Michigan Skilled Nursing Facilities, the Minimum Data Set, and the MI Choice Waiver Program: An Analysis and Implications for Policy

Michigan Skilled Nursing Facilities, the Minimum Data Set, and the MI Choice Waiver Program: An Analysis and Implications for Policy May 2011 Prepared for Health Care Association of Michigan Lansing, Michigan www.hcam.org Prepared by Public Sector Consultants Inc. Lansing, Michigan www.pscinc.com

Contents BACKGROUND OF THE STUDY...1 INTRODUCTION... 1 TODAY S NURSING AND REHABILITATION FACILITY... 1 DEMOGRAPHIC CHANGES IN MICHIGAN... 3 CURRENT ISSUES IN LONG-TERM CARE... 7 THE MI CHOICE WAIVER PROGRAM... 11 PSC RESEARCH ON MICHIGAN S SKILLED NURSING FACILITY POPULATION... 13 ANALYSIS OF MICHIGAN MINIMUM DATA SET... 13 SURVEY OF MI CHOICE WAIVER PROGRAM PARTICIPANTS... 14 FINDINGS AND IMPLICATIONS FOR POLICY... 16 APPENDIX A ANALYSIS OF MDS DATA SET... 21 INTRODUCTION... 21 METHODOLOGY... 22 SELECTED DEMOGRAPHICS... 23 ANALYSIS OF DEMOGRAPHIC TRAITS... 24 MODELING THE DATA... 25 CANONICAL DISCRIMINANT ANALYSIS... 26 INDICATOR VARIABLES: A CLOSER LOOK AT THE MDS GROUPS... 27 APPENDIX B SURVEY OF MI CHOICE PARTICIPANTS IN MICHIGAN: REPORT OF FINDINGS... 35 INTRODUCTION AND METHODOLOGY... 35 DISTRIBUTION OF RESPONSES... 36 DISCUSSION OF RESPONSES... 40

Background of the Study INTRODUCTION The provision and financing of long-term care services is in a dynamic state of change as a result of several factors, including: (1) the aging of the United States population; (2) budgetary pressures in both state and federal governments; (3) competing plans at the national level on how to address the rising costs of both Medicaid and Medicare, which finance a significant portion of long-term care services; (4) rising consumer demand for smaller, more home-like, and specialized long-term care settings; and (5) the passage of the Patient Protection and Affordable Care Act, known as the Affordable Care Act (ACA), which contains a number of provisions that affect long-term care services. It is within this context that the Health Care Association of Michigan (HCAM) has engaged Public Sector Consultants (PSC) to explore trends and issues affecting long term care services in Michigan and to discuss implications for the state s long-term care policy. HCAM is particularly interested in the question of whether significant numbers of skilled nursing facility residents could or should be transitioned to home or other community-based care, either because this would provide a lower cost to Medicaid or for other reasons. This study contains several components: (1) a statistical and quantitative analysis of data collected from the Centers for Medicare and Medicaid Services (CMS) about people who transitioned from a skilled nursing facility into home or other community-based care settings and others who remained in nursing facilities; (2) qualitative information gathered from a survey that identifies key characteristics of persons who have transitioned into home and community-based care using the services of MI Choice; (3) an analysis of the recently released Michigan 2010 Census data related to age; and (4) a brief discussion outlining some of the current issues in longterm care policy. Results from the survey and Minimum Data Set analysis were then analyzed in the context of the demographic changes in Michigan and the various trends in long-term care, summarized in key findings along with implications for long-term care policy in Michigan. TODAY S NURSING AND REHABILITATION FACILITY Skilled nursing facilities today provide both long term medical care and short term rehabilitation services. Long-term medical care is provided for residents with complex or chronic health conditions that require a safe environment and all-inclusive medical and personal care 24 hours a day. These residents are generally older and fit the description of what many people historically think of as residents of a nursing home; they represent about 77 percent of skilled nursing facility residents. A significant number of individuals in skilled nursing facilities, however, are there to receive intensive short-term rehabilitation services after a stroke, joint replacement, heart attack or other major health episode. These residents are typically in a facility for a short period of time less than 100 days and represent about 23 percent of residents. In fact, skilled nursing 1

facilities are now the largest provider of post-acute care in the nation. 1 Exhibit 1 illustrates this distribution of short-term and long-term residents. EXHIBIT 1. Percentage of Nursing/Rehabilitation Facility Residents Receiving Long-term vs. Short-term Post-acute Care SOURCE: Michigan Department of Community Health Worksheets for Provider Tax Computation and Variable Cost Limit Reports for State FY2011. Skilled nursing facilities have changed significantly during the past several years in response to consumer demand for more individualized and specialized settings, as well as federal law that requires all residents of a nursing facility have an individualized care plan that changes as their needs change. New constructions, replacement facilities, and numerous major renovations answer the demand for single occupancy rooms and more homelike settings. Smaller facilities are being built, as well as special units for residents suffering from such illnesses as dementia. Retirement communities that allow seniors to move from one level of assisted living to another in a single location are also being built. Skilled nursing facilities are regulated by both federal and state governments. The Centers for Medicare and Medicaid Services is the federal governing agency that ensures compliance with more than 1,000 regulations dealing with facility environment, resident rights, quality of life, and quality of care. The Michigan Department of Licensing and Regulatory Affairs (LARA) has federal delegation to ensure compliance with all federal regulations, as well as additional state regulations 2. LARA conducts an unannounced on-site visit of all facilities in the state each year, and it investigates all complaints related to nursing facility care. Skilled nursing facilities are also regulated by Michigan s Certificate of Need (CON) law: CON review and approval is required for new facilities, the replacement of existing facilities, change in ownership, and major improvements above an established dollar level. Care provided in skilled nursing facilities is funded by Medicare, Medicaid, and private resources. Private resources provide only 14 percent of nursing home care in Michigan. Medicare provides 19 percent, and is typically the source of funds for individuals 65 and older who are discharged from a hospital to a nursing facility for short-term rehabilitative care (often referred to 1 American Health Care Association and Alliance for Quality Nursing Home Care, 2010 Annual Quality Report, available online at http://www.ahcancal.org/quality_improvement/documents/2010quality Report.pdf (accessed 5/10/11). 2 These responsibilities were recently transferred from the Michigan Department of Community Health (MDCH). 2

