9 Chronic Condition Groups
Chronic Condition GroupChronic Condition
Cardiovascular Chronic ConditionsAcute Myocardial InfarctionAtrial FibrillationChronic Heart FailureIschemic Heart DiseaseStroke
CancerBreast CancerColorectal CancerLung CancerEndometrial CancerProstate Cancer
Endocrine & RenalChronic Kidney DiseaseDiabetesHypothyroidism
Alzheimer's DiseaseAlzheimer's DiseaseAlzheimer's and Alzheimer's-Related Disorders or Senile Dementia
DepressionDepression
MusculoskeletalHip FractureOsteoporosisRheumatoid Arthritis
PulmonaryChronic Obstructive Pulmonary DiseaseAsthma
OphthalmicCataractGlaucoma
OtherAnemiaHyperlipidemiaHypertensionBenign Prostatic Hyperplasia

6. Health Care Utilization

Health care utilization metrics were generally defined as available in the Medicare Beneficiary Summary File. In a few instances, we combined existing metrics to create a summation of related services. Table A3 includes the utilization metrics created in our dataset and the corresponding Medicare variable or Medicare data-based logic. Variables were created for each year available in the Medicare (2007, 2008, 2009 and 2010) data.

TABLE A3. Medicare Health Care Utilization Variable Definitions
Utilization MetricVariable as Named inMedicare Data or AlgorithmUsing Medicare Data VariablesVariable Definitions1
NOTES:
  1. See https://www.ccwdata.org/web/guest/data-dictionaries.
  2. Limited to beneficiaries with Part D coverage.
Total AdmissionsACUTE_STAYS + OIP_STAYS 
   Acute stay admissionsACUTE_STAYSCount of hospital stays (unique admissions, which may span more than one facility) in the acute inpatient setting for a given year. An acute stay is defined as a set of one or more consecutive acute claims where the beneficiary is only discharged on the most recent claim in the set.Acute care settings include a hospital, ER, and short-stay facilities for shorter-term treatment.
   Other inpatient admissionsOIP_STAYSCount of hospital stays (unique admissions, which may span more than 1 facility) in the non-acute inpatient setting for a given year. A non-acute inpatient stay is defined as a set of 1 or more consecutive non-acute inpatient claims where the beneficiary is only discharged on the most recent claim in the set.Non-acute inpatient settings are SNFs.
Hospital ReadmissionsREADMISSIONSCount of hospital readmissions in the acute inpatient setting for a given year.
Medicare SNF daysSNF_STAYSCount of SNF setting stays (unique admissions, which may span more than 1 facility) for a given year. A SNF stay is defined as a set of 1 or more consecutive SNF claims where the beneficiary is only discharged on the most recent claim in the set.
Medicare HH visitsHH_VISITSCount of HH visits for a given year.
Medicare hospice daysHOS_COV_DAYSCount of covered days in the hospice setting for a given year.
Hospital Outpatient visitsHOP_VISITSCount of unique revenue center dates (as a proxy for visits) in the hospital outpatient setting for a given year.
Total ER visits (total)HOP_ER_VISITS + IP_ER_VISITS 
   ER visits without an admissionHOP_ER_VISITSCount of unique ED revenue center dates (as a proxy for an ED visit) in the hospital outpatient data file for a given year.
   ER visits resulting in an admissionIP_ER_VISITSCount of ED claims in the inpatient setting for a given year.
Physician office visitsEM_EVENTS + PHYS_EVENTSThe EM_EVENTS variable is the count of events for the Part B E&M services for a given year. E&M claims are a subset of the claims in the Part B Carrier and DME data files, and a subset of physician claims.The PHYS_EVENTS variable is the count of events in the Part B PHYS for a given year. Physician office claims are a subset of the claims in the Part B Carrier and DME data files, and a subset of physician E&M claims.
ASC visitsASC_EVENTSCount of events in the Part B ASC setting for a given year.
Dialysis eventsDIALYS_EVENTSCount of events for Part B dialysis services (primarily the professional component since treatments are covered in hospital outpatient) for a given year.
Anesthesia eventsANES_EVENTSCount of events for Part B ANES for a given year.
Imaging eventsIMG_EVENTSCount of events for IMG for a given year.
Test eventsTEST_EVENTSCount of events in for Part B tests for a given year.
Other proceduresOPROC_EVENTSCount of events for Part B other procedures for a given year.
DMEDME_EVENTSCount of events in the Part B DME for a given year.
Part BPTB_DRUG_EVENTSCount of events in the Part B drug setting for a given year.
Part D2PTD_EVENTS where PLNCOVMO not equal to 0Count of events for Part D drugs for a given year (i.e., a unique count of the PDE_IDs). An event is a dispensed (filled) drug prescription covered by the Part D benefit.PLNCOVMO is Part D Plan Coverage Months.

For MMEs, the following three variables were defined using the Medicaid Person Summary data for 2007 and 2008.

  • Medicaid HCBS. These services help Medicaid beneficiaries stay in their communities and are implemented by states. As stated above, they can cover a range of services including DME and transportation. Sum of claims (FFS_CLM_CNT) when type of service represents one of the following: personal care services, residential care, DME, private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice. (TOS=13, 30, 31, 33, 35, 26, 38, 51, 52, 54).

  • Medicaid Nursing Facility Days. Given this is a Medicaid-covered service, it is only included in the Medicaid person summary file and not the Medicare claims data. This value is equal to the variable named "TOT_LTC_CVR_DAY_CNT_NF", which represents total covered nursing facility days, including long-term stays.

  • Other Long-Term Days (excluding Nursing Facility). This is the count of covered days for any of the following types of service: mental hospital, inpatient psych (age <21), and intermediate care facility for persons with mental retardation covered days.

7. Health Care Spending

Medicare plan paid variables were created for various types of services (e.g., total admissions PMPM). Payments are not standardized due to geographic differences in payment rates. Therefore, care must be used when comparing average spending for a particular service across the geographies.30

Table A4 includes the health care spending metrics created in our dataset and the corresponding Medicare variable or Medicare data-based logic. Variables were created for each year available in the Medicare (2007, 2008, 2009 and 2010) data. Definitions for the services included in each variable are included in Table A3.

TABLE A4. Medicare Health Care Spending Variable Definitions
Health Care Spending MetricMedicare Plan Paid Definition
NOTE:
  1. Limited to beneficiaries with Part D coverage.
Total medical and pharmacy paymentsTOT_MED_MDCR + PTD_MDCR_PMT
Total medical paymentsSum of all individual values below excluding long-term care and Part D drugs (not double counting categories that have subcategories (e.g., total admissions))
Total admissionsACUTE_MDCR_PMT + OIP_MDCR_PMT
   Acute stay admissionsACUTE_MDCR_PMT
   Other inpatient admissionsOIP_MDCR_PMT
Medicare SNF daysSNF_MDCR_PMT
Medicare HH visitsHH_MDCR_PMT
Medicare hospice daysHOS_MDCR_PMT
Hospital outpatient visitsHOP_MDCR_PMT
Physician office visitsEM_MDCR_PMT + PHYS_MDCR_PMT
ASC visitsASC_MDCR_PMT
Dialysis eventsDIALYS_MDCR_PMT
Anesthesia eventsANES_MDCR_PMT
Imaging eventsIMG_MDCR_PMT
Test eventsTEST_MDCR_PMT
Other proceduresOPROC_MDCR_PMT
DMEDME_MDCR_PMT
Part BPTB_DRUG_MDCR_PMT
Part D1PTD_MDCR_PMT

 

APPENDIX B. STUDY SAMPLES AND ANALYTIC APPROACH

A. Study Samples

1. Objective A, B, and C study samples

Our sample of HUD-assisted individuals consisted of all individuals that appear in the TRACS or PIC data at any point in time during 2008. This sample was the basis for Objective A (Understanding demographic characteristics of HUD-assisted individuals). The "TRACS/PIC base" sample was created by combining the TRACS and PIC data into one dataset. For individuals that appear in both datasets (across all three years this was 38,326 individuals, or 1.54%), we created only one observation per individual based on the PIC data. Individuals were excluded if an individual's move out date in a given year occurred prior to the current data year31 (N=194,674). The resulting total sample size was 2,057,893 unique individuals and 967,557 unique heads of households.

The sample of HUD-assisted individuals for Objective B (Calculate match rate between HUD and Medicare and/or Medicaid data) and Objective C (Estimate HUD-assisted Medicare beneficiaries with enrollment in Part D subsidy assistance programs or Medical Savings Program) was a subset of this sample. The Medicare and Medicaid data can have duplicate observations for a given SSN due to a variety of reasons (e.g., Medicaid beneficiary who moved to a different state mid-year, data error, etc.). We excluded any individual who had more than one observation in the Medicare or Medicaid data from Objective B in order to calculate an accurate Participation and Match Rate without duplicate beneficiaries. HUD-assisted individuals who appeared multiple times in the Medicare or Medicaid data files were therefore excluded.

Table B1 presents the HUD-assisted individuals 2008 sample size for Objective A and the reduced sample for Objectives B and C.

TABLE B1. HUD-Assisted Individuals 2008 Unique Observations
Objective SampleIndividuals
Objective A sample: HUD-assisted individuals 2008 unique observations2,057,893
Objectives B & C sample: HUD-assisted individuals 2008 unique observations, removing Medicare/Medicaid duplicates  2,025,126

2. Objective D Study Samples

As discussed in the Summary Report and Appendix C, the sample was refined in order to compare HUD-assisted Medicare beneficiaries to unassisted beneficiaries in the community with regards to enrollment, chronic conditions, health care payment and utilization. Although our overall study focused on individuals age 62 and older receiving housing assistance, we narrowed this analysis to individuals age 65 and older to result in the most representative comparison because the vast majority of individuals age 65 and older have Medicare coverage.

Before comparing groups of beneficiaries, we first refined our study sample of Medicare beneficiaries (both HUD-assisted and unassisted beneficiaries in the 12 study jurisdictions). Figure B1 depicts the inclusion criteria. Medicare beneficiaries had to be enrolled in Part A Hospital and Part B Physician Services for all 12 months of 2008 or up until death and not have any Medicare managed care (i.e., Medicare Part C or Medicare Advantage)32 enrollment in 2008. This reduced the sample of Medicare beneficiaries 38%, down to 3.6 million. Approximately 7% of the Medicare beneficiaries were HUD-assisted individuals and 93% were unassisted in the community.

FIGURE B1. Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison, All Ages

FIGURE B1, Inverted Pyramid: Top layer, All Medicare beneficiaries (N=5,752,662); Middle layer, Enrolled in A&B for entire 12 month period or up until death (N=4,799,074 or 83%); Lower layer, Medicaid Fee-for-service only (N=3,562,032 or 62%).


B. Analytic Approach

This section outlines our analytic approach by research question.

1. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions

We calculated the number and percentage of HUD-assisted individuals by characteristic (e.g., race, living arrangement). Objective A results were stratified by year (2007, 2008, 2009), age group (<62, 62-64, 65+ years of age), the 12 geographic areas, subsidy type, and property type. For certain characteristics, such as living arrangement, we calculated the metric on both an individual and household level. In addition to individual characteristics, we also looked at the number and proportion of properties defined as an elderly property (see Appendix A for variable definitions).

2. Calculate HUD and Medicare and/or Medicaid Match Rate in 12 Study Jurisdictions

The main objective of this study task was to determine the ability to link HUD-assisted individuals to Medicare and Medicaid administrative data (i.e., enrolled in Medicare or Medicaid health care programs) in the 12 study jurisdictions. We focus the results on 2008 -- the most recent year for which we had data on HUD, CMS Medicare, and CMS Medicaid enrollment. We accomplished this objective by calculating two measures that represent the ability to link the HUD and CMS data sources:

  1. SSN Match Rate. The proportion of HUD-assisted individuals enrolled in Medicare and/or Medicaid programs (determined solely on a matched SSN between the datasets).

  2. Restrictive Match Rate. A fairly conservative method on which we established whether the individuals receiving housing assistance were definitively included in the Medicare and/or Medicaid data by requiring a direct match on SSN, gender, and date of birth.

The SSN Match Rate gives a fairly good estimate on the proportion of HUD-assisted individuals that had any Medicare and/or Medicaid enrollment during the observation period. The more conservative Restrictive Match Rate creates a sample for which we can be very confident the HUD-assisted individual was in fact enrolled in Medicare and/or Medicaid. This method is important for identifying a sample of beneficiaries with HUD housing assistance for which we would then compare to other Medicare and/or Medicaid beneficiaries' payment and utilization.

The following outlines the steps we completed to calculate the Match Rates:

  1. Lewin sent Buccaneer/General Dynamics Information Technology (GDIT) HUD data that included the individuals with a valid SSN receiving housing assistance at any point in time during 2007, 2008, and 2009.

  2. Buccaneer/GDIT linked these files to Medicare and Medicaid data (not limited to any particular years) based on SSN. Buccaneer/GDIT sent Lewin a "crosswalk" file that contained the list of all valid SSNs that linked to Medicare or Medicaid data and the corresponding Medicare/Medicaid beneficiary identification number (Bene_ID).

    • This crosswalk contained HUD individuals with valid SSNs who had a Medicare or Medicaid Bene_ID at any point in time.
  3. Of the individuals in the crosswalk file, Lewin limited it those who had a Bene_ID present in our years of study: 2007, 2008, 2009, or 2010.

  4. Calculate Participation rate: same SSN only.

  5. Calculate Match rate: Same SSN, date of birth, and gender.

To compare the HUD-assisted individuals who linked to those that did not link to the Medicare or Medicaid data (based on the Match Rate definition), we calculated the number and percentages of individuals with certain characteristics. We calculated the chi-square statistic for categorical variables (e.g., gender) and the t-test for continuous variables (e.g., age) to test if the two samples were statistically different. We also created summary statistics to examine the Medicare enrollment characteristics among those that linked, including the participation rates in Medicare Part D LIS and Medicaid programs.

3. Estimate the Proportion of HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program (i.e., Medicaid)

We calculated the number and percentage of HUD-assisted Medicare beneficiaries by Part D subsidy assistance program types and the Medicaid coverage status categories.

4. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions

Similar to the above, to compare demographic and coverage characteristics, we calculated the number and percentages of individuals with certain characteristics along with the chi-square statistic to compare the distribution of categorical variables (e.g., reason for entitlement) and means and standard deviations along with the t-test to compare continuous variables (e.g., months of Medicare coverage) across three strata: (1) HUD-assisted beneficiaries partial-year; (2) HUD-assisted beneficiaries full year; and (3) unassisted beneficiaries in the community. The results of this comparison then informed whether or not to combine HUD partial-year and HUD full year into one group for the comparison of chronic conditions, payment and utilization.

To compare prevalence of chronic conditions, we calculated the number and percentage of individuals with nine different categories of conditions (See Appendix A for variable definitions) and the chi-square statistic to test if the groups were statistically different. To examine health care utilization and payments, we used descriptive techniques that account for length of enrollment time: utilization metrics were calculated per 1000 member months, and we used PMPM for health care payments. Means, medians, and standard deviations were calculated for these variables and t-tests were calculated for significance testing.

Note that with large sample sizes (such as those of our study samples) tests of association tend to be statistically significant.

 

APPENDIX C. 2008 DETAILED RESULTS

Introduction

The results presented below are organized by study objective and based on 2008--the most recent year for which we had data on HUD, CMS Medicare, and CMS Medicaid enrollment. We also produced results for Objectives A, B, and C by year (2007, 2008, and 2009), which are included in Appendix D.

A. Understand Demographic Characteristics of HUD-Assisted Individuals in the 12 Study Jurisdictions

The HUD individual tenant-level data for 2008 included 2,057,893 unique individuals and 967,557 unique household heads that received housing assistance at some point during 2008. The tables below summarize characteristics of these individuals based on the data available in the HUD data sources.

1. Demographic Characteristics of HUD-Assisted Individuals

Table C1 presents the demographic characteristics for HUD-assisted individuals in the 12 study jurisdictions.

TABLE C1. Demographic Characteristics of HUD-Assisted Individuals, 2008, 12 Study Jurisdictions
 Unique Individuals(N=2,057,893)
MeanSD
NOTES:
  1. Thepurpose of the property type variable was to assign individuals to the actual type of property in which they live whereas the subsidy type variable assigned individuals to a category based on their subsidy type regardless of the type of property in which they live. While there is much overlap between the two variables, the main difference is for individuals receiving a project-based Section 8 subsidy. The property type variable distributes individuals receiving a project-based Section 8 subsidy to the actual property in which they live -- Other multifamily, Section 202, Section 811 and 202/162, or Other. The majority of individuals with project-based Section 8 subsidies were mapped to "Other multifamily". However, a larger portion of individuals age 65+ with a project-based Section 8 subsidy were mapped to "Section 202" property.
  2. Assisted households are categorized by property type in which they live, despite minor variations in subsidy type. "Housing choice vouchers" include all types of buildings in which housing choice voucher holders live, from a single-family home to a large apartment complex.
Age
All Persons34.925.3
Household Heads54.218.8
 N%
Gender
Male765,16937.2%
Female1,291,14062.7%
Unknown/declined to report1,5840.1%
Age Group
<18764,54337.2%
18-44566,05927.5%
45-61327,06715.9%
62-6446,7002.3%
65+352,09117.1%
Unknown/declined to report1,4330.1%
Race/Ethnicity
White non-Hispanic331,94416.1%
Hispanic489,33723.8%
Black non-Hispanic773,55037.6%
Asian56,9342.8%
American Indian/Alaskan3,6990.2%
Hawaiian/Pacific Islander2,7620.1%
Other non-Hispanic2590.0%
Mixed non-Hispanic2750.0%
Unknown/declined to report399,13319.4%
Property Type1
Public housing635,90630.9%
Housing Choice Vouchers2960,38746.7%
Other multifamily381,07818.5%
Section 20263,0333.1%
Section 811 & 202/1623,9930.2%
Other13,4960.7%
Subsidy Type
Public housing638,58431.0%
Housing Choice Vouchers969,14047.1%
Project-Based Section 8376,55218.3%
Section 20221,3861.0%
Section 202/1621000.0%
Section 8113,8390.2%
Other48,2922.3%
Disability or Handicapped Status
Yes376,33918.3%
No1,472,67971.6%
Unknown/declined to report208,87510.1%
Geographic Area
Vermont22,6351.1%
New Haven- Milford50,2012.4%
Bridgeport- Stamford- Norwalk35,3261.7%
Milwaukee- Waukesha- West Allis51,5552.5%
San Francisco- Oakland- Fremont167,7958.2%
Boston- Cambridge- Quincy243,60111.8%
Durham- Chapel Hill17,9450.9%
Richmond43,1962.1%
New York- Northern New Jersey- Long Island1,213,20159.0%
Columbus72,3323.5%
Akron35,4091.7%
Cleveland104,6975.1%

Over half of the sample (59%) was located in the New York-Northern New Jersey-Long Island area. The remaining ten geographies each represent 5% or less of the sample. Given the New York City area is the majority of our sample, the individuals living in this area heavily influence the other characteristics presented in the result tables.

The majority of individuals were female (63%) with an average age of 34.9 years. Approximately 19% of the sample was older than 62 years of age ("elderly") with 37% younger than age 18. As expected, among household heads, the average age was much higher (54.2 years). While 19% of the sample declined to report race/ethnicity, among the individuals where the information was available, 38% were non-Hispanic Black, 24% were Hispanic, 16% were White non-Hispanic, 3% were Asian, and less than 1% were American Indian/Alaskan, Hawaiian/Pacific Islander, Other non-Hispanic or Mixed non-Hispanic. Only 18% of individuals were designated as living in a household identified as a disability status or handicapped household (11% declined to report so 20% among those reporting).

The largest proportion of subsidy type received was Housing Choice Vouchers (47%). Approximately 31% resided in public housing and another 18% lived in multifamily project-based Section 8 housing. Among individuals age 62 years or older, public housing, other multifamily, and Housing Choice Vouchers were the most common (31%, 29%, and 25%) with an additional 14% residing in Section 202 properties (data not shown, see Appendix D).

TABLE C2. Living Arrangement, All Ages and Age 62+ Years
 All Ages(N=2,057,893)Age 62+ Years(N=398,791)
N%N%
Living arrangement
Lives Alone450,41021.9%267,75167.1%
Lives with Spouse89,0664.3%68,69417.2%
Other Adults144,7057.0%30,6037.7%
Minor(s) in Household1,345,19065.4%24,1306.1%
Live-in Aide7,0390.3%1,8770.5%
Other2830.0%380.0%
Unknown/declined to report21,2001.0%5,6981.4%
Living with minor(s) in household
Age 62+ with minor(s) in household    24,1306.1%

Table C2 presents living arrangement on an individual level by age (all ages and 62+ years only) and Table C3 presents living arrangement on a household level. Approximately 65% of individuals of all ages were in households with children and 22% lived alone. In contrast, among individuals age 62 years or older, over 67% live alone and 6% had minors in the home. On a household level, 41% of households had children while 47% of households consisted of a single individual. Among households with more than one person, there was an average of 3.1 people (data not shown).

  TABLE C3. Living Arrangement, Household Level, All Ages  
 Households(N=967,557)
N%
Living arrangement
Lives Alone448,55646.4%
Lives with Spouse45,9854.8%
Other Adults68,7517.1%
Minor(s) in Household394,26040.7%
Live-in Aide3,1310.3%
Other780.0%
Unknown/declined to report6,7960.7%

Table C4 presents the number and proportion of HUD-assisted individuals who received housing assistance for the entire year of 2008 or for only a portion of the year by age (all ages, 62+ years; 65+ years). For those individuals that only received housing assistance for a portion of the year, the reason is also included. Approximately 87% of all individuals received housing assistance for all of 2008. Among the 13% who only received housing assistance for a portion of the year, the large majority (60%) had moved into the property after January 1stof the calendar year. Approximately 40% of partial-year participants terminated the program or moved out of the property. Among individuals age 65 years or older, 79% received housing assistance for the full year. Similarly, the reason for partial-year assistance was also different among this older sample -- 54% had terminated the program or moved out of the property compared with 40% of partial-year recipients across all ages. This difference was expected because older individuals would be more likely to move to another facility (such as a nursing home) or die.

TABLE C4. Full and Partial HUD Housing Assistance, 2008, by Age Group
 All Ages(N=2,057,893)Age 62+(N=398,791)Age 65+(N=352,091)
N%N%N%
Full year participants1,783,20286.7%353,78988.7%313,03588.9%
Partial-year participants274,69113.3%45,00211.3%39,05611.1%
Reason for partial-year
Moved Out/Terminated110,37040.2%22,81550.7%20,87653.5%
Moved in after Jan 1164,32159.8%22,18749.3%18,18046.5%

2. HUD Elderly Properties

Table C5 presents the proportion of properties that were defined as elderly33 among all public housing properties (from PIC) and private assisted multifamily properties (from TRACS) within the 12 study jurisdictions. Half of all multifamily properties had an individual age 62 years or older in over 50% of their units. Approximately 42% of all public housing properties were either designated as elderly by PHAs or had an individual age 62 years or older in over 50% of their units. Both of these proportions varied greatly by the 12 study jurisdictions, from 30% of multifamily properties in Durham-Chapel Hill, North Carolina to 68% of multifamily properties in Bridgeport-Stamford-Norwalk, Connecticut and only 19% of public housing properties in Cleveland compared to 90% of public housing properties in Vermont (data not shown, see Appendix D).

TABLE C5. Proportion of HUD-Assisted Properties in 12 Study Jurisdictions Defined as "Elderly"
Property TypesNumber of PropertiesPercent of Properties
Multifamily elderly properties (reported in TRACS)1,81150.0%
Public Housing elderly properties (reported in PIC)33542.0%

B. Calculate HUD and Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008

1. Match Rate: HUD-Assisted Individuals to Medicare or Medicaid Enrollment Data

The main objective of this study task was to assess the feasibility and potential for successful (i.e., accurate and complete) matching to Medicare and Medicaid administrative data (i.e., enrolled in Medicare or Medicaid health care programs) in the 12 study jurisdictions to support research and policy analysis. We accomplished this objective by calculating two measures that represent the ability to link the HUD and CMS data sources:

  1. "SSN Match Rate". The proportion of HUD-assisted individuals enrolled in Medicare and/or Medicaid programs (determined solely on a matched SSN between the datasets).

  2. "Restrictive Match Rate". A fairly conservative method on which we established whether the individuals receiving housing assistance were truly included in the Medicare and/or Medicaid data by requiring a direct match on SSN, gender, and date of birth.34

Table C6 presents the Medicare and Medicaid SSN match rate and restrictive match rate for the 2008 sample of HUD-assisted individuals, excluding duplicates. SSN match rate and restrictive match rate are presented separately for individuals less than age 65 years and those age 65 years or older to align with Medicare eligibility at age 65. Before calculating SSN and restrictive match rates, we excluded individuals from the HUD sample that do not have valid SSNs. This results in a less than 1% reduction in sample size for both age groups.

Among HUD-assisted individuals age 65 years or older with a valid SSN, the Medicare SSN Match Rate was 93%. This decreased by 8 percentage points to 85% when requiring a match on gender and date of birth in addition to SSN. Among HUD-assisted individuals less than age 65 with a valid SSN, the Medicaid SSN match rate was 73%. This decreased to 69% when requiring a match on gender and date of birth in addition to SSN. Approximately 63% of HUD-assisted individuals age 65 years or older with a valid SSN linked to Medicare and Medicaid in 2008 based on SSN alone. The restrictive match rate for Medicare and Medicaid among this sample was 58%.

TABLE C6. Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals,1 2008, by Age Group
 Age <65 Years(N=1,663,348)Age 65+ Years(N=349,208)
N%N%
NOTE:
  1. Excluded individuals with valid SSNs (drop invalid SSNs such as "999" of "T.."). Only reduced sample by approximately 1%.
Medicare
SSN Match Rate (SSN only)116,8037.0%325,58993.2%
Restrictive Match Rate (SSN, gender, date of birth)110,3956.7%297,22785.1%
Medicaid
SSN Match Rate (SSN only)1,200,98172.2%229,09365.6%
Restrictive Match Rate (SSN, gender, date of birth)1,131,13768.5%209,67960.0%
Medicare and Medicaid
SSN Match Rate (SSN only)96,5095.8%220,13963.0%
Restrictive Match Rate (SSN, gender, date of birth)91,5645.5%201,75357.8%

We also calculated the restrictive match rate for individuals age 65 and older by property type and geography and found little variation (data not shown, see Appendix D). The percentages that linked to Medicare or Medicaid are consistently high across property types, ranging from 84.2% (public housing) to 89.4% (Section 202). These percentages are also high across geographic area (86.2% in New York-Northern New Jersey-Long Island to 94.2% in Vermont) with the exception of Richmond, where only 31.2% linked.

TABLE C7. Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI,1 2008, by Age Group
 Age <65 Years(N=158,323)Age 65+ Years(N=131,335)
N%N%
NOTE:
  1. Excluded individuals with valid SSNs (drop invalid SSNs such as "999" of "T.."). Only reduced sample by approximately 1%.
Medicare
SSN Match Rate (SSN only)49,36731.2%123,75994.2%
Restrictive Match Rate (SSN, gender, date of birth)46,87529.6%112,78385.9%
Medicaid
SSN Match Rate (SSN only)143,44390.6%124,73095.0%
Restrictive Match Rate (SSN, gender, date of birth)134,32084.8%113,71186.6%
Medicare and Medicaid
SSN Match Rate (SSN only)46,71729.5%120,81192.0%
Restrictive Match Rate (SSN, gender, date of birth)44,38928.0%110,25483.9%

Table C7 presents the Medicare and Medicaid SSN match rate and restrictive match rate among the 2008 sample of HUD-assisted household heads who were receiving SSI in 2008. The purpose of calculating the SSN match rate and restrictive match rate for this subsample is that individuals eligible for SSI are automatically eligible for Medicaid. Therefore, we would expect to see close to 100% SSN match rate, understanding a perfect Match Rate is not expected given limitations in the dataset. Among head of households receiving SSI under age 65 with a valid SSN in the HUD data, 91% were linked to Medicaid enrollment files on SSN alone. Among head of households receiving SSI age 65 and older, the Medicare SSN Match Rate was 94% and the SSN match rate for Medicaid and Medicare was 92%. While the proportions are not 100%, over 90% of individuals receiving SSI were enrolled in SSI in 2008. Additional analysis would need to be conducted in order to better estimate program enrollment among individuals that are eligible.

From this point forward, all HUD-assisted individuals identified as Medicare beneficiaries are based on the restrictive match rate -- individuals that linked based on gender, date of birth, and SSN. The more conservative Restrictive Match Rate creates a sample for which we can be very confident the HUD-assisted housing recipient was in fact enrolled in Medicare and/or Medicaid.

2. HUD-Assisted Individuals Matched to Medicare Administrative Data Compared to HUD-Assisted Individuals Who Did Not Match

Next we wanted to better understand the characteristics of HUD-assisted individuals that matched to the CMS administrative data relative to the individuals that did not match.

Table C8 presents a comparison of HUD-assisted individuals who were linked to Medicaid or Medicare enrollment files to those that were not linked (using the restrictive match based on gender, date of birth, and SSN). Among individuals less than 65 years of age, we compared those who linked to Medicaid compared to those who did not. Individuals who linked to Medicaid files were more likely to be White (17% vs. 11%); much more likely to have received Housing Choice Vouchers (56% vs. 42%); disabled (19% vs. 10%); and received SSI (18% vs. 6%).

Similar results were found when comparing individuals age 65 years or older who linked to either Medicare or Medicaid enrollment files. Individuals who linked to Medicare or Medicaid were more likely to be White (22% vs. 15% of individuals who did not link); more likely to receive Housing Choice Vouchers (24.5% vs. 20.2%) or live in Section 202 (elderly supportive housing) property (15.9% vs. 12.9%); and more likely to receive SSI (44% vs. 38%).

TABLE C8. Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid, 2008, by Age Group
 HUD-Assisted Individuals(<65 Years)HUD-Assisted Individuals(65+ Years)
Linked to 2008Medicaid Enrollment Files(N=1,131,137)Did No Link to2008 Medicaid Enrollment Files(N=542,905)Linked to 2008Medicaid Enrollment Files(N=305,153)Did No Link to2008 Medicaid Enrollment Files(N=44,505)
MeanSDMeanSDMeanSDMeanSD
NOTES:
  1. As defined in the HUD data source.

