Over 22 million nonelderly adults, or 12% of adults ages 18 to 64, living in the community in the United States have a disability as of 2015.¹   They include people with physical disabilities, such as cerebral palsy, multiple sclerosis, and traumatic brain or spinal cord injuries; intellectual or developmental disabilities (I/DD), such as Down syndrome and autism; and mental illness.  Some people live with a disability from birth or childhood, while others have a disability onset later in life resulting from an injury or chronic illness. Medicaid plays an important role by providing health insurance coverage for more than one in three nonelderly adults with disabilities (Figure 1).

The House American Health Care Act bill calls for changes to Medicaid that could affect nonelderly adults with disabilities.  Some nonelderly adults with disabilities have gained Medicaid eligibility through the ACA’s expansion, and the House bill calls for eliminating the expansion’s enhanced federal matching funds.²   The House bill also contains changes to Medicaid’s financing structure in ways that would limit federal funding through a per capita cap.  While states could be offered increased flexibility beyond what is available under current law in future legislation, states may look to cuts in Medicaid eligibility, benefits, and provider reimbursement as a result of the associated federal funding reductions.  People with disabilities have higher Medicaid spending relative to other populations due to their greater health needs and more intensive service use.  Nonelderly adults with disabilities could be particularly affected by potential cuts, as many of them rely on Medicaid coverage pathways and services that are offered at state option.  This issue brief describes the role that Medicaid plays for nonelderly adults with disabilities.  The appendix table contains per enrollee spending and take-up of selected eligibility and service options relevant to nonelderly adults with disabilities by state.

Background

The U.S. Census Bureau defines a disability as a limitation in vision, hearing, mobility, cognitive functioning, self-care, and/or independent living.  Self-care includes activities such as bathing and dressing.  Independent living includes the ability to move around the community, such as visiting a doctor’s office or going shopping.  People with disabilities have a range of medical and long-term care needs as a result of functional limitations.  They may need attendant care for help with getting out of bed, bathing, dressing, and eating; assistive technology such as power wheelchairs and speech generating devices; medical equipment and supplies; mental health counseling; and/or supportive services to maintain community-based housing or employment.

What Type of Health Insurance Do Nonelderly Adults with Disabilities Have?

Among people who are working, nonelderly adults with disabilities are significantly less likely to have a job that offers health insurance compared to those without disabilities (Figure 2).  Overall, nonelderly adults with disabilities are about half as likely to be working for pay than those without disabilities.  The ability of nonelderly adults with disabilities to work likely is limited as a result of their greater health needs and functional limitations, although with appropriate supports, many people with disabilities can and do work.  However, among those who are working, nonelderly adults with disabilities are significantly less likely to be working full-time and to have access to employer-sponsored health insurance.

Non-elderly adults with disabilities are significantly more likely to have Medicaid and significantly less likely to have private insurance than those without disabilities (Figure 3).  Those with disabilities are about three times as likely to be covered by Medicaid and about half as likely to have private insurance compared to those without disabilities. These differences are influenced by the greater health needs of people with disabilities and the fact that they are less likely to have access to employer-sponsored coverage.  Nonelderly adults with disabilities also are significantly less likely to be uninsured than those without disabilities, reflecting the important role that Medicaid plays in covering this population.

How Do Nonelderly Adults with Disabilities Qualify for Medicaid?

Some nonelderly adults with disabilities are eligible for Medicaid through the ACA’s Medicaid expansion, based solely on their low income, in the 32 states (including DC) that have adopted the expansion as of 2017.³   The ACA expands Medicaid eligibility to nearly all nonelderly adults with income up to 138% FPL ($16,643/year for an individual in 2017) without an asset limit and provides enhanced federal matching funds for states to cover this group.⁴  Although having a disability is not required to qualify for Medicaid under the ACA’s expansion or other poverty-related pathways, more than three in 10 (35%) Medicaid adults who are not working report that they have an disability or illness.  People with disabilities may receive Medicaid as expansion adults while they are waiting for a disability determination, which can take a substantial amount of time, or they may have a disability that does not meet Supplemental Security Income (SSI) medical criteria but is nonetheless limiting, or they may not financially qualify for SSI.⁵   Some expansion adults with disabilities are working, and a recent study found that working-age adults with disabilities are significantly more likely to be employed if their state has adopted the Medicaid expansion, compared to states that have not expanded.⁶ 