as post-acute care) prior to going back home. Medicaid funds 67 percent of all nursing home care in Michigan (see Exhibit 2). EXHIBIT 2. Percentage of Funding Provided by Medicare, Medicaid, and Private Resources SOURCE: Michigan Department of Community Health Worksheets for Provider Tax Computation and Variable Cost Limit Reports for State FY2011. Michigan currently has 47,379 licensed skilled nursing facility beds 3. Michigan has fewer beds per 1,000 people over age 65, fewer nursing facility residents per 1,000 people over age 65, and a higher occupancy rate than the national averages (see Exhibit 3). 4 EXHIBIT 3. Skilled Nursing Facility Capacity (Per 1,000 People Aged 65 and Over) Michigan Number of beds 38 45 Number of residents 33 38 Rate of occupancy 87% 85% SOURCE: Houser, Fox-Grage, and Gibson, Across the States Profiles of Long-Term Care and Independent Living, 2009. U.S. DEMOGRAPHIC CHANGES IN MICHIGAN The aging of the baby boomer generation will not only significantly increase the need for longterm care services, but it will also impact the kind of long-term care services required. Nationally, from 2007 to 2030, the population age 65+ is projected to grow by 89 percent, more than four times as fast as the country s population as a whole. Most of this growth will be among the young old (aged 65 74) because of the aging of the baby boomers. An even better indicator of the potential demand for long-term care services is the growth in the population aged 85 or 3 Michigan Department of Community Health CON Bed Inventory Report November 2010. 4 Ari Houser, Wendy Fox-Grage, and Mary Jo Gibson, Across the States Profiles of Long-Term Care and Independent Living, 8 th ed. (N.p.: AARP, 2009), available online at http://assets.aarp.org/ rgcenter/il/d19105_2008_ats.pdf (accessed 5/5/11). 3

older, which is expected to increase by 74 percent between 2007 and 2030, after which it is expected to grow by 118 percent (see Exhibit 4). 5 EXHIBIT 4. The Aging of the United States SOURCE: Houser, Fox-Grage, and Gibson, Across the States Profiles of Long-Term Care and Independent Living, 2009. Reprinted with permission of AARP Public Policy Institute. Within these demographic changes occurring nationally, three points should be emphasized about what is happening in Michigan: 1. While Michigan is experiencing these same changes, they are occurring even faster than predictions. The Houser, Fox-Grage, and Gibson study, Across the States Profiles of Long- Term Care and Independent Living, predicted that the population of people aged 65 and over in Michigan in 2007 was 12.7 percent of the overall Michigan population, slightly above the national average that year of 12.6 percent. A 2009 update from the U.S. Census Bureau 6, however, estimated that the percentage of the Michigan population in this age group had increased to 13.4 percent, well above the actual national average of 12.9 percent). As Exhibit 5 shows, Census data released on May 5, 2011, now documents that 13.7 percent of the Michigan population was aged 65 or over in 2010. 5 Houser, Fox-Grage, and Gibson, Across the States, p. 6. Note: The U.S. Census Bureau intends to release nationwide demographic data, including age, in late May 2011. 6 U.S. Census Bureau, State and County QuickFacts, Michigan, available online at: http://quickfacts. census.gov/qfd/states/26000.html (accessed 5/6/11). 4

EXHIBIT 5. Michigan 2010 Census by Age 2000 Census 2010 Census 2000 2010 Change Age groups* Count Percentage 2010 Percentage Count Percentage Under 5 years 672,005 6.8% 596,286 6 75,719-11.3% 5 to 9 years 745,181 7.5 637,784 6.5 107,397 14.4 10 to 14 years 747,012 7.5 675,216 6.8 71,796 9.6 15 to 19 years 719,867 7.2 739,599 7.5 19,732 2.7 20 to 24 years 643,839 6.5 669,072 6.8 25,233 3.9 25 to 34 years 1,362,171 13.7 1,164,149 11.8 198,022 14.5 35 to 44 years 1,598,373 16.1 1,277,974 12.9 320,399 20.0 45 to 54 years 1,367,939 13.8 1,510,033 15.2 142,094 10.4 55 to 59 years 485,895 4.9 683,186 6.9 197,291 40.6 60 to 64 years 377,144 3.8 568,811 5.8 191,667 50.8 65 to 74 years 642,880 6.5% 724,709 7.3% 81,829 12.7% 75 to 84 years 433,678 4.4 444,940 4.5 11,262 2.6% 85 years and over 142,460 1.4 191,881 1.9 49,421 34.7% Totals 9,938,444 100% 9,883,640 100% 54,804 0.6% SOURCE: U.S. Census http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk. * Based on 2000 Census age groupings 2. The increase in the 85+ population group is particularly significant, because they utilize longterm care services differently than the 65 74 age group. The oldest old not only have much higher rates of disability, but they also are much more likely to be widowed and without someone to provide assistance with daily activities. 7 The recently released census data shown in Exhibit 5 indicates that the 85+ age group is now 1.9 percent of the overall Michigan population, an increase of 35 percent almost 50,000 individuals since 2000. Moreover, it continues to increase; this age group, as a percentage of the overall Michigan population, will be higher than the national average in the years ahead. 3. Another significant change occurring in the Michigan population that will impact the use of long-term care services is the dynamics of the population of people aged 45 64. This group as a percentage of the overall Michigan population has increased during the past decade from 22.5 percent in 2000 to 27.9 percent in 2010 (see Exhibit 5). It is the key group of informal family caregivers that has been at least partly responsible for the fact that Michigan has utilized skilled nursing facility services at a lower rate than the national average during the past several years. As will be discussed later in this report, this is also the group that many individuals who utilize home and community-based services rely upon for assistance. As this age group moves into the 65+ age bracket they will themselves need long-term care services in the near future but they are being replaced by a much smaller group. The replacement group aged 35 44 has declined by a dramatic 20 percent 320,000 people in the past ten years (see Exhibit 5). The decline of this group as a percentage of 7 Houser, Fox-Grage, and Gibson, Across the States, p. 6. 5