* Cell sizes less than 11 are not displayed.

Age--All Persons24.617.930.418.476.37.776.98.3
Age--Household Heads42.611.845.511.576.57.877.28.3
 N%N%N%N%
Gender
Male429,97238.0%216,91840.0%91,24129.9%12,41630.15%
Female701,16562.0%325,90360.0%213,91270.1%30,96569.6%
Unknown/declined to report00.0%840.0%00.0%1240.3%
Age Group
<18531,04246.9%168,35631.0%    
18-44383,52233.9%220,23140.6%    
45-61194,29617.2%130,22024.0%    
62-6422,2772.0%24,0984.5%    
65+    305,153100.0%44,505100.0%
Unknown/declined to report00.0%00.0%00.0%00.0%
Race/Ethnicity
White non-Hispanic196,34417.4%59,77311.0%65,94221.6%6,46214.5%
Hispanic293,36725.9%131,00724.1%46,76015.3%8,52819.2%
Black non-Hispanic456,22640.3%241,90744.6%48,78516.0%10,72924.1%
Asian29,1992.6%14,1992.6%11,3733.7%1,2502.8%
American Indian/Alaskan2,2220.2%9380.2%4280.1%490.1%
Hawaiian/Pacific Islander1,4920.1%7900.1%3510.1%390.1%
Other non-Hispanic1460.0%660.0%330.0%*0.0%
Mixed non-Hispanic1320.0%560.0%730.0%*0.0%
Unknown/declined to report152,00913.4%94,16917.3%131,40843.1%17,43339.2%
Property Type
Public housing314,20127.8%206,23238.0%89,01329.2%16,70937.5%
Housing Choice vouchers630,61055.8%226,88441.8%74,79624.5%9,00520.2%
Other multifamily170,34015.1%102,26418.8%91,46930.0%12,83628.8%
Section 2026,3300.5%2,8170.5%48,42415.9%5,75012.9%
Section 811 & 202/1622,7360.2%8930.2%2980.1%370.1%
Other7,9200.7%3,8150.7%1,1530.4%1680.4%
Subsidy Type
Public housing315,98527.9%206,87438.1%89,29029.3%16,75337.6%
Housing Choice vouchers636,52056.3%228,80042.1%75,35124.7%9,09420.4%
Project-based Section 8155,41013.7%88,14316.2%113,58237.2%15,25934.3%
Section 2024670.0%4850.1%18,2586.0%1,9964.5%
Section 202/162310.0%110.0%510.0%*0.0%
Section 8112,6270.2%8550.2%2940.1%370.1%
Other20,0971.8%17,7373.3%8,3272.7%1,3593.1%
Elderly Designated Property1
Yes59,7885.3%40,5237.5%160,39852.6%20,10245.2%
No438,54138.8%274,45750.6%69,37722.7%15,32834.4%
Unknown632,80855.9%227,92542.0%75,37824.7%9,07520.4%
Disability or Handicapped Status
Yes213,84318.9%53,5579.9%90,97429.8%13,51130.4%
No793,73970.2%407,93875.1%213,60370.0%30,83369.3%
Unknown/declined to report123,55510.9%81,41015.0%5760.2%1610.4%
Living Arrangement
Lives alone123,75910.9%82,82915.3%209,45068.6%29,61266.5%
Lives with spouse12,8901.1%13,2672.4%55,41918.2%6,62614.9%
Other adults58,8165.2%59,54511.0%20,7136.8%4,2779.6%
Children in household925,45581.8%375,23469.1%14,5374.8%3,0826.9%
Live-in aide2,6550.2%2,7790.5%1,3160.4%2240.5%
Other1050.0%1480.0%190.0%*0.0%
Unknown/declined to report7,4570.7%9,1031.7%3,6991.2%6771.5%
SSI
Receives SSI198,45717.5%34,1406.3%134,48444.1%16,94438.1%
Geographic Area
Vermont14,9561.3%2,9100.5%4,3501.4%2700.6%
New Haven- Milford31,4512.8%10,3961.9%7,0162.3%7671.7%
Bridgeport- Stamford- Norwalk19,8311.8%8,6511.6%5,5381.8%7161.6%
Milwaukee- Waukesha- West Allis34,0943.0%7,7871.4%8,3732.7%7061.6%
San Francisco- Oakland- Fremont82,5867.3%48,0478.8%29,5429.7%3,1737.1%
Boston- Cambridge- Quincy147,75013.1%46,3868.5%43,12414.1%4,79910.8%
Durham- Chapel Hill11,8771.1%4,2710.8%1,3800.5%2060.5%
Richmond11,0051.0%28,1145.2%1,1850.4%2,6145.9%
New York- Northern New Jersey- Long Island628,30455.5%349,78064.4%182,50759.8%29,14165.5%
Columbus51,1594.5%12,4172.3%7,1072.3%7391.7%
Akron25,5982.3%5,5361.0%3,4671.1%2600.6%
Cleveland72,5266.4%18,6103.4%11,5643.8%1,1142.5%

C. Estimate Proportion of HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs or Medical Savings Program (i.e., Medicaid)

As detailed in an October 2012 CBPP memo,35 HUD-assisted individuals may have unreimbursed medical expenses deducted from their income for the purpose of determining the level of assistance provided by HUD. According to the CBPP, approximately 700,000 assisted households claim such expenses, which increase HUD rental subsidy costs by $400 to $500 million per year. Therefore, HUD is interested in understanding what proportion of HUD-assisted Medicare beneficiaries are enrolled in programs that could reduce their out-of-pocket health care expenses and therefore reduce HUD's costs. These programs include:

  1. Medicare Part D (drug coverage) LIS. Medicare Part D provides coverage for prescription medications and the LIS program assists beneficiaries with paying the premium and out-of-pocket expenses for their medication.

  2. Medicare Savings Program. This is more commonly referred to as being a MME. Medicare beneficiaries may be eligible for Medicaid to assist with out-of-pocket expenses and to receive Medicaid-covered services that are not covered by Medicare (e.g., transportation).

Table C9 and Table C10 present the proportion of assisted Medicare beneficiaries, age 65 years or older enrolled in Part D LIS and Medicaid. Approximately 90% of the Medicare beneficiaries had Part D coverage. Among those with Part D coverage, 82% received premium and out-of-pocket assistance. This proportion increased to 90% among those under age 65. Approximately 72% of Medicare beneficiaries were dually enrolled in Medicare and Medicaid (83% of beneficiaries under age 65; 68% of beneficiaries age 65 and over). The majority (68%) of MMEs were QMB plus (receiving Medicaid assistance with Medicare expenses and fully eligible for Medicaid-covered services). Almost all of the remaining beneficiaries were "Other dual" (QDWIs not eligible for SLMB or QMB). These proportions were similar in both individuals under age 65 and age 65 or older. As expected, the Medicaid basis for eligibility was largely blind/disabled (95%) for beneficiaries under age 65 while for beneficiaries age 65 years or older it was aged (77%) followed by blind/disabled (23%). Among the 8% of all Medicare and Medicaid beneficiaries who had 1915c waiver, 78% was for aged or aged/disabled. Among beneficiaries less than 65 years of age with a 1915c waiver, almost 60% received the mental retardation and other developmental disabilities (MR/DD) waiver.

TABLE C9. HUD-Assisted Medicare Beneficiary Enrollment in Part D LIS, 2008
Medicare BeneficiariesTotalUnder Age 65Age 65+
Total407,622110,395297,227
Part D Beneficiaries365,812102,927262,885
   Percent of Medicare Beneficiaries89.7%93.2%88.5%
Part D LIS Participants301,53792,142209,395
   Percent of Part D with LIS82.4%89.5%79.7%
   Percent of Medicare with Part D LIS74.0%83.5%70.4%
   Percent of Medicare without Part D LIS26.0%16.5%29.6%

It is important to note that the results presented in Table C9 and Table C10 are the proportion of all HUD-assisted Medicare beneficiaries who match based on the restrictive match criteria. It is not the proportion of HUD-assisted Medicare beneficiaries who are eligible for these programs based on national criteria for Part D LIS and criteria established by each state for Medicaid benefits. In other words, we would not expect the proportion to be 100% as not all HUD-assisted Medicare beneficiaries are eligible for these programs because the income criteria for HUD-assisted housing can be higher than that for health insurance assistance.

TABLE C10. HUD-Assisted Medicare Beneficiary Enrollment in Medicaid and Medicaid Coverage Characteristics, 2008
Medicare BeneficiariesTotalUnder Age 65Age 65+
Total Medicare beneficiaries407,622110,395297,227
Medicare beneficiaries dually enrolled with Medicaid (MMEs)293,31791,564201,753
Percent of Medicare with Medicaid72.0%82.9%67.9%
Percent of Medicare without Medicaid28.0%17.1%32.1%
Medicare and Medicaid eligibility status293,31791,564201,753
   QMB only3.7%4.9%3.1%
   QMB plus66.9%64.8%67.8%
   SLMB only3.7%3.4%3.8%
   SLMB Plus1.6%1.8%1.5%
   Other dual23.3%23.6%23.2%
   NA0.8%1.4%0.5%
Medicaid basis for eligibility293,31791,564201,753
   Aged53.2%0.1%77.3%
   Blind/disabled45.3%95.3%22.6%
   Child0.0%0.0%0.0%
   Adult1.4%4.5%0.1%
   Other child of unemployed adult0.0%0.1%0.0%
Medicaid maintenance of assistance293,31791,564201,753
   Receiving cash or eligible under Section 1931 of the Act58.2%48.4%62.7%
   Medically needy13.7%14.9%13.2%
   Poverty-related (includes children eligible under SCHIP expansion)18.3%23.9%15.8%
   Other7.6%9.5%6.8%
   Foster care child0.0%0.0%0.0%
   1115 demonstration expansion eligible2.1%3.2%1.6%
   Unknown0.0%0.0%0.0%
   Missing0.0%0.1%0.0%
Beneficiaries with 1915c waiver status code23,3566,39916,957
   Percent of Medicare/Medicaid beneficiaries with 1915c waiver status code8.0%7.0%8.4%
Type of 1915c waiver status code (among beneficiaries with a waiver)
   Aged and disabled54.3%25.0%65.4%
   Aged23.6%1.5%31.9%
   Disabled3.0%10.4%0.1%
   Brain injured0.3%1.1%0.01%
   HIV-AIDS0.7%2.2%0.1%
   MR/DD18.1%59.7%2.4%
   Mentally ill/severely emotionally disturbed0.0%0.0%0.0%
   Technology-dependent/medically fragile0.0%0.02%0.0%

D. Compare Medicare and Medicaid Payment and Utilization for HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions

After we identified the HUD-assisted individuals who were Medicare and/or Medicaid beneficiaries in 2008, the next objective was to compare HUD-assisted Medicare beneficiaries to unassisted beneficiaries in the community with regards to enrollment, chronic conditions, health care payment and utilization. Before comparing groups of beneficiaries, we first refined our study sample of Medicare beneficiaries (both HUD-assisted and unassisted beneficiaries in the 12 study jurisdictions). Figure C1 presents a graphical depiction of the inclusion criteria. Medicare beneficiaries had to be enrolled in Part AHospital and Part B Physician Services for all 12 months of 2008 or up until death and not have any Medicare managed care (i.e., Medicare Part C or Medicare Advantage)36 enrollment in 2008. This reduced the sample of Medicare beneficiaries 38%, down to 3.6 million. Approximately 7% of the Medicare beneficiaries were HUD-assisted individuals and 93% were unassisted in the community (data not shown).

  FIGURE C1. Inclusion Criteria for HUD-Assisted and Unassisted Medicare Beneficiaries for Payment and Utilization Comparison

FIGURE C1, Inverted Pyramid: Top layer, All Medicare beneficiaries (N=5,752,662); Middle layer, Enrolled in A&B for entire 12 month period or up until death (N=4,799,074 or 83%); Lower layer, Medicaid Fee-for-service only (N=3,562,032 or 62%).


We expected HUD-assisted beneficiaries to differ dramatically from unassisted beneficiaries in the community with regards to health status, utilization, and payment given the differences in demographic factors (e.g., race, income) and Medicare and Medicaid program eligibility due to low income (e.g., Part D LIS and Medicaid participation defined above in Section C). This was supported by preliminary comparisons of HUD-assisted Medicare beneficiaries age 65+ and unassisted Medicare beneficiaries age 65+ in the community. We found:

  • After limiting our sample to HUD-assisted Medicare beneficiaries with continuous enrollment in Medicare Parts A and B with no Medicare managed care during 2008, approximately 70% of HUD-assisted Medicare FFS beneficiaries age 65+ were dually enrolled in Medicaid compared to only 13% of unassisted Medicare FFS beneficiaries in the community.

  • The average total medical and prescription drug PMPM Medicare paid was $1,479 for HUD-assisted Medicare FFS beneficiaries age 65+ compared to only $937 for unassisted Medicare FFS beneficiaries age 65+ in the community.

The higher expenditures for HUD-assisted Medicare beneficiaries likely reflects the much higher proportion of MMEs in the HUD-assisted sample. According to a recent study by the Kaiser Family Foundation, spending for MMEs was 1.8 times higher than for Medicare/No Medicaid counterparts.37 Based on this knowledge, we stratified the Medicare beneficiaries into five sub-groups in an attempt to group similar beneficiaries into distinct samples to reduce variability between the HUD-assisted beneficiaries to the unassisted beneficiaries when comparing their Medicare enrollment characteristics and health care utilization and payment. We stratified the sample into five sub-groups:

  • Sub-groupA (MMEs, all ages): MMEs, with no SNF or nursing facility38 use in 2008.

  • Sub-group A1 (MMEs, age 65+): MMEs, Age 65+, with no SNF or nursing facility use in 2008.

  • Sub-group B (MMEs, all ages, SNF or nursing facility use): MMEs, with >0 and <365 days of SNF or nursing facility use in 2008 (results presented in Appendix D only).

  • Sub-group C (Medicare/No Medicaid): Medicare beneficiaries with no Medicaid enrollment, no SNF use in 2008.

  • Sub-group C1 (Medicare/No Medicaid, age 65+): Medicare beneficiaries with no Medicaid enrollment, age 65+, no SNF use in 2008.

Table C11 presents the sample sizes for each of these five sub-groups on which we conducted comparisons of HUD-assisted Medicare beneficiaries to unassisted beneficiaries. We further stratified the HUD-assisted individuals by length of time receiving assistance during 2008 -- full year or part of the year.

TABLE C11. Five Subgroups for Comparison of Health and Health Care Utilization Among HUD-Assisted Beneficiaries and Unassisted Beneficiaries in the Community
 HUDFullHUDPartialTotal HUD-Assisted BeneficiariesUnassisted Beneficiariesin the CommunityTotal
Exclusion Criteria
All Medicare Beneficiaries362,16245,460407,6225,345,0405,752,662
Medicare benes enrolled in A&B for entire 12 month period or up until death326,70736,450363,1574,435,9174,799,074
No managed care/HMO (Medicare FFS only)232,63026,807259,4373,302,5953,562,032
Sub-group sample sizes
Sub-group A: Medicare/ Medicaid benes, no SNF or NF days162,01015,046177,056446,751623,807
   Sub-group A1: Age 65+104,4167,629112,045249,490361,535
Sub-group B: Medicare/ Medicaid benes with >0 and <365 SNF/NF days12,6843,87516,55986,626103,185
Sub-group C: Medicare/No Medicaid, 65+45,5665,40650,9722,453,7632,504,735
   Sub-group C1: Medicare/ No Medicaid, 65+, No SNF days42,2864,52046,8062,318,3942,365,200

1. Comparison of HUD-assisted fee-for-service MMEs and unassisted fee-for-service MMEs in the community, age 65+ in 12 study jurisdictions, 2008 (sub-group A1)

In 2008 in the 12 study jurisdictions, there were 112,045 HUD-assisted MMEs and 249,490 unassisted MMEs in the community age 65+. Tables C12-C17 present the findings when comparing HUD-assisted MMEs and unassisted MMEs in the community. Sub-group A1 consists of Medicare beneficiaries who were also enrolled in Medicaid, age 65 years or older without any utilization of SNF or nursing facility.

Table C12 compares the demographic characteristics between the HUD-assisted MMEs and unassisted MMEs. The two samples were statistically significantly different across all characteristics. However, the statistical difference is expected given the large sample size and it does not always equate to a meaningful difference. HUD-assisted MMEs had a higher proportion of Blacks (18% vs. 14%) and a lower proportion of Asian/Pacific Islander (13% vs. 18%). The distribution of beneficiaries differs slightly by geography. A higher proportion of HUD-assisted MMEs reside in Boston than unassisted beneficiaries (14% vs. 10%) and a lower proportion in Vermont (2% vs. 4%) and Milwaukee (3% vs. 5%). A higher proportion of unassisted MMEs died in 2008 than HUD-assisted MMEs (3% vs. 2%).

TABLE C12. HUD-Assisted and Unassisted MME Demographic Characteristics
Demographic CharacteristicsHUD-Assisted MMEs(N=112,045)Unassisted MMEs(N=249,490)
N%N%
* Cell sizes less than 11 are not displayed.
Race (based on RTI variable)
Non-Hispanic White49,78144.4%109,98144.1%
Hispanic23,94821.4%50,56320.3%
Black20,11618.0%35,83714.4%
Asian/Pacific Islander14,94913.3%45,77318.3%
American Indian/Alaska Native1070.1%4180.2%
Other2,7962.5%6,2892.5%
Unknown3480.3%6290.3%
Age Group
65 to 7451,25345.7%110,12644.1%
75 to 7924,72822.1%50,50520.2%
80 to 8419,71517.6%42,35717.0%
85+16,34914.6%46,50218.6%
Geographic Area
Vermont2,6212.3%10,8814.4%
New Haven- Milford2,3482.1%4,4081.8%
Bridgeport- Stamford- Norwalk2,0231.8%3,5411.4%
Milwaukee- Waukesha- West Allis3,5653.2%12,1304.9%
San Francisco- Oakland- Fremont14,14412.6%32,72613.1%
Boston- Cambridge- Quincy15,70214.0%25,41910.2%
Durham- Chapel Hill4670.4%2,1340.9%
Richmond2900.3%2,9331.2%
New York- Northern New Jersey- Long Island64,34557.4%137,64255.2%
Columbus1,9301.7%4,5721.8%
Akron9790.9%1,9300.8%
Cleveland3,6313.2%7,5583.0%
Missing*0.0%3,6161.4%
Died during the year1,7191.5%6,4562.6%

Table C13 presents the comparison of Medicare and Medicaid coverage and enrollment characteristics. Although our sample is limited to those dually eligible for Medicare and Medicaid, the basis for Medicaid eligibility differs for those receiving HUD assistance compared to those not receiving HUD assistance. Approximately 24% of HUD-assisted MMEs' basis for Medicaid eligibility was blind/disabled compared to only 15% of unassisted MMEs. Similarly, a higher proportion of HUD-assisted MMEs' Medicaid maintenance of assistance was due to receiving cash or eligibility under Section 1931 (68% vs. 55%) and a lower proportion was due to being medically needy39 (10% vs. 15%) or poverty-related (15% vs. 19%). Finally, HUD-assisted Medicare and Medicaid beneficiaries are more likely to be QMB plus (78% vs. 65%) than unassisted beneficiaries.

TABLE C13. HUD-Assisted and Unassisted MME Medicare and Medicaid Enrollment Characteristics
Topic AreaHUD-Assisted MMEs(N=112,045)Unassisted MMEs in the Community(N=249,490)
N%N%
* Cell sizes less than 11 are not displayed.
Medicaid Basis for Eligibility
Aged85,63576.4%212,43185.1%
Blind/disabled26,34923.5%36,63614.7%
Child*0.0%*0.0%
Adult490.0%2340.1%
Other/missing/unknown*0.0%1870.1%
Medicaid Maintenance of Assistance
Receiving cash or eligible under Section 1931 of the Act76,08367.9%136,91454.9%
Medically needy11,34610.1%36,41414.6%
Poverty-related (includes children eligible under SCHIP expansion)16,49714.7%46,34518.6%
Other5,9405.3%16,7396.7%
Foster care child*0.0%*0.0%
1115 demonstration expansion eligible2,1691.9%12,8965.2%
Unknown/missing*0.0%1820.1%
Medicare and Medicaid Eligibility Status
QMB only3,1382.8%8,8293.5%
QMB plus87,11177.7%163,14665.4%
SLMB only3,5673.2%11,5404.6%
SLMB Plus1,3171.2%3,0201.2%
Other dual16,81715.0%62,37425.0%
Unknown*0.0%1740.1%
NA850.1%4070.2%

Note: We initially separated HUD-assisted MMEs into those receiving HUD assistance all year vs. a portion of the year. There were no meaningful differences between the two samples so we combined them. See Appendix D for results separated by partial vs. all year receiving HUD assistance.

Table C14 compares the presence and count of chronic conditions among the Sub-group A1 HUD-assisted MMEs and unassisted MMEs in the community. HUD-assisted MMEs are, on average, more chronically ill than the unassisted MMEs. First, HUD-assisted MMEs are more likely to have multiple chronic diseases; 55% have five or more compared to 43% of unassisted MMEs. The prevalence of eight of the nine chronic condition categories is also higher among HUD-assisted MMEs. The one exception is Alzheimer's which is 1 percentage point higher (16% vs. 15%) among unassisted MMEs which is most likely due to the fact that individuals with Alzheimer's and related diseases tend to require more care than can be provided in a public housing or related property.

TABLE C14. MMEs with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
CCW ConditionHUD-Assisted MMEs(N=112,045)Unassisted MMEs in the Community(N=249,490)
N%N%
Prevalence of Chronic Conditions
0 conditions6,0315.4%24,4049.8%
1 condition6,5715.9%20,6868.3%
2 conditions9,7658.7%27,67811.1%
3 conditions13,39812.0%34,38713.8%
4 conditions15,21013.6%34,87814.0%
5 or more conditions61,07054.5%107,45743.1%
Prevalence of Chronic Conditions by Category
Cardiovascular62,01255.3%120,34448.3%
Cancer9,1628.2%17,2696.9%
Endocrine and Renal60,07653.6%118,12447.4%
Alzheimer's-related16,42714.7%39,69115.9%
Depression18,60316.6%28,77411.5%
Musculoskeletal55,12949.2%94,35437.9%
Pulmonary21,35219.0%39,20915.7%
Opthalmic42,91038.3%76,17230.6%
Other (Anemia, Hyperlipidemia, Hyperplasia, Hypertension)96,64286.2%196,46578.8%

Table C15 provides a comparison of health care service utilization between HUD-assisted MMEs and unassisted MMEs in Sub-group A1 (dually eligible for Medicaid, age 65 years or older, with no SNF/nursing facility use in 2008). With the exception of dialysis, the difference in utilization (as defined as per 1,000 member months) was statistically significant for all health care services. HUD-assisted MMEs were higher utilizers of health care services. The higher utilization was most notable for home health visits (31% higher), ASC visits (45% higher), other procedures (78% higher), DME (22% higher), and Part B and D drugs (22% and 24% higher, respectively). HUD-assisted MMEs also had a 26% higher rate of physician office visits and 13% higher rate of total emergency room visits. In addition, the proportion of HUD-assisted MMEs with any use was statistically significantly higher across all services but dialysis (data not shown, see Appendix D). As expected, this higher utilization resulted in higher average medical and pharmacy PMPM payments for HUD-assisted MMEs than unassisted MMEs ($1,222 vs. $1,054) (data not shown, see Appendix D).

TABLE C15. Comparison of HUD-Assisted MME and Unassisted MME Medicare Health Care Utilization
Medicare Services Annual Utilization per 1000 Member Months1HUD-Assisted MMEs(N=112,045)Unassisted MMEs(N=249,490)Ratio of HUD-Assistedvs. Unassisted
MeanMean
NOTES:
  1. Utilization per 1000 member months calculated as number of total services across all MMEs divided by number of months enrolled for all MMEs in 2008 multiplied by 1000.
  2. Non-acute inpatient settings include long-term care settings, like nursing facilities.
  3. This variable is the count of unique revenue center dates (as a proxy for visits) in the hospital outpatient setting for a given year.
  4. Limited to Medicare beneficiaries with Part D coverage.

** p<0.01; * p<0.05

Total Admissions32.830.91.06**
   Acute stay admissions31.429.41.07**
   Other inpatient admissions21.41.60.90**
Hospital Readmissions5.24.91.06*
Medicare HH visits581.5445.51.30*
Medicare hospice days36.9208.10.18*
Hospital Outpatient visits3686.5590.51.16**
Total ER visits58.451.61.13**
   ER visits without an admission36.630.61.19**
   ER visits resulting in an admission21.921.01.04**
Physician office visits1,652.31,307.91.26**
ASC visits14.510.01.45**
Dialysis events17.017.30.98
Anesthesia events27.823.71.17**
Imaging events510.5420.11.21**
Test events1,762.21,428.51.23**
Other procedures1,464.6821.61.78**
DME369.2301.41.22**
Part B Drugs296.1241.81.22**
Part D Drugs45,080.14,094.31.24**

The higher utilization of health care services among MMEs receiving HUD assistance than beneficiaries without HUD assistance is even more apparent when comparing utilization and payments of Medicaid-covered services40 (Table C16 and Table C17). HUD-assisted MMEs used over 100% more Personal Care services,41 80% more other HCBS, and 67% more DME services covered by Medicaid than unassisted MMEs. This suggests that the poorer health status of HUD-assisted MMEs increases their likelihood of needing HCBS and/or relationships between the property management and community supports make these services more available to HUD-assisted individuals.

TABLE C16. Comparison of HUD-Assisted MME and Unassisted MME Medicaid Health Care Utilization per 1000 Member Months1
Topic AreaHUD-Assisted MMEsUnassisted MMEsRatio of HUD-Assisted vs. Unassisted
N%N%
NOTE:
  1. Limited to MMEs with Medicaid FFS only. In other words, MMEs with any Medicaid managed care enrollment during 2008 are excluded from this table and the Medicaid utilization and cost comparison.

** p<0.01; * p<0.05

Limited to individuals with Medicaid FFS106,76495%227,18691% 
 MeanSDMeanSD 
Personal care services4,512.410,047.12,149.17,128.22.09**
Residential care38.9683.063.71,164.30.61**
DME380.0946.2227.7746.91.67**
Other HCBS (private duty nursing, adult day care, HH, rehab, targeted case management, transportation, and hospice)3,309.88,321.71,840.66,675.41.79**

On the other hand, unassisted MMEs used 60% more Residential Care services than unassisted MMEs. Residential care settings, including assisted living facilities, have traditionally provided assistance and oversight to persons with physical and mental impairments who cannot live at home alone, but do not require a nursing home level of care. As such, residential care lies on the long-term care continuum between home care and nursing facility care. Since residential care settings generally require licensure by the state, HUD properties generally do not provide Medicaid reimbursed residential care. This higher utilization drove the 32% higher average Medicaid PMPM for HUD-assisted MMEs compared to unassisted MMEs ($1,180 vs. $895). (Data not shown,see Appendix D.)

TABLE C17. Comparison of HUD-Assisted MME and Unassisted MME Medicaid Health Care Utilization, Proportion of MMEs with Any Use
 HUD-Assisted MMEsUnassisted MMEsRatio of HUD-Assistedvs. Unassisted
N%N%
NOTE:
  1. Limited to MMEs with Medicaid FFS only. In other words, MMEs with any Medicaid managed care enrollment during 2008 are excluded from this table and the Medicaid utilization and cost comparison.

** p<0.01; * p<0.05

Personal Care services30,18628%32,88514%0.92**
Residential care1,9312%2,6961%0.72**
DME53,92751%79,59835%0.67**
Other HCBS43,92941%57,81525%0.76**

Sub-group A: Medicare/Medicaid Beneficiaries, No SNF or Nursing Facility Days

We also compared demographic, enrollment, and health care payment and utilization for all Medicare/Medicaid beneficiaries with no SNF or nursing facility days, regardless of age (Sub-group A1 in Table C11 above). This almost doubled the sample size from 361,535 when limited to individuals age 65 years or older to 623,807. The tables providing this comparison are included in Appendix D. In summary, the findings are similar to those highlighted for Sub-group A1 (when limited to individuals 65 years or older) with a couple of exceptions. Most notably, although the HUD-assisted Medicare/Medicaid beneficiaries have much higher Medicare and Medicaid health care service utilization, the Medicaid spending PMPM was actually 6% lower for the unassisted Medicare/Medicaid beneficiaries. This is largely driven by the lower payment for residential services ($47 for HUD-assisted beneficiaries vs. $165 for unassisted beneficiaries).

2. Comparison of HUD-Assisted FFS Medicare/No Medicaid42 Beneficiaries and Unassisted FFS Medicare/No Medicaid Beneficiaries in the Community, Age 65+ in 12 Study Jurisdictions, 2008 (Sub-group C1)

In 2008 in the 12 study jurisdictions, there were 46,806 FFS HUD-assisted Medicare/No Medicaid beneficiaries and 2,318,394 unassisted beneficiaries in the community age 65+. Tables C18-C20 compare HUD-assisted to unassisted Medicare/No Medicaid beneficiaries. Similar to the results for Sub-group A1 above, in this section the tables provide comparisons of demographic characteristics, Medicare enrollment characteristics, chronic conditions, and Medicare health care utilization and payments. Unlike the previous comparison of MMEs, estimates below do not control for income and, therefore, we would expect that the HUD-assisted Medicare/No Medicaid beneficiaries would generally have lower income and greater social determinants of health than their community counterparts.

Table C18 provides a comparison of the demographic and Medicare enrollment characteristics for HUD-assisted Medicare/No Medicaid beneficiaries relative to unassisted beneficiaries. HUD-assisted beneficiaries are much more likely to be female (71% vs. 57%), Black (24% vs. 7%), and 85 years and over (23% vs. 16%) than unassisted beneficiaries. HUD-assisted beneficiaries are also more likely to have received Medicare originally due to disability (15% vs. 6%) and receive Part D LIS (32% vs. 6%) than unassisted beneficiaries.