Some nonelderly adults with disabilities qualify for Medicaid through a disability-related pathway based on both their low income and functional limitations (Figure 4).⁷   All states generally must provide Medicaid to SSI beneficiaries, although this pathway does not include all people with disabilities.  Only 36% of nonelderly Medicaid adults with disabilities receive SSI,⁸  meaning that nearly 2/3 of nonelderly Medicaid adults with disabilities qualify through another pathway.  States can choose to extend Medicaid financial eligibility for people with disabilities up to 100% of the federal poverty level (FPL, $12,060 for an individual in 2017).  In addition, states can expand financial eligibility for people who need long-term care services, in nursing homes and/or in the community, up to three times the SSI benefit rate (equivalent to 219% FPL or $26,460/year for an individual in 2017).  States also can allow working people with disabilities at higher incomes to buy into Medicaid coverage by paying a premium.  The disability-related pathways also generally have asset limits set by the state, which typically are at the SSI level of $2,000 for an individual.

What are the Characteristics of Non-Elderly Adults with Disabilities Receiving Medicaid?

Nonelderly adults with disabilities who receive Medicaid have very low incomes, with more than half living below the federal poverty level (Figure 5).  Nearly 85% of nonelderly Medicaid adults with disabilities have incomes below 200% FPL (less than $24,120/year for an individual in 2017).  Their low incomes reflect the fact that people with disabilities generally are poorer than those without disabilities as well as the low income eligibility limits associated with the Medicaid program. Nearly six in 10 nonelderly adults with disabilities receiving Medicaid are white, about one-quarter are black, and 16% are Hispanic.  Just under three in 10 are between ages 55 to 64, and another three in 10 are between ages 45 to 54.  About one-fifth are between ages 35 to 44, and one-quarter are ages 18 to 34.

Nonelderly adults with disabilities who receive Medicaid are significantly more likely to have multiple functional limitations compared to people with disabilities who have private insurance or those who are uninsured (Figure 6).  About one-third of nonelderly Medicaid adults with disabilities have three or more functional limitations, more than two and one-half times the rate of privately insured nonelderly adults with disabilities and more than double the rate for those who are uninsured.  By contrast, nonelderly Medicaid adults with disabilities are significantly less likely to have limited functioning in only one area compared to those who are privately insured.  For an example of a Medicaid beneficiary with multiple functional limitations, see Nicholas’s story below.

Nicholas, age 33, North Carolina

Nicholas was diagnosed with multiple sclerosis at age 29, and the disease is advancing.  He cannot walk more than a few feet, his hearing and vision are impaired, and he has difficulty using his hands.  He experiences a lot of pain on a daily basis.  Medicaid covers the medications he takes to control symptoms, doctor visits, and his power wheelchair.  Nicholas lives in an apartment that is physically inaccessible, and he was injured after a fall when he was trying to transfer from the bathroom to his wheelchair.  His Medicaid case plan includes help with making the shower accessible, home delivered meals, and a car hook-up for his power wheelchair so that he can more easily go out into the community.

Nonelderly adults with disabilities receiving Medicaid are significantly more likely to use health care services than those with private insurance or those who are uninsured (Figure 7).  They are four times as likely to receive nursing or other health care at home, nearly twice as likely to be hospitalized overnight, and more than one and one-half times as likely to have 10 or more health care visits in a year compared to those with private insurance.  This increased utilization likely reflects their greater health needs and functional limitations as well as the fact that certain services, such as care at home, are more likely to be covered by Medicaid than by private insurance.

Nonelderly adults with disabilities receiving Medicaid are significantly more likely to have lower out-of-pocket medical costs compared to people with disabilities who are privately insured (Figure 8).  Those with Medicaid are nearly three times as likely to incur costs of less than $500 per year compared to those with private insurance.  Those with Medicaid also were more than three times less likely to have the highest out-of-pocket costs (more than $5,000/year) compared to those with private insurance.  These differences are likely due to Medicaid’s cost-sharing rules, which are designed to protect people with low incomes from burdensome out-of-pocket costs, which can create financial barriers to receiving needed care.  The lower out-of-pocket costs among people with Medicaid also reflect their low incomes.  People with out-of-pocket medical expenses whose income exceeds Medicaid financial eligibility limits may become eligible for Medicaid through a “spend down’ by subtracting incurred medical expenses from their income.

What Services Does Medicaid Provide for Nonelderly Adults with Disabilities?