the Michigan population a decline larger than the national average will significantly impact both the need for long-term care services and the kind of services required. More young adults leaving Michigan, fewer young adults moving to Michigan, and fewer retirees leaving the state all the result of a decade-long recession appear to be accelerating the aging trends predicted a few years ago by the U.S. Census Bureau for Michigan. 6

CURRENT ISSUES IN LONG-TERM CARE State budget challenges, budget and deficit issues at the federal level, the recent passage of the Affordable Care Act (ACA), demographic changes noted earlier, as well as other issues make this a dynamic time in both the provision and financing of long-term care services. Although it is not this report s purpose to examine current issues in long term care in depth, several should be mentioned to help put the results of this study in the larger context of both the policy discussions and demographic changes that are occurring in Michigan and throughout the country: 1. Medicaid and State Budgets. Medicaid expenditures in Michigan as in other states have increased significantly during the past several years, driven primarily by the increase in caseloads. The overall Medicaid caseload in Michigan, for example, has grown by 80 percent during the past decade. Although federal matching rates have increased during that same period, and Michigan has greatly increased the use of restricted fund sources to support Medicaid, Medicaid continues to be a significant piece of state budgets 20 percent of Michigan s general fund budget in FY 2010. Total Medicaid expenditures in Michigan were $11.1 billion in FY 2010, representing an increase of 45.7 percent over the past five years. Medicaid long-term care expenditures represented 19 percent of that total, or $2.1 billion for FY 2010 in Michigan: $1.6 billion for skilled nursing facility care and $475 million for a variety of community-based long-term care programs (see Exhibit 6). EXHIBIT 6. Long-term Care Expenditures, FY 2010 SOURCE: Senate Fiscal Agency, Michigan Medicaid Program (Physical Health and Mental Health) Expenditure History from FY 1999-2000 to FY 2010-11. Provided by Steve Angelotti, May 2011. The trend in Medicaid long-term care expenditures is shown in Exhibit 7. While total Medicaid expenditures have increased 45.7 percent during the past five years in Michigan, long-term care expenditures have increased 20.6 percent during the same period. Facility-based long-term care expenditures have increased by 9.4 percent, while community-based long-term care expenditures have increased by 42.8 percent (see Exhibit 7). 7

EXHIBIT 7. Percentage Funding Increase over Past Five Years in Michigan SOURCE: Senate Fiscal Agency, Michigan Medicaid Program (Physical Health and Mental Health) Expenditure History from FY 1999-2000 to FY 2010-11. Provided by Steve Angelotti, May 2011. *Home and community -based services, for the purposes of this exhibit, include MI Choice, Adult Home Help, and Personal Care Services. PACE, hospice, or the services offered by the Office of Services to the Aging are not included. Two points should be emphasized about Medicaid and Medicaid long-term care expenditures: Michigan spends a lower percentage of its Medicaid dollar on long-term care services than the national average 22.7 percent in 2009 compared to the national average of 33.3 percent. 8 In fact, only three states have a lower percentage than does Michigan. Long-term care costs are not driving the increased Medicaid expenditures in most states. 9 Medicaid expenditures are driven primarily by family caseload increases and the resulting increase in the utilization of all health care services financed by Medicaid. 2. Medicaid and FMAP. Medicaid is jointly funded by state and federal governments, and the federal contribution called the Federal Medical Assistance Percentage (FMAP) is an important issue to the states. Temporary increases as part of the federal economic stimulus legislation are being phased out, and states are already experiencing a decline in federal Medicaid funding as a result. Average Michigan FMAP rates will decline 5.1 percent from FY 2011 to FY 2012, costing Michigan over $500 million in federal Medicaid assistance. 3. Medicaid and the Federal Budget. Medicaid is a key part of the discussions occurring at the national level regarding the federal budget and the federal deficit. To put those discussions in context, it is important to remember that the recently enacted health care reform law (ACA) mandates a significant expansion of Medicaid to cover more people, requires that states not reduce the coverage that existed when the new law was enacted ( maintenance of effort ), and provides additional funding to the states to 8 Kaiser Family Foundation, Kaiser statehealthfacts.org. Distribution of Medicaid Spending on Long Term Care, FY 2009, available online at http://www.statehealthfacts.org/comparetable.jsp?ind=180&cat=4 (accessed 5/10/11). 9 Houser, Fox-Grage, and Gibson, Across the State, p. 16. Total Medicaid spending increased by 28% 2002-2007. Medicaid LTC spending for Older People and Adults with Disabilities increased by 18%. 8