TABLE C18. HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Beneficiaries, Age 65+, 2008, by Demographic and Coverage Characteristics
 HUD-Assisted Medicare/No Medicaid Beneficiaries(N=46,806)Unassisted Medicare/No Medicaid Beneficiaries(N=2,318,394)
N%N%
Demographic Characteristics
Gender
Male13,42828.7%997,87743.0%
Female33,37871.3%1,320,51757.0%
Missing00.0%00.0%
Race (based on RTI variable)
Non-Hispanic White30,11564.3%2,023,13087.3%
Hispanic4,4609.5%75,0903.2%
Black11,33824.2%153,3396.6%
Asian/Pacific Islander5701.2%45,3542.0%
American Indian/Alaska Native280.1%6790.0%
Other2580.6%18,9940.8%
Unknown370.1%1,8080.1%
Age Group
65 to 698,20717.5%531,47222.9%
70 to 749,34720.0%546,59923.6%
75 to 799,19219.6%468,47820.2%
80 to 849,41520.1%398,83617.2%
85+10,64522.7%373,00916.1%
Medicare Coverage Characteristics
Original reason for entitlement
   Old age and survivor's insurance39,82985.1%2,181,22894.1%
   DIB6,92114.8%134,8565.8%
   ESRD280.1%1,3160.1%
   Disability Insurance and ESRD280.1%9940.0%
Medicare Part D coverage at any point during the year25,78255.1%866,62937.4%
Average months of Part D coverage (among those with Part D coverage)11.75 11.77 
Proportion with Part D LIS
   No Part D LIS (premium-subsidy nor cost sharing)16,27963.1%774,36389.4%
   Part D LIS (Any premium-subsidy or cost sharing8,36332.4%51,3015.9%
   Missing1,1404.4%40,9654.7%

Table C19 shows the proportion of HUD-assisted and unassisted Medicare/No Medicaid beneficiaries with chronic conditions -- by count and type. Unlike the HUD-assisted MMEs age 65 or older (Sub-group A1 above), the HUD-assisted Medicare/No Medicaid beneficiaries of the same age group are more similar to the unassisted Medicare/No Medicaid beneficiaries with regards to health status. Approximately 14% of HUD-assisted Medicare/No Medicaid beneficiaries had zero chronic conditions compared to 13% of unassisted Medicare/No Medicaid beneficiaries. As highlighted in Table C14 above, only 5% of HUD-assisted had zero chronic conditions compared to 10% of unassisted Medicare/Medicaid beneficiaries. Not only were the proportions of the HUD-assisted and unassisted Medicare/No Medicaid beneficiaries with chronic conditions more similar than for the HUD and unassisted Medicare/Medicaid beneficiaries, the prevalence of chronic conditions was not consistently higher for the HUD-assisted Medicare/No Medicaid beneficiaries. In fact, the unassisted Medicare/No Medicaid beneficiaries had a higher proportion of cancer (10% vs. 9%) and ophthalmic (37% vs. 32%).

TABLE C19. HUD-Assisted Medicare/No Medicaid and Unassisted Beneficiaries with Chronic Condition Count and Prevalence, Age 65+, 2008, 12 Study Jurisdictions
CCW ConditionHUD-Assisted Medicare/No Medicaid Beneficiaries(N=46,806)Unassisted Medicare/No Medicaid Beneficiaries(N=2,318,394)
N%N%
Prevalence of Chronic Conditions
0 conditions6,62414.2%299,93612.9%
1 condition3,9728.5%249,96210.8%
2 conditions5,24511.2%312,46213.5%
3 conditions6,47213.8%358,21915.5%
4 conditions6,66914.2%339,58614.6%
5 or more conditions17,82438.1%758,22932.7%
Prevalence of Chronic Conditions by Category
Cardiovascular20,63844.1%952,06141.1%
Cancer4,1058.8%238,41610.3%
Endocrine and renal19,31341.3%800,92434.5%
Alzheimer's-related3,5437.6%179,3047.7%
Depression3,9308.4%167,1197.2%
Musculoskeletal15,49433.1%719,06531.0%
Pulmonary6,77714.5%246,61910.6%
Opthalmic15,09832.3%847,51236.6%
Other (anemia, hyperlipidemia, hyperplasia, hypertension)35,03074.8%1,696,31973.2%

Table C20 provides a comparison of Medicare health care utilization between the HUD-assisted Medicare/No Medicaid and unassisted Medicare/No Medicaid beneficiaries age 65 or older. Similarly to the variability between the higher prevalence of chronic conditions between the two groups, the utilization by HUD-assisted beneficiaries was higher for some services and lower for others when compared to unassisted beneficiaries. In particular, the HUD-assisted beneficiaries used 42% more emergency room visits per 1000 member months than unassisted beneficiaries. The admission rate was 19% higher and the readmission rate was 17% higher for HUD-assisted Medicare/No Medicaid beneficiaries. In addition, 29% of HUD-assisted Medicare/No Medicaid beneficiaries visited the emergency room at least one during 2008 compared to only 23% of unassisted Medicare/No Medicaid beneficiaries. Interestingly, the utilization of office visits was 1% lower for the HUD-assisted beneficiaries and 13% of HUD-assisted beneficiaries did not have any office visits during 2008 compared to 9% of unassisted beneficiaries (data not shown, see Appendix D). The HUD-assisted beneficiaries also had lower utilization of hospice services, ambulatory care surgery, tests, other services, and Part B pharmacy. Overall, the Medicare PMPM was 8% higher for HUD-assisted Medicare/No Medicaid beneficiaries than unassisted Medicare/No Medicaid beneficiaries ($617 vs. $574) (data not shown, see Appendix D).

TABLE C20. Comparison of HUD-Assisted Medicare/No Medicaid Beneficiaries and Unassisted Medicare/No Medicaid Beneficiaries' Medicare Health Care Utilization
Medicare Services Annual Utilization per 1000 Member MonthsHUD-AssistedMedicare Beneficiaries(N=46,806)UnassistedMedicare Beneficiaries(N=2,318,394)Ratio of HUD-Assistedvs. Unassisted
MeanMean
NOTE:
  1. Non-acute inpatient settings include long-term care settings, like inpatient rehabilitation facilities.

** p<0.01; * p<0.05

Total Admissions23.019.31.19**
   Acute stay admissions21.718.21.19**
   Other inpatient admissions11.31.11.20**
Hospital Readmissions2.72.31.17**
Medicare HH visits220.5136.01.62**
Medicare hospice days36.577.60.47**
Total outpatient visits444.8409.51.09**
Total ED visits43.030.41.42**
   ED visits without an admission27.218.81.45**
   ED visits resulting in an admission15.811.61.37**
Physician office visits1,145.21,129.11.01*
ASC visits12.716.00.79**
Dialysis events6.54.81.36**
Anesthesia events25.929.20.89**
Imaging events378.2382.20.99*
Test events1,347.01,416.90.95**
Other procedures527.2635.40.83**
DME180.7139.01.30**
Part B Drugs205.7242.20.85**
Part D Drugs2983.92496.01.20**

Sub-group C: Medicare Only Beneficiaries, Age 65+ Years

We also compared demographic, enrollment, and health care payment and utilization for all Medicare/No Medicaid beneficiaries age 65+ years, regardless of SNF use during 2008. Approximately 8% of HUD-assisted Medicare/No Medicaid beneficiaries utilized SNF services in 2008 compared to 6% of unassisted Medicare/No Medicaid beneficiaries (see Table C11). Despite the slightly different proportion of SNF users in the two groups, we also conducted the comparison between HUD-assisted and unassisted Medicare/No Medicaid beneficiaries including individuals with SNF utilization in 2008. As expected, given the higher proportion of individuals with SNF utilization among the HUD-assisted group, the differences in utilization and payments widened. HUD-assisted Medicare/No Medicaid beneficiariesadmission rate was 29% higher than unassisted beneficiaries and the readmission utilization was 37% higher. Not surprisingly, HUD-assisted Medicare/No Medicaid beneficiaries had 18% higher payments than unassisted Medicare/No Medicaid beneficiaries (difference was only 8% when individuals with SNF utilization were excluded). See Appendix D for result tables.

 

APPENDIX D. SUPPLEMENTAL RESULTS, ALL YEARS

Objective A: Understand Demographic Characteristics of HUD-Assisted Individuals/Households in the 12 Study Jurisdictions

TABLE D-A1. HUD-Assisted Individuals Full vs. Partial Year, by Year(All Ages)
All Ages2007(N=1,976,053)2008(N=2,057,893)2009(N=2,075,168)
N%N%N%
Full year participants1,703,86386.2%1,783,20286.7%1,812,86987.4%
Partial year participants272,19013.8%274,69113.3%262,29912.6%
Reason for Partial Year
Moved Out/Terminated110,76740.7%110,37040.2%114,35243.6%
Moved in after Jan 1161,42359.3%164,32159.8%147,94756.4%
Total272,190 274,691 262,299 
Property Type
Public housing579,60929.3%635,90630.9%628,24730.3%
Housing choice vouchers933,95047.3%960,38746.7%978,71747.2%
Other multifamily381,44219.3%381,07818.5%388,28718.7%
Section 20263,8603.2%63,0333.1%61,2823.0%
Section 811 & 202/1623,9700.2%3,9930.2%4,1250.2%
Other13,2220.7%13,4960.7%14,5100.7%
Subsidy Type
Public housing581,14529.4%638,58431.0%631,41930.4%
Housing choice vouchers939,23147.5%969,14047.1%987,16047.6%
Project-Based Section 8382,63419.4%376,55218.3%378,66218.2%
Section 20221,4991.1%21,3861.0%21,9411.1%
Section 202/1621020.0%1000.0%1000.0%
Section 8113,8920.2%3,8390.2%3,9530.2%
Other47,5502.4%48,2922.3%51,9332.5%

 

TABLE D-A2. HUD-Assisted Individuals Full vs. Partial Year, by Year(Age 62+)
All Ages2007(N=383,576)2008(N=398,791)2009(N=407,428)
N%N%N%
Full year participants335,47887.5%353,78988.7%359,70288.3%
Partial year participants48,09812.5%45,00211.3%47,72611.7%
Reason for Partial Year
Moved Out/Terminated23,23048.3%22,81550.7%26,84656.3%
Moved in after Jan 124,86851.7%22,18749.3%20,88043.7%
Total48,098 45,002 47,726 
Property Type
Public housing113,60129.6%123,55731.0%124,53630.6%
Housing choice vouchers94,70424.7%100,50425.2%106,16826.1%
Other multifamily115,89230.2%115,73329.0%118,76929.2%
Section 20257,16814.9%56,60414.2%55,18213.5%
Section 811 & 202/1624510.1%5170.1%5820.1%
Other1,7600.5%1,8760.5%2,1910.5%
Subsidy Type
Public housing113,77829.7%123,93031.1%124,96130.7%
Housing choice vouchers95,19824.8%101,29225.4%106,87126.2%
Project-Based Section 8141,87337.0%140,80435.3%141,07734.6%
Section 20221,0565.5%20,9695.3%21,5325.3%
Section 202/162630.0%660.0%620.0%
Section 8114510.1%5070.1%5670.1%
Other11,1572.9%11,2232.8%12,3583.0%

 

TABLE D-A3. HUD-Assisted Individuals Full vs. Partial Year, by Year(Age 65+)
All Ages2007(N=340,297)2008(N=352,091)2009(N=357,207)
N%N%N%
Full year participants298,22787.6%313,03588.9%316,00888.5%
Partial year participants42,07012.4%39,05611.1%41,19911.5%
Reason for Partial Year
Moved Out/Terminated21,30650.6%20,87653.5%24,36159.1%
Moved in after Jan 120,76449.4%18,18046.5%16,83840.9%
Total42,070 39,056 41,199 
Property Type
Public housing98,19828.9%106,30630.2%106,50229.8%
Housing choice vouchers80,37623.6%84,63524.0%88,42124.8%
Other multifamily105,21330.9%104,91429.8%107,20830.0%
Section 20254,88516.1%54,56015.5%53,15914.9%
Section 811 & 202/1623310.1%3360.1%3640.1%
Other1,2940.4%1,3400.4%1,5530.4%
Subsidy Type
Public housing98,35228.9%106,63430.3%106,88029.9%
Housing choice vouchers80,79223.7%85,29624.2%89,01524.9%
Project-Based Section 8130,65138.4%129,60836.8%129,30836.2%
Section 20220,3196.0%20,3995.8%20,9325.9%
Section 202/162560.0%580.0%560.0%
Section 8113220.1%3330.1%3620.1%
Other9,8052.9%9,7632.8%10,6543.0%
Number of Household Heads  303,695 311,657 314,014 

 

TABLE D-A4. HUD-Assisted Individuals Across Consecutive Years Full vs. Partial
 2007-20082008-20092007-2009
N%N%N%
All ages2,277,570100.0%2,292,744100.0%2,490,435100.0%
Full year participants1,478,13364.9%1,556,23867.9%1,299,35252.2%
Partial year participants799,43735.1%736,50632.1%1,191,08347.8%
62+454,076100.0%445,035100.0%481,175100.0%
Full year participants324,97471.6%328,15173.7%289,88560.2%
Partial year participants129,10228.4%116,88426.3%191,29039.8%
65+402,714100.0%391,434100.0%424,729100.0%
Full year participants288,87971.7%289,66774.0%256,85160.5%
Partial year participants113,83528.3%101,76726.0%167,87839.5%

 

< class="active"td>2,057,893
TABLE D-A5. Living Arrangement, HUD-Assisted Individuals, All Ages and Age 62+ Years
 200720082009
N%N%N%
Individuals, All Ages1,976,053100.0%100.0%2,075,168100.0%
   Living Arrangement
      Lives Alone437,10122.1%450,41021.9%456,08522.0%
      Lives with Spouse85,4534.3%89,0664.3%90,2914.4%
      Other Adults129,0666.5%144,7057.0%154,8627.5%
      Minor(s) in Household1,299,63665.8%1,345,19065.4%1,341,70064.7%
      Live-in Aide5,8700.3%7,0390.3%7,8840.4%
      Other3270.0%2830.0%2820.0%
      Unknown/declined to report18,6000.9%21,2001.0%24,0641.2%
Individuals, Age 62+383,576100.0%398,791100.0%407,428100.0%
   Living Arrangement
      Lives Alone260,84968.0%267,75167.1%271,56166.7%
      Lives with Spouse65,85117.2%68,69417.2%70,17817.2%
      Other Adults27,3987.1%30,6037.7%32,3517.9%
      Minor(s) in Household22,8836.0%24,1306.1%24,2826.0%
      Live-in Aide1,5530.4%1,8770.5%2,1430.5%
      Other360.0%380.0%360.0%
      Unknown/declined to report5,0061.3%5,6981.4%6,8771.7%
   Living with Minor(s) in the Household
      Age 62+ with minor(s) in the home22,8836.0%24,1306.1%24,2826.0%
      Age 62+ without minor(s) in the home360,69394.0%374,66193.9%383,14694.0%

 

TABLE D-A6. Living Arrangement, HUD-Assisted Households, All Ages
 200720082009
N%N%N%
Household Heads, All Ages939,775100.0%967,557100.0%971,094100.0%
Living arrangement--By Household
   Lives Alone435,14646.3%448,55646.4%454,31346.8%
   Lives with Spouse45,7944.9%45,9854.8%45,1314.6%
   Other Adults63,4086.7%68,7517.1%71,2277.3%
   Children in Household386,51341.1%394,26040.7%388,65340.0%
   Live-in Aide2,6900.3%3,1310.3%3,4770.4%
   Other940.0%780.0%740.0%
   Unknown/declined to report6,1300.7%6,7960.7%8,2190.8%

 

TABLE D-A7a. HUD-Assisted Individual Demographics, 2008
 Unique Individuals Receiving PublicHousing Subsidy During 2008(N=2,057,893)
MeanSD
Age
   Age--All Persons34.8925.32
   Age--Household Heads54.2218.77
 N%
Gender
   Male765,16937.2%
   Female1,291,14062.7%
   Unknown/declined to report1,5840.1%
Age Group
   <18764,54337.2%
   18-44566,05927.5%
   45-61327,06715.9%
   62-6446,7002.3%
   65+352,09117.1%
   Unknown/declined to report1,4330.1%
Race/Ethnicity
   White non-Hispanic331,94416.1%
   Hispanic489,33723.8%
   Black non-Hispanic773,55037.6%
   Asian56,9342.8%
   American Indian/Alaskan3,6990.2%
   Hawaiian/Pacific Islander2,7620.1%
   Other non-Hispanic2590.0%
   Mixed non-Hispanic2750.0%
   Unknown/declined to report399,13319.4%
Property Type
   Public housing635,90630.9%
   Housing choice vouchers960,38746.7%
   Other multifamily381,07818.5%
   Section 20263,0333.1%
   Section 811 & 202/1623,9930.2%
   Other13,4960.7%
Subsidy Type
   Public housing638,58431.0%
   Housing choice vouchers969,14047.1%
   Project-Based Section 8376,55218.3%
   Section 20221,3861.0%
   Section 202/1621000.0%
   Section 8113,8390.2%
   Other48,2922.3%
Disability or Handicapped Status
   Yes376,33918.3%
   No1,472,67971.6%
   Unknown/declined to report208,87510.1%
Geographic Area
   Vermont22,6351.1%
   New Haven- Milford50,2012.4%
   Bridgeport- Stamford- Norwalk35,3261.7%
   Milwaukee- Waukesha- West Allis51,5552.5%
   San Francisco- Oakland- Fremont167,7958.2%
   Boston- Cambridge- Quincy243,60111.8%
   Durham- Chapel Hill17,9450.9%
   Richmond43,1962.1%
   New York- Northern New Jersey- Long Island1,213,20159.0%
   Columbus72,3323.5%
   Akron35,4091.7%
   Cleveland104,6975.1%

 

TABLE D-A7b. HUD-Assisted Individuals Demographics, 2007-2009*
 Unique Individuals Receiving Public HousingSubsidy During 2007, 2008, or 2009*(N=2,490,435)
MeanSD
* Each individual only included once, regardless of appearing in more than one year of data. For those individuals that appeared in multiple years, we defined the demographic variables based on their most recent year variable value. Earlier analyses showed very strong consistency across years (over 99% of individuals had consistent values for a given variable across multiple years). For the small percentage of individuals who were present in all three years, if an individual had a demographic variable the same for two years, but different in a third year, the third year was changed to match the other two.
Age
   Age--All Persons34.9625.34
   Age--Household Heads54.8019.06
 N%
Gender
   Male944,75948.31%
   Female1,544,01978.95%
   Unknown/declined to report1,6570.07%
Age group
   <18888,052192.81%
   18-44735,274159.64%
   45-61384,50083.48%
   62-6456,44612.26%
   65+424,72992.22%
   Unknown/declined to report1,4340.31%
Race/Ethnicity
   White non-Hispanic411,74315.92%
   Hispanic590,80722.84%
   Black non-Hispanic953,14136.85%
   Asian69,6222.69%
   American Indian/Alaskan4,5450.18%
   Hawaiian/Pacific Islander3,6570.14%
   Other non-Hispanic9490.04%
   Mixed non-Hispanic7410.03%
   Unknown/declined to report455,23017.60%
Subsidy Type
   Public housing734,28433.69%
   Housing choice vouchers1,190,53354.63%
   Project-based Section 8468,14521.48%
   Section 20225,9441.19%
   Section 202/1621170.01%
   Section 8115,0090.23%
   Other66,4033.05%
Disability or Handicapped Status
   Yes434,01317.43%
   No1,787,47671.77%
   Unknown/declined to report268,94610.80%
Geographic Area
   Vermont29,40123.08%
   New Haven- Milford60,68747.64%
   Bridgeport- Stamford- Norwalk42,47233.34%
   Milwaukee- Waukesha- West Allis65,88251.71%
   San Francisco- Oakland- Fremont205,888161.61%
   Boston- Cambridge- Quincy291,571228.87%
   Durham- Chapel Hill23,77918.67%
   Richmond57,40645.06%
   New York- Northern New Jersey- Long Island1,437,4151128.29%
   Columbus97,51176.54%
   Akron46,95736.86%
   Cleveland131,466103.19%

 

TABLE D-A8. HUD-Assisted Individual Demographics by Geographic Area
 Vermont(N=29,401)New Haven-Milford(N=60,687)Bridgeport- Stamford-Norwalk(N=42,472)Milwaukee- Waukesha-West Allis(N=65,882)San Francisco-Oakland- Fremont(N=205,888)Boston- Cambridge-Quincy(N=291,571)
MeanSDMeanSDMeanSDMeanSDMeanSDMeanSD
Age37.2126.3933.1325.2834.8425.6833.7125.6837.40225.6737.7226.08
 N%N%N%N%N%N%
Gender
   Male11,59239.4%22,27936.7%15,53636.6%23,49235.7%80,60739.2%110,77538.0%
   Female17,80360.6%38,37663.2%26,91763.4%42,36864.3%125,17160.8%180,58261.9%
   Did not report**320.1%190.0%220.0%1100.1%2140.1%
Age Group
   <189,56932.5%23,94839.5%15,58136.7%26,03039.5%66,55532.3%94,21432.3%
   18-448,50128.9%17,89329.5%12,50329.4%18,05427.4%58,96928.6%81,86428.1%
   45-614,75916.2%8,09613.3%5,78013.6%8,78713.3%35,32417.2%49,72017.1%
   62-647472.5%1,2352.0%9112.1%1,6062.4%4,8512.4%7,2682.5%
   65+5,81819.8%9,48315.6%7,67918.1%11,38417.3%40,03219.4%58,29120.0%
   Unknown/ declined to report**320.1%180.0%210.0%1570.1%2140.1%
Race/Ethnicity
   White non-Hispanic20,25068.9%8,87614.6%4,70011.1%8,64613.1%29,01414.1%83,75028.7%
   Hispanic3301.1%18,29330.1%14,60934.4%2,9284.4%19,9649.7%60,60520.8%
   Black non-Hispanic1,3694.7%20,33233.5%16,24238.2%34,55352.4%86,70542.1%57,78819.8%
   Asian2841.0%940.2%2750.6%1920.3%33,13616.1%9,8733.4%
   American Indian/Alaskan890.3%760.1%580.1%1480.2%8020.4%8000.3%
   Hawaiian/Pacific Islander**160.0%400.1%170.0%1,7370.8%8390.3%
   Other non-Hispanic**200.0%****1180.1%1880.1%
   Mixed non-Hispanic180.1%190.0%**530.1%850.0%1170.0%
   Declined to report7,05224.0%12,96121.4%6,54315.4%19,34129.4%34,32716.7%77,61126.6%
Property Type
   Public housing4,25514.5%10,52117.3%13,45731.7%10,48615.9%30,81715.0%54,14918.6%
   Housing choice vouchers16,90357.5%33,38855.0%20,54748.4%30,78146.7%130,99463.6%140,71448.3%
   Other multifamily6,93723.6%14,77724.3%6,31014.9%21,46832.6%31,27715.2%82,91528.4%
   Section 2029113.1%1,2882.1%1,5093.6%2,6364.0%9,0974.4%10,2923.5%
   Section 811 & 202/162940.3%1290.2%430.1%2140.3%6580.3%8560.3%
   Other3011.0%5841.0%6061.4%2970.5%3,0451.5%2,6450.9%
Subsidy Type
   Public housing4,29414.6%10,60517.5%13,51231.8%10,58116.1%30,92915.0%54,50318.7%
   Housing choice vouchers17,08058.1%33,82855.7%20,66348.7%31,20547.4%131,67664.0%142,07548.7%
   Project-based Section 87,35825.0%14,87924.5%7,11116.7%22,64434.4%35,46717.2%82,96028.5%
   Section 2022690.9%6361.0%3210.8%9231.4%3,5111.7%3,3971.2%
   Section 202/162********140.0%**
   Section 811920.3%1060.2%430.1%2110.3%6480.3%8680.3%
   Other3081.0%6331.0%8221.9%3180.5%3,6431.8%7,7682.7%
Disability or Handicapped Status
   Yes7,55325.7%10,25416.9%6,24814.7%12,63719.2%39,80619.3%61,25921.0%
   No18,70263.6%42,99170.8%34,46881.2%42,71464.8%151,15073.4%187,27564.2%
   Did not report3,14610.7%7,44212.3%1,7564.1%10,53116.0%14,9327.3%43,03714.8%

 

TABLE D-A8 (continued)
 Durham-Chapel Hill(N=23,779)Richmond(N=57,406)New York- NorthernNew Jersey- Long Island(N=1,437,415)Columbus(N=97,511)Akron(N=46,957) Cleveland(N=131,466)
MeanSDMeanSDMeanSDMeanSDMeanSDMeanSD
* Cell sizes of less than 11 are not displayed.
Age27.4622.472622.1735.7425.2627.64823.4128.48823.4230.5224.1
 N%N%N%N%N%N%
Gender
   Male8,54335.9%20,26135.3%546,89738.0%37,78338.7%17,00336.2%49,99138.0%
   Female15,22864.0%37,12564.7%889,44361.9%59,66761.2%29,93863.8%81,40161.9%
   Did not report**200.0%1,0750.1%610.1%160.0%740.1%
Age Group
   <1811,12146.8%28,10649.0%490,85434.1%45,48246.6%20,49143.6%56,10142.7%
   18-447,13630.0%17,55530.6%430,78830.0%29,22630.0%14,71131.3%38,07429.0%
   45-613,01612.7%6,10010.6%226,68215.8%11,04011.3%6,15713.1%19,03914.5%
   62-644381.8%8261.4%33,6772.3%1,5621.6%8791.9%2,4461.9%
   65+2,0608.7%4,7998.4%254,57017.7%10,14310.4%4,70910.0%15,76112.0%
   Unknown/ declined to report**200.0%8440.1%580.1%**450.0%
Race/Ethnicity
   White non-Hispanic1,9028.0%2,6194.6%198,08213.8%20,86121.4%16,14434.4%16,89912.9%
   Hispanic4271.8%5921.0%463,15732.2%8190.8%4701.0%8,6136.6%
   Black non-Hispanic18,49777.8%39,65069.1%527,08736.7%50,94252.2%19,96442.5%80,01260.9%
   Asian500.2%1340.2%25,0491.7%1740.2%1910.4%1700.1%
   American Indian/Alaskan360.2%750.1%2,1200.1%1030.1%410.1%1970.1%
   Hawaiian/Pacific Islander**120.0%8680.1%710.1%150.0%330.0%
   Other non-Hispanic**510.1%3010.0%1930.2%240.1%330.0%
   Mixed non-Hispanic120.1%150.0%2010.0%930.1%400.1%870.1%
   Declined to report2,84212.0%14,25824.8%220,55015.3%24,25524.9%10,06821.4%25,42219.3%
Property Type
   Public housing7,91333.3%15,91327.7%528,71736.8%11,80012.1%12,59826.8%30,31123.1%
   Housing choice vouchers12,01750.5%21,47237.4%636,80744.3%50,88652.2%19,68141.9%65,82650.1%
   Other multifamily3,32714.0%18,15631.6%223,27415.5%28,71829.5%12,53426.7%29,63022.5%
   Section 2024061.7%1,4982.6%37,6072.6%4,8134.9%1,0672.3%4,5643.5%
   Section 811 & 202/1621160.5%1490.3%1,9200.1%5090.5%1720.4%3360.3%
   Other**2180.4%9,0900.6%7850.8%9051.9%7990.6%
Subsidy Type
   Public housing7,99233.6%16,28128.4%529,78836.9%12,10712.4%12,77927.2%30,91323.5%
   Housing choice vouchers12,12651.0%22,08338.5%640,84644.6%52,13753.5%20,07342.7%66,74150.8%
   Project-Based Section 83,39814.3%17,03829.7%204,17714.2%29,58930.3%12,80327.3%30,72123.4%
   Section 2021600.7%3470.6%12,8290.9%1,9372.0%2060.4%1,4081.1%
   Section 202/162230.1%****770.1%****
   Section 811800.3%1460.3%1,8580.1%4740.5%1620.3%3210.2%
   Other**1,5112.6%47,9143.3%1,1901.2%9342.0%1,3621.0%
Disability or Handicapped Status
   Yes4,06817.1%8,18914.3%237,71816.5%14,75515.1%8,03117.1%23,49517.9%
   No17,52873.7%36,82064.1%1,070,41674.5%62,01763.6%30,99366.0%92,40270.3%
   Did not report2,1839.2%12,39721.6%129,2819.0%20,73921.3%7,93316.9%15,56911.8%

 

TABLE D-A9. Demographics by Age Group
 <62(N=2,007,826) 62 to 64(N=56,446)65+(N=424,729)
N%N%N%
* Cell sizes of less than 11 are not displayed.
Gender
Male797,18339.7%17,92031.7%129,63830.5%
Female1,210,53860.3%38,51868.2%294,92369.4%
Declined to report1050.0%**1680.0%
Race/Ethnicity
White non-Hispanic306,96515.3%12,41422.0%92,36021.7%
Hispanic508,70125.3%13,96924.7%68,13516.0%
Black non-Hispanic863,13743.0%15,52227.5%74,45817.5%
Asian50,2682.5%2,3724.2%16,9554.0%
American Indian/Alaskan3,8010.2%1180.2%6250.1%
Hawaiian/Pacific Islander3,0310.2%870.2%5390.1%
Other non-Hispanic7850.0%300.1%1340.0%
Mixed non-Hispanic5390.0%130.0%1890.0%
Declined to report270,59913.5%11,92121.1%171,33440.3%
Property Type
Public housing586,55929.2%19,82335.1%124,50129.3%
Housing choice vouchers1,055,77152.6%20,12235.6%104,12224.5%
Other multifamily337,12016.8%13,10523.2%127,99030.1%
Section 2027,3900.4%2,3404.1%65,69115.5%
Section 811 & 202/1624,4730.2%2620.5%4580.1%
Other16,5130.8%7941.4%1,9670.5%
Subsidy Type
Public housing589,45429.4%19,87335.2%124,90329.4%
Housing choice vouchers1,065,29453.1%20,28335.9%104,95524.7%
Project-Based Section 8296,95214.8%13,38123.7%156,50736.8%
Section 2024510.0%6461.1%24,8185.8%
Section 202/162450.0%**650.0%
Section 8114,3100.2%2460.4%4520.1%
Other51,3202.6%2,0103.6%13,0293.1%
Disability or Handicapped Status
Yes275,55813.7%30,78954.5%127,65030.1%
No1,468,80273.2%22,76940.3%295,86469.7%
Unknown/Declined to report263,46613.1%2,8885.1%1,2150.3%