Medicaid covers a broad range of medical and long-term care services to meet the diverse needs of nonelderly adults with disabilities (Figure 9).   Through Medicaid, nonelderly adults with disabilities have access to regular preventive care as well as medical care for illnesses and chronic conditions.  States must provide certain minimum services for adults, such as inpatient and outpatient hospital, physician, lab and x-ray, and nursing home services.  States also can choose to provide a broad range of optional services, many of which are important to people with disabilities, such as prescription drugs, physical therapy, private duty nursing, personal care, rehabilitative services, and case management.  Most home and community-based services (HCBS) are provided at state option.⁹ 

Medicaid provides long-term care services that support people with disabilities’ ability to live independently and safely in the community.  Medicaid is the primary payer for long-term services and supports and plays a key role in helping states finance services to meet their community integration obligations under the Americans with Disabilities Act and the Supreme Court’s Olmstead decision.¹⁰  Many of these services are unavailable through private insurance, and they are too costly for people to afford out-of-pocket, especially given the low incomes of most people with disabilities.  For example, Medicaid covers personal and attendant care services that assist people with disabilities with the tasks necessary for daily living, such as eating, bathing, dressing, preparing meals, and going grocery shopping.¹¹   Medicaid also covers habilitative services that help people with disabilities learn independent living skills; assistive technology, such as lifts, wheelchairs, and speech-generating devices; supportive housing services that help people with disabilities obtain and retain community housing; and community-based mental health services, which help people with mental illness remain out of institutions.  For an example of a Medicaid beneficiary relying on long-term mental health services, see Bill’s story below.

Bill, age 42, Arizona

Bill was going to school and working as a nurse when he had his first mental break at age 27.  While he was hospitalized, he was diagnosed with bipolar disorder and found out that he qualified for Medicaid. Medicaid covers the day treatment program that he attends for 30 hours a week, along with medication and doctor visits, to manage his condition.  Bill says that Medicaid means he has “been given a second chance.”

Medicaid covers services that enable people with disabilities to work.  In addition to providing personal care and transportation services that help people with disabilities get ready for the day and get to work, states also can cover supported employment services, such as job coaching, to help people with disabilities work in the community.

The ACA offers states new and expanded options to provide community-based long-term care services.  The ACA created the Community First Choice option to provided attendant care services and supports with 6% enhanced federal matching funds; eight states elect this option as of 2016.  The ACA also expanded states’ ability to offer HCBS through the Section 1915 (i) option, which allows states to serve people with functional limitations that do not yet rise to an institutional level of care.  This enables states to offer services in efforts to retain people in the community and prevent the need for costlier and more intensive services in the future.  As of 2015, 17 states are using this option to provide targeted services to particular populations, such as people with mental health needs and those with I/DD.  The ACA also created the Medicaid health homes option, which enables states to provide care coordination services for people with chronic conditions at a 90% enhanced federal match for the first two years; 22 states offer health homes as of 2016.  For an example of a beneficiary receiving Medicaid community-based services, see Curtis’s story below.

Curtis, age 20, Kansas

Curtis lives with his mother and is diagnosed with autism, intellectual disabilities, and sensory integration issues.  He functions on the level of a 2nd or 3rd grader and recently has started to read. While he has a very easy-going personality, he cannot be left alone and needs help with shaving, bathing, and taking medication. Medicaid provides attendant care services that help him to learn basic life skills at home, such as making his bed and dusting his room, while his mother is at work. His attendant also accompanies him to the library, to get his hair cut, to community events, and to the book store where his favorite activity is looking at picture books.

Medicaid helps make coverage affordable for nonelderly adults with disabilities.  Federal minimum standards limit Medicaid beneficiaries’ out-of-pocket costs to prevent them from encountering financial barriers to accessing necessary care.  States can charge premiums to those with incomes above 150% FPL.  Copayment levels vary based on income and generally are limited to nominal amounts for those below poverty.  People who receive long-term care services, such as those in nursing homes, may be required to contribute most of their income toward the cost of their care, except for a small personal needs allowance to pay for items that Medicaid does not cover, such as clothing.

How Much Does Medicaid Spend on People with Disabilities?

Medicaid spending on people with disabilities is disproportionate to their enrollment in the program (Figure 10).  Unlike the data presented earlier in this paper, which are limited to nonelderly adults with disabilities, the Medicaid enrollment and spending data for people with disabilities in this section include both nonelderly adults and children with disabilities.  As of 2011, people with disabilities accounted for 15% of total Medicaid enrollment but 42% of program spending.  This discrepancy is likely due to their greater health needs and more intensive use of services compared to those without disabilities.