pay for the additional coverage. That additional federal funding would, however, be phased out over time. The House Republican budget proposal recently introduced by Congressman Paul Ryan (R- Wisconsin) and passed by the House takes a dramatically different approach. It assumes the repeal of the ACA and would convert the federal share of the Medicaid program into block grants to the states. Each grant would be adapted to meet the individual needs of the state, and would be indexed both for inflation and population growth. States would have significant flexibility as current program requirements and enrollment criteria would be eliminated. The projected savings target is $750 billion over ten years. 4. Medicare and the Federal Budget. Like Medicaid, Medicare is a focus of deficit reduction at the federal level. The ACA assumes a hard cap on Medicare expenditures with actual reductions in Medicare services or reimbursements rates made by the newly created Independent Payment Advisory Board (IPAB) designed to achieve $150 billion in savings over ten years including, in the case of skilled nursing facilities, a $14.6 billion cut in Medicare rates over ten years. The ACA also requires the creation of Accountable Care Organizations (ACOs), which will have an impact upon Medicare payment to health care providers. An ACO is a network of health care providers principally doctors and hospitals, but potentially including other providers such as nursing facilities and home health agencies that share responsibility for providing care to patients. An ACO will have incentives to cooperate and save money by avoiding unnecessary tests, procedures, and hospitalizations for their patients. Those that save money while also meeting quality targets would keep a portion of the savings, but the providers in an ACO could also be at risk of losing money. Congressional House Republicans, as part of the budget proposal and long-term deficit reduction plan recently passed by the House, have proposed a new model for Medicare. Beneficiaries older than 55 would not see any changes. Beginning in 2022, however, Medicare would follow a premium-support model, similar to Part D, so that recipients would have the ability to choose private plans best suited to them. Beneficiaries would have to opt in to the plans. Medicare would subsidize the cost of the plan, adjusted for income and acuity, and pay the plan directly. Plans that elect to participate must offer coverage to all Medicare beneficiaries. 5. Home and Community-Based Services (HCBS). Michigan has a number of programs and options that are called home and community-based services (HCBS), which are designed to support seniors and other adults living at home or in small residential settings like adult foster care. The best known is MI Choice, which provides home and community-based services to physically disabled adults and elderly people who need a nursing facility level of care. Initially established in 1992 and expanded statewide in 1998, MI Choice operates under a waiver granted by the federal Centers for Medicare and Medicaid Services (CMS). In FY 2010, 10,876 individuals were enrolled in MI Choice. 10 MI Choice, however, is not the only HCBS program in Michigan. Adult Home Help, Personal Care Services, the Program of All-inclusive Care for the Elderly (PACE), hospice services, and various services provided by the Office of Service to the Aging (OSA) all have the same goal of allowing seniors and other adults to live at home by providing various kinds and levels of 10 Enrollment data supplied by Michael Daeschlein, State Administrative Manager, Michigan Department of Community Health, Home and Community-Based Services Section. 9

support. Home Help, for example, was established in 1982 and is Michigan s largest HCBS program, serving over 48,000 people in FY 2008. State budget documents generally list MI Choice, Adult Home Help, and Personal Care Services as HCBS programs as noted above, they totaled $475 million in FY 2010. If PACE, hospice services, and OSA programs are included, $675 million was spent on HCBS programs in FY 2010. The CMS provides special funding in the form of enhanced FMAP rates referred to as Money Follows the Person grants as an incentive for states to move transition people from skilled nursing facilities to home or other community-based settings. The policy goal is often termed rebalancing, referring to the attempt to balance Medicaid long-term care spending between facility and non-facility settings. ACA contains provisions to extend these programs and the increased FMAP rates--until 2020. Michigan has participated in the Money Follows the Person program for a number of years. 6. Coordinating the Care and Financing of the Dual Eligibles The term dual eligibles refers to individuals who qualify for both Medicare and Medicaid i.e., they are 65 years of age or older (Medicare), as well as low income (Medicaid). In Michigan this applies to about 200,000 residents, including nearly all approximately 25,000 Medicaid recipients in skilled nursing facilities, as well as individuals enrolled in MI Choice. Given that Medicaid and Medicare provide different kinds of health care and have different funding streams and levels, coordinating the care of these dual eligibles has been a challenge for both state and federal governments. Michigan was recently awarded a federal grant one of several funded by the new Federal Office of Coordinated Care created by ACA intended to design a demonstration project that would fully integrate health care for these individuals. 7. Reimbursement Systems for Medicaid Long-Term Care Services. Changing the reimbursement system for both facility and community-based long-term care providers has been the subject of discussions in Michigan for a number of years. The current system is essentially a cost-based reimbursement system. Much of the discussion for possible change has focused on attempting to include a variety of quality and efficiency incentives, to create what is often referred to as a reimbursement system based on case mix or an acuity. The basic concept is that the reimbursement system should reflect the fact that different levels of care merit different payment levels. A number of states have had these kinds of reimbursement systems and are, in fact, in the process of refining and updating them. A few states Arizona, Minnesota, and New Mexico have developed some form of an integrated managed care system in an attempt to provide better care coordination and program savings. The impact of these systems on the health care delivery to the client has not been documented. 8. Community Living Assistance Services and Supports (CLASS) Act. CLASS enacted as a part of ACA is an attempt to reduce the demand for Medicaid-financed long-term care services by creating a voluntary long-term care insurance program. Financed by employee payroll deductions, individuals could use it to finance services if they become functionally disabled after paying premiums for at least five years. CLASS would be repealed by the House Republican budget and deficit reduction proposal. 10