 

TABLE D-A10. Demographics by Property Type
 Public Housing(N=730,937)Housing ChoiceVouchers(N=1,180,016)Other Multifamily(N=479,323)Section 202(N=75,688)Section 811 &Section 202/162(N=5,196)Other(N=19,275)
N%N%N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Gender
Male282,28938.6%452,07138.3%173,58736.2%23,87131.5%5,19668.4%10,14952.7%
Female448,64761.4%727,94461.7%304,47563.5%51,43168.0%2,39831.6%9,12447.3%
Declined to report****1,2610.3%3860.5%****
Race/Ethnicity
White non-Hispanic80,79611.1%300,96125.5%20,2504.2%4,9946.6%68513.2%4,05721.0%
Hispanic251,64634.4%309,34026.2%22,9194.8%1,7582.3%811.6%5,06326.3%
Black non-Hispanic366,45650.1%525,94844.6%49,47410.3%2,4473.2%3667.0%8,45043.8%
Asian28,4783.9%35,7363.0%3,6090.8%1,4331.9%270.5%3391.8%
American Indian/Alaskan1,4540.2%2,5310.2%4500.1%470.1%**570.3%
Hawaiian/Pacific Islander1,1090.2%2,2970.2%1600.0%260.0%**650.3%
Other non-Hispanic****8570.2%640.1%**220.1%
Mixed non-Hispanic****6400.1%860.1%110.2%**
Declined to report9980.1%3,2030.3%380,96479.5%64,83385.7%4,01477.3%1,2186.3%
Age Group
<18237,84132.5%501,21742.5%144,39930.1%1880.2%2514.8%4,15621.6%
18-44223,88930.6%375,69631.8%126,24826.3%1,9802.6%1,80534.7%5,65629.3%
45-61124,82917.1%178,85815.2%66,47313.9%5,2226.9%2,41746.5%6,70134.8%
62-6419,8232.7%20,1221.7%13,1052.7%2,3403.1%2625.0%7944.1%
65+124,50117.0%104,1228.8%127,99026.7%65,69186.8%4588.8%1,96710.2%
Unknown/declined to report540.0%**1,1080.2%2670.4%****
Disability or Handicapped Status
Yes147,07420.1%199,86116.9%64,60313.5%10,84914.3%4,65789.6%6,96936.2%
No583,86379.9%980,15583.1%149,62731.2%62,96483.2%1933.7%10,67455.4%
Unknown/declined to report****265,09355.3%1,8752.5%3466.7%1,6328.5%

 

TABLE D-A11. Demographics by Subsidy Type
 Public Housing(N=734,284)Housing ChoiceVouchers(N=1,190,533)Other Multifamily(N=468,144)Section 202(N=25,945)Section 811 &Section 202/162(N=5,126)Other(N=66,403)
N%N%N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Gender
Male283,45538.6%455,77638.3%166,84635.6%7,94930.6%2,74253.5%27,99142.2%
Female450,82861.4%734,75661.7%299,74664.0%17,95669.2%2,38046.4%38,35357.8%
Declined to report****1,5530.3%390.2%**590.1%
Race/Ethnicity
White non-Hispanic81,19411.1%302,62125.4%19,6054.2%1,7376.7%60911.9%5,9779.0%
Hispanic252,35434.4%311,72126.2%18,3103.9%7042.7%841.6%7,63411.5%
Black non-Hispanic368,61950.2%532,26244.7%36,4977.8%9093.5%3366.6%14,51821.9%
Asian28,5463.9%35,8313.0%3,8930.8%6242.4%210.4%7071.1%
American Indian/Alaskan1,4590.2%2,5480.2%3970.1%340.1%**1030.2%
Hawaiian/Pacific Islander1,1100.2%2,3170.2%1440.0%110.0%**750.1%
Other non-Hispanic****8010.2%330.1%**1090.2%
Mixed non-Hispanic****6480.1%290.1%110.2%530.1%
Declined to report1,0020.1%3,2330.3%387,85082.8%21,86384.3%4,05579.1%37,22756.1%
Age Group
<18239,35832.6%506,21242.5%124,46026.6%**2514.9%17,76326.8%
18-44224,97330.6%379,16731.8%109,57323.4%540.2%1,75634.3%19,75129.7%
45-61125,12317.0%179,91515.1%62,91913.4%3891.5%2,34845.8%13,80620.8%
62-6419,8732.7%20,2831.7%13,3812.9%6462.5%2534.9%2,0103.0%
65+124,90317.0%104,9558.8%156,50733.4%24,81895.7%51710.1%13,02919.6%
Unknown/declined to report540.0%**1,3050.3%290.1%**440.1%
Disability or Handicapped Status
Yes147,61620.1%201,39716.9%68,93814.7%1,2144.7%4,54188.6%10,30715.5%
No586,66879.9%989,13683.1%162,36634.7%24,26093.5%2284.4%24,81837.4%
Unknown/declined to report****236,84150.6%4701.8%3577.0%31,27847.1%

 

TABLE D-A12. Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+
Proportion of Households with at Least One Individual Age 62+Number of Properties 2007Percent of Properties 2007Number of Properties 2008Percent of Properties 2008Number of Properties 2009Percent of Properties 2009Number of Properties 2007-2009Percent of Properties 2007-2009
0-9%763625.4%706727.8%663861.0%802524.2%
10-19%372304.9%375386.6%372483.1%378247.1%
20-29%362296.7%372383.5%374485.7%380248.4%
30-39%226185.2%221227.8%250324.7%253165.4%
40-49%130106.6%140144.3%137177.9%14091.5%
50-59%10888.5%115118.6%137177.9%14192.2%
60-69%10182.8%103106.2%99128.6%9964.7%
70-79%137112.3%144148.5%139180.5%14091.5%
80-89%262214.8%236243.3%238309.1%239156.2%
90-100%1,213994.3%1,2131250.5%1,2091570.1%1,390908.5%
Elderly (Age 62+) Properties
Properties with 50%+ Households designated as Elderly1,82149.6%1,81150.0%1,82250.4%200950.7%
Properties with under 50% Households designated as Elderly1,85350.4%1,81450.0%1,79649.6%195349.3%

 

TABLE D-A13. Number and Proportion of TRACS Properties by Proportion of Households with at Least One Individual Age 62+ by Geographic Location
Proportion of Households with at Least One Individual Age 62+VermontNew Haven-MilfordBridgeport- Stamford-NorwalkMilwaukee- Waukesha-West AllisSan Francisco-Oakland- FremontBoston- Cambridge-Quincy
N%N%N%N%N%N%
Proportion of Households with at Least One Individual Age 62+ (2007)
0-9%4326.9%2016.7%67.4%4422.1%6417.6%9414.6%
10-19%106.3%1210.0%78.6%147.0%318.5%9314.5%
20-29%106.3%119.2%44.9%136.5%226.0%6510.1%
30-39%53.1%32.5%56.2%63.0%226.0%385.9%
40-49%53.1%43.3%22.5%94.5%82.2%213.3%
50-59%31.9%32.5%33.7%115.5%82.2%274.2%
60-69%74.4%32.5%56.2%2110.6%41.1%243.7%
70-79%159.4%75.8%89.9%189.0%30.8%325.0%
80-89%2817.5%119.2%78.6%199.5%174.7%7511.7%
90-100%3421.3%4638.3%3442.0%4422.1%18550.8%17427.1%
Proportion of Households with at Least One Individual Age 62+ (2008)
0-9%3823.8%2017.1%67.2%4020.7%4913.6%9114.3%
10-19%138.1%1412.0%910.8%147.3%359.7%8813.8%
20-29%138.1%97.7%44.8%136.7%267.2%7311.4%
30-39%21.3%43.4%22.4%73.6%205.6%345.3%
40-49%85.0%43.4%44.8%84.1%92.5%264.1%
50-59%42.5%32.6%33.6%147.3%102.8%253.9%
60-69%116.9%32.6%56.0%157.8%41.1%203.1%
70-79%127.5%97.7%78.4%2211.4%30.8%396.1%
80-89%2314.4%108.5%1012.0%157.8%143.9%6610.3%
90-100%3622.5%4135.0%3339.8%4523.3%18952.6%17627.6%
Proportion of Households with at Least One Individual Age 62+ (2009)
0-9%3321.2%2219.1%67.3%4020.1%4111.7%8413.1%
10-19%127.7%87.0%78.5%157.5%298.3%9414.7%
20-29%127.7%1210.4%44.9%136.5%298.3%7311.4%
30-39%74.5%54.3%33.7%73.5%246.8%365.6%
40-49%42.6%32.6%67.3%94.5%72.0%203.1%
50-59%53.2%43.5%33.7%147.0%133.7%335.2%
60-69%106.4%32.6%67.3%136.5%30.9%152.3%
70-79%127.7%76.1%33.7%2311.6%61.7%446.9%
80-89%2616.7%119.6%1214.6%147.0%133.7%6410.0%
90-100%3522.4%4034.8%3239.0%5125.6%18653.0%17727.7%
Proportion of Households with at Least One Individual Age 62+ (2007-2009)
0-9%3923.6%2418.6%78.2%4521.0%5413.9%9614.3%
10-19%127.3%86.2%78.2%167.5%297.5%9514.2%
20-29%127.3%129.3%44.7%136.1%307.7%7310.9%
30-39%74.2%53.9%33.5%73.3%246.2%365.4%
40-49%42.4%32.3%67.1%94.2%71.8%203.0%
50-59%53.0%53.9%33.5%146.5%133.3%334.9%
60-69%106.1%32.3%67.1%136.1%30.8%152.2%
70-79%127.3%75.4%33.5%2310.7%61.5%446.6%
80-89%2615.8%118.5%1214.1%146.5%133.3%659.7%
90-100%3823.0%5139.5%3440.0%6028.0%21054.0%19328.8%
Elderly (Age 62+) Properties
Properties with 50%+ Households designated as Elderly9155.2%7759.7%5868.2%12457.9%24563.0%35052.2%
Properties with under 50% Households designated as Elderly7444.8%5240.3%2731.8%9042.1%14437.0%32047.8%

 

TABLE D-A13 (continued)
Proportion of Households with at Least One Individual Age 62+Durham-Chapel HillRichmondNew York- NorthernNew Jersey- Long IslandColumbusAkronCleveland
 N%N%N%N%N%N%
NOTE: N represents the number of properties with X% of households with at least one individual age 62+.
Proportion of Households with at Least One Individual Age 62+ (2007)
0-9%3055.6%4449.4%26117.8%6731.5%2940.8%6129.0%
10-19%713.0%89.0%1429.7%2210.3%57.0%2110.0%
20-29%23.7%55.6%20814.1%125.6%45.6%62.9%
30-39%23.7%44.5%1288.7%62.8%34.2%41.9%
40-49%00.0%00.0%664.5%41.9%22.8%94.3%
50-59%00.0%11.1%422.9%62.8%11.4%31.4%
60-69%00.0%22.2%171.2%31.4%57.0%104.8%
70-79%23.7%55.6%231.6%115.2%34.2%104.8%
80-89%47.4%44.5%684.6%104.7%912.7%104.8%
90-100%713.0%1618.0%51535.0%7233.8%1014.1%7636.2%
Proportion of Households with at Least One Individual Age 62+ (2008)
0-9%2955.8%4650.0%23216.1%6932.1%2939.2%5727.8%
10-19%59.6%77.6%1379.5%2813.0%45.4%2110.2%
20-29%23.8%44.3%20414.2%115.1%45.4%94.4%
30-39%00.0%33.3%1349.3%73.3%56.8%31.5%
40-49%00.0%11.1%674.7%41.9%22.7%73.4%
50-59%23.8%11.1%443.1%52.3%00.0%42.0%
60-69%00.0%22.2%241.7%41.9%68.1%94.4%
70-79%35.8%66.5%241.7%83.7%34.1%83.9%
80-89%23.8%22.2%594.1%104.7%1013.5%157.3%
90-100%917.3%2021.7%51235.6%6932.1%1114.9%7235.1%
Proportion of Households with at Least One Individual Age 62+ (2009)
0-9%2850.9%4044.0%21414.9%7032.1%3142.5%5427.0%
10-19%712.7%1213.2%1349.3%2913.3%45.5%2110.5%
20-29%35.5%77.7%19913.8%115.0%34.1%84.0%
30-39%00.0%11.1%14410.0%104.6%68.2%73.5%
40-49%11.8%11.1%795.5%31.4%11.4%31.5%
50-59%11.8%11.1%473.3%52.3%34.1%84.0%
60-69%00.0%44.4%251.7%94.1%34.1%84.0%
70-79%35.5%44.4%171.2%31.4%79.6%105.0%
80-89%35.5%66.6%624.3%94.1%79.6%115.5%
90-100%916.4%1516.5%51736.0%6931.7%811.0%7035.0%
Proportion of Households with at Least One Individual Age 62+ (2007-2009)
0-9%3151.7%4544.6%28317.8%8032.8%3240.0%6628.6%
10-19%711.7%1211.9%1388.7%2911.9%45.0%219.1%
20-29%35.0%76.9%20312.7%124.9%33.8%83.5%
30-39%00.0%22.0%1469.2%104.1%67.5%73.0%
40-49%11.7%11.0%825.1%31.2%11.3%31.3%
50-59%11.7%11.0%503.1%52.0%33.8%83.5%
60-69%00.0%44.0%251.6%93.7%33.8%83.5%
70-79%35.0%44.0%181.1%31.2%78.8%104.3%
80-89%35.0%65.9%623.9%93.7%78.8%114.8%
90-100%1118.3%1918.8%58736.8%8434.4%1417.5%8938.5%
Elderly (Age 62+) Properties
Properties with 50%+ Households designated as Elderly1830.0%3433.7%74246.5%11045.1%3442.5%12654.5%
Properties with under 50% Households designated as Elderly4270.0%6766.3%85253.5%13454.9%4657.5%10545.5%

 

TABLE D-A14. Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+
Proportion of Households with at Least One Individual Age 62+ or In a Property Designated as ElderlyNumber of Properties 2007 Percent of Properties 2007 Number of Properties 2008 Percent of Properties 2008 Number of Properties 2009 Percent of Properties 2009
0-9%22815.8%9111.4%10112.1%
10-19%22115.3%11013.8%11914.3%
20-29%19113.2%10413.0%11613.9%
30-39%15210.5%9912.4%9311.2%
40-49%896.2%587.3%536.4%
50-59%483.3%445.5%435.2%
60-69%251.7%202.5%202.4%
70-79%171.2%232.9%232.8%
80-89%141.0%141.8%131.6%
90-100%46231.9%23429.4%25330.3%
Elderly (Age 62+) Properties
Properties with 50%+ Households designated as Elderly or a Property Designated as Elderly56639.1%33542.0%35242.2%
Properties with under 50% Households designated as Elderly and no Elderly Property Designation88160.9%46258.0%48257.8%

 

TABLE D-A15. Number and Proportion of PIC Properties by Proportion of Households with at Least One Individual Age 62+ or Designated Elderly by Geographic Location
 VermontNew Haven-MilfordBridgeport- Stamford-NorwalkMilwaukee- Waukesha-West AllisSan Francisco-Oakland- FremontBoston- Cambridge-Quincy
N%N%N%N%N%N%
Proportion of Households with at Least One Individual Age 62+ (2007)
0-9%726.9%1926.0%1924.7%1634.8%2315.0%189.4%
10-19%311.5%1419.2%1823.4%817.4%4328.1%2412.5%
20-29%00.0%79.6%1215.6%48.7%2415.7%2613.5%
30-39%00.0%00.0%11.3%12.2%138.5%73.6%
40-49%00.0%11.4%22.6%00.0%42.6%73.6%
50-59%00.0%00.0%22.6%00.0%42.6%10.5%
60-69%00.0%00.0%00.0%00.0%32.0%10.5%
70-79%00.0%00.0%00.0%00.0%21.3%10.5%
80-89%00.0%00.0%00.0%00.0%21.3%10.5%
90-100%1661.5%3243.8%2329.9%1737.0%3522.9%10655.2%
Proportion of Households with at Least One Individual Age 62+ (2008)
0-9%19.1%527.8%412.9%829.6%46.5%107.0%
10-19%00.0%15.6%929.0%311.1%2337.1%128.4%
20-29%00.0%15.6%516.1%00.0%1117.7%2316.1%
30-39%00.0%15.6%26.5%00.0%11.6%64.2%
40-49%00.0%00.0%00.0%00.0%23.2%32.1%
50-59%327.3%15.6%13.2%00.0%23.2%32.1%
60-69%19.1%00.0%00.0%00.0%34.8%21.4%
70-79%19.1%00.0%00.0%00.0%23.2%64.2%
80-89%00.0%00.0%00.0%00.0%23.2%32.1%
90-100%545.5%950.0%1032.3%1659.3%1219.4%7552.4%
Proportion of Households with at Least One Individual Age 62+ (2009)
0-9%19.1%1021.7%619.4%725.0%68.0%85.7%
10-19%00.0%715.2%722.6%517.9%2330.7%1611.3%
20-29%00.0%48.7%516.1%00.0%1216.0%2014.2%
30-39%00.0%12.2%13.2%00.0%11.3%53.5%
40-49%00.0%12.2%13.2%00.0%34.0%21.4%
50-59%327.3%12.2%13.2%00.0%34.0%42.8%
60-69%218.2%00.0%00.0%00.0%34.0%21.4%
70-79%00.0%00.0%00.0%00.0%34.0%53.5%
80-89%00.0%00.0%00.0%00.0%11.3%42.8%
90-100%545.5%2247.8%1032.3%1657.1%2026.7%7553.2%
Elderly (Age 62+) Properties
Properties with 50%+ Households Designated as Elderly1090.9%1055.6%1135.5%1659.3%2133.9%8962.2%
Properties with under 50% Households Designated as Elderly19.1%844.4%2064.5%1140.7%4166.1%5437.8%

 

TABLE D-A15 (continued)
 Durham-Chapel HillRichmondNew York- NorthernNew Jersey- Long IslandColumbusAkronCleveland
N%N%N%N%N%N%
NOTE: N represents the number of properties with X% of households with at least one individual age 62+.
Proportion of Households with at Least One Individual Age 62+ (2007)
0-9%1344.8%1029.4%264.0%2054.1%1743.6%4044.0%
10-19%413.8%1132.4%6910.6%616.2%820.5%1314.3%
20-29%413.8%38.8%10516.2%25.4%00.0%44.4%
30-39%26.9%00.0%11718.0%00.0%25.1%99.9%
40-49%13.4%12.9%7110.9%00.0%00.0%22.2%
50-59%00.0%38.8%284.3%00.0%820.5%22.2%
60-69%00.0%12.9%182.8%00.0%25.1%00.0%
70-79%00.0%00.0%91.4%25.4%12.6%22.2%
80-89%00.0%38.8%60.9%00.0%12.6%11.1%
90-100%517.2%25.9%20130.9%718.9%00.0%1819.8%
Proportion of Households with at Least One Individual Age 62+ (2008)
0-9%842.1%635.3%112.8%1343.3%1348.1%830.8%
10-19%421.1%529.4%4110.6%723.3%27.4%311.5%
20-29%15.3%15.9%5815.0%13.3%13.7%27.7%
30-39%210.5%00.0%7820.2%00.0%13.7%830.8%
40-49%00.0%00.0%5213.5%00.0%13.7%00.0%
50-59%00.0%15.9%246.2%00.0%725.9%27.7%
60-69%00.0%00.0%123.1%13.3%13.7%00.0%
70-79%00.0%00.0%112.8%00.0%13.7%27.7%
80-89%00.0%317.6%61.6%00.0%00.0%00.0%
90-100%421.1%15.9%9324.1%826.7%00.0%13.8%
Proportion of Households with at Least One Individual Age 62+ (2009)
0-9%945.0%738.9%123.1%1344.8%1346.4%934.6%
10-19%420.0%527.8%3910.2%724.1%414.3%27.7%
20-29%15.0%15.6%6817.8%13.4%13.6%311.5%
30-39%210.0%00.0%7519.7%00.0%13.6%726.9%
40-49%00.0%15.6%4411.5%00.0%13.6%00.0%
50-59%00.0%00.0%236.0%00.0%621.4%27.7%
60-69%00.0%00.0%112.9%13.4%13.6%00.0%
70-79%00.0%00.0%123.1%00.0%13.6%27.7%
80-89%00.0%316.7%51.3%00.0%00.0%00.0%
90-100%420.0%15.6%9224.1%724.1%00.0%13.8%
Elderly (Age 62+) Properties
Properties with 50%+ Households Designated as Elderly421.1%529.4%14637.8%930.0%933.3%519.2%
Properties with under 50% Households Designated as Elderly1578.9%1270.6%24062.2%2170.0%1866.7%2180.8%

Objective B: Calculate Number of HUD-Assisted Individual Medicare and/or Medicaid Match Rate in the 12 Study Jurisdictions, 2008

  TABLE D-B1. Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Individuals, 2008, by Age Group  
2008 HUD-Assisted IndividualsAge <65 Years(N=1,663,348)Age 65+ Years(N=349,208)
N%N%
Medicare
SSN Mate Rate (SSN only)116,8037.0%325,58993.2%
Restrictive Match Rate (SSN, gender, date of birth)110,3956.7%297,22785.1%
Medicaid
SSN Mate Rate (SSN only)1,200,98172.2%229,09365.6%
Restrictive Match Rate (SSN, gender, date of birth)1,131,13768.5%209,67960.0%
Medicare & Medicaid
SSN Mate Rate (SSN only)96,5095.8%220,13963.0%
Restrictive Match Rate (SSN, gender, date of birth)91,5645.5%201,75357.8%

 

  TABLE D-B2. Medicare and Medicaid SSN and Restrictive Match Rates for HUD-Assisted Household Heads Receiving SSI, 2008  
2008 HUD-Assisted Household Heads with SSIAge <65 Years(N=158,323)Age 65+ Years(N=131,335)
N%N%
Medicare
SSN Mate Rate (SSN only)49,36731.2%123,75994.2%
Restrictive Match Rate (SSN, gender, date of birth)46,87529.6%112,78385.9%
Medicaid
SSN Mate Rate (SSN only)143,44390.6%124,73095.0%
Restrictive Match Rate (SSN, gender, date of birth)134,32084.8%113,71186.6%
Medicare & Medicaid
SSN Mate Rate (SSN only)46,71729.5%120,81192.0%
Restrictive Match Rate (SSN, gender, date of birth)44,38928.0%110,25483.9%

 

TABLE D-B3. Restrictive Match Rate: All Ages
 HUD (any year)(N=2,449,591)HUD 2007(N=1,945,719)HUD 2008(N=2,025,126)HUD 2009(N=2,041,413)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)1,793,28373.2%1,444,00674.2%1,526,08775.4%1,537,94975.3%
Any Medicare enrollment (07, 08, or 09)485,55019.8%419,10121.5%426,23921.0%420,06120.6%
   Medicare 2007 enrollment445,32918.2%389,00820.0%391,97419.4%382,87318.8%
   Medicare 2008 enrollment453,05218.5%391,78320.1%407,62220.1%401,44319.7%
   Medicare 2009 enrollment455,56918.6%389,88120.0%410,85120.3%417,41620.4%
Any Medicaid enrollment (07 or 08)1,661,58467.8%1,333,80968.6%1,412,23469.7%1,423,39569.7%
   Medicaid 2007 enrollment1,559,05763.6%1,277,89965.7%1,338,72666.1%1,334,37765.4%
   Medicaid 2008 enrollment1,559,27763.7%1,245,50764.0%1,340,81666.2%1,357,37066.5%
Both Medicare & Medicaid 2007 enrollment313,63312.8%278,16814.3%277,73213.7%270,08813.2%
Both Medicare & Medicaid 2008 enrollment324,19213.2%283,56814.6%293,31714.5%286,82914.1%

 

TABLE D-B4. Restrictive Match Rate: <62 Years of Age
 HUD (any year)(N=1,970,548)HUD 2007(N=1,563,245)HUD 2008(N=1,627,667)HUD 2009(N=1,635,510)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)1,392,19470.7%1,115,78871.4%1,185,20572.8%1,192,52272.9%
Any Medicare enrollment (07, 08, or 09)115,6895.9%104,5736.7%104,7916.4%101,1106.2%
   Medicare 2007 enrollment99,0055.0%91,6265.9%90,4635.6%86,0915.3%
   Medicare 2008 enrollment106,0315.4%96,7916.2%97,3366.0%93,3635.7%
   Medicare 2009 enrollment112,4465.7%101,0856.5%102,7846.3%100,4246.1%
Any Medicaid enrollment (07 or 08)1,375,48469.8%1,100,71170.4%1,170,47471.9%1,178,33572.0%
   Medicaid 2007 enrollment1,286,97465.3%1,054,41267.5%1,108,34268.1%1,100,60967.3%
   Medicaid 2008 enrollment1,292,35865.6%1,028,38065.8%1,108,86068.1%1,121,57268.6%
Both Medicare & Medicaid 2007 enrollment82,8784.2%77,0004.9%76,1964.7%72,4664.4%
Both Medicare & Medicaid 2008 enrollment88,7264.5%80,8965.2%81,8995.0%78,7424.8%

 

TABLE D-B5. Restrictive Match Rate: Age 62 to 64
 HUD (any year)(N=56,057)HUD 2007(N=42,974)HUD 2008(N=46,375)HUD 2009(N=49,885)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)35,27962.9%33,25377.4%32,65970.4%31,64863.4%
Any Medicare enrollment (07, 08, or 09)16,06228.7%26,72362.2%20,87745.0%14,21728.5%
   Medicare 2007 enrollment14,49725.9%12,17828.3%12,43926.8%12,74225.5%
   Medicare 2008 enrollment15,03026.8%19,38445.1%13,05928.2%13,49327.0%
   Medicare 2009 enrollment15,37927.4%25,96360.4%20,51244.2%14,15728.4%
Any Medicaid enrollment (07 or 08)31,44056.1%24,80257.7%26,02356.1%28,28456.7%
   Medicaid 2007 enrollment29,54152.7%22,83853.1%24,36852.5%26,55453.2%
   Medicaid 2008 enrollment26,99148.1%21,55850.2%22,27748.0%24,58049.3%
Both Medicare & Medicaid 2007 enrollment10,74319.2%9,02521.0%9,23119.9%9,47119.0%
Both Medicare & Medicaid 2008 enrollment11,08119.8%13,85332.2%9,66520.8%9,98120.0%

 

TABLE D-B6. Restrictive Match Rate: Age 65+
 HUD (any year)(N=421,559)HUD 2007(N=338,128)HUD 2008(N=349,658)HUD 2009(N=354,609)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)365,81086.8%294,96487.2%308,22388.1%313,77988.5%
Any Medicare enrollment (07, 08, or 09)353,79983.9%287,80585.1%300,57186.0%304,73485.9%
   Medicare 2007 enrollment331,82778.7%285,20484.3%289,07282.7%284,04080.1%
   Medicare 2008 enrollment331,99178.8%275,60881.5%297,22785.0%294,58783.1%
   Medicare 2009 enrollment327,74477.7%262,83377.7%287,55582.2%302,83585.4%
Any Medicaid enrollment (07 or 08)254,66060.4%208,29561.6%215,73761.7%216,77661.1%
   Medicaid 2007 enrollment242,54257.5%200,64859.3%206,01658.9%207,21458.4%
   Medicaid 2008 enrollment239,92856.9%195,56857.8%209,67960.0%211,21859.6%
Both Medicare & Medicaid 2007 enrollment220,01252.2%192,14356.8%192,30555.0%188,15153.1%
Both Medicare & Medicaid 2008 enrollment224,38553.2%188,81955.8%201,75357.7%198,10655.9%

 

TABLE D-B7. Restrictive Match Rate: Heads of Household, All Ages
 HUD (any year)(N=1,122,000)HUD 2007(N=931,294)HUD 2008(N=958,115)HUD 2009(N=960,903)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)843,18775.2%705,02075.7%732,78676.5%737,49176.7%
Any Medicare enrollment (07, 08, or 09)428,19138.2%372,19640.0%377,05439.4%370,68138.6%
   Medicare 2007 enrollment395,17435.2%347,42037.3%348,81936.4%339,78535.4%
   Medicare 2008 enrollment399,71035.6%347,89237.4%361,45937.7%355,22737.0%
   Medicare 2009 enrollment400,08735.7%344,75937.0%362,23537.8%368,19438.3%
Any Medicaid enrollment (07 or 08)724,67964.6%605,22265.0%629,98665.8%634,24466.0%
   Medicaid 2007 enrollment684,56961.0%576,77261.9%597,43162.4%599,53862.4%
   Medicaid 2008 enrollment676,40660.3%563,14960.5%598,07662.4%603,93962.9%
Both Medicare & Medicaid 2007 enrollment275,78424.6%246,32326.4%244,85325.6%237,37824.7%
Both Medicare & Medicaid 2008 enrollment283,50125.3%249,76526.8%257,81126.9%251,43126.2%

 

TABLE D-B8. Restrictive Match Rate: Heads of Household, <62 Years of Age
 HUD (any year)(N=702,852)HUD 2007(N=592,293)HUD 2008(N=608,852)HUD 2009(N=606,692)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)489,56769.7%412,64569.7%431,27470.8%433,67771.5%
Any Medicare enrollment (07, 08, or 09)99,81114.2%91,10215.4%91,21915.0%88,16614.5%
   Medicare 2007 enrollment86,21012.3%80,52713.6%79,41013.0%75,61412.5%
   Medicare 2008 enrollment91,80113.1%84,55614.3%85,06914.0%81,72713.5%
   Medicare 2009 enrollment96,89013.8%87,93114.8%89,35514.7%87,56714.4%
Any Medicaid enrollment (07 or 08)475,21367.6%399,55667.5%418,50368.7%421,34769.4%
   Medicaid 2007 enrollment446,99663.6%379,54464.1%395,76265.0%396,12465.3%
   Medicaid 2008 enrollment444,39963.2%372,22362.8%395,25164.9%399,08765.8%
Both Medicare & Medicaid 2007 enrollment72,15510.3%67,68111.4%66,86111.0%63,60310.5%
Both Medicare & Medicaid 2008 enrollment76,83710.9%70,70011.9%71,58211.8%68,92211.4%