Per enrollee spending for people with disabilities is substantially higher than for those without disabilities (Figure 11).  Per enrollee spending for people with disabilities totaled $16,643 in 2011, more than five times higher than for adults without disabilities ($3,247) and nearly seven times higher than for children without disabilities ($2,463).  Some of this difference is due to people with disabilities’ greater use of both institutional and community-based long-term care services ($6,137) compared to those without disabilities, who have per enrollee long-term care spending at negligible amounts.  This is because some people enrolled in Medicaid through a pathway based solely on their low income also have disabilities and use long-term care services.  In addition to having higher per enrollee long-term care spending, people enrolled through a disability-related pathway also have higher per enrollee spending for acute care services ($10,505), more than eight times that of adults without disabilities ($3,234) and more than four times that of children without disabilities ($2,399).

Per enrollee spending for people with disabilities varies substantially by state, ranging from $10,142 in AL to $33,808 in NY in 2011 (Figure 12 and Appendix Table).  Over half of states spend between $15,000 to $19,999 per enrollee for people with disabilities, and another third of states spend between $20,000 to $34,999 per enrollee for people with disabilities.  This variation is due to state choices about eligibility and services, as many disability-related coverage pathways and most home and community-based long-term care services are offered at state option.

As of 2011, 80% of nonelderly people with disabilities who use Medicaid long-term care are served in the community, with the remaining 20% in institutions (Figure 13).  Over the last several decades, states have made substantial progress in serving more nonelderly adults with disabilities in the community instead of in nursing homes.  HCBS typically are less expensive than nursing homes and are preferred by many nonelderly adults with disabilities.  However, state Medicaid programs must cover nursing home services, while nearly all HCBS are optional, making HCBS vulnerable to cuts if states are faced with funding reductions.

Looking Ahead

Medicaid covers more than three in 10 nonelderly adults with disabilities, providing a broad range of medical and long-term care services to meet their diverse needs and making coverage affordable.  Over half of nonelderly Medicaid adults with disabilities live below the federal poverty level, and nearly 85% have incomes of less than 200% FPL, or $24,120 per year for an individual in 2017.  They are significantly more likely to use health care services, including care at home, and to have lower out-of-pocket medical costs than nonelderly adults with disabilities who are privately insured, reflecting their lower incomes and greater functional limitations.  Medicaid spending for this population is disproportionate to their program enrollment, due to their greater health needs and reliance on Medicaid for expensive but necessary services, especially long-term care in the community and nursing homes, that are generally unavailable through private insurance and too costly to afford out-of-pocket.

The American Health Care Act, as approved by the House Energy and Commerce Committee, would fundamentally change Medicaid’s financing structure to a per capita cap, resulting in an estimated $880 billion reduction in federal Medicaid spending from 2017 to 2026.  People with disabilities may be especially affected by a per capita cap as most disability-related coverage pathways and many services important to people with disabilities, such as community-based long-term care, are provided at state option, making them subject to cuts if states are faced with federal funding reductions.  States’ delivery system reform efforts that affect people with disabilities, such as those that seek to improve care coordination; integrate physical, behavioral health, and long-term care services; and serve more people in the community instead of institutions also could be curtailed if federal Medicaid funding is limited.

While per enrollee Medicaid spending for both acute and long-term care services is substantially higher for people with disabilities compared to those without disabilities, it also varies substantially across states.  A per capita cap, as proposed in the American Health Care Act, could lock in historical state differences in scope of coverage and spending for people with disabilities. Changing Medicaid financing to a per capita cap also ties spending levels to a base year, which does not account for future spending increases due to new drug therapies or other medical advances yet to be developed, which could offer important new treatments to improve the lives and functioning of people with disabilities.

The American Health Care Act also ends enhanced federal funding for the ACA’s Medicaid expansion and the Community First Choice (CFC) option, both of which can be important to people with disabilities.  The ACA’s Medicaid expansion provides a pathway through which some nonelderly adults with disabilities newly gained coverage, while CFC allows states to offer attendant care services and supports that help people with disabilities live and work in the community.  As approved by the House Energy and Commerce Committee, the AHCA repeals the enhanced federal matching funds for the ACA’s expansion enrollees as of January 1, 2020, except for those enrolled as of December 31, 2019, who do not have a break in eligibility of more than one month, and ends the enhanced federal matching funds for CFC as of January 1, 2020.  These reductions in federal funding could jeopardize states’ ability to continue to finance the expansion coverage pathway and Community First Choice services.