9. Regulatory Reform. Providers across the health care spectrum are now operating in an environment in which they must demonstrate quality and improve both outcomes and efficiency, and long-term care providers are no exception. This environment and pressure on public budgets is helping to drive a reexamination of how best to achieve compliance with regulations whose goal is quality and safety. For example, could compliance costs be reduced and outcomes improved if the survey and complaint investigation process for skilled nursing facilities were changed to rely more on incentives for compliance and quality, rather than an approach that relies solely on punitive enforcement for lower performance? THE MI CHOICE WAIVER PROGRAM MI Choice is often referred to as the home and community-based waiver program. It is, however, only one of several optional home and community-based long-term care programs funded by Medicaid in Michigan. 11 Authorized by federal law under Medicaid Section 1915 and administered by the MDCH, it uses Medicaid long-term care dollars to finance a variety of personal care services 12 that allow individuals to reside either at home or in community-based settings such as adult foster care facilities. Approval for the program comes from the federal Centers for Medicare and Medicaid Services (CMS). It is called a waiver, because a federal provision is waived to allow adults who are eligible for Medicaid-covered nursing home services to remain in their own home or other residential setting in their community. Individuals eligible for MI Choice are the same as those who are eligible for Medicaid-covered skilled nursing facility services; -they must meet certain income and asset requirements, as well as be medically eligible for nursing home placement. The MDCH, although responsible for administrating the program, contracts it out to 20 waiver service providers ( waiver agents ) that are located in the 16 Area Agency on Aging (AAA) regions in Michigan in fact, 13 of the 16 regional AAAs are waiver agents. The MDCH monitors the waiver agents for program compliance and quality of care. Waiver agents receive referrals from a variety of sources families, physicians, skilled nursing facilities, home health agencies, and others. These referrals go on a waiting list, which is used by the waiver agents to assess individuals for a variety of eligibility factors, including financial status, level of care needed, and what services would be required to enable an individual to function in the community or at home. The use of the term waiting list has been controversial over the years, since waiver agents find many duplications, inaccuracies, and changes in individual situations once they actually begin the assessment process that may disqualify applicants. The statewide waiting list currently numbers approximately 7,900 individuals. In 2007, 7,889 persons were enrolled in MI Choice, including 406 who were transitioned from skilled nursing facilities; the remaining 7,493 individuals entered the program directly without 11 MDCH budget breakdown: Home and Community Based Waiver (MI-Choice); Adult Home Help; In- Home Services High Cost AHH; Personal Care-AFC; PACE; Other. 12 Personal care services include: adult day care, chore services, community living supports, counseling services, fiscal intermediary, goods and services, home modifications, home delivered meals, homemaker, medical equipment and supplies, nursing facility transition, personal care, private duty nursing, residential services, respite care, specialized medical equipment and supplies, training, and non-medical transportation. 11

first going into a skilled nursing facility. 13 In FY 2010, 10,876 persons were enrolled in MI Choice, and 1,013 transitions occurred that year. 14 The budget for this program has risen steadily over time, rising from $115.8 million in FY 2000 to $185.8 million in FY 2010. 15 13 Pages 27 and 28 of Michigan Profile of Publicly-Funded Long-Term Care Services, prepared and submitted by the Michigan Department of Community Health Office of Long-Term Care Supports and Services, June 2009. 14 information supplied by Michael Daeschlein, State Administrative Manager, Michigan Department of Community Health, Home and Community-Based Services Section. 15 Senate Fiscal Agency, Michigan Medicaid Program (Physical Health and Mental Health) Expenditure History from FY 1999-2000 to FY 2010-11. Provided by Steve Angelotti, May 2011. 12

PSC Research on Michigan s Skilled Nursing Facility Population ANALYSIS OF MICHIGAN MINIMUM DATA SET PSC analyzed Minimum Data Set (MDS) 2.0 assessments conducted on residents of skilled nursing facilities (SNFs) in the State of Michigan during 2008. The data set contains records for 95,628 individuals who had a full MDS assessment during that year. Among these are 35,378 Medicaid recipients, of whom 29,414 (83 percent) did not transition out of the SNF during the year, while 5,964 (17 percent) transitioned in one of several ways. A total of 567 individuals transitioned into the MI Choice waiver program, while 5,397 individuals transitioned, but not into the waiver program, including 81 who transitioned from the waiver program into a SNF. The study focuses on five groups within the data set: All Medicaid recipients with full MDS assessments during 2008 (n = 35,378) (Group 1) All those in Group 1 who remained in SNFs during the year (n = 29,414) (Group II) All those in Group 1 who transitioned out of SNFs during the year, but not into the MI Choice waiver program (n = 5,397) (Group III) All those in Group 1 who transitioned into the MI Choice waiver program during the year (n = 567) (Group IV) MI Choice waiver program participants who transitioned from the community into a SNF (n = 81) (Group V) PSC s investigation of the MDS 2.0 data set for Michigan SNFs for 2008 supports the following conclusions. Michigan s SNFs house two quite different populations: an older population that generally prefers not to return to the community and is not preparing to return; and a younger population that generally prefers to return to the community but is experiencing an acute or emergent injury or medical condition from which it is recovering with treatment. The latter group is likely to transition out of the SNF; the former group is not. As a group, transitioning SNF residents are distinguished from non-transitioning residents in Michigan primarily by their greater interest in and potential for discharge; the fact that they receive therapy (physical, occupational, or clinical); their stronger cognitive acuity; their younger age; and their shorter length of stay in the SNF. We did not find that transitioning residents are strongly distinguished from non-transitioning residents by greater mobility or other measures of their capacity to perform the activities of daily living (ADLs). Discharge potential is the single most important predictor of transition. This potential is reflected in the resident s desire to return to the community, the existence of support for the resident s return to the community, a projected SNF stay of short duration, the necessity of the resident acquiring training before return to the community, a belief by resident that he or she is capable of increased independence, and a belief by staff that the resident is capable of increased independence. Participation in therapy is the second most important predictor of transition. In the model PSC developed it is reflected in the duration and frequency of physical therapy, the duration and frequency of occupational therapy, the presence of IV medication or some other form of clinical therapy, the fact that the resident is being monitored for an acute condition, a staff 13