 

TABLE D-B9. Restrictive Match Rate: Heads of Household, Age 62 to 64
 HUD (any year)(N=47,374)HUD 2007(N=37,145)HUD 2008(N=39,593)HUD 2009(N=42,274)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)30,62964.7%29,09478.3%28,44871.9%27,53065.1%
Any Medicare enrollment (07, 08, or 09)14,65430.9%23,47363.2%18,53646.8%13,00630.8%
   Medicare 2007 enrollment13,27328.0%11,23330.2%11,44428.9%11,70227.7%
   Medicare 2008 enrollment13,72429.0%17,26746.5%11,99130.3%12,37129.3%
   Medicare 2009 enrollment14,01429.6%22,76761.3%18,18045.9%12,94730.6%
Any Medicaid enrollment (07 or 08)27,25257.5%21,86758.9%22,76757.5%24,56558.1%
   Medicaid 2007 enrollment25,72854.3%20,27654.6%21,42554.1%23,16554.8%
   Medicaid 2008 enrollment23,44649.5%19,04951.3%19,59149.5%21,43350.7%
Both Medicare & Medicaid 2007 enrollment9,95421.0%8,42222.7%8,58421.7%8,80120.8%
Both Medicare & Medicaid 2008 enrollment10,23521.6%12,44733.5%8,97422.7%9,25821.9%

 

TABLE D-B10. Restrictive Match Rate: Heads of Household, Age 65+
 HUD (any year)(N=371,728)HUD 2007(N=301,856)HUD 2008(N=309,626)HUD 2009(N=311,891)
N%N%N%N%
NOTE: Shaded cells are of most interest for the column.
Any Medicare or Medicaid enrollment (07, 08, or 09)322,99186.9%263,28187.2%273,06488.2%276,28488.6%
Any Medicare enrollment (07, 08, or 09)313,72684.4%257,62185.3%267,29986.3%269,50986.4%
   Medicare 2007 enrollment295,69179.5%255,66084.7%257,96583.3%252,46980.9%
   Medicare 2008 enrollment294,18579.1%246,06981.5%264,39985.4%261,12983.7%
   Medicare 2009 enrollment289,18377.8%234,06177.5%254,70082.3%267,68085.8%
Any Medicaid enrollment (07 or 08)222,21459.8%183,79960.9%188,71660.9%188,33260.4%
   Medicaid 2007 enrollment211,84557.0%176,95258.6%180,24458.2%180,24957.8%
   Medicaid 2008 enrollment208,56156.1%171,87756.9%183,23459.2%183,41958.8%
Both Medicare & Medicaid 2007 enrollment193,67552.1%170,22056.4%169,40854.7%164,97452.9%
Both Medicare & Medicaid 2008 enrollment196,42952.8%166,61855.2%177,25557.2%173,25155.5%

 

TABLE D-B11. Restrictive Match Rate: Heads of Household, Receiving SSI
 HUD (any year)(N=289,907)HUD 2007(N=141,073)HUD 2008(N=17,385)HUD 2009(N=131,434)
N%N%N%N%
Any Medicare or Medicaid enrollment (07, 08, or 09)258,43189.1%126,54589.7%14,63984.2%117,24789.2%
Any Medicare enrollment (07, 08, or 09)166,40557.4%44,71731.7%7,65544.0%114,03386.8%
   Medicare 2007 enrollment153,06452.8%39,06227.7%4,49425.8%109,50883.3%
   Medicare 2008 enrollment159,65855.1%42,14329.9%4,73227.2%112,78385.8%
   Medicare 2009 enrollment161,12555.6%44,01531.2%7,53943.4%109,57183.4%
Any Medicaid enrollment (07 or 08)253,17887.3%124,43988.2%13,82379.5%114,91687.4%
   Medicaid 2007 enrollment249,69186.1%122,38486.8%13,41877.2%113,88986.7%
   Medicaid 2008 enrollment248,03185.6%122,42586.8%11,89568.4%113,71186.5%
Both Medicare & Medicaid 2007 enrollment147,98451.0%36,98526.2%4,21524.2%106,78481.2%
Both Medicare & Medicaid 2008 enrollment154,64353.3%39,94428.3%4,44525.6%110,25483.9%

 

TABLE D-B12. Comparison of HUD-Assisted Individuals Who Linked to Medicare or Medicaid by Age Group, 2008
 HUD-Assisted Individuals (<65 years)(N=1,970,702)HUD-Assisted Individuals (>65 years)(N=56,060)Restrictive Match Rate by Characteristic
Linked to 2008 Medicaid Enrollment Files(N=1,121,137)Did Not Link to 2008 Medicaid Enrollment Files(N=542,905)p-ValueLinked to 2008 Medicaid Enrollment Files(N=305,153)Did Not Link to 2008 Medicaid Enrollment Files(N=44,505)p-Value
MeanSDMeanSDMeanSDMeanSD
NOTE: TOTAL N lower than Table 1 N due to individuals with unknown age.* Cell sizes of less than 11 are not displayed.
Age--All Persons24.5617.9030.4218.39<0.000176.277.6976.868.28<0.0001 
Age--Household Heads42.6111.8145.4911.50<0.000176.527.7777.178.34<0.0001 
 N%N% N%N%  
Gender
   Male429,97238.0%216,91840.0%<0.000191,24129.9%13,41630.1%<0.000187.2%
   Female701,16562.0%325,90360.0% 213,91270.1%30,96569.6% 87.4%
   Unknown/declined to report**840.0% **1240.3%  
Age Group
   0-18531,04246.9%168,35631.0%<0.0001      
   18-44383,52233.9%220,23140.6%       
   45-61194,29617.2%130,22024.0%       
   62-6422,2772.0%24,0984.4%       
   65-69     70,52923.1%10,46623.5%<0.000187.1%
   70-74     70,83823.2%9,33221.0% 88.4%
   75-79     62,31920.4%8,34118.7% 88.2%
   80-84     51,79817.0%7,60017.1% 87.2%
   85+     49,66916.3%8,76619.7% 85.0%
Race/Ethnicity
   White non-Hispanic196,34417.4%59,77311.0%<0.000165,94221.6%6,46214.5%<0.000191.1%
   Hispanic293,36725.9%131,00724.1% 46,76015.3%8,52819.2% 84.6%
   Black non-Hispanic456,22640.3%241,90744.6% 48,78516.0%10,72924.1% 82.0%
   Asian29,1992.6%14,1992.6% 11,3733.7%1,2502.8% 90.1%
   American Indian/Alaskan2,2220.2%9380.2% 4280.1%490.1% 89.7%
   Hawaiian/Pacific Islander1,4920.1%7900.1% 3510.1%390.1% 90.0%
   Other non-Hispanic1460.0%660.0% 330.0%** 82.5%
   Mixed non-Hispanic1320.0%560.0% 730.0%** 90.1%
   Unknown/declined to report152,00913.4%94,16917.3% 131,40843.1%17,43339.2% 88.3%
Property Type
   Public housing314,20127.8%206,23238.0%<0.000189,01329.2%16,70937.5%<0.000184.2%
   Housing choice vouchers630,61055.8%226,88441.8% 74,79624.5%9,00520.2% 89.3%
   Other multifamily170,34015.1%102,26418.8% 91,46930.0%12,83628.8% 87.7%
   Section 2025,3300.5%2,8170.5% 48,42415.9%5,75012.9% 89.4%
   Section 811 & 202/1622,7360.2%8930.2% 2980.1%370.1% 89.0%
   Other7,9200.7%3,8150.7% 1,1530.4%1680.4% 87.3%
Subsidy Type
   Public housing315,98527.9%206,87438.1%<0.000189,29029.3%16,75337.6%<0.000184.2%
   Housing choice vouchers636,52056.3%228,80042.1% 75,35124.7%9,09420.4% 89.2%
   Project-Based Section 8155,41013.7%88,14316.2% 113,58237.2%15,25934.3% 88.2%
   Section 2024670.0%4850.1% 18,2586.0%1,9964.5% 90.1%
   Section 202/162310.0%110.0% 510.0%** 87.9%
   Section 8112,6270.2%8550.2% 2940.1%370.1% 88.8%
   Other20,0971.8%17,7373.3% 8,3272.7%1,3593.1% 86.0%
Elderly Designated Property
   Yes59,7885.3%40,5237.5%<0.0001160,39852.6%20,10245.2%<0.000188.9%
   No438,54138.8%274,45750.6% 69,37722.7%15,32834.4% 81.9%
   Unknown632,80855.9%227,92542.0% 75,37824.7%9,07520.4% 89.3%
Disability or Handicapped Status
   Yes213,84318.9%53,5579.9%<0.000190,97429.8%13,51130.4%<0.000187.1%
   No793,73970.2%407,93875.1% 213,60370.0%30,83369.3% 87.4%
   Unknown/declined to report123,55510.9%81,41015.0% 5760.2%1610.4% 78.2%
Living Arrangement
   Lives Alone123,75910.9%82,82915.3%<0.0001209,45068.6%29,61266.5%<0.000187.6%
   Lives with Spouse12,8901.1%13,2672.4% 55,41918.2%6,62614.9% 89.3%
   Other Adults58,8165.2%59,54511.0% 20,7136.8%4,2779.6% 82.9%
   Children in Household925,45581.8%375,23469.1% 14,5374.8%3,0826.9% 82.5%
   Live-in Aide2,6550.2%2,7790.5% 1,3160.4%2240.5% 85.5%
   Other1050.0%1480.0% 190.0%** 73.1%
   Unknown/declined to report7,4570.7%9,1031.7% 3,6991.2%6771.5% 84.5%
SSI
   Receives SSI198,45717.5%34,1406.3% 134,48444.1%16,94438.1%  
Geographic Area
   Vermont14,9561.3%2,9100.5%<0.00014,3501.4%2700.6%<0.000194.2%
   New Haven- Milford31,4512.8%10,3961.9% 7,0162.3%7671.7% 90.1%
   Bridgeport- Stamford- Norwalk19,8311.8%8,6511.6% 5,5381.8%7161.6% 88.6%
   Milwaukee- Waukesha- West Allis34,0943.0%7,7871.4% 8,3732.7%7061.6% 92.2%
   San Francisco- Oakland- Fremont82,5867.3%48,0478.8% 29,5429.7%3,1737.1% 90.3%
   Boston- Cambridge- Quincy147,75013.1%46,3868.5% 43,12414.1%4,79910.8% 90.0%
   Durham- Chapel Hill11,8771.1%4,2710.8% 1,3800.5%2060.5% 87.0%
   Richmond11,0051.0%28,1145.2% 1,1850.4%2,6145.9% 31.2%
   New York- Northern New Jersey- Long Island628,30455.5%349,78064.4% 182,50759.8%29,14165.5% 86.2%
   Columbus51,1594.5%12,4172.3% 7,1072.3%7391.7% 90.6%
   Akron25,5982.3%5,5361.0% 3,4671.1%2600.6% 93.0%
   Cleveland72,5266.4%18,6103.4% 11,5643.8%1,1142.5% 91.2%

Objective C: Estimate HUD-Assisted Medicare Beneficiaries with Enrollment in Part D Subsidy Assistance Programs of Medical Savings Program

TABLE D-C1. Health Insurance Characteristics of Those That Link by Year by Age
 All Ages<62 Years of Age
2007 Linked to Medicare or Medicaid 20072008 Linked to Medicare or Medicaid 20082009 Linked to Medicare 20092007 Linked to Medicare or Medicaid 2007>2008 Linked to Medicare or Medicaid 20082009 Linked to Medicare 2009
N%N%N%N%N%N%
Individuals Linked to Medicare389,008 407,622 417,416 91,626 97,336 100,424 
Proportion with Part A coverage377,19097.0%394,90496.9%407,19997.55%91,614100.0%97,326100.0%100,41099.99%
Average months of Medicare Part A coverage (among those with Part A cov)11.56 11.59 11.59 11.55 11.54 11.54 
Proportion with Part B coverage379,66297.6%398,48097.8%408,53597.87%88,18896.2%94,19096.8%97,41897.01%
Average months of Medicare Part B coverage (among those with Part B cov)11.56 11.57 11.58 11.45 11.45 11.46 
Proportion enrolled in Medicare Part A & B during entire year338,86187.1%357,08287.6%367,88188.13%80,03387.3%85,49387.8%88,59188.22%
Proportion in Medicare managed care for at least one month86,72222.3%108,49926.6%115,23127.61%11,64312.7%16,66817.1%18,97418.89%
Average months of Medicare HMO coverage (among those with HMO cov)10.28 10.31 10.78 9.68 9.61 10.30 
Original Reason for Entitlement
Old age & survivor's insurance240,77761.9%249,54761.2%253,11260.64%420.0%340.0%**
Disability insurance benefits144,76137.2%154,45237.9%160,59838.47%88,94197.1%94,57597.2%97,64897.24%
ESRD1,3930.4%1,4350.4%1,4450.35%1,0101.1%1,0291.1%1,0221.02%
Disability Insurance & ESRD2,0770.5%2,1880.5%2,2610.54%1,6331.8%1,6981.7%1,7441.74%
Current Reason for Entitlement
Old age & survivor's insurance282,56972.6%293,23871.9%297,45171.26%550.1%440.0%120.01%
Disability insurance benefits103,21126.5%111,05227.2%116,61827.94%89,19497.3%94,87197.5%97,98797.57%
ESRD1,5890.4%1,6270.4%1,6320.39%9571.0%9781.0%9680.96%
Disability Insurance & ESRD1,6390.4%1,7050.4%1,7150.41%1,4201.5%1,4431.5%1,4571.45%
Medicare Coverage Status
   Part A only7,9462.0%7,6861.9%7,4221.78%2,9173.2%2,6592.7%2,4882.48%
   Part B only10,3762.7%11,1782.7%8,3131.99%****110.01%
   Both Part A & B or combination of coverage370,68695.3%388,75895.4%401,68196.23%88,69996.8%94,66897.3%97,92597.51%
Medicare Part D coverage at any point during the year344,94588.7%365,81289.7%376,86490.28%84,77492.5%91,00793.5%94,31493.92%
Average months of Part D coverage (among those with Part D cov)11.43 11.50 11.520.00%11.40 11.41 11.43 
Proportion with Part D Coverage by Cost Share Group Code
   Beneficiary deemed with 100% premium-subsidy & no copayment11,1483.2%10,7902.9%11,5503.1%2,4472.9%2,3112.5%2,3842.5%
   Beneficiary deemed with 100% premium-subsidy & low copayment221,01064.1%232,15963.5%244,17664.8%60,32871.2%64,31070.7%67,36771.4%
   Beneficiary deemed with 100% premium-subsidy & high copayment34,48810.0%36,76810.1%37,54110.0%10,12511.9%11,00912.1%11,09511.8%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy14,3874.2%15,3684.2%15,6384.1%2,9663.5%3,3863.7%3,5823.8%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy & 15% copayment  1,2130.4%1,1050.3%1,0440.3%750.1%830.1%780.1%
   Beneficiary with LIS, 75% premium-subsidy & 15% copayment1,9090.6%1,9840.5%2,0740.6%2770.3%3330.4%3500.4%
   Beneficiary with LIS, 50% premium-subsidy & 15% copayment1,8100.5%1,8660.5%1,8510.5%2990.4%3030.3%3010.3%
   Beneficiary with LIS, 25% premium-subsidy & 15% copayment1,7980.5%1,4970.4%1,5710.4%2800.3%2120.2%2210.2%
   No premium subsidy nor cost sharing42,05012.2%45,76112.5%48,54712.9%3,7344.4%4,0414.4%4,3224.6%
   Missing15,1324.4%18,5145.1%12,8723.4%4,2435.0%5,0195.5%4,6144.9%
Individuals Linked to Medicaid1,277,899 1,340,816   1,054,412 1,108,860   
Average months of Medicaid coverage (FFS or Managed care)10.92 10.97   10.86 10.92   
Proportion enrolled in Medicaid during entire year (FFS or Managed care)1,010,82179.1%1,071,61979.9%  817,09077.5%868,21878.3%  
In Medicaid managed care for at least one month218,95917.1%247,62018.5%  106,38210.1%131,01911.8%  
Average months of Medicaid managed care (among those with managed care)8.43 8.34   8.47 8.33   
Medicaid Basis for Eligibility
Aged155,63312.2%162,21212.1%  690.0%570.0%  
Blind/disabled280,80822.0%291,81221.8%  216,45820.5%225,46620.3%  
Child525,09741.1%546,17840.7%  525,07249.8%546,16749.3%  
Adult299,41423.4%324,30424.2%  296,15928.1%320,99328.9%  
Child of unemployed adult9790.1%7500.1%  9790.1%7500.1%  
Unemployed adult8150.1%6320.0%  8120.1%6310.1%  
Foster care child6,7780.5%6,4610.5%  6,7780.6%6,4610.6%  
Covered under Breast & Cervical Cancer Prevention Act750.0%970.0%  610.0%810.0%  
Unknown7830.1%3360.0%  5380.1%2540.0%  
90117,5170.6%8,0340.6%  7,4860.7%8,0000.7%  
Medicaid Maintenance of Assistance
Receiving cash or eligible under Section 1931 of the Act665,50352.1%716,48553.4%  527,98650.1%573,89851.8%  
Medically needy120,3569.4%114,5748.5%  91,1408.6%84,0667.6%  
Poverty related (includes children eligible under SCHIP expansion)201,93115.8%204,65215.3%  168,43916.0%168,83215.2%  
Other150,07011.7%147,39611.0%  133,19812.6%130,50211.8%  
Foster care child6,7780.5%6,4610.5%  6,7780.6%6,4610.6%  
1115 demonstration expansion eligible124,9619.8%142,87810.7%  118,84711.3%136,84712.3%  
Unknown7830.1%3360.0%  5380.1%2540.0%  
Missing7,5170.6%8,0340.6%  7,4860.7%8,0000.7%  
Proportion with 1915c waiver status code28,1367.2%30,0252.2%  10,2981.0%11,2371.0%  
Type of 1915c Waiver Status Code
Unknown--missing eligibility7830.1%5330.0%  5380.1%4510.0%  
Aged & disabled13,5751.1%14,7721.1%  1,9120.2%2,3410.2%  
Aged5,5680.4%5,6830.4%  360.0%290.0%  
Disabled1,2330.1%1,2450.1%  1,1970.1%1,2090.1%  
Brain injured1040.0%1270.0%  990.0%1220.0%  
HIV-AIDS2790.0%2410.0%  2480.0%2130.0%  
MR/DD7,3660.6%7,9490.6%  6,7950.6%7,3150.7%  
Mentally ill/severely emotionally disturbed****  ****  
Technology-dependent/ medically fragile110.0%**  110.0%**  
Never enrolled in a 1915c waiver during the year1,248,98097.7%1,310,25897.7%  1,043,57699.0%1,097,17298.9%  
Individuals Linked to Both Medicare & Medicaid in Given Year278,168 293,317   77,000 81,899   
Average months of Medicare enrollment10.91 10.96   11.20 11.25   
Average months of Medicaid enrollment11.29 11.34   11.31 11.36   
Average months of both Medicare & Medicaid enrollment10.99 11.06   10.78 10.86   
Proportion enrolled in Medicaid & Medicare during entire year233,04883.8%247,85684.5%  62,5805.9%66,92681.7%  
Medicare & Medicaid Eligibility Status
QMB only10,1913.7%10,8053.7%  3,6684.8%3,9494.8%  
QMB plus187,82667.5%196,16766.9%  51,31766.6%53,46265.3%  
SLMB only11,1884.0%10,7843.7%  2,6363.4%2,6493.2%  
SLMB Plus3,7701.4%4,6611.6%  1,1231.5%1,4571.8%  
Other dual62,30722.4%68,47123.3%  16,86221.9%19,12823.4%  
Unknown2090.1%830.0%  390.1%210.0%  
NA2,6771.0%2,3460.8%  1,3551.8%1,2331.5%  

 

TABLE D-C1 (continued)
 Age 62 - 64Age 65+
2007 Linked to Medicare or Medicaid 20072008 Linked to Medicare or Medicaid 20082009 Linked to Medicare 20092007 Linked to Medicare or Medicaid 2007>2008 Linked to Medicare or Medicaid 20082009 Linked to Medicare 2009
N%N%N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Individuals linked to Medicare12,178 13,059 14,157 285,204 297,227 302,835 
Proportion with Part A coverage12,159100.00%13,03899.84%14,13399.83%273,41795.87%284,54095.73%292,65696.64%
Average months of Medicare Part A coverage (among those with Part A cov)11.61 11.61 11.62 11.57 11.61 11.60 
Proportion with Part B coverage11,80297.06%12,65896.93%13,76397.22%279,67298.06%291,63298.12%297,35498.19%
Average months of Medicare Part B coverage (among those with Part B cov)11.55 11.56 11.55 11.60 11.61 11.62 
Proportion enrolled in Medicare Part A & B during entire year10,90689.69%11,70689.64%12,72989.91%247,92286.93%259,88387.44%266,56188.02%
Proportion in Medicare managed care for at least one month2,65121.80%3,40926.10%3,80326.86%72,42825.40%88,42229.75%92,45430.53%
Average months of Medicare HMO coverage (among those with HMO cov)10.00 10.07 10.69 10.38 10.45 10.88 
Original Reason for Entitlement
Old age & survivor's insurance740.61%720.55%530.37%240,66184.38%249,44183.92%253,04983.56%
Disability insurance benefits11,80096.90%12,65096.87%13,75797.17%44,02015.43%47,22715.89%49,19316.24%
ESRD1070.88%1220.93%1300.92%2760.10%2840.10%2930.10%
Disability Insurance & ESRD1971.62%2151.65%2171.53%2470.09%2750.09%3000.10%
Current Reason for Entitlement
Old age & survivor's insurance1000.82%910.70%640.45%282,41499.02%293,10398.61%297,37598.20%
Disability insurance benefits11,79096.81%12,66696.99%13,78597.37%2,2270.78%3,5151.18%4,8461.60%
ESRD1070.88%1170.90%1260.89%5250.18%5320.18%5380.18%
Disability Insurance & ESRD1811.49%1851.42%1821.29%380.01%770.03%760.03%
Medicare Coverage Status
   Part A only3392.78%3692.83%3692.61%4,6901.64%4,6581.57%4,5651.51%
   Part B only******10,3623.63%11,1663.76%8,3002.74%
   Both Part A & B or combination of coverage11,83597.18%12,68797.15%13,78697.38%270,15294.72%281,40394.68%289,97095.75%
Medicare Part D coverage at any point during the year10,92789.73%11,92091.28%12,93991.40%249,24487.39%262,88588.45%269,61189.03%
Average months of Part D coverage (among those with Part D cov)11.49 11.49 11.52 11.44 11.52 11.55 
Proportion with Part D Coverage by Cost Share Group Code
   Beneficiary deemed with 100% premium-subsidy & no copayment3513.2%3162.7%3242.5%8,3503.4%8,1633.1%8,8423.3%
   Beneficiary deemed with 100% premium-subsidy & low copayment6,78762.1%7,29861.2%8,11162.7%153,89561.7%160,55161.1%168,69862.6%
   Beneficiary deemed with 100% premium-subsidy & high copayment1,51313.8%1,71814.4%1,79513.9%22,8509.2%24,0419.1%24,6519.1%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy5254.8%6135.1%6925.3%10,8964.4%11,3694.3%11,3644.2%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy & 15% copayment  180.2%190.2%200.2%1,1200.4%1,0030.4%9460.4%
   Beneficiary with LIS, 75% premium-subsidy & 15% copayment840.8%700.6%790.6%1,5480.6%1,5810.6%1,6450.6%
   Beneficiary with LIS, 50% premium-subsidy & 15% copayment620.6%760.6%830.6%1,4490.6%1,4870.6%1,4670.5%
   Beneficiary with LIS, 25% premium-subsidy & 15% copayment920.8%850.7%860.7%1,4260.6%1,2000.5%1,2640.5%
   No premium subsidy nor cost sharing1,0569.7%1,1309.5%1,2769.9%37,26014.9%40,59015.4%42,94915.9%
   Missing4394.0%5955.0%4733.7%10,4504.2%12,9004.9%7,7852.9%
Individuals Linked to Medicaid22,838 22,277   200,648 209,679   
Average months of Medicaid coverage (FFS or Managed care)11.23 11.28   11.26 11.32   
Proportion enrolled in Medicaid during entire year (FFS or Managed care)19,63485.97%19,35786.89%  174,09686.77%184,04487.77%  
In Medicaid managed care for at least one month4,12618.07%4,84721.76%  108,45154.05%111,75453.30%  
Average months of Medicaid managed care (among those with managed care)7.96 8.24   8.22 8.47   
Medicaid Basis for Eligibility
Aged630.28%450.20%  155,50177.50%162,11077.31%  
Blind/disabled19,68386.19%19,06785.59%  44,66722.26%47,27922.55%  
Child****  240.01%**  
Adult3,04113.32%3,11914.00%  2140.11%1920.09%  
Child of unemployed adult****  ****  
Unemployed adult****  ****  
Foster care child****  ****  
Covered under Breast & Cervical Cancer Prevention Act120.05%140.06%  ****  
Unknown210.09%130.06%  2240.11%690.03%  
9011150.07%170.08%  160.01%170.01%  
Medicaid Maintenance of Assistance
Receiving cash or eligible under Section 1931 of the Act13,07957.27%12,30655.24%  124,43862.02%130,28162.13%  
Medically needy2,66911.69%2,56911.53%  26,54713.23%27,93913.32%  
Poverty related (includes children eligible under SCHIP expansion)2,80212.27%3,08613.85%  30,68915.29%32,73415.61%  
Other1,5316.70%1,4986.72%  15,3417.65%15,3967.34%  
Foster care child****  ****  
1115 demonstration expansion eligible2,72111.91%2,78812.52%  3,3931.69%3,2431.55%  
Unknown210.09%130.06%  2240.11%690.03%  
Missing150.07%170.08%  160.01%170.01%  
Proportion with 1915c waiver status code1,0914.78%1,1485.15%  16,7478.35%17,6408.41%  
Type of 1915c Waiver Status Code
Unknown--missing eligibility210.09%130.06%  2240.11%690.03%  
Aged & disabled7843.43%8313.73%  10,8795.42%11,6005.53%  
Aged1170.51%1080.48%  5,4152.70%5,5462.64%  
Disabled180.08%130.06%  180.01%230.01%  
Brain injured****  ****  
HIV-AIDS****  230.01%240.01%  
MR/DD1620.71%1890.85%  4090.20%4450.21%  
Mentally ill/severely emotionally disturbed****  ****  
Technology-dependent/ medically fragile****  ****  
Never enrolled in a 1915c waiver during the year21,72695.13%21,11694.79%  183,67791.54%191,97091.55%  
Individuals Linked to Both Medicare & Medicaid in Given Year9,025 9,665   192,143 201,753   
Average months of Medicare enrollment11.38 11.41   10.78 10.83   
Average months of Medicaid enrollment11.15 11.28   11.29 11.34   
Average months of both Medicare & Medicaid enrollment10.81 10.98   11.08 11.14   
Proportion enrolled in Medicaid & Medicare during entire year7,25380.37%8,01982.97%  163,21584.94%172,91185.70%  
Medicare & Medicaid Eligibility Status
QMB only5125.67%5675.87%  6,0113.13%6,2893.12%  
QMB plus5,54561.44%5,88060.84%  130,96468.16%136,82567.82%  
SLMB only4655.15%4855.02%  8,0874.21%7,6503.79%  
SLMB Plus1521.68%2012.08%  2,4951.30%3,0031.49%  
Other dual2,23724.79%2,45025.35%  43,20822.49%46,89323.24%  
Unknown110.12%**  1590.08%560.03%  
NA1031.14%760.79%  1,2190.63%1,0370.51%  

 