Unlike nursing home services which are required, many home and community-based services important to people with disabilities are provided through optional waivers, which put them at risk for cuts as states look for ways to limit spending in the face of substantial federal funding reductions. In addition, new flexibility granted to states under Section 1115 Medicaid expansion waivers could further complicate coverage for nonelderly adults with disabilities. Existing Medicaid expansion waivers include provisions not otherwise permitted under current Medicaid rules, such as premiums, copayments above statutory limits, healthy behavior incentives, and health savings accounts.  People who are considered “medically frail” are exempt from some of these rules, which can create additional administrative burdens for states, health plans, and beneficiaries in making sure that people with disabilities are appropriately recognized.

Under a per capita cap and the elimination of enhanced federal financing as proposed in the AHCA, states will be faced with budgetary pressures resulting from the substantial reduction in federal Medicaid funds.  At the same time, states will continue to be required to meet their community integration obligations for people with disabilities under the Americans with Disabilities Act and the Supreme Court’s Olmstead decision.  Medicaid has played a key role in financing community-based long-term care services, but its ability to continue to do so may be limited by the loss of the CFC enhanced funding for attendant care services, along with the broader reductions in federal Medicaid funds under a per capita cap, thereby increasing budgetary pressures on states.  Because the AHCA could have significant consequences for enrollees and states, its potential implications warrant careful consideration for their impact on nonelderly Medicaid adults with disabilities.

1. Kaiser Family Foundation analysis of 2015 National Health Interview Survey (NHIS) data. The total number of nonelderly adults with disabilities is likely higher as people living in long-term care facilities are excluded from NHIS. ↩︎
2. As approved by the House Energy and Commerce Committee, the bill would eliminate the ACA’s enhanced federal matching funds for expansion enrollees as of January 1, 2020, except for those enrolled as of December 31, 2019, who do not have a break in eligibility of more than one month. ↩︎
3. Applicants have their Medicaid eligibility assessed solely on the basis of income as the first step in the eligibility determination process. To the extent that different coverage groups receive different benefit packages, people who qualify for Medicaid both as an expansion adult and in a disability-related group can choose to remain in the expansion group or change their enrollment to the disability-related group so that they can access the benefit package that best meets their needs.  42 C.F.R. § 435.911(c)(2). ↩︎
4. The expansion is optional for states, as a result of the Supreme Court’s ruling on the ACA’s constitutionality.  Prior to the ACA, there was not a Medicaid coverage pathway for childless adults regardless of how low their income was, and this remains the case in the 19 non-expansion states, except for Wisconsin, which covers childless adults up to 100% FPL. ↩︎
5. To be eligible for SSI, beneficiaries must have low incomes, limited assets, and a significant disability that impairs their ability to work at a substantial gainful level. ↩︎
6. Jean P. Hall, et al., Effect of Medicaid Expansion on Workforce Participation for People with Disabilities, 107 Amer. J. Pub. Health 262-264 (Feb. 2017), doi: 10.2105/AJPH.2016.303543. ↩︎
7. These pathways were not changed by the ACA. ↩︎
8. Kaiser Family Foundation analysis of 2015 NHIS data. ↩︎
9. States also can provide HCBS through waivers which allow them to target particular populations and limit enrollment. ↩︎
10. In Olmstead, the Supreme Court held that the unjustified institutionalization of people with disabilities is illegal discrimination under the ADA. ↩︎
11. States have the option to allow beneficiaries to self-direct their services by selecting or dismissing personal care attendants and/or allocating their service budgets among available services. ↩︎

KEY FINDINGS:

• The latest Kaiser Health Tracking Poll examines the public’s early attitudes towards the House Republican plan to replace the Affordable Care Act and finds that more expect the new plan will make things worse rather than better when it comes to the number of people with coverage and costs for those buying insurance on their own. Republicans are more optimistic, with nearly half expecting the plan to decrease costs and one-third saying it will increase the number of people covered.
• More expect the law to increase rather than decrease costs in the individual insurance market for various groups, including younger and older people, those living in urban and rural areas, and those with lower incomes. The public is more divided on whether costs will increase or decrease for higher-income people who buy their own insurance.
• Three-quarters of the public support continuing current federal Medicaid funding to Planned Parenthood to pay for non-abortion services received by people on Medicaid, while 22 percent support cutting off all federal funding to the group. Majorities of both Republican women and Republican men support continuing current funding.
• At the same time, only one-third of the public is aware that there is currently a ban on federal Medicaid funds being used to pay for abortions, while almost half say there is not a ban and one in five say they don’t know.
• After shifting in a positive direction in February, the public’s overall views of the ACA hold steady this month, with 49 percent holding a favorable view of the 2010 law and 44 percent holding an unfavorable view. The public is split on the question of repeal, with 51 percent saying Congress should not vote to repeal the law and 45 percent saying they should. As previous Kaiser surveys have found, more say lawmakers should wait to vote on repeal until the details of a replacement plan are worked out (24 percent) than say they should repeal the law immediately and work out the details of a replacement plan later (19 percent).

Public Opinion on ACA Replacement Plan

On March 6, 2017, House Republicans announced their plan to replace the Affordable Care Act (ACA) – the American Health Care Act (AHCA). While Americans are still learning about the details of this plan, the latest Kaiser Health Tracking Poll (which was in the field before the Congressional Budget Office released its analysis of the plan) examines the public’s first impressions of its likely effects on the number of uninsured, those who buy insurance on their own, and those with pre-existing conditions.

About half (48 percent) the public thinks the Republican replacement plan will decrease the number of people who have health insurance, while one in five (18 percent) say the number of insured people will increase and three in ten say it will stay about the same. Notably, the public is confused about the effect the current health care law has had on coverage rates in the U.S.; while four in ten (41 percent) know that the share of people who are uninsured has decreased since the ACA was passed, nearly as many (31 percent) believe it has increased and 26 percent think it has remained about the same.

When it comes to costs, about half (48 percent) expect the new plan to increase costs for people who buy insurance on their own, while about a quarter each say costs for these people will decrease (23 percent) or stay about the same (25 percent). Similarly, four in ten (41 percent) expect the replacement plan to provide health insurance with higher deductibles, while a quarter each think deductibles will decrease (25 percent) or stay the same (23 percent). When it comes to protections for people with pre-existing conditions, nearly half the public (45 percent) expects the replacement plan would provide similar protections to the current law, but about twice as many say the plan would offer fewer protections as say it would offer more (32 percent versus 15 percent).

Not surprisingly, partisans hold very different views of how the replacement plan would impact coverage and costs. While majorities of Democrats believe the plan will decrease the number of people with insurance and increase costs, Republicans’ views are more mixed. Nearly half of Republicans (46 percent) think the plan will decrease costs for people who buy their own insurance, four in ten (41 percent) think it will provide insurance with lower deductibles, and one-third (34 percent) think it will increase the number of people with health insurance.

Expected Changes in Costs for Different Groups Under Replacement Plan

A Kaiser Family Foundation analysis has shown that the tax credit structure under the replacement plan would generally increase costs for certain groups of people buying insurance in the non-group market and decrease costs for others. For example, compared with current law, lower-income individuals, older adults, and those living in rural areas would pay more for insurance on average under the replacement plan, while younger adults, those with higher incomes, and those living in urban areas would pay less.¹ 

However, the public does not see a distinction in how the replacement plan would impact costs for these different groups of individuals, except when it comes to income. About four in ten Americans expect costs to increase for younger, older, urban-dwelling, and rural-dwelling people buying insurance on the individual market, while between about two in ten and one quarter expect them to decrease. Similarly, almost half (46 percent) think costs would increase for lower-income people buying their own insurance, while one quarter think they would decrease. Views are more divided on how higher-income people buying their own insurance would be affected, with over four in ten (45 percent) saying their costs would be about the same, and about a quarter each saying they would increase (25 percent) or decrease (27 percent).

Public Opinion on the ACA

These views of the replacement plan are set against a backdrop of a public that continues to be divided in its view of the ACA. After shifting in a more favorable direction last month, public opinion on the current health care law mostly holds steady this month, with 49 percent having a favorable view and 44 percent unfavorable. Partisanship continues to underlie these views, with 77 percent of Republicans expressing an unfavorable view, 73 percent of Democrats expressing a favorable one, and independents evenly divided (46 percent favorable, 47 percent unfavorable).