judgment that the resident needs training to return to the community, and a greater tendency for the resident to have experienced one or more hospital stays during the previous 90 days. Cognitive acuity is the third most important predictor of transition. In PSC s analysis this is reflected in cognitive skills, short-term memory, and long-term memory. PSC used a relatively small set of variables from three domains (discharge potential, participation in therapy and cognitive acuity), plus age and length of stay. The analysis achieves 94 percent success in distinguishing between SNF residents who actually transitioned and those who did not. There is little difference in the profiles of residents who transition with support from the MI Choice waiver program and residents who transition but without support from the waiver program. Waiver program participants made up a small proportion (9 percent) of all SNF residents who transitioned in 2008. These two facts suggest that increasing the size of the waiver program would not proportionately increase the total number of transitioning SNF residents, but would simply decrease the number of those who transition without support. PSC s survey of MI Choice waiver program participants showed that many of them receive strong support from family members for their financial (31 percent), emotional (64 percent) and care-giving (56 percent) needs, and a majority (53 percent) participate in social activities with family members at least weekly. Data from the MDS show, however, that SNF residents who transitioned out are more likely to have lived alone before they entered the SNF than those who did not transition out. Further, the data show that this was especially the case with waiver program participants. In PSC s survey of MI Choice waiver program participants, more than two out of five said they had visited an emergency room (43 percent) or had been admitted to a hospital (42 percent) in the previous six months. PSC s estimate of the rate of re-hospitalization for waiver program participants based on survey results may be low, since it is about 15 percentage points below the rate found in the MDS (58 percent). Further, MDS data show that the rate of re-hospitalization among MI Choice waiver program participants is much higher than the rate among non-transitioning SNF residents almost four times as high. These elevated rates of hospital use among waiver program participants raise concerns about the coordination of care among their treating providers. For details of PSC s MDS data set analysis, please see Appendix A. SURVEY OF MI CHOICE WAIVER PROGRAM PARTICIPANTS PSC constructed and fielded a survey of MI Choice waiver program participants in Michigan (as of late 2010) to assess their satisfaction with the experience of home-based care. Working with the Michigan Department of Community Health, Home and Community Based Services Section, PSC created a random sample of 1,000 MI Choice participants who were sent pen-and-pencil questionnaires, together with an enclosed self-addressed stamped envelope marked for return to PSC. In all, 293 completed questionnaires were received, for a 29.3 percent response rate. A sample of this size yields a sampling error of 5.7 percent. PSC s analysis of the survey results suggests the following areas of concern for policy makers: The implied costs of the MI Choice waiver program: Waiver program participants receive a number of services an average of four apiece, as documented by the survey. The services imply direct and indirect costs that are borne by a variety of federal, state, local, and private resources. These costs need to be accounted for if a direct comparison is to be made between the cost of care under the waiver program and the cost of care within a nursing facility. 14

Coordination of care: Among the survey respondents, significant proportions of waiver program participants report recent visits to hospital emergency rooms (43 percent) and recent admissions to hospitals (42 percent). This suggests that many participants experience acute or emergent clinical issues, and that as a result, multiple providers are involved in their care. In a non-facility environment, this challenges on many levels the capacity of primary care providers to coordinate care. Social isolation: Waiver program participants report varying frequencies with which they have social contact with family, friends, church groups, and other community groups. At the low end of the scale, one in four respondents has little or no social contact with anyone. Social isolation poses a challenge that facilities dedicate resources to overcome; in a nonfacility setting, however, individuals themselves must make the effort to overcome it, and it appears that many waiver program participants are not successful. Dependence on private safety nets: Many waiver program participants report receiving strong support from family members for their financial (31 percent), emotional (64 percent) and care-giving (56 percent) needs, with smaller proportions reporting similar support from friends. Some observers have suggested this family burden should be accounted for if an accurate comparison is to be made between the cost of care incurred when Medicaid recipients live in a facility setting and the cost of care incurred when they live in the community. In any case, to the extent that waiver program participants enjoy and depend on such support, they depend on a safety net that is inherently fragile. Family members ability to continue to provide financial, emotional, or care-giving support is contingent on circumstances that cannot be controlled; such support cannot be guaranteed. Quality of life issues: The survey of waiver program participants reveals that there is a strong relationship between participants overall quality of life and their perception of the quality of services they receive under the waiver program. Those who are satisfied with the quality of the services are three times as likely to say their quality of life is good or very good as those who are not. Those who are satisfied with the timeliness of the services are 2.4 times as likely to say their quality of life is good or very good. Conclusions of causality cannot be made, but policymakers, managers of services, care providers, and others concerned with the quality of life outcomes of waiver program participants should be alerted to the relationship. For a detailed discussion of the survey s methodology and findings, please see Appendix B. 15