TABLE D-C2. Health Insurance Characteristics of HUD Heads of Households Receiving SSI that Link in 2008 by Age
  All AgesAge <62Age 62-64Age 65+
N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Individuals Linked to 2008 Medicare159,658 42,143 4,732 112,783 
Proportion with Part A coverage151,06294.6%42,140100.0%4,71699.7%104,20692.4%
Average months of Medicare Part A coverage (among those with Part A cov)11.59 11.57 11.64 11.60 
Proportion with Part B coverage158,71799.4%41,56698.6%4,67398.8%112,47899.7%
Average months of Medicare Part B coverage (among those with Part B cov)11.59 11.50 11.59 11.63 
Proportion enrolled in Medicare Part A and B during entire year139,32587.3%37,96390.1%4,34391.8%97,01986.0%
Proportion in Medicare managed care for at least one month34,17021.4%6,41915.2%1,05322.3%26,69823.7%
Average months of Medicare HMO coverage (among those with HMO cov)9.67 9.16 9.64 9.79 
Original Reason for Entitlement
Old age and survivor's insurance97,81261.3%110.0%310.7%97,77086.7%
Disability insurance benefits60,30037.8%40,92897.1%4,55996.3%14,81313.1%
ESRD8100.5%5991.4%771.6%1340.1%
Disability Insurance and ESRD7360.5%6051.4%651.4%660.1%
Current Reason for Entitlement
Old age and survivor's insurance111,47069.8%160.0%400.8%111,41498.8%
Disability insurance benefits46,73429.3%41,02597.3%4,55896.3%1,1511.0%
ESRD8510.5%5771.4%761.6%1980.2%
Disability Insurance and ESRD6030.4%5251.2%581.2%200.0%
Medicare Coverage Status
   Part A only8020.5%4981.2%541.1%2500.2%
   Part B only7,7074.8%****7,7026.8%
   Both Part A and B or combination of coverage151,14994.7%41,64398.8%4,67598.8%104,83192.9%
Medicare Part D coverage at any point during the year157,31498.5%41,00697.3%4,62897.8%111,68099.0%
Average months of Part D coverage (among those with Part D cov)11.56 11.44 11.53 11.61 
Proportion with Part D Coverage by Cost Share Group Code
   Beneficiary deemed with 100% premium-subsidy and no copayment4,4892.9%1,1592.8%1453.1%3,1852.9%
   Beneficiary deemed with 100% premium-subsidy and low copayment141,00189.6%34,68584.6%3,90584.4%102,41191.7%
   Beneficiary deemed with 100% premium-subsidy and high copayment3,6502.3%2,1555.3%2164.7%1,2791.1%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy8160.5%4201.0%621.3%3340.3%
   Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy and 15% copayment  420.0%****300.0%
   Beneficiary with LIS, 75% premium-subsidy and 15% copayment680.0%390.1%**210.0%
   Beneficiary with LIS, 50% premium-subsidy and 15% copayment700.0%340.1%**300.0%
   Beneficiary with LIS, 25% premium-subsidy and 15% copayment410.0%280.1%****
   No premium subsidy nor cost sharing1,2360.8%4641.1%851.8%6870.6%
   Missing5,9013.8%2,0134.9%1954.2%3,6933.3%
Individuals Linked to Medicaid248,031 122,425 11,895 113,711 
Average months of Medicaid coverage (FFS or Managed care)11.70 11.69 11.69 11.71 
Proportion enrolled in Medicaid during entire year (FFS or Managed care)233,70894.2%114,74493.7%11,19394.1%107,77194.8%
In Medicaid managed care for at least one month16,0286.5%7,3996.0%6615.6%7,9687.0%
Average months of Medicaid managed care (among those with managed care)8.31 8.20 7.78 8.46 
Medicaid Basis for Eligibility
Aged75,53130.5%150.0%190.2%75,49766.4%
Blind/disabled162,75465.6%112,89192.2%11,69298.3%38,17133.6%
Child740.0%740.1%****
Adult9,5053.8%9,2997.6%1761.5%300.0%
Child of Unemployed Adult********
Unemployed Adult********
Foster Care Child********
Covered under Breast and Cervical Cancer Prevention Act********
Unknown260.0%130.0%****
Missing1300.1%1230.1%****
Medicaid Maintenance of Assistance
Receiving cash or eligible under section 1931 of the Act221,47789.3%105,47286.2%10,24486.1%105,76193.0%
Medically needy6,0232.4%2,8012.3%4323.6%2,7902.5%
Poverty related (includes children eligible under S-CHIP expansion)6,7422.7%4,2673.5%4844.1%1,9911.8%
Other10,7664.3%7,0955.8%5724.8%3,0992.7%
Foster Care Child********
1115 Demonostration expansion eligible2,8651.2%2,6522.2%1561.3%570.1%
Unknown260.0%130.0%****
Missing1300.1%1230.1%****
Proportion with 1915c waiver status code13,7905.6%4,2993.5%6395.4%8,8527.8%
Type of 1915c Waiver Status Code
Unknown--missing eligibility260.0%130.0%****
Aged and disabled7,7183.1%1,2191.0%4794.0%6,0205.3%
Aged2,6801.1%210.0%670.6%2,5922.3%
Disabled7560.3%7460.6%****
Brain injured430.0%420.0%****
HIV-AIDS1290.1%1180.1%****
MR/DD2,4641.0%2,1531.8%890.7%2220.2%
Mentally ill/Severely emotionally disturbed********
Technology-dependent/ medically fragile********
Never enrolled in a 1915c waiver during the year234,21594.4%118,11396.5%11,25194.6%104,85192.2%
Individuals Linked to Both Medicare and Medicaid in Given Year154,643 39,944 4,445 110,254 
Average months of Medicare enrollment10.89 11.39 11.48 10.69 
Average months of Medicaid enrollment11.71 11.67 11.69 11.73 
Average months of both Medicare & Medicaid enrollment11.45 11.20 11.36 11.55 
Proportion enrolled in Medicaid & Medicare during entire year140,74691.0%34,82687.2%3,99689.9%101,92492.4%
Medicare and Medicaid Eligibility Status
QMB only2,1111.4%1,1853.0%1333.0%7930.7%
QMB plus128,03582.8%31,34978.5%3,44577.5%93,24184.6%
SLMB only6360.4%3821.0%431.0%2110.2%
SLMB Plus4960.3%2580.6%250.6%2130.2%
Other dual22,71914.7%6,39616.0%76917.3%15,55414.1%
Unknown********
NA6370.4%3700.9%290.7%2380.2%

Objective D: Compare Medicare and Medicaid Cost and Utilization for HUD-Assisted Medicare Beneficiaries and Unassisted Beneficiaries in the Community in the 12 Study Jurisdictions

TABLE D-D1. Final 2008 Samples for Utilization and Cost Analysis
 HUD-Assisted  Beneficiaries  (full year)HUD-Assisted  Beneficiaries  (partial year)Total HUD-  Assisted Beneficiaries    Unassisted Beneficiaries  in the CommunityTotal
Exclusion Criteria
All Medicare Beneficiaries362,16245,460407,6225,345,0405,752,662
Medicare benes enrolled in A&B for entire 12 month period or up until death  326,70736,450363,1574,435,9174,799,074
No managed care/HMO (Medicare FFS only)232,63026,807259,4373,302,5953,562,032
Subgroup Sample Sizes
Subgroup A: Medicare-Medicaid benes, no SNF or NF days162,01015,046177,056446,751623,807
   Subgroup A1: Age 65+104,4167,629112,045249,490361,535
Subgroup B: Medicare-Medicaid benes with >0 and <365 SNF/NF days12,6843,87516,55986,626103,185
Subgroup C: Medicare only, 65+45,5665,40650,9722,453,7632,504,735
   Subgroup C1: Medicare only, 65+, No SNF days42,2864,52046,8062,318,3942,365,200
   92%94% 

 

TABLE D-D2. Final 2008 Samples Limited to Individuals with Medicaid Fee-for-Service (FFS) -- No Managed Care
 2008
HUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Total HUD-Assisted BeneficiariesUnassisted Beneficiariesin the CommunityTotal
Subgroup A: Medicare-Medicaid benes, no SNF or NF days153,51414,126167,640405,908573,548
   Subgroup A1: Age 65+99,5687,196106,764227,186333,950
Subgroup B: Medicare-Medicaid benes with >0 and <365 SNF/NF days11,6203,10314,72364,63179,354

Sub-group A: Medicare-Medicaid Enrollees, with No Skilled Nursing Facility or Nursing Facility Use in 2008

TABLE D-D.A1. Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries, 2008
 HUD-AssistedBeneficiaries(full year)(N=162,010)HUD-AssistedBeneficiaries(partial year)(N=15,046)Total HUD-Assisted Beneficiaries(N=177,056)Unassisted Beneficiariesin the Community(N=446,751)Assisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-ValueTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Demographic Characteristics
Gender<0.0001<0.0001<0.0001<0.0001
Male53,47033.0%5,73538.1%59,20533.4%185,70541.6%    
Female108,54067.0%9,31161.9%117,85166.6%261,04658.4%    
Race (based on RTI variable)<0.0001<0.0001<0.0001<0.0001
Non-Hispanic White72,06544.5%6,77845.0%78,84344.5%225,41750.5%    
Hispanic33,68620.8%2,78218.5%36,46820.6%77,64717.4%    
Black37,25423.0%4,03726.8%41,29123.3%79,70317.8%    
Asian/Pacific Islander15,1049.3%1,0947.3%16,1989.1%52,45111.7%    
American Indian/Alaska Native2190.1%260.2%2450.1%9110.2%    
Other3,2192.0%3022.0%3,5212.0%9,4302.1%    
Unknown4630.3%270.2%4900.3%1,1920.3%    
Age Group<0.0001<0.0001<0.0001<0.0001
Under 18 years140.0%**150.0%820.0%    
18 to 4417,57410.8%2,81618.7%20,39011.5%75,60016.9%    
45 to 6440,00624.7%4,60030.6%44,60625.2%121,57927.2%    
65 to 7447,01729.0%4,23628.2%51,25328.9%110,12624.7%    
75 to 7923,25614.4%1,4729.8%24,72814.0%50,50511.3%    
80 to 8418,66411.5%1,0517.0%19,71511.1%42,3579.5%    
85+15,4799.6%8705.8%16,3499.2%46,50210.4%    
Geographic Area<0.0001<0.0001<0.0001<0.0001
Vermont4,5452.8%5583.7%5,1032.9%18,6844.2%    
New Haven- Milford3,3402.1%1,1917.9%4,5312.6%10,0922.3%    
Bridgeport- Stamford- Norwalk3,1231.9%3122.1%3,4351.9%7,0171.6%    
Milwaukee- Waukesha- West Allis5,4353.4%7795.2%6,2143.5%22,3105.0%    
San Francisco- Oakland- Fremont18,64511.5%1,1367.6%19,78111.2%53,09011.9%    
Boston- Cambridge- Quincy29,95118.5%2,63017.5%32,58118.4%60,73113.6%    
Durham- Chapel Hill1,1860.7%1701.1%1,3560.8%4,7701.1%    
Richmond6010.4%1100.7%7110.4%6,0001.3%    
New York- Northern New Jersey- Long Island82,44950.9%6,20241.2%88,65150.1%214,84048.1%    
Columbus4,0252.5%6914.6%4,7162.7%13,9873.1%    
Akron2,2301.4%3672.4%2,5971.5%5,7421.3%    
Cleveland6,4804.0%9006.0%7,3804.2%19,8684.4%    
Missing******9,6202.2%    
Died during the year1,1150.7%1,1057.3%2,2201.3%8,3611.9%<0.0001<0.0001<0.0001<0.0001
Medicare Coverage Characteristics
Original Reason for Entitlement<0.0001<0.0001<0.0001<0.0001
Old age and survivor's insurance86,44053.4%6,26241.6%92,70252.4%212,15447.5%    
Disability insurance benefits73,50745.4%8,47356.3%81,98046.3%226,97050.8%    
ESRD7900.5%1010.7%8910.5%2,8720.6%    
Disability Insurance and ESRD1,2730.8%2101.4%1,4830.8%4,7551.1%    
Current reason for entitlement<0.0001<0.0001<0.0001<0.0001
Old age and survivor's insurance102,55363.3%7,53150.1%110,08462.2%245,20454.9%    
Disability insurance benefits57,60835.6%7,23748.1%64,84536.6%194,66443.6%    
ESRD8500.5%1150.8%9650.5%3,0940.7%    
Disability Insurance and ESRD9990.6%1631.1%1,1620.7%3,7890.8%    
Medicare Part D coverage at any point during the year160,72099.2%14,89599.0%175,61599.2%438,81698.2%<0.0001<0.0001<0.0001<0.0001
Average months of Part D coverage (among those with Part D cov)11.92 11.64 11.90 11.81 <0.0001<0.0001<0.0001<0.0001
Proportion with Part D Coverage by Cost Share Group Code<0.0001<0.0001<0.0001<0.0001
Beneficiary deemed with 100% premium-subsidy and no copayment2,5621.6%2641.8%2,8261.6%17,0133.9%    
Beneficiary deemed with 100% premium-subsidy and low copayment139,49786.8%12,36983.0%151,86686.5%348,88979.5%    
Beneficiary deemed with 100% premium-subsidy and high copayment16,95010.5%2,01313.5%18,96310.8%62,34014.2%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy3940.2%680.5%4620.3%2,1130.5%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy and 15% copayment310.0%**340.0%1120.0%    
Beneficiary with LIS, 75% premium-subsidy and 15% copayment530.0%**580.0%1930.0%    
Beneficiary with LIS, 50% premium-subsidy and 15% copayment640.0%**690.0%2710.1%    
Beneficiary with LIS, 25% premium-subsidy and 15% copayment490.0%**530.0%1950.0%    
No premium subsidy nor cost sharing2850.2%320.2%3170.2%2,4090.5%    
Missing8350.5%1320.9%9670.6%5,2811.2%    
Medicaid Coverage Status
Average months of Medicaid coverage (FFS or Managed care)11.72 11.12 11.66 11.27 <0.0001<0.0001<0.0001<0.0001
Proportion enrolled in Medicaid during entire year (FFS or Managed care)151,84293.7%12,06380.2%163,90592.6%382,47185.6%<0.0001<0.0001<0.0001<0.0001
In Medicaid managed care for at least one month8,4795.2%9136.1%9,3925.3%40,5469.1%<0.0001<0.0001<0.0001<0.0001
Average months of Medicaid managed care (among those with managed care)8.27 7.33 8.18 8.69 <0.0001<0.0001<0.0001<0.0001
Medicaid Basis for Eligibility<0.0001<0.0001<0.0001<0.0001
Aged79,56149.1%6,08940.5%85,65048.4%212,48547.6%    
Blind/disabled80,92349.9%8,68057.7%89,60350.6%227,89251.0%    
Child******310.0%    
Adult1,5020.9%2671.8%1,7691.0%6,0261.3%    
Child of unemployed adult********    
Unemployed adult********    
Foster care child********    
Covered under Breast and Cervical Cancer Prevention Act******160.0%    
Unknown160.0%**230.0%2680.1%    
Missing******290.0%    
Medicaid Maintenance of Assistance<0.0001<0.0001<0.0001<0.0001
Receiving cash or eligible under section 1931 of the Act100,29261.9%7,87152.3%108,16361.1%221,48049.6%    
Medically needy18,37011.3%1,68411.2%20,05411.3%64,67514.5%    
Poverty related (includes children eligible under SCHIP expansion)29,34318.1%3,43922.9%32,78218.5%98,46822.0%    
Other10,4226.4%1,65211.0%12,0746.8%41,7109.3%    
Foster Care Child********    
1115 Demonstration expansion eligible3,5642.2%3932.6%3,9572.2%20,1174.5%    
Unknown160.0%**230.0%2680.1%    
Missing******290.0%    
Proportion with 1915c Waiver Status Code13,9398.6%1,2318.2%15,1708.6%41,1189.2%<0.0001<0.0001<0.0001<0.0001
Type of 1915c Waiver Status Code<0.0001<0.0001<0.0001<0.0001
Unknown--missing eligibility160.0%**230.0%2680.1%    
Aged and disabled7,1964.4%5773.8%7,7734.4%12,6452.8%    
Aged2,9761.8%2972.0%3,2731.8%3,7570.8%    
Disabled4280.3%580.4%4860.3%9460.2%    
Brain injured580.0%**610.0%5610.1%    
HIV-AIDS1260.1%**1300.1%2610.1%    
MR/DD3,1541.9%2921.9%3,4461.9%22,9425.1%    
Mentally ill/Severely emotionally disturbed********    
Technology-dependent/ medically fragile********    
Never enrolled in a 1915c waiver during the year148,05591.4%13,80891.8%161,86391.4%405,36590.7%    
Average Months of both Medicare & Medicaid Enrollment11.69 11.06 11.63 11.22 <0.0001<0.0001<0.0001<0.0001
Proportion Enrolled in Medicaid & Medicare During Entire Year151,22293.3%12,00079.8%163,22292.2%380,03785.1%<0.0001<0.0001<0.0001<0.0001
Medicare & Medicaid Eligibility Status<0.0001<0.0001<0.0001<0.0001
QMB only5,3773.3%8865.9%6,2633.5%18,8074.2%    
QMB plus123,12676.0%10,45369.5%133,57975.4%286,04964.0%    
SLMB only4,9173.0%5553.7%5,4723.1%18,6124.2%    
SLMB Plus2,2311.4%2831.9%2,5141.4%7,2801.6%    
Other dual26,20216.2%2,82318.8%29,02516.4%114,92925.7%    
Unknown160.0%**230.0%2680.1%    
NA1410.1%390.3%1800.1%8060.2%    

 

TABLE D-D.A2. Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
CCW ConditionTotal HUD-AssistedBeneficiariesUnassisted Beneficiariesin the CommunityP-Value
N%N%
Prevalence of Chronic Conditions<0.0001
0 Conditions19,02210.7%75,51916.9% 
1 Condition18,52610.5%59,82713.4% 
2 Conditions19,85011.2%58,40713.1% 
3 Conditions22,16912.5%58,46113.1% 
4 Conditions22,07512.5%53,15511.9% 
5 or More Conditions75,41442.6%141,38231.6% 
Prevalence of Chronic Conditions by Category
Cardiovascular77,22643.6%160,60335.9%<0.0001
Cancer10,9206.2%21,5154.8%<0.0001
Endocrine and Renal83,30747.1%179,73440.2%<0.0001
Alzheimer's-related18,20610.3%46,56110.4%0.1088
Depression41,19423.2%85,19619.1%<0.0001
Musculoskeletal70,01539.5%128,18728.7%<0.0001
Pulmonary33,66019.0%67,92815.2%<0.0001
Opthalmic51,16128.9%96,82321.7%<0.0001
Other134,17675.8%299,14966.9%<0.0001

 

TABLE D-D.A3. Medicare Health Care Utilization per 1000 Member Months
 HUD-Assisted Beneficiaries Unassisted Beneficiaries in the Community p-Value
MeanSDMedianMeanSDMedian
Total Admissions35.494.4035.2399.2500.5912
   Acute stay admissions31.3986.67029.6187.380<0.0001
   Other inpatient admissions428.6705.6236.920<0.0001
Hospital Readmissions5.4939.2605.4741.4700.8163
Medicare home health visits450.442752.270317.872303.140<0.0001
Medicare hospice days27.95693.460126.791754.930<0.0001
Hospital Outpatient visits896.072063.01333.33822.822179.92250<0.0001
Total emergency room visits (total)80.26216.39077.85209.050<0.0001
   Hospital outpatient ER visits57.42190.47055.29175.570<0.0001
   Hospital inpatient ER visits22.8469.9022.5772.8900.1923
Physician office visits1605.791877.551083.331356.051940.13833.33<0.0001
Ambulatory surgery center visits12.0385.2808.5159.640<0.0001
Dialysis events25.69229.91027.42234.2300.0098
Anesthesia events27.884.41024.5684.330<0.0001
Imaging events467.35654.78250385.9646.42166.67<0.0001
Test events1523.42284.72666.671299.092121.31500<0.0001
Other procedures1089.993168.87166.67625.62187.9983.33<0.0001
Durable medical equipment (DME)351.86743.070270.44695.60<0.0001
Part B Drugs261.44680.983.33217.56629.950<0.0001
Part D* Drugs4846.773962.124083.334037.483610.163250<0.0001
Any Use of Medicare Service During Year(% with any use) N%  N%  p-Value
Total Admissions3850921.7% 9342720.9% <0.0001
   Acute stay admissions3596620.3% 8446418.9% <0.0001
   Other inpatient admissions52102.9% 167923.8% <0.0001
Hospital Readmissions65043.7% 157623.5% 0.0058
Medicare home health1965111.1% 379838.5% <0.0001
Medicare hospice7340.4% 41560.9% <0.0001
Hospital Outpatient13577876.7% 31576370.7% <0.0001
Total emergency room6999439.5% 16465036.9% <0.0001
   Hospital outpatient ER5605131.7% 13040729.2% <0.0001
   Hospital inpatient ER2894216.3% 7021515.7% <0.0001
Physician office16468393.0% 39877289.3% <0.0001
Ambulatory surgery center97645.5% 199234.5% <0.0001
Dialysis35192.0% 95422.1% 0.0003
Anesthesia3190918.0% 6927315.5% <0.0001
Imaging13077573.9% 29385565.8% <0.0001
Test14023879.2% 33117274.1% <0.0001
Other procedures11029562.3% 24230654.2% <0.0001
Durable medical equipment (DME)7303941.3% 14149231.7% <0.0001
Part B Drugs9230052.1% 21374847.8% <0.0001
Part D* Drugs16460093.7% 39159787.7% <0.0001

 

TABLE D-D.A3 (continued)
 HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
N% N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to Those with Medicaid Fee-for-Service (FFS)--No Managed Care167,64095% 405,90891%  
Medicaid Service Utilization per 1000 Member MonthsMeanSDMedianMeanSDMedianp-Value
Personal Care services3244.438680.3101434.045868.360<0.0001
Residential care55.15849.470130.891607.230<0.0001
DME368.861330.650238.731153.170<0.0001
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)3081.628567.0102311.067827.640<0.0001
Any Use of Medicare Service During Year(% with any use)N% N% p-Value
Personal Care services35,85821% 41,90410% <0.0001
Residential care3,4012% 12,4303% <0.0001
DME76,91846% 131,42132% <0.0001
Other HCBS services64,46938% 116,20529% <0.0001

 

TABLE D-D.A4. Medicare and Beneficiary Health Care Costs
Medicare Payment PMPM ($)Total HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
MeanSDMedianMeanSDMedian
Total Admissions PMPM$375$1,356$0$373$1,408$00.5962
   Acute stay admission PMPM$329$1,248$0$311$1,259$0<0.0001
   Other inpatient admission PMPM$46$400$0$62$495$0<0.0001
Medicare home health PMPM (Medicare only)$52$235$0$39$203$0<0.0001
Medicare hospice PMPM (Medicare only)$5$122$0$20$277$0<0.0001
Hospital Outpatient PMPM$153$519$38$142$486$24<0.0001
Physician office PMPM$100$133$62$83$138$45<0.0001
Ambulatory surgery center PMPM$4$26$0$3$23$0<0.0001
Dialysis PMPM$4$32$0$4$32$00.0156
Anesthesia PMPM$4$11$0$3$11$0<0.0001
Imaging PMPM$31$56$8$23$46$4<0.0001
Test PMPM$34$64$11$26$57$8<0.0001
Other procedures PMPM$64$158$10$43$128$3<0.0001
Durable medical equipment (DME) PMPM$29$95$0$22$92$0<0.0001
Part B Drugs PMPM$23$260$0$23$642$00.8782
Part D* Drugs PMPM$398$547$243$367$570$195<0.0001
Total Medical and Rx PMPM (sum of all above)$1,269$2,043$648$1,161$2,178$508<0.0001
Total Medical PMPM (sum of all above except part D PMPM)$854$1,832$272$783$1,876$184<0.0001

 

TABLE D-D.A4 (continued)
Medicaid Service Utilization per 1000 Member MonthsTotal HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
N% N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to Those with Medicaid Fee-for-Service (FFS)--No Managed Care167,64095% 405,90891%  
Medicaid Payment PMPM--Divide by Total Medicaid MonthsMeanSDMedianMeanSDMedianp-Value
Total Medicaid payment amount$1,057$1,892$193$1,127$2,667$105<0.0001
Fee-for-service Medicaid payment amount$947$1,844$156$1,044$2,648$83<0.0001
Personal Care services$392$1,176$0$204$902$0<0.0001
Residential care$47$474$0$165$1,153$0<0.0001
DME$22$122$0$18$128$0<0.0001
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)$254$996$0$220$919$0<0.0001

Sub-group A1: Medicare-Medicaid Enrollees (MMEs), Age 65+, with No SNF or NF Use in 2008

TABLE D-D.A1.1. Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
 HUD-AssistedBeneficiaries(full year)(N=104,416)HUD-AssistedBeneficiaries(partial year)(N=7,629)Total HUD-Assisted Beneficiaries(N=112,045)Unassisted Beneficiariesin the Community(N=249,490)Assisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-ValueTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Demographic Characteristics
Gender<0.00010.0249<0.0001<0.0001
Male30,47929.2%2,52633.1%33,00529.5%79,56531.9%    
Female73,93770.8%5,10366.9%79,04070.5%169,92568.1%    
Race (based on RTI variable)<0.0001<0.00010.0008<0.0001
Non-Hispanic White46,34344.4%3,43845.1%49,78144.4%109,98144.1%    
Hispanic22,44721.5%1,50119.7%23,94821.4%50,56320.3%    
Black18,65917.9%1,45719.1%20,11618.0%35,83714.4%    
Asian/Pacific Islander13,95913.4%99013.0%14,94913.3%45,77318.3%    
American Indian/Alaska Native970.1%100.1%1070.1%4180.2%    
Other2,5822.5%2142.8%2,7962.5%6,2892.5%    
Unknown3290.3%190.2%3480.3%6290.3%    
Age Group<0.0001<0.0001<0.0001<0.0001
Under 18 years********    
18 to 44********    
45 to 64********    
65 to 7447,01745.0%4,23655.5%51,25345.7%110,12644.1%    
75 to 7923,25622.3%1,47219.3%24,72822.1%50,50520.2%    
80 to 8418,66417.9%1,05113.8%19,71517.6%42,35717.0%    
85+15,47914.8%87011.4%16,34914.6%46,50218.6%    
Geographic Area<0.0001<0.0001<0.0001<0.0001
Vermont2,3622.3%2593.4%2,6212.3%10,8814.4%    
New Haven- Milford1,8711.8%4776.3%2,3482.1%4,4081.8%    
Bridgeport- Stamford- Norwalk1,8881.8%1351.8%2,0231.8%3,5411.4%    
Milwaukee- Waukesha- West Allis3,2063.1%3594.7%3,5653.2%12,1304.9%    
San Francisco- Oakland- Fremont13,39712.8%7479.8%14,14412.6%32,72613.1%    
Boston- Cambridge- Quincy14,76914.1%93312.2%15,70214.0%25,41910.2%    
Durham- Chapel Hill4280.4%390.5%4670.4%2,1340.9%    
Richmond2630.3%270.4%2900.3%2,9331.2%    
New York- Northern New Jersey- Long Island60,31057.8%4,03552.9%64,34557.4%137,64255.2%    
Columbus1,6721.6%2583.4%1,9301.7%4,5721.8%    
Akron8900.9%891.2%9790.9%1,9300.8%    
Cleveland3,3603.2%2713.6%3,6313.2%7,5583.0%    
Missing******3,6161.4%    
Died during the year8990.9%82010.7%1,7191.5%6,4562.6%<0.0001<0.0001<0.0001<0.0001
Medicare Coverage Characteristics
Original Reason for Entitlement<0.0001<0.00010.0013<0.0001
Old age and survivor's insurance86,43982.8%6,26282.1%92,70182.7%212,13285.0%    
Disability insurance benefits17,70117.0%1,33317.5%19,03417.0%36,58714.7%    
ESRD1490.1%120.2%1610.1%4080.2%    
Disability Insurance and ESRD1270.1%220.3%1490.1%3630.1%    
Current Reason for Entitlement0.0138<0.0001<0.00010.1429
Old age and survivor's insurance102,54998.2%7,53098.7%110,07998.2%245,19298.3%    
Disability insurance benefits1,5721.5%640.8%1,6361.5%3,4931.4%    
ESRD2560.2%300.4%2860.3%7210.3%    
Disability Insurance and ESRD390.0%**440.0%840.0%    
Medicare Part D Coverage at Any Point During the Year103,31498.9%7,51098.4%110,82498.9%242,66497.3%<0.0001<0.0001<0.0001<0.0001
Average Months of Part D Coverage(among those with Part D cov)11.94 11.51 11.91 11.80 <0.0001<0.0001<0.0001<0.0001
Proportion with Part D Coverage by Cost Share Group Code<0.0001<0.0001<0.0001<0.0001
Beneficiary deemed with 100% premium-subsidy and no copayment1,2921.3%1171.6%1,4091.3%6,8022.8%    
Beneficiary deemed with 100% premium-subsidy and low copayment91,55088.6%6,32284.2%97,87288.3%196,47481.0%    
Beneficiary deemed with 100% premium-subsidy and high copayment9,4679.2%91512.2%10,3829.4%32,73613.5%    
Beneficiary with Low Income Subsidy(LIS), 100% premium-subsidy2580.2%390.5%2970.3%1,2300.5%    
Beneficiary with Low Income Subsidy(LIS), 100% premium-subsidy and 15% copayment270.0%**300.0%900.0%    
Beneficiary with LIS, 75% premium-subsidy and 15% copayment380.0%**400.0%1290.1%    
Beneficiary with LIS, 50% premium-subsidy and 15% copayment480.0%**520.0%2000.1%    
Beneficiary with LIS, 25% premium-subsidy and 15% copayment320.0%**340.0%1460.1%    
No premium subsidy nor cost sharing2350.2%280.4%2630.2%2,0680.9%    
Missing3670.4%781.0%4450.4%2,7891.1%    
Medicaid Coverage Status
Average Months of Medicaid Coverage(FFS or Managed care)11.76 11.05 11.71 11.33 <0.0001<0.0001<0.0001<0.0001
Proportion Enrolled in Medicaid During Entire Year(FFS or Managed care)98,98794.8%6,01778.9%105,00493.7%216,69686.9%<0.0001<0.0001<0.0001<0.0001
In Medicaid Managed Care for at Least One Month4,8404.6%4315.6%5,2714.7%22,1228.9%<0.0001<0.0001<0.0001<0.0001
Average Months of Medicaid Managed Care(among those with managed care)8.55 7.79 8.48 8.89 <0.0001<0.00010.0003<0.0001
Medicaid Basis for Eligibility<0.0001<0.0001<0.0001<0.0001
Aged79,55076.2%6,08579.8%85,63576.4%212,43185.1%    
Blind/disabled24,81623.8%1,53320.1%26,34923.5%36,63614.7%    
Child********    
Adult410.0%**490.0%2340.1%    
Child of Unemployed Adult********    
Unemployed Adult********    
Foster Care Child********    
Covered under Breast and Cervical Cancer Prevention Act********    
Unknown******1740.1%    
Missing********    
Medicaid Maintenance of Assistance<0.0001<0.0001<0.0001<0.0001
Receiving cash or eligible under section 1931 of the Act71,52968.5%4,55459.7%76,08367.9%136,91454.9%    
Medically needy10,58410.1%76210.0%11,34610.1%36,41414.6%    
Poverty related(includes children eligible under S-CHIP expansion)15,14414.5%1,35317.7%16,49714.7%46,34518.6%    
Other5,1985.0%7429.7%5,9405.3%16,7396.7%    
Foster Care Child********    
1115 Demonstration expansion eligible1,9531.9%2162.8%2,1691.9%12,8965.2%    
Unknown******1740.1%    
Missing********    
Proportion with 1915c Waiver Status Code9,4609.1%78810.3%10,2489.1%15,8866.4%<0.0001<0.00010.0003<0.0001
Type of 1915c Waiver Status Code<0.0001<0.00010.00010.0001
Unknown--missing eligibility******1740.1%    
Aged and disabled6,2416.0%4736.2%6,7146.0%10,6564.3%    
Aged2,9132.8%2923.8%3,2052.9%3,6601.5%    
Disabled110.0%**120.0%190.0%    
Brain injured******220.0%    
HIV-AIDS150.0%**170.0%170.0%    
MR/DD2800.3%200.3%3000.3%1,5110.6%    
Mentally ill/Severely emotionally disturbed********    
Technology-dependent/ medically fragile********    
Never enrolled in a 1915c waiver during the year94,94890.9%6,83989.6%101,78790.8%233,43093.6%    
Average Months of both Medicare & Medicaid Enrollment11.74 11.00 11.69 11.29 <0.0001<0.0001<0.0001<0.0001
Proportion Enrolled in Medicaid & Medicare During Entire Year98,79794.6%5,99278.5%104,78993.5%215,78386.5%<0.0001<0.0001<0.0001<0.0001
Medicare & Medicaid Eligibility Status<0.0001<0.0001<0.0001<0.0001
QMB only2,8122.7%3264.3%3,1382.8%8,8293.5%    
QMB plus81,66778.2%5,44471.4%87,11177.7%163,14665.4%    
SLMB only3,2673.1%3003.9%3,5673.2%11,5404.6%    
SLMB Plus1,1941.1%1231.6%1,3171.2%3,0201.2%    
Other dual15,40314.8%1,41418.5%16,81715.0%62,37425.0%    
Unknown******1740.1%    
NA650.1%200.3%850.1%4070.2%    