The public also remains divided in their views of repeal, with half (51 percent) saying lawmakers should not vote to repeal the law and 45 percent saying they should vote to repeal it. Of those who want to see Congress vote to repeal the law, more say they want lawmakers to wait to vote to repeal the law until the details of a replacement plan have been worked out (24 percent) than say Congress should vote to repeal the law immediately and work out the details of a replacement plan later (19 percent).

Again, these views are largely driven by party identification, with the vast majority of Republicans (81 percent) saying Congress should vote to repeal the law and a similar share of Democrats (78 percent) saying they do not want the law repealed. Republicans are about evenly divided between those who want Congress to wait to vote on repeal until the details of a replacement plan have been worked out (41 percent) and those who say they should vote to repeal the ACA immediately and work out the details of a replacement plan later (36 percent).

Views on Planned Parenthood Funding

The House replacement bill includes a provision that would block Planned Parenthood for one year from receiving federal funds to provide services to low-income people on Medicaid, including non-abortion services like contraception, testing and treatment for sexually transmitted infections, and cancer screenings. Seventy-five percent of the public thinks Medicaid should continue to pay Planned Parenthood for non-abortion services, while 22 percent think all federal payments to the organization should be stopped.

While there are partisan and gender differences on this question, majorities across groups, including just over half of Republican men (55 percent) and Republican women (57 percent) say that the federal government should continue to pay Planned Parenthood for non-abortion services received by people on Medicaid.

Americans’ attitudes towards federal payments to Planned Parenthood through the Medicaid program are largely stable, with few being persuaded from their opinions after hearing arguments for and against continued funding. Among those who think all federal Medicaid payments to Planned Parenthood should be stopped, about one-third (7 percent of the public overall) change their mind after hearing that this would make it difficult for many lower-income women to access certain health services, including treatment for sexually transmitted infections, cancer screenings, and birth control. On the other hand, a small share of those who initially support continuing current federal Medicaid payments to Planned Parenthood (10 percent overall) shift their opinion to oppose these payments after hearing that even though no federal funding goes directly to abortion services, Planned Parenthood does provide and refer women for abortions.

When asked about services currently provided by Planned Parenthood, most are aware of the variety of health care services provided by the organization including contraception (87 percent), treatment and testing for sexually transmitted infections (80 percent), cancer screenings (68 percent), and abortions (68 percent). Similar shares of women and men are aware that Planned Parenthood provides each of these services.

Awareness is lower when it comes to the question of whether federal funds can be used to pay for abortion services. While the Hyde Amendment currently prohibits federal Medicaid funds from being used to pay for abortions, just one-third (33 percent) of the public is aware of this ban, while almost half (45 percent) say there is not a ban and one in five (21 percent) say they don’t know.

Republicans are less likely than Democrats to be aware that there is a ban on federal funds being used to pay for abortions. One-quarter of Republican women (25 percent) and Republican men (26 percent) are aware of the ban, compared with about four in ten women and about a third of men who identify as Democrats and independents.

Government Role in Women’s Health

Beyond the question of federal funding for Planned Parenthood, more generally, a majority of the public says it is important that the federal government provides funding for reproductive health services including family planning and birth control for lower-income women. While a large majority of Democrats (84 percent) and over half of independents (56 percent) say this is “very” important, the share is much smaller among Republicans (36 percent).

Large majorities of the public also say it is important (either “very” or “somewhat”) for several of the ACA’s requirements affecting women and children to be kept in place in any replacement plan. Roughly eight in ten say that if lawmakers decide to repeal and replace the 2010 health care law, it is “very” important that insurance companies cannot deny coverage to pregnant women (85 percent), that all private health plans must cover the full cost of preventive care with no cost-sharing for women (81 percent) and children (80 percent), and that insurance companies cannot charge women more than men for the same policy (79 percent). Fewer – 54 percent –  say it is very important to keep in place the requirement for health plans to provide birth control with no cost-sharing, but three-quarters (78 percent) say this is at least “somewhat” important.

Large majorities of women, regardless of political party, say keeping most of these requirements in place is very important. One exception is the requirement that private health plans cover the cost of birth control with no cost-sharing, which majorities of women who are Democrats and independents view as very important compared with a smaller share of Republican women (38 percent). Men’s views are more divided by partisanship, with substantially higher shares of men who identify as Democrats saying each requirement is very important compared with Republican men. Three in ten Republican men (31 percent) say it is very important that the requirement for private health plans to cover the full cost of birth control is kept in place if the ACA is repealed.