Findings and Implications for Policy 1. Dependence on private safety nets in MI Choice. Survey results show that many waiver program participants report receiving strong support from family members for their financial (31 percent), emotional (64 percent), and care-giving (56 percent) needs, with smaller proportions reporting similar support from friends. This dependence upon family members for care and support rather than friends is consistent with national studies 16 and might not seem particularly surprising. It does, however, have significant policy implications for the future of long-term care in Michigan. This is a safety net that is inherently fragile. Family members ability to continue to provide financial, emotional, or care-giving support is contingent on circumstances that cannot be controlled; such support cannot be guaranteed. More importantly, this key group of informal caregivers working age adults aged 50 64 is declining in Michigan, at the same time Michigan is expected to see a significant growth in the 65 and over age group. Recipients of MI Choice and other home and community-based programs rely significantly upon younger family members for success, and replacing the support they provide if it is even possible will be a challenge. 2. Age. The MDS profile sharply distinguishes between users of skilled nursing facilities and those who are transitioned back home or into a community-based setting based on age. The median age of SNF residents (no transition) is 85.0; the median age of MI Choice participants is 76.0; and the median age of others who transitioned is 71.2. This difference is further illustrated by the fact that the subgroup of MI Choice participants who transitioned back into a skilled nursing facility from the MI Choice program is also older than those who stayed in MI Choice. While this might seem self-evident and intuitive, it has significant implications for Michigan s long-term care policy. As a percentage of the overall population, the 85+ group is growing faster in Michigan than the national average. Considering the fact that Michigan is already below the national average in terms of such key indicators of skilled nursing facility capacity as beds, and above the national average in such indicators of intensity of use such as occupancy rate, it is difficult to see how based on demographic changes alone Michigan will rely less on skilled nursing facilities in the future than it does today. 3. Skilled nursing facilities and home and community-based care: people are in appropriate settings. The MDS analysis supports the results currently being achieved in Michigan regarding placement in the most appropriate setting. The overall accuracy rate of the analysis is 94 percent, meaning that 94 percent of all Medicaid recipients are being placed with the group they most strongly resemble on the basis of data recorded in the MDS. The 6 percent who are not are more readily described as exceptions to the rule because of individual circumstances, rather than as evidence of error in the placement system. 16 Houser, Fox-Grage, and Gibson, Across the State, p. 9. 16

The MDS analysis does not support the position that many current Medicaid nursing home residents should not be there, but rather should be in a home or other community-based setting. Nor does it support the position that many individuals in home or community-based settings are there inappropriately. This conclusion from the MDS analysis is consistent with other data related to the use of skilled nursing facilities Michigan i.e., compared to other states, Michigan does not over-utilize skilled nursing facilities. The MDS analysis also provides clear evidence that Michigan skilled nursing facilities are acting as secondary/tertiary care providers, transitional between the hospital and home or other community settings. This is demonstrated, in part, by the high number of Medicaid recipients transitioning independent of the MI Choice program almost 5,367 compared to 567 transitioning to MI Choice. In addition, MDS variables that indicate a resident is receiving one or more forms of clinical, physical, or occupational therapy are among the strongest predictors that the resident will transition. Stated another way, the MDS data indicates that skilled nursing facilities in Michigan are aggressively preparing residents for transitioning to home and community-based settings in order to help achieve successful transitioning. More research is needed to determine how Michigan compares with other states in the relationship between the data tracked by the MDS and the transitioning of SNF residents. A similar study was conducted in 2006 in Virginia, 17 for example. However, a detailed comparison between that analysis and our analysis of the Michigan data is beyond the scope of this report. 4. Quality-of-life issues and MI Choice. The survey of waiver program participants reveals that there is a strong relationship between participants overall quality of life and their perception of the quality of services they receive under the waiver program. Those who are satisfied with the quality of the services are three times as likely to say their quality of life is good as those who are not. Those who are satisfied with the timeliness of the services are 2.4 times as likely to say their quality of life is good. Conclusions of causality cannot be made, but policy makers, managers of services, care providers, and others concerned with the quality-of-life outcomes of waiver program participants should be alert to the relationship. As the state considers new federal incentives to expand home and community-based services, maintaining both the quality and timeliness of services is critical. Moreover, how to do so when the group relied upon for support family care givers is declining in Michigan will be a challenge. 5. Coordination of care. In the survey, significant proportions of waiver program participants report recent visits to hospital emergency rooms (43 percent) and recent admissions to hospitals (42 percent). This is supported by the analysis of the MDS data, which showed that a majority (58 percent) of individuals transitioned into MI Choice had at least one hospital admission within the previous 90 days, and 12 percent had at least one emergency room (ER) visit. 17 Fiscal Analytics Ltd., Comparing Long Term Care Settings The Potential for and Cost of Discharging Nursing Facility Residents to Home and Community-Based Care, prepared by for the Virginia Health Care Association (Richmond, Va.: Fiscal Analytics Ltd., September 2006), available online at www.vhca.org/ whatsnewapp/files/24e3312c9.pdf (accessed 5/9/11). 17