 

TABLE D-D.A1.2. Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
CCW ConditionHUD-Assisted MMEsUnassisted MMEs in the CommunityTotal HUD vs. None
N%N%
Prevalence of Chronic Conditions<0.0001
0 Conditions6,0315.4%24,4049.8% 
1 Condition6,5715.9%20,6868.3% 
2 Conditions9,7658.7%27,67811.1% 
3 Conditions13,39812.0%34,38713.8% 
4 Conditions15,21013.6%34,87814.0% 
5 or More Conditions61,07054.5%107,45743.1% 
Prevalence of Chronic Conditions by Category
Cardiovascular62,01255.3%120,34448.3%<0.0001
Cancer9,1628.2%17,2696.9%<0.0001
Endocrine and Renal60,07653.6%118,12447.4%<0.0001
Alzheimer's-related16,42714.7%39,69115.9%<0.0001
Depression18,60316.6%28,77411.5%<0.0001
Musculoskeletal55,12949.2%94,35437.9%<0.0001
Pulmonary21,35219.0%39,20915.7%<0.0001
Opthalmic42,91038.3%76,17230.6%<0.0001
Other96,64286.2%196,46578.8%<0.0001

 

TABLE D-D.A1.3. Medicare Health Care Utilization per 1000 Member Months
Medicare Services Annual Utilization per 1000 Member MonthsTotal HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
MeanSDMedianMeanSDMedian
Total Admissions32.885.52030.9383.050  <0.0001
   Acute stay admissions31.3982.08029.3879.160<0.0001
   Other inpatient admissions1.4113.9801.5615.200.0064
Hospital Readmissions5.1535.3104.8734.1100.0238
Medicare home health visits581.483158.270445.462724.070<0.0001
Medicare hospice days36.93791.190208.132258.930<0.0001
Hospital Outpatient visits686.511607.06250590.491631.05166.67<0.0001
Total emergency room visits (total)58.43137.9051.57119.050<0.0001
   Hospital outpatient ER visits36.55108.59030.6186.910<0.0001
   Hospital inpatient ER visits21.8864.57020.9663.540<0.0001
Physician office visits1652.341811.571166.671307.91820833.33<0.0001
Ambulatory surgery center visits14.4895.1101056.760<0.0001
Dialysis events16.96187.04017.3181.7800.6564
Anesthesia events27.8173.6023.7370.90<0.0001
Imaging events510.5659.82333.33420.07650.69250<0.0001
Test events1762.152384.11916.671428.512127.43666.67<0.0001
Other procedures1464.613748.74250821.562641.483.33<0.0001
Durable medical equipment (DME)369.16714.710301.43692.360<0.0001
Part B Drugs296.11718.58166.67241.84638.5283.33<0.0001
Part D* Drugs5080.113905.364333.334094.333514.213333.33<0.0001
Any Use of Medicare Service During Year (% with any use) N%  N%  p-Value
Total Admissions2389521% 5027320% <0.0001
   Acute stay admissions2361121% 4947320% <0.0001
   Other inpatient admissions14521% 34921% 0.0138
Hospital Readmissions40614% 86803% 0.0327
Medicare home health1532114% 2858211% <0.0001
Medicare hospice6231% 36551% <0.0001
Hospital Outpatient8093272% 16316165% <0.0001
Total emergency room3845134% 7836231% <0.0001
   Hospital outpatient ER2808725% 5519122% <0.0001
   Hospital inpatient ER1835816% 3899316% <0.0001
Physician office10558894% 22443590% <0.0001
Ambulatory surgery center73657% 136115% <0.0001
Dialysis15151% 34351% 0.596
Anesthesia2126319% 3984916% <0.0001
Imaging8721778% 17391370% <0.0001
Test9350883% 19412878% <0.0001
Other procedures7657368% 14642959% <0.0001
Durable medical equipment (DME)5125346% 9126537% <0.0001
Part B Drugs6588059% 13561154% <0.0001
Part D* Drugs10491094% 21825287% <0.0001

 

TABLE D-D.A1.3 (continued)
 Total HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
N% N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to those with no Medicaid managed care106,76495% 227,18691%  
Medicaid Service Utilization per 1000 Member MonthsMeanSDMedianMeanSDMedianp-Value
Personal Care services4512.3710047.1202149.117128.170<0.0001
Residential care38.99682.95063.681164.330<0.0001
DME380.04946.1883.33227.71746.90<0.0001
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)3309.838321.7101840.566675.40<0.0001
Any Use of Medicaid Services During Year(% with any use) N% N% p-Value
Personal Care services30,18628% 32,88514% <0.0001
Residential care1,9312% 2,6961% <0.0001
DME53,92751% 79,59835% <0.0001
Other HCBS services43,92941% 57,81525% <0.0001

 

TABLE D-D.A1.4. Medicare and Beneficiary Health Care Costs
Medicare Payment PMPM($)Total HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
MeanSDMedianMeanSDMedian
Total Admissions PMPM$361$1,326$0$349$1,355$00.0124
   Acute stay admission PMPM$336$1,252$0$320$1,237$00.0002
   Other inpatient admission PMPM$25$282$0$29$385$00.0005
Medicare home health PMPM (Medicare only)$66$267$0$54$238$0<0.0001
Medicare hospice PMPM (Medicare only)$7$139$0$33$353$0<0.0001
Hospital Outpatient PMPM$121$397$29$105$370$16<0.0001
Physician office PMPM$107$134$71$85$137$49<0.0001
Ambulatory surgery center PMPM$5$28$0$4$22$0<0.0001
Dialysis PMPM$3$26$0$3$26$00.9825
Anesthesia PMPM$4$11$0$3$10$0<0.0001
Imaging PMPM$37$61$12$27$50$7<0.0001
Test PMPM$39$66$15$27$49$10<0.0001
Other procedures PMPM$79$172$18$51$141$7<0.0001
Durable medical equipment (DME) PMPM$28$77$0$22$77$0<0.0001
Part B Drugs PMPM$22$235$2$19$183$1<0.0001
Part D* Drugs PMPM$349$437$238$282$398$177<0.0001
Total Medical and Rx PMPM (sum of all above)$1,222$1,924$648$1,054$1,935$463<0.0001
Total Medical PMPM (sum of all above except part D PMPM)$856$1,749$304$764$1,795$196<0.0001

 

TABLE D-D.A1.4 (continued)
 Total HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
N%  N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to those with no Medicaid managed care106,76495% 227,18691%  
Medicaid Payment PMPM--Divide by Total Medicaid MonthsMeanSDMedianMeanSDMedianp-Value
Total Medicaid payment amount$1,180$1,938$227$895$2,017$70<0.0001
Fee-for-service Medicaid payment amount$1,035$1,884$169$771$1,961$54<0.0001
Personal Care services$533$1,336$0$305$1,0990<0.0001
Residential care$23$246$0$40$5060<0.0001
DME$19$77$0$14$700<0.0001
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)$295$1,122$0$195$9760<0.0001

Sub-group B: Medicare-Medicaid Enrollees (MMEs), with >0 and <365 Days of SNF or NF Use in 2008

TABLE D-D.B1. Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
 HUD-AssistedBeneficiaries(full year)(N=12,684)HUD-AssistedBeneficiaries(partial year)(N=3,875)Unassisted Beneficiariesin the Community(N=86,626)Assisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
N%N%N%
* Cell sizes of less than 11 are not displayed.
Demographic Characteristics (available in Medicare data)
Gender<0.0001<0.00010.3205
Male3,49727.6%1,10028.4%28,27532.7%   
Female9,18772.4%2,77571.6%58,35167.4%   
Race (based on RTI variable)<0.0001<0.0001<0.0001
Non-Hispanic White7,11256.1%2,42462.6%59,03168.2%   
Hispanic1,66313.1%3859.9%7,0568.1%   
Black2,94223.2%89523.1%15,80418.3%   
Asian/Pacific Islander6224.9%1173.0%3,4994.0%   
American Indian/Alaska Native130.1%**940.1%   
Other2702.1%441.1%9121.1%   
Unknown620.5%**2300.3%   
Age Group<0.00010.0009<0.0001
Under 18 years******   
18 to 442421.9%782.0%1,6391.9%   
45 to 641,69213.3%47412.2%10,46812.1%   
65 to 742,96623.4%72718.8%14,21716.4%   
75 to 792,33918.4%55114.2%10,67812.3%   
80 to 842,57720.3%68617.7%15,17017.5%   
85+2,86822.6%1,35935.1%34,45439.8%   
Geographic Area<0.0001<0.0001<0.0001
Vermont2552.0%1303.4%2,2012.5%   
New Haven- Milford3332.6%1894.9%3,1433.6%   
Bridgeport- Stamford- Norwalk2542.0%1213.1%2,0182.3%   
Milwaukee- Waukesha- West Allis3472.7%1684.3%2,9663.4%   
San Francisco- Oakland- Fremont8526.7%2015.2%6,1497.1%   
Boston- Cambridge- Quincy1,79614.2%62616.2%9,28510.7%   
Durham- Chapel Hill530.4%330.9%8511.0%   
Richmond180.1%270.7%1,0121.2%   
New York- Northern New Jersey- Long Island7,67560.5%1,73644.8%46,00853.1%   
Columbus3292.6%1884.9%3,1433.6%   
Akron1971.6%1323.4%1,5971.8%   
Cleveland5754.5%3248.4%6,4207.4%   
Missing****1,8332.1%   
Died During the Year7425.8%1,04226.9%15,53317.9%<0.0001<0.0001<0.0001
Medicare Coverage Characteristics
Original Reason for Entitlement<0.00010.00200.1071
Old age and survivor's insurance8,57867.6%2,67469.0%61,43671.0%   
Disability insurance benefits3,92030.9%1,13929.4%24,15727.9%   
ESRD670.5%150.4%3860.4%   
Disability Insurance and ESRD1190.9%471.2%6470.7%   
Current Reason for Entitlement<0.00010.00360.0017
Old age and survivor's insurance10,60283.6%3,30385.2%73,75785.2%   
Disability insurance benefits1,89815.0%50813.1%11,81913.6%   
ESRD970.8%220.6%5310.6%   
Disability Insurance and ESRD870.7%421.1%5190.6%   
Medicare Part D Coverage at Any Point During the Year12,54898.9%3,82698.7%84,93098.1%<0.00010.00310.2573
Average Months of Part D Coverage (among those with Part D cov)11.73 10.69 11.00 <0.0001<0.0001<0.0001
Proportion with Part D Coverage by Cost Share Group Code<0.0001<0.0001<0.0001
Beneficiary deemed with 100% premium-subsidy and no copayment2,64921.1%1,63342.7%55,82265.7%   
Beneficiary deemed with 100% premium-subsidy and low copayment8,48967.7%1,53740.2%19,61123.1%   
Beneficiary deemed with 100% premium-subsidy and high copayment1,0878.7%2967.7%4,0974.8%   
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy680.5%511.3%4990.6%   
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy and 15% copayment****640.1%   
Beneficiary with LIS, 75% premium-subsidy and 15% copayment****770.1%   
Beneficiary with LIS, 50% premium-subsidy and 15% copayment160.1%110.3%540.1%   
Beneficiary with LIS, 25% premium-subsidy and 15% copayment****530.1%   
No premium subsidy nor cost sharing150.1%**1420.2%   
Missing2011.6%2687.0%4,5115.3%   
Medicaid Coverage Status
Average Months of Medicaid Coverage(FFS or managed care)11.40 9.97 10.41 <0.0001<0.0001<0.0001
Proportion Enrolled in Medicaid During Entire Year(FFS or managed care)11,02186.9%2,11754.6%57,81366.7%<0.0001<0.0001<0.0001
In Medicaid Managed Care for at Least One Month1,0598.3%76919.8%21,95625.3%<0.0001<0.0001<0.0001
Average Months of Medicaid Managed Care (among those with managed care)7.87 6.74 8.29 0.0009<0.0001<0.0001
Medicaid Basis for Eligibility<0.00010.4871<0.0001
Aged9,04471.3%3,10880.2%70,62381.5%   
Blind/disabled3,61728.5%76019.6%15,87118.3%   
Child******   
Adult170.1%**910.1%   
Child of Unemployed Adult******   
Unemployed Adult******   
Foster Care Child******   
Covered under Breast and Cervical Cancer Prevention Act******   
Unknown****350.0%   
Missing******   
Medicaid Maintenance of Assistance<0.0001<0.0001<0.0001
Receiving cash or eligible under section 1931 of the Act6,48751.1%98625.4%13,62515.7%   
Medically needy2,47119.5%76419.7%32,90838.0%   
Poverty related(includes children eligible under S-CHIP expansion)1,66213.1%53013.7%7,8219.0%   
Other1,88914.9%1,54839.9%30,95035.7%   
Foster Care Child******   
1115 demonstration expansion eligible1701.3%441.1%1,2831.5%   
Unknown****350.0%   
Missing******   
Proportion with 1915c waiver status code2,32918.4%70418.2%7,0368.1%<0.0001<0.00010.7865
Type of 1915c Waiver Status Code<0.0001<0.00010.0441
Unknown--missing eligibility****350.0%   
Aged and disabled1,38310.9%43111.1%4,3215.0%   
Aged7566.0%2275.9%1,6781.9%   
Disabled520.4%250.6%1460.2%   
Brain injured****750.1%   
HIV-AIDS110.1%**320.0%   
MR/DD1221.0%180.5%7840.9%   
Mentally ill/Severely emotionally disturbed******   
Technology-dependent/ medically fragile******   
Never enrolled in a 1915c waiver during the year10,35281.6%3,16881.8%79,55591.8%   
Average Months of Both Medicare & Medicaid Enrollment11.360.1%9.890.3%10.240.0%<0.0001<0.0001<0.0001
Proportion Enrolled in Medicaid & Medicare During Entire Year10,96586.4%2,09754.1%56,86365.6%<0.0001<0.0001<0.0001
Medicare & Medicaid Eligibility Status<0.0001<0.0001<0.0001
QMB only2311.8%661.7%8941.0%   
QMB plus8,95470.6%2,25658.2%36,67542.3%   
SLMB only2812.2%671.7%9091.0%   
SLMB Plus3893.1%1874.8%2,9023.4%   
Other dual2,81122.2%1,28133.1%44,41851.3%   
Unknown****350.0%   
NA150.1%150.4%7930.9%   
Average SNF days30.51 44.57 30.86 0.2681<0.0001<0.0001
Average NF days57.4 93.11 151.91 <0.0001<0.0001<0.0001
Average SNF stays1.13 1.43 1 <0.0001<0.0001<0.0001

 

TABLE D-D.B2. Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
CCW ConditionHUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Unassisted Beneficiariesin the CommunityAssisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
N%N%N%
Prevalence of Chronic Conditions<0.0001<0.00010.4779
0 Conditions540.4%160.4%4720.5%   
1 Condition980.8%270.7%11661.3%   
2 Conditions1991.6%751.9%23472.7%   
3 Conditions4363.4%1263.3%44065.1%   
4 Conditions7025.5%1935.0%67787.8%   
5 or More Conditions1119588.3%343888.7%7145782.5%   
Prevalence of Chronic Conditions by Category
Cardiovascular1035881.7%321583.0%6921879.9%<0.0001<0.00010.0596
Cancer154512.2%54814.1%867610.0%<0.0001<0.00010.0009
Endocrine and Renal946174.6%288074.3%5983269.1%<0.0001<0.00010.7701
Alzheimer's-related508340.1%223757.7%5878567.9%<0.0001<0.0001<0.0001
Depression526841.5%186548.1%3698942.7%0.0095<0.0001<0.0001
Musculoskeletal862468.0%230959.6%4369150.4%<0.0001<0.0001<0.0001
Pulmonary479537.8%154839.9%2796432.3%<0.0001<0.00010.0171
Opthalmic474737.4%119230.8%2998934.6%<0.0001<0.0001<0.0001
Other1230397.0%373396.3%8044292.9%<0.0001<0.00010.0413

 

TABLE D-D.B3. Medicare Health Care Utilization per 1000 Member Months
Medicare Services Annual Utilization per 1000 Member MonthsHUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Unassisted Beneficiariesin the CommunityAssisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
MeanSDMedianMeanSDMedianMeanSDMedian
Total Admissions191.12201.61166.67234.26218.67166.67161.64188.7383.33<0.0001<0.0001<0.0001
   Acute stay admissions178.02190.6783.33216.72205.65166.67148.95176.7683.33<0.0001<0.0001<0.0001
   Other inpatient admissions13.144.52017.5457.79012.6948.4200.377<0.0001<0.0001
Hospital Readmissions53.59122.6070.69136.01041.71106.560<0.0001<0.0001<0.0001
Medicare Skilled Nursing Facility (SNF) days2583.192786.261666.674058.463515.163333.332717.573257.261416.67<0.0001<0.0001<0.0001
Medicare Skilled Nursing Facility (SNF) stays96.09100.0183.33135.27123.3783.3390.03109.6783.33<0.0001<0.0001<0.0001
Medicare home health visits2207.54764.78333.331454.133515.170783.792785.920<0.0001<0.0001<0.0001
Medicare hospice days98.851086.970601.592778.440782.593498.680<0.00010.0015<0.0001
Hospital Outpatient visits1504.462828.445002041.023253.52833.331850.873162.88666.67<0.00010.0004<0.0001
Total emergency room visits (total)218.65277.36166.67280.44293.73222.22193.29257.65125<0.0001<0.0001<0.0001
   Hospital outpatient ER visits84.17179.39098.72198.81070.25168.650<0.0001<0.0001<0.0001
   Hospital inpatient ER visits134.48163.9683.33181.72184.48166.67123.04156.4983.33<0.0001<0.0001<0.0001
Physician office visits4970.814732.243666.676036.065729.764272.734733.544860.063250<0.0001<0.0001<0.0001
Ambulatory surgery center visits12.7484.0607.1348.4605.1450.230<0.00010.0146<0.0001
Dialysis events108.07585.110124.47650.27081.31501.780<0.0001<0.00010.1394
Anesthesia events85.7151.78077.65152.39059.68144.390<0.0001<0.00010.0035
Imaging events1514.961404.61166.671713.881738.6512501193.141373.51833.33<0.0001<0.0001<0.0001
Test events2866.483091.721833.332689.742991.271714.292540.63115.871500<0.00010.00350.0015
Other procedures1975.764114.62583.33976.462317.45428.57785.821852.82416.67<0.0001<0.0001<0.0001
Durable medical equipment (DME)774.51063.86333.33579.32952.52166.67411.25850.040<0.0001<0.0001<0.0001
Part B Drugs458.82887.35250385.49624.7166.67345.64606.23166.67<0.0001<0.0001<0.0001
Part D* Drugs6574.834191.015916.676372.834271.1257506563.334229.445916.670.89780.01090.0148
Any Use of Medicare Service During Year(% with any use)MeanSDMedianMeanSDMedianMeanSDMedian   
Total Admissions10,17580% 3,34786% 62,18372% <0.0001<0.0001<0.0001
   Acute stay admissions10,04679% 3,27785% 60,44670% <0.0001<0.0001<0.0001
   Other inpatient admissions1,40911% 49713% 8,35210% <0.0001<0.00010.0023
Hospital Readmissions3,81330% 1,38736% 20,37624% <0.0001<0.0001<0.0001
Medicare home health6,72653% 1,60942% 19,92123% <0.0001<0.0001<0.0001
Medicare hospice2802% 45112% 8,90010% <0.00010.0066<0.0001
Hospital Outpatient10,75285% 3,48790% 75,50887% <0.0001<0.0001<0.0001
Total emergency room9,74677% 3,38387% 63,47473% <0.0001<0.0001<0.0001
   Hospital outpatient ER5,61044% 1,92850% 33,59239% <0.0001<0.0001<0.0001
   Hospital inpatient ER8,60868% 3,06379% 55,29864% <0.0001<0.0001<0.0001
Physician office12,629100% 3,870100% 85,81899% <0.0001<0.00010.005
Ambulatory surgery center7066% 1414% 2,2803% <0.0001<0.0001<0.0001
Dialysis7066% 2496% 3,6924% <0.0001<0.00010.0453
Anesthesia5,83346% 1,44237% 26,23030% <0.0001<0.0001<0.0001
Imaging12,25097% 3,73096% 77,49489% <0.0001<0.00010.324
Test12,26397% 3,73896% 78,81291% <0.0001<0.00010.4747
Other procedures11,54591% 3,49390% 76,41988% <0.00010.00020.1301
Durable medical equipment (DME)9,08472% 2,19157% 37,61643% <0.0001<0.0001<0.0001
Part B Drugs9,46975% 2,80472% 60,09469% <0.0001<0.00010.004
Part D* Drugs12,02096% 3,42890% 76,21190% <0.00010.50490.0014

 

TABLE D-D.B3 (continued)
 HUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Unassisted Beneficiariesin the CommunityAssisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
N% N% N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to those with no Medicaid managed care11,62092% 3,10380% 64,63175%    
Medicaid Service Utilization per 1000 Member Months MeanSDMedianMeanSDMedianMeanSDMedian   
Personal Care services5222.889443.5702190.986237.2801078.94657.870<0.0001<0.0001<0.0001
Residential care75.611047.980225.51784.330203.331765.910<0.00010.3818<0.0001
DME685.341306.46250518.961289.683.33327.78887.810<0.0001<0.0001<0.0001
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)6705.7610444.021916.674139.238454.95888.893019.37735.14416.67<0.0001<0.0001<0.0001
Nursing facility days per member month5094.478601.3708624.1710301.082916.6713860.5712766.8814500<0.0001<0.0001<0.0001
Other long-term days (excluding NF)10.51447.33045.31221.11090.851653.290<0.00010.11640.0249
Any Use of Medicaid Service During Year (% with any use)N%  N%  N%        
Personal Care services4,44438% 71523% 6,44510% <0.0001<0.0001<0.0001
Residential care3133% 1445% 2,3444% <0.00010.0008<0.0001
DME7,81467% 1,64953% 26,89742% <0.0001<0.0001<0.0001
Other HCBS services9,48982% 2,42578% 44,92470% <0.0001<0.0001<0.0001

 

TABLE D-D.B4. Medicare and Beneficiary Health Care Cost
Medicare Payment PMPM($)HUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Unassisted Beneficiariesin the CommunityAssisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
MeanSDMedianMeanSDMedianMeanSDMedian
Total Admissions PMPM$2,527$3,652$1,331$2,916$4,361$1,483$2,027$3,584$858<0.0001<0.0001<0.0001
   Acute stay admission PMPM$2,256$3,351$1,211$2,601$4,095$1,310$1,800$3,278$762<0.0001<0.0001<0.0001
   Other inpatient admission PMPM$271$1,037$0$316$1,137$0$228$1,034$0<0.0001<0.00010.0176
Medicare Skilled Nursing Facility (SNF) PMPM$1,059$1,104$779$1,497$1,336$1,260$991$1,197$587<0.0001<0.0001<0.0001
Medicare home health PMPM (Medicare only)$273$423$51$192$349$0$110$294$0<0.0001<0.0001<0.0001
Medicare hospice PMPM (Medicare only)$19$193$0$111$482$0$128$555$0<0.00010.0684<0.0001
Hospital Outpatient PMPM$268$619$75$297$610$107$236$545$72<0.0001<0.00010.0116
Physician office PMPM$362$383$254$440$472$289$329$390$213<0.0001<0.0001<0.0001
Ambulatory surgery center PMPM$5$29$0$3$20$0$2$17$0<0.00010.00020.0006
Dialysis PMPM$13$60$0$13$58$0$9$51$0<0.0001<0.00010.8861
Anesthesia PMPM$15$24$0$12$25$0$9$21$0<0.0001<0.0001<0.0001
Imaging PMPM$62$73$39$47$55$30$33$46$18<0.0001<0.0001<0.0001
Test PMPM$57$78$29$41$56$23$34$48$19<0.0001<0.0001<0.0001
Other procedures PMPM$145$214$76$96$173$33$72$150$21<0.0001<0.0001<0.0001
Durable medical equipment (DME) PMPM$71$154$18$58$160$7$38$128$0<0.0001<0.0001<0.0001
Part B Drugs PMPM$34$252$4$23$110$5$21$123$5<0.00010.35580.0096
Part D* Drugs PMPM$485$607$339$414$468$300$435$466$320<0.00010.0108<0.0001
Total Medical & Rx PMPM (sum of all above)$5,378$4,906$4,008$6,134$5,599$4,653$4,444$4,806$3,143<0.0001<0.0001<0.0001
Total Medical PMPM(sum of all above except Part D PMPM)$4,873$4,827$3,530$5,724$5,572$4,244$4,019$4,768$2,721<0.0001<0.0001<0.0001

 

TABLE D-D.B4 (continued)
 HUD-AssistedBeneficiaries(full year)HUD-AssistedBeneficiaries(partial year)Unassisted Beneficiariesin the CommunityAssisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-Value
N%  N%  N% 
* Limited to those who have Part D coverage for entire year or up until death.
Limited to those with no Medicaid managed care11,62092% 3,10380% 64,63175%    
Medicaid Payment PMPM--Divide by Total Medicaid Months MeanSDMedianMeanSDMedianMeanSDMedian     
Total Medicaid payment amount2789.892763.942174.463053.572428.682733.74202.643244.144129.33<0.0001<0.0001<0.0001
Fee-for-service Medicaid payment amount2653.552783.661914.632947.462466.852595.144137.063278.994081.42<0.0001<0.0001<0.0001
Personal care services575.751174.360218.14651.210138.39645.690<0.0001<0.0001<0.0001
Residential care36.19359.06052.2364.89074.07616.50<0.00010.05970.0094
DME42.68132.085.2539.67139.21.7529.53114.80<0.0001<0.00010.0028
Other HCBS services (*private duty nursing, adult day care, home health, rehab, targeted case management, transportation, and hospice)602.781490.4389.08299.18906.2134256.62896.2318.5<0.0001<0.0001<0.0001
Nursing facility989.461772.41921954.432269.351096.913305.683089.313156.83<0.0001<0.0001<0.0001
Other long-term days0.010.2600000.010.480<0.00010.11640.0249

Sample C: Medicare Beneficiaries with No Medicaid Enrollment, Regardless of SNF Use in 2008

TABLE D-D.C1. Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
 HUD-AssistedBeneficiaries(full year)(N=45,566)HUD-AssistedBeneficiaries(partial year)(N=5,406)Total HUD-Assisted Beneficiaries(N=50,972)Unassisted Beneficiariesin the Community(N=2,453,763)Assisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-ValueTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Demographic Characteristics (available in Medicare data)
Gender<0.0001<0.0001<0.0001<0.0001
Male12,74528.0%1,67230.9%14,41728.3%1,046,13642.6%    
Female32,82172.0%3,73469.1%36,55571.7%1,407,62757.4%    
Race (based on RTI variable)<0.0001<0.0001<0.0001<0.0001
Non-Hispanic White29,65565.1%3,83370.9%33,48865.7%2,148,12387.5%    
Hispanic4,2119.2%3716.9%4,5829.0%77,1043.1%    
Black10,87123.9%1,10820.5%11,97923.5%160,1166.5%    
Asian/Pacific Islander5291.2%601.1%5891.2%46,3671.9%    
American Indian/Alaska Native240.1%**280.1%7270.0%    
Other2350.5%300.6%2650.5%19,3750.8%    
Unknown410.1%**410.1%1,9510.1%    
Age Group<0.0001<0.0001<0.0001<0.0001
65 to 697,38616.2%1,08120.0%8,46716.6%539,61422.0%    
70 to 748,74219.2%1,04819.4%9,79019.2%560,48522.8%    
75 to 798,84219.4%1,00218.5%9,84419.3%490,65920.0%    
80 to 849,39520.6%99918.5%10,39420.4%431,96917.6%    
85+11,20124.6%1,27623.6%12,47724.5%431,03617.6%    
Geographic Area<0.0001<0.0001<0.0001<0.0001
Vermont5741.3%1142.1%6881.3%55,9622.3%    
New Haven- Milford1,5453.4%2975.5%1,8423.6%67,3972.7%    
Bridgeport- Stamford- Norwalk1,1872.6%1552.9%1,3422.6%71,2402.9%    
Milwaukee- Waukesha- West Allis1,0622.3%2013.7%1,2632.5%108,7164.4%    
San Francisco- Oakland- Fremont1,2622.8%1312.4%1,3932.7%165,8246.8%    
Boston- Cambridge- Quincy9,74621.4%1,10720.5%10,85321.3%309,45412.6%    
Durham- Chapel Hill2330.5%400.7%2730.5%31,8511.3%    
Richmond2450.5%450.8%2900.6%71,6972.9%    
New York- Northern New Jersey- Long Island26,61858.4%2,76351.1%29,38157.6%1,194,28448.7%    
Columbus9752.1%2053.8%1,1802.3%94,5773.9%    
Akron4961.1%871.6%5831.1%48,0372.0%    
Cleveland1,6233.6%2614.8%1,8843.7%164,6896.7%    
Missing******70,0352.9%    
Died During the Year4501.0%77814.4%1,2282.4%57,5132.3%<0.0001<0.0001<0.0001<0.0001
Medicare Coverage Characteristics
Original Reason for Entitlement<0.0001<0.0001<0.0001<0.0001
Old age and survivor's insurance38,74785.0%4,66586.3%43,41285.2%2,306,85694.0%    
Disability insurance benefits6,75414.8%73113.5%7,48514.7%144,3515.9%    
ESRD260.1%**320.1%1,4420.1%    
Disability Insurance and ESRD390.1%**430.1%1,1140.0%    
Current Reason for Entitlement<0.00010.0613<0.0001<0.0001
Old age and survivor's insurance44,92598.6%5,36799.3%50,29298.7%2,435,99899.3%    
Disability insurance benefits5691.2%290.5%5981.2%15,3640.6%    
ESRD630.1%**730.1%2,1220.1%    
Disability Insurance and ESRD******2790.0%    
Medicare Part D Coverage at Any Point During the Year25,42155.8%2,82452.2%28,24555.4%922,02137.6%<0.0001<0.0001<0.0001<0.0001
Average Months of Part D Coverage (among those with Part D cov)11.79 11.15 11.72 11.75 <0.0001<0.0001<0.00010.0011
Proportion with Part D Coverage by Cost Share Group Code<0.0001<0.0001<0.0001<0.0001
Beneficiary deemed with 100% premium-subsidy and no copayment130.1%**210.1%3400.0%    
Beneficiary deemed with 100% premium-subsidy and low copayment3291.3%551.9%3841.4%2,3950.3%    
Beneficiary deemed with 100% premium-subsidy and high copayment4611.8%401.4%5011.8%3,1530.3%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy4,88519.2%46916.6%5,35419.0%34,1653.7%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy and 15% copayment5202.0%592.1%5792.0%4,2120.5%    
Beneficiary with LIS, 75% premium-subsidy and 15% copayment7322.9%632.2%7952.8%3,6650.4%    
Beneficiary with LIS, 50% premium-subsidy and 15% copayment7002.8%431.5%7432.6%3,5070.4%    
Beneficiary with LIS, 25% premium-subsidy and 15% copayment5352.1%441.6%5792.0%2,9710.3%    
No premium subsidy nor cost sharing16,26564.0%1,60256.7%17,86763.3%818,19188.7%    
Missing9813.9%44115.6%1,4225.0%49,4225.4%    

 

TABLE D-D.C2. Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
CCW ConditionTotal HUD-Assisted BeneficiariesUnassisted Beneficiaries in the CommunityTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%
Prevalence of Chronic Conditions<0.0001
0 Conditions6,62513.0%300,02012.2% 
1 Condition3,9857.8%250,43010.2% 
2 Conditions5,29310.4%314,21912.8% 
3 Conditions6,60313.0%362,62414.8% 
4 Conditions6,93113.6%348,24714.2% 
5 or More Conditions21,53542.2%878,22335.8% 
Prevalence of Chronic Conditions by Category
Cardiovascular24,05847.2%1,061,55143.3%<0.0001
Cancer4,8539.5%263,76010.7%<0.0001
Endocrine and Renal22,25743.7%891,22736.3%<0.0001
Alzheimer's-related4,9359.7%234,6179.6%0.3695
Depression5,28110.4%213,3458.7%<0.0001
Musculoskeletal18,21235.7%808,81833.0%<0.0001
Pulmonary8,39216.5%291,71011.9%<0.0001
Opthalmic16,47332.3%891,80936.3%<0.0001
Other39,12776.8%1,828,46974.5%<0.0001

 

TABLE D-D.C3. Medicare Health Care Utilization per 1000 Member Months
Medicare Services Annual Utilization per 1000 Member MonthsTotal HUD-Assisted BeneficiariesUnassisted Beneficiaries in the CommunityTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
MeanSDMedianMeanSDMedian
* Limited to those who have Part D coverage for entire year or up until death.
Total Admissions37.4391.53028.9178.680<0.0001
   Acute stay admissions35.3686.36027.2573.810<0.0001
   Other inpatient admissions2.0716.6101.6614.410<0.0001
Hospital Readmissions6.4337.6304.6931.230<0.0001
Medicare home health visits368.541710.970221.311321.70<0.0001
Medicare hospice days50.59879.67091.471285.450<0.0001
Hospital Outpatient visits490.011233.45166.67449.421150.54166.67<0.0001
Total emergency room visits (total)57.01116.67039.5692.820<0.0001
   Hospital outpatient ER visits30.3775.88021.159.090<0.0001
   Hospital inpatient ER visits26.6471.65018.4658.570<0.0001
Physician office visits1430.192002.1916.671331.891903.06833.33<0.0001
Ambulatory surgery center visits12.6357.77015.964.730<0.0001
Dialysis events10.0713606.91122.210<0.0001
Anesthesia events31.3583.99033.0584.170<0.0001
Imaging events471.2696250447.59671.91250<0.0001
Test events1427.722001.967501474.862011.49833.33<0.0001
Other procedures566.221493.82166.67667.791654.15166.67<0.0001
Durable medical equipment (DME)204.33514.420159.36473.40<0.0001
Part B Drugs217.97549.0583.33251.43613.44166.67<0.0001
Part D* Drugs3075.982489.132583.332590.612214.882083.33<0.0001
Any Use of Medicare Service During Year (% with any use) N%  N%  p-Value
Total Admissions11,87623% 467,05919% <0.0001
   Acute stay admissions11,74023% 462,48919% <0.0001
   Other inpatient admissions9922% 39,4162% <0.0001
Hospital Readmissions2,3405% 84,8633% <0.0001
Medicare home health7,32314% 230,8629% <0.0001
Medicare hospice5041% 32,0121% <0.0001
Hospital Outpatient33,85666% 1,605,64965% <0.0001
Total emergency room17,44834% 645,19526% <0.0001
   Hospital outpatient ER11,64123% 429,52318% <0.0001
   Hospital inpatient ER9,70519% 339,23314% <0.0001
Physician office45,02788% 2,253,63892% <0.0001
Ambulatory surgery center3,7347% 233,62010% <0.0001
Dialysis4191% 14,0971% <0.0001
Anesthesia10,08220% 527,05421% <0.0001
Imaging36,06771% 1,794,38573% <0.0001
Test39,90578% 2,025,52383% <0.0001
Other procedures31,82662% 1,663,03668% <0.0001
Durable medical equipment (DME)14,95329% 582,13024% <0.0001
Part B Drugs28,35856% 1,590,17765% <0.0001
Part D* Drugs24,91988% 836,93291% <0.0001

 

TABLE D-D.C4. Medicare and Beneficiary Health Care Costs
Medicare Payment PMPM($)HUD-Assisted BeneficiariesUnassisted Beneficiaries in the CommunityTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
MeanSDMedianMeanSDMedian
Total Admissions PMPM$379$1,259$0$313$1,199$0<0.0001
Acute stay admission PMPM$343$1,116$0$283$1,074$0<0.0001
Other inpatient admission PMPM$36$346$0$30$317$0<0.0001
Medicare home health PMPM (Medicare only)$55$189$0$38$167$0<0.0001
Medicare hospice PMPM (Medicare only)$10$154$0$16$210$0<0.0001
Hospital Outpatient PMPM$91$303$15$86$311$110.0007
Physician office PMPM$96$157$53$86$146$48<0.0001
Ambulatory surgery center PMPM$5$22$0$6$26$0<0.0001
Dialysis PMPM$2$19$0$1$15$0<0.0001
Anesthesia PMPM$4$12$0$4$12$00.0007
Imaging PMPM$26$43$9$27$47$9<0.0001
Test PMPM$25$41$11$27$43$14<0.0001
Other procedures PMPM$47$114$10$55$130$14<0.0001
Durable medical equipment (DME) PMPM$16$61$0$13$55$0<0.0001
Part B Drugs PMPM$21$194$2$28$220$3<0.0001
Part D* Drugs PMPM$137$297$97$105$208$84<0.0001
Total Medical and Rx PMPM (sum of all above)$953$1,959$282$808$1,830$226<0.0001
Total Medical PMPM (sum of all above except part D PMPM)$859$1,902$197$741$1,778$176<0.0001

Sample C1: Medicare Beneficiaries with No Medicaid Enrollment, Age 65+, No SNF Use in 2008

TABLE D-D.C1.1. Demographic and Medicare Health Insurance Characteristics of HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries 2008
 HUD-AssistedBeneficiaries(full year)(N=42,286)HUD-AssistedBeneficiaries(partial year)(N=4,520)Total HUD-Assisted Beneficiaries(N=46,806)Unassisted Beneficiariesin the Community(N=2,318,394)Assisted FullYear vs. Unassistedp-ValueAssisted PartialYear vs. Unassistedp-ValueAssisted Full Year vs.Assisted Partial Yearp-ValueTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%N%N%
* Cell sizes of less than 11 are not displayed.
Demographic Characteristics (available in Medicare data)
Gender<0.0001<0.0001<0.0001<0.0001
Male11,98628.3%1,44231.9%13,42828.7%997,87743.0%    
Female30,30071.7%3,07868.1%33,37871.3%1,320,51757.0%    
Race (based on RTI variable)<0.00010.0043<0.0001<0.0001
Non-hispanic white27,03463.9%3,08168.2%30,11564.3%2,023,13087.3%    
Hispanic4,1089.7%3527.8%4,4609.5%75,0903.2%    
Black10,34124.5%99722.1%11,33824.2%153,3396.6%    
Asian/pacific Islander5141.2%561.2%5701.2%45,3542.0%    
American Indian/Alaska Native240.1%**280.1%6790.0%    
Other2280.5%300.7%2580.6%18,9940.8%    
Unknown370.1%**370.1%1,8080.1%    
Age Group<0.0001<0.0001<0.0001<0.0001
65 to 697,17317.0%1,03422.9%8,20717.5%531,47222.9%    
70 to 748,37619.8%97121.5%9,34720.0%546,59923.6%    
75 to 798,31519.7%87719.4%9,19219.6%468,47820.2%    
80 to 848,61820.4%79717.6%9,41520.1%398,83617.2%    
85+9,80423.2%84118.6%10,64522.7%373,00916.1%    
Geographic Area<0.0001<0.0001<0.0001<0.0001
Vermont5471.3%982.2%6451.4%53,9542.3%    
New Haven- Milford1,4043.3%2605.8%1,6643.6%62,4652.7%    
Bridgeport- Stamford- Norwalk1,0872.6%1312.9%1,2182.6%66,8822.9%    
Milwaukee- Waukesha- West Allis9922.3%1743.8%1,1662.5%102,7954.4%    
San Francisco- Oakland- Fremont1,2142.9%1112.5%1,3252.8%158,5236.8%    
Boston- Cambridge- Quincy8,82620.9%87319.3%9,69920.7%289,11412.5%    
Durham- Chapel Hill2200.5%360.8%2560.5%30,5851.3%    
Richmond2270.5%400.9%2670.6%69,3163.0%    
New York- Northern New Jersey- Long Island24,93859.0%2,33151.6%27,26958.3%1,132,74748.9%    
Columbus8852.1%1703.8%1,0552.3%89,0723.8%    
Akron4591.1%731.6%5321.1%45,1691.9%    
Cleveland1,4873.5%2234.9%1,7103.7%153,5536.6%    
Missing******64,2192.8%    
Died During the Year2770.7%4109.1%6871.5%37,4241.6%<0.0001<0.0001<0.0001<0.0001
Medicare Coverage Characteristics
Original Reason for Entitlement<0.0001<0.00010.1714<0.0001
Old age and survivor's insurance35,94185.0%3,88886.0%39,82985.1%2,181,22894.1%    
Disability insurance benefits6,29414.9%62713.9%6,92114.8%134,8565.8%    
ESRD240.1%**280.1%1,3160.1%    
Disability Insurance and ESRD270.1%**280.1%9940.0%    
Current Reason for Entitlement<0.00010.80830.0015<0.0001
Old age and survivor's insurance41,67498.6%4,48699.2%46,16098.6%2,301,30299.3%    
Disability insurance benefits5551.3%290.6%5841.2%14,9600.6%    
ESRD520.1%**570.1%1,8830.1%    
Disability Insurance and ESRD******2490.0%    
Medicare Part D Coverage at Any Point During the Year23,44755.4%2,33551.7%25,78255.1%866,62937.4%<0.0001<0.0001<0.00010.0031
Average Months of Part D Coverage (among those with Part D cov)11.79 11.35 11.75 11.77 0.0223<0.0001<0.0001<0.0001
Proportion with Part D Coverage by Cost Share Group Code<0.0001<0.0001<0.0001<0.0001
Beneficiary deemed with 100% premium-subsidy and no copayment110.0%**170.1%2480.0%    
Beneficiary deemed with 100% premium-subsidy and low copayment3161.3%512.2%3671.4%2,2630.3%    
Beneficiary deemed with 100% premium-subsidy and high copayment4381.9%381.6%4761.8%3,0190.3%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy4,60419.6%41917.9%5,02319.5%32,4253.7%    
Beneficiary with Low Income Subsidy (LIS), 100% premium-subsidy and 15% copayment4862.1%421.8%5282.0%3,9170.5%    
Beneficiary with LIS, 75% premium-subsidy and 15% copayment6872.9%532.3%7402.9%3,4110.4%    
Beneficiary with LIS, 50% premium-subsidy and 15% copayment6372.7%371.6%6742.6%3,2680.4%    
Beneficiary with LIS, 25% premium-subsidy and 15% copayment4972.1%411.8%5382.1%2,7500.3%    
No premium subsidy nor cost sharing14,90263.6%1,37759.0%16,27963.1%774,36389.4%    
Missing8693.7%27111.6%1,1404.4%40,9654.7%    

 

TABLE D-D.C1.2. Prevalence of CCW Conditions Among HUD-Assisted Beneficiaries vs. Unassisted Beneficiaries
CCW ConditionTotal HUD-Assisted BeneficiariesUnassisted Beneficiaries in the CommunityTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
N%N%
Prevalence of Chronic Conditions<0.0001
0 Conditions6,62414.2%299,93612.9% 
1 Condition3,9728.5%249,96210.8% 
2 Conditions5,24511.2%312,46213.5% 
3 Conditions6,47213.8%358,21915.5% 
4 Conditions6,66914.2%339,58614.6% 
5 or More Conditions17,82438.1%758,22932.7% 
Prevalence of Chronic Conditions by Category
Cardiovascular20,63844.1%952,06141.1%<0.0001
Cancer4,1058.8%238,41610.3%<0.0001
Endocrine and Renal19,31341.3%800,92434.5%<0.0001
Alzheimer's-related3,5437.6%179,3047.7%0.1825
Depression3,9308.4%167,1197.2%<0.0001
Musculoskeletal15,49433.1%719,06531.0%<0.0001
Pulmonary6,77714.5%246,61910.6%<0.0001
Opthalmic15,09832.3%847,51236.6%<0.0001
Other35,03074.8%1,696,31973.2%<0.0001

 

TABLE D-D.C1.3. Medicare Health Care Utilization per 1000 Member Months
Medicare Services Annual Utilization per 1000 Member MonthsHUD-Assisted BeneficiariesUnassisted Beneficiaries in the CommunityTotal HUD-Assistedvs. Unassisted Beneficiariesp-Value
MeanSDMedianMeanSDMedian
* Limited to those who have Part D coverage for entire year or up until death.
Total Admissions2365.83019.2758.890<0.0001
   Acute stay admissions21.6962.11018.1755.280<0.0001
   Other inpatient admissions1.3113.4501.0911.520<0.0001
Hospital Readmissions2.6923.4202.2920.590<0.0001
Medicare home health visits220.491433.710135.961100.290<0.0001
Medicare hospice days36.49778.4077.551229.740<0.0001
Hospital Outpatient visits444.771140.2166.67409.471069.6483.33<0.0001
Total emergency room visits (total)43.0494.73030.3874.760<0.0001
   Hospital outpatient ER visits27.2171.16018.8354.530<0.0001
   Hospital inpatient ER visits15.8349.15011.5541.50<0.0001
Physician office visits1145.161406.1833.331129.111378.647500.0128
Ambulatory surgery center visits12.6557.14016.0164.450<0.0001
Dialysis events6.5196.9204.7794.330<0.0001
Anesthesia events25.8774.98029.2277.380<0.0001
Imaging events378.19530.96250382.24544.222500.1145
Test events1346.961938.25666.671416.921952.61833.33<0.0001
Other procedures527.231438.3883.33635.381616.68166.67<0.0001
Durable medical equipment (DME)180.65478.620139.02435.640<0.0001
Part B Drugs205.65536.0783.33242.2597.12166.67<0.0001
Part D* Drugs2983.912461.0925002495.972143.142000<0.0001
Any Use of Medicare Service During Year (% with any use) N%  N%   
Total Admissions7,84917% 336,50915% <0.0001
   Acute stay admissions7,73017% 332,64414% <0.0001
   Other inpatient admissions5881% 25,3271% <0.0001
Hospital Readmissions1,0052% 43,5272% <0.0001
Medicare home health4,4229% 145,9746% <0.0001
Medicare hospice2741% 21,7421% <0.0001
Hospital Outpatient30,20365% 1,485,54964% <0.0001
Total emergency room13,76829% 529,30823% <0.0001
   Hospital outpatient ER9,88321% 374,63416% <0.0001
   Hospital inpatient ER6,27213% 232,23410% <0.0001
Physician office40,86487% 2,118,41691% <0.0001
Ambulatory surgery center3,4407% 223,23210% <0.0001
Dialysis2851% 10,2460% <0.0001
Anesthesia7,98017% 455,29820% <0.0001
Imaging31,95268% 1,660,48072% <0.0001
Test35,81877% 1,893,01482% <0.0001
Other procedures28,04860% 1,537,44866% <0.0001
Durable medical equipment (DME)12,37026% 496,46821% <0.0001
Part B Drugs25,26054% 1,484,60964% <0.0001
Part D* Drugs22,60488% 784,92491% <0.0001

 

TABLE D-D.C1.4. Medicare and Beneficiary Health Care Costs
Medicare Payment PMPM($)HUD-Assisted BeneficiariesUnassisted Beneficiaries in the Communityp-Value
MeanSDMedianMeanSDMedian
Total Admissions PMPM$219$875$0$199$886$0<0.0001
Acute stay admission PMPM$196$760$0$181$801$0<0.0001
Other inpatient admission PMPM$22$280$0$18$229$0<0.0001
Medicare home health PMPM (Medicare only)$32$145$0$23$130$0<0.0001
Medicare hospice PMPM (Medicare only)$7$135$0$13$199$0<0.0001
Hospital Outpatient PMPM$82$283$12$80$299$100.1331
Physician office PMPM$74$105$46$71$101$45<0.0001
Ambulatory surgery center PMPM$5$23$0$6$26$0<0.0001
Dialysis PMPM$1$16$0$1$13$0<0.0001
Anesthesia PMPM$3$10$0$4$11$0<0.0001
Imaging PMPM$24$42$7$25$46$8<0.0001
Test PMPM$24$40$10$27$43$13<0.0001
Other procedures PMPM$41$106$8$50$126$12<0.0001
Durable medical equipment (DME) PMPM$14$57$0$11$49$0<0.0001
Part B Drugs PMPM$20$195$1$27$218$3<0.0001
Part D* Drugs PMPM$135$303$94$102$205$81<0.0001
Total Medical and Rx PMPM (sum of all above)$617$1,263$240$574$1,284$205<0.0001
Total Medical PMPM (sum of all above except part D PMPM)$525$1,181$163$510$1,218$1570.0057

Part D LIS Enrollment

TABLE D1. Proportion of Beneficiaries with Part D Coverage, Enrolled in LIS
 HUD-AssistedUnassisted inthe Community
Subgroup A : Medicare-Medicaid benes, no SNF or NF days99%98%
   Subgroup  A1 : Age 65+99%98%
Subgroup  B : Medicare-Medicaid benes with >0 and <365 SNF/NF days98%95%
Subgroup  C : Medicare only, 65+32%6%
   Subgroup  C1 : Medicare only, 65+, No SNF days32%6%

 

NOTES

  1. U.S. Department of Housing and Urban Development.(2013). Worst Case Housing Needs 2011: Report to Congress. Washington, D.C. Available at: http://www.huduser.org/portal/publications/affhsg/wc_HsgNeeds11_report.html.

  2. Summit on Aging in Place in Public Housing. (2011). Hosted by Enterprise Community Partners, Inc., LeadingAge. Supported by the Atlantic Philanthropies.

  3. Gibler, K. (2003). Aging Subsidized Housing Residents: A Growing Problem in U.S. Cities. Journal of Real Estate Research 25, (4) 395-420.

  4. Long-term services and supports include institutional (e.g., nursing facility) and home and community-based services (HCBS) (e.g., personal care to assist with bathing, dressing, eating and other ADLs in an individual's home or residential group setting, homemaker services, and transportation).

  5. Assisted households are categorized by property type in which they live, despite minor variations in subsidy type. "Housing choice vouchers" include all types of buildings in which housing choice voucher holders live, from a single-family home to a large apartment complex.

  6. For this report, "elderly properties" include properties designated as elderly by PHAs and properties in which 50 percent or more of households had an individual aged 62 years or older.

  7. Before calculating the match rates, we excluded all Medicare or Medicaid beneficiaries that were duplicates in the CMS administrative data. We excluded these individuals from both the CMS administrative data and the HUD data, for those that were also individuals receiving HUD-assisted housing. This way, the rates were accurate and did not overestimate the rates due to an individual being counted multiple times. We also excluded any individuals receiving HUD-assisted housing without a valid SSN from the denominator (invalid SSNs are those that begin with "999" or a letter).

  8. We examined HUD-assisted individuals age 65 and older since most individuals over age 65 would be eligible for Medicare based on age (more than 97% of Americans age 65 and older are enrolled in Medicare, http://www.ssa.gov/history/pdf/WhatMedicareMeant.pdf).

  9. For the sample included in Figure 6, HUD-assisted individuals age 65 and older, over 40% of the sample has unknown/declined to report race/ethnicity.

  10. Center on Budget and Policy Priorities. October 2012. Reducing HUD Program Costs Associated with the Medical Deduction Policy.

  11. We excluded individuals with any managed care enrollment as the CMS administrative data only has enrollment for individuals in managed care, but does not have their claims data (i.e., health care cost and utilization data).

  12. Kaiser Family Foundation, Issue Brief, "Medicare's Role for Dual Eligible Beneficiaries", Gretchen Jacobson, Tricia Neuman, and Anthony Damico, April 2012.

  13. Given that beneficiaries residing in a nursing facility, for even a portion of the year, have poorer health status and therefore utilize more health care services compared to beneficiaries not in a nursing facility, accounting for that use becomes an important control factor. Due to incomplete information regarding the use of nursing facilities in the data available to us for this study and the fact that a higher proportion of individuals in the community sample used nursing facilities than in the HUD-assisted sample based on what we could observe (15% of unassisted beneficiaries in the community used Medicare SNF services compared to 8% HUD-assisted beneficiaries), we elected to remove all individuals with any nursing facility use that we could identify from both samples.

  14. Services only available through Medicaid health insurance, not covered by Medicare. We excluded MMEs with Medicaid managed care enrollment from the calculation of Medicaid cost and utilization because we do not have their complete health care cost and utilization data (similarly to those with Medicare managed care or Medicare Advantage.

  15. Personal Care Services include a range of human assistance provided to persons with disabilities and chronic conditions of all ages, enabling them to accomplish tasks they would normally do for themselves if they did not have a disability, including ADLs (such as eating, bathing, dressing, and bladder and bowel requirements) or IADLs (such as taking medications and shopping for groceries).

  16. Althought Medicaid does not cover room and board services provided in residential care facilities, other components of residential care -- for example, personal care, 24-hour services, and chore services -- can be covered. Residential care includes group, family or individual home residential care; cluster residential care; and therapeutic residential care services, assisted living, supported living, and nigh supervision. (DME and services for mental health or substance abuse treatment are excluded.) We would expect HUD-assisted beneficiaries to be less likely to use Medicaid residential services because: (1) only under limited circumstances would HUD properties be considered residential for Medicaid purposes -- assisted living conversion and 811 group homes; and (2) few housing voucher recipients use them for assisted living.

  17. No Medicaid eligibility; in other words, excluding individuals who are dually eligible for both Medicare and Medicaid.

  1. For purposes of determining income and rent contributions, HUD's public and assisted housing programs define "elderly" as 62 years of age and over, as provided by authorizing statutes.

  2. For further information about types of public and assisted housing, see "Programs of HUD 2011," (HUD 2013). http://www.huduser.org/portal/publications/progs_of_hud.html.

  3. Note that some tables refer to both "Section 202" and "Section 202/162". A small number of 202 properties contain units for non-elderly disabled that were funded with rental assistance created by Section 162 of the Housing and Community Development Act of 1987. Section 162 has been superseded by the Section 811 program, which supports housing for disabled individuals.

  4. Ended participation in HUD, death of sole family member.

  5. Before calculating the match rates, we excluded all Medicare or Medicaid beneficiaries that were duplicates in the CMS administrative data. We excluded these individuals from both the CMS administrative data and the HUD data, for those that were also HUD-assisted housing recipients. This way, the Rates were accurate and did not overestimate the rates due to an individual being counted multiple times. We also excluded any HUD-assisted housing recipient without a valid SSN from the denominator (invalid SSNs are those that begin with "999" or a letter.

  6. See http://www.medicare.gov/what-medicare-covers/part-b/what-medicare-part-b-covers.html.

  7. See https://www.ccwdata.org/cs/groups/public/documents/document/ccw_partddata_userguide.pdf.

  8. See https://www.ccwdata.org/cs/groups/public/documents/document/ccw_partddata_userguide.pdf.

  9. See http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/MedicareEnrpts/Downloads/Buy-InDefinitions.pdf.

  10. See http://www.cms.gov/Research-Statistics-Data-and-Systems/Research/MedicareMedicaidStatSupp/Downloads/07Glossary.pdf.

  11. See https://www.ccwdata.org/summary-statistics/demographics/a2-assistance-1999-2007.htm.

  12. See http://aspe.hhs.gov/health/reports/transition/welfare.htm.

  13. For additional information see Dartmouth Atlas: http://www.dartmouthatlas.org/.

  14. For example, if current year of analysis is 2008 and their move out date was 2007, we removed them from 2008 dataset.

  15. We excluded individuals with any managed care enrollment as the CMS administrative data only has enrollment for individuals in managed care but does not have their claims data (i.e., health care cost and utilization data).

  16. For this report, "elderly properties" include properties designated as elderly by PHAs and properties in which 50% or more of households had an individual aged 62 years or older.

  17. Before calculating the match rates, we excluded all Medicare or Medicaid beneficiaries that were duplicates in the CMS administrative data. We excluded these individuals from both the CMS administrative data and the HUD data, for those that were also individuals receiving HUD-assisted housing. This way, the rates were accurate and did not overestimate the rates due to an individual being counted multiple times. We also excluded any individuals receiving HUD-assisted housing without a valid SSN from the denominator (invalid SSNs are those that begin with "999" or a letter).

  18. Center on Budget and Policy Priorities. October 2012. Reducing HUD Program Costs Associated with the Medical Deduction Policy.

  19. We excluded individuals with any managed care enrollment as the CMS administrative data only has enrollment for individuals in managed care but does not have their claims data (i.e., health care cost and utilization data).

  20. Kaiser Family Foundation, Issue Brief, "Medicare's Role for Dual Eligible Beneficiaries", Gretchen Jacobson, Tricia Neuman, and Anthony Damico, April 2012.

  21. Given that beneficiaries residing in a nursing facility, for even a portion of the year, have poorer health status and therefore utilize more health care services compared to beneficiaries not in a nursing facility, accounting for that use becomes an important control factor. Due to incomplete information regarding the use of nursing facilities in the data available to us for this study and the fact that a higher proportion of individuals in the community sample used nursing facilities than in the HUD-assisted sample based on what we could observe 15% of unassisted beneficiaries in the community used Medicare SNF services compared to 8% HUD-assisted beneficiaries), we elected to remove all individuals with any nursing facility use that we could identify from both samples.

  22. Those determined "medically needy" have high medical expenditures, but have an income that exceeds maximum Medicaid income threshold. These individuals would otherwise be eligible for the program.

  23. Services only available through Medicaid health insurance, not covered by Medicare. We excluded MMEs with Medicaid managed care enrollment from the calculation of Medicaid cost and utilization as we do not have their complete health care cost and utilization data (similarly to those with Medicare managed care).

  24. Personal Care Services include a range of human assistance provided to persons with disabilities and chronic conditions of all ages, enabling them to accomplish tasks they would normally do for themselves if they did not have a disability, including ADLs (such as eating, bathing, dressing, and bladder and bowel requirements) or IADLs (such as taking medications and shopping for groceries.

  25. No Medicaid enrollment.


DESIGN OF A DEMONSTRATION OF COORDINATED HOUSING, HEALTH AND LONG-TERM CARE SERVICES AND SUPPORTS FOR LOW INCOME OLDER ADULTS -- Reports Available

Design of a Demonstration of Coordinated Housing, Health and Long-Term Care Services and Supports for Low-Income Older Adults
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2011/lioaDemo.shtml
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2011/lioaDemo.pdf
Picture of Housing and Health: Medicare and Medicaid Use Among Older Adults in HUD-Assisted Housing
Executive Summary   http://aspe.hhs.gov/daltcp/reports/2014/HUDpices.shtml
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2014/HUDpic.shtml
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2014/HUDpic.pdf
The "Value Added" of Linking Publicly Assisted Housing for Low-Income Older Adults with Enhanced Services: A Literature Syntheses and Environmental Scan
Executive Summary   http://aspe.hhs.gov/daltcp/reports/2012/ValueAddes.shtml
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2012/ValueAdd.shtml
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2012/ValueAdd.pdf

To obtain a printed copy of this report, send the full report title and your mailing information to:

U.S. Department of Health and Human ServicesOffice of Disability, Aging and Long-Term Care PolicyRoom 424E, H.H. Humphrey Building200 Independence Avenue, S.W.Washington, D.C. 20201FAX:  202-401-7733Email:  webmaster.DALTCP@hhs.gov


Preview
Download

"HUDpic.pdf" (pdf, 3.66Mb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®