The MDS data clearly shows, moreover, that the rate of rehospitalization among MI Choice waiver program participants is much higher than the rate among non-transitioning nursing facility residents almost four times as high. Similarly, the rate of ER visits among waiver program participants is more than three times as high. In addition to these rates of rehospitalization, a number of key predictor variables shows that transitioned residents are very likely to have ongoing, acute medical conditions that require continuing treatment. At least two implications for policy makers arise from these results: Many participants experience acute or emergent clinical issues, and as a result, multiple providers are involved in their care. In a non-facility environment, this challenges on many levels the capacity of primary care providers to coordinate care. Efforts to reduce the incidences of rehospitalizations to control health care costs and improve outcomes, present in ACA and state-level health care initiatives, need to focus on both facility and non-facility settings. Both the survey results and the MDS data about hospital visits and ER visits is consistent with the Report to the Medicaid Long-Term Care Task Force from the University of Michigan that documented higher use of hospital admissions and ER visits from MI Choice Waiver participants than those from nursing facilities. 18 6. Social Isolation. Waiver program participants report varying frequencies with which they have social contact with family, friends, church groups, and other community groups. At the low end of the scale, a significant proportion of participants (25 percent) have little or no social contact with anyone. Social isolation poses a challenge that facilities dedicate resources to overcome; in a non-facility setting, however, individuals themselves must make the effort to overcome such isolation, and it appears that many waiver program participants are not successful. The severity of this problem will increase because of the population decline of the key group of informal family caregivers in Michigan. 7. The implied costs of the MI Choice waiver program. Waiver program participants engage a number of services an average of four apiece, as documented by the survey. Those who are heavy consumers of waiver services are more likely to live at home, rather than in an assisted living arrangement. In addition to waiver services, 79 percent use a variety of services funded by other state and federal assistance programs, as well as receiving support from informal family caregivers. Any comparison between the cost of care under the waiver program and care within a nursing facility would need to account for both the direct costs of other publicly funded programs and the value of informal family care giving. Given the population decline of the age group that typically provides the informal family support, long-term care policy in Michigan needs to address the challenge financial and other of replacing this support. 18 Brant Fries, Mary James, and Pablo Aliaga, A Report to the Medicaid Long-Term Care Task Force: Persons Using Medicaid-Funded Nursing Home and Community-Based Waiver Services Fiscal Year 2002 (Ann Arbor, Mich.: Institute of Gerontology, University of Michigan), August 2004. 18

8. Diversion vs.. This study did not directly examine the issue of whether efforts devoted to home and communitybased placements should focus more on transitioning individuals from skilled nursing facilities or keeping individuals at home who are already there. For the following reasons, however, it suggests that expanding the program would not be likely to result in a proportional increase in total transitions: The MI Choice waiver program population is not statistically different, from a clinical point of view, from the population that transitions without waiver program assistance. The MI Choice waiver program population in this 2008 data was a small proportion (9.5 percent) of the total transitioning population. The MDS analysis suggests that skilled nursing facilities are aggressively preparing people for transitioning. It also indicates that those who are able to leave facilities are, in fact, doing so. Further study is needed, but this analysis suggests that it is not likely that (for example) a 100 percent increase in waiver program funding would result in increasing the total number of transitions by an amount equal to the current number of waiver program transitions. It is more likely that an increase in funding for the waiver program would increase the number of waiver program participants by cannibalizing from the population that would transfer without assistance, and do so at public expense. If this were the case, what would be achieved would be a marginal increase in total transitions, at the cost of a disproportionate increase in outlay from public funds. 19

20

Appendix A Analysis of MDS Data Set INTRODUCTION This report includes findings from a study of Minimum Data Set (MDS) assessments conducted on residents of nursing facilities (SNFs) in the State of Michigan during 2008, compiled by the Michigan Department of Community Health. The data set contains records for 95,628 unique individuals who had a full MDS assessment during that year. Among these are 35,378 Medicaid recipients, of whom 29,414 (83.1%) did not transition out of the SNF during the year, while 5,964 (16.9%) transitioned in one of several ways. A total of 567 individuals transitioned into the MI Choice waiver program, while 5,397 individuals transitioned, but not into the waiver program, including 81 who transitioned from the waiver program into a SNF. 19 The study focuses on five groups within the data set: 1. All Medicaid recipients with full MDS assessments during 2008 (n = 35,378) 2. All those in Group 1 who remained in SNFs during the year (n = 29,414) 3. All those in Group 1 who transitioned out of SNFs during the year, but not into the MI Choice waiver program (n = 5,397) 4. All those in Group 1 who transitioned into the MI Choice waiver program during the year (n = 567) 5. MI Choice waiver program participants who transitioned from the community into a SNF (n = 81) 19 The relationship between MI Choice status and transition status is complicated, in that transition took place from SNFs to the waiver program, from the waiver program into SNFs, and in both directions. See Appendix A. 21

The chart below illustrates the relationship among these groups. METHODOLOGY The study provides descriptive information about the members of each of the groups. This is of two kinds: Demographic information, including age, gender, marital status, and race Information that is significantly related to discharge, including potential to transition, therapy usage, cognitive acuity, age, and length of stay The data set amounts to a census of Michigan skilled nursing facility (SNF) residents with full MDS assessments taken during 2008. However, if it is regarded as a sample of Michigan SNF residents with MDS assessments without regard to the year in which they were assessed, or more generally, all Michigan SNF residents, then the findings can be expected to have sampling error, and differences between groups may or may not be statistically significant. The sampling error for the five groups is: 1. Medicaid recipients: 0.5% 2. Medicaid recipients who did not transition out of SNFs: 0.6% 3. Medicaid recipients who transitioned but not into the MI Choice waiver program: 1.3% 22