A large majority of the public (89 percent) also supports the ACA’s requirement that all private health plans must include coverage for maternity care,  including 86 percent of Republicans and 94 percent of Democrats.

Views on this requirement are somewhat malleable. When those who initially support the requirement are read the argument that it means some people have to pay for benefits they do not use, about a quarter (22 percent of the public overall) change their mind so that 31 percent overall oppose the requirement. Still, a solid majority of 65 percent continues to support the requirement for all private health plans to include coverage for maternity care after this argument is read.

Buying Insurance Across State Lines

President Trump and Republican leaders in Congress have supported the idea of allowing insurers to sell their plans across state lines as a way to drive down insurance costs, though this provision is not included in the current House repeal and replace legislation.²  Overall, nearly two-thirds of Americans (64 percent) generally favor this proposal while three in ten (28 percent) oppose it. These attitudes are fairly malleable with both those who favor and oppose this policy being persuaded after hearing counter-messages.

Among the 28 percent who originally oppose this policy, four in ten (12 percent of the public overall) change their opinion after hearing that allowing the sale of insurance across state lines could encourage competition and decrease costs for consumers, increasing the total share who favor the policy to 76 percent.

On the other side, among the 64 percent who originally said they favor the policy, about half (30 percent of the public overall) change their opinion after hearing the argument that allowing the sale of insurance across state lines could encourage insurers to operate from states with fewer consumer protections and make it harder for people to get help if they have a problem with their plans, increasing the total share who oppose selling insurance across state lines to 57 percent.

Awareness of Key ACA Provisions

As the focus of debate moves away from the ACA itself and on to potential replacement plans, the public’s awareness of what the 2010 health care law actually does is somewhat uneven. While large majorities are aware of many of the ACA’s key provisions, others are less well-known. A large majority of the public is aware that the ACA allows young adults to stay on their parents’ insurance until age 26 (83 percent) and that it requires nearly all Americans to have health insurance or pay a fine (82 percent). Between six and seven in ten are aware the law requires insurance plans purchased by individuals to include a minimum package of benefits (71 percent), that it provides financial help to low-and moderate-income Americans who do not get coverage through an employer (69 percent), that it gives states the option of expanding Medicaid (65 percent), that it prohibits insurers from denying coverage to those with pre-existing conditions (60 percent), and that it requires health insurance plans to cover the cost of children’s preventive services, including immunizations, with no cost-sharing (59 percent).

Smaller shares – about half the public – know that the law eliminates out-of-pocket costs for preventive health care services (53 percent) and birth control (47 percent). Fewer still – 42 percent – are aware that the law prohibits insurance companies from charging women more than men.

At the same time, misconceptions about the law remain. Half (50 percent) incorrectly believe the ACA allows undocumented immigrants to receive financial help from the government to buy health insurance, and four in ten (40 percent) mistakenly think it cuts benefits for people in the traditional Medicare program.

This Kaiser Health Tracking Poll was designed and analyzed by public opinion researchers at the Kaiser Family Foundation (KFF). The survey was conducted March 6-12, 2017, among a nationally representative random digit dial telephone sample of 1,206 adults ages 18 and older, living in the United States, including Alaska and Hawaii (note: persons without a telephone could not be included in the random selection process). Computer-assisted telephone interviews conducted by landline (421) and cell phone (785, including 465 who had no landline telephone) were carried out in English and Spanish by Princeton Data Source under the direction of Princeton Survey Research Associates International (PSRAI). Both the random digit dial landline and cell phone samples were provided by Survey Sampling International, LLC. For the landline sample, respondents were selected by asking for the youngest adult male or female currently at home based on a random rotation. If no one of that gender was available, interviewers asked to speak with the youngest adult of the opposite gender. For the cell phone sample, interviews were conducted with the adult who answered the phone. KFF paid for all costs associated with the survey.

The combined landline and cell phone sample was weighted to balance the sample demographics to match estimates for the national population using data from the Census Bureau’s 2015 American Community Survey (ACS) on sex, age, education, race, Hispanic origin, and region along with data from the 2010 Census on population density. The sample was also weighted to match current patterns of telephone use using data from the January-June 2016 National Health Interview Survey. The weight takes into account the fact that respondents with both a landline and cell phone have a higher probability of selection in the combined sample and also adjusts for the household size for the landline sample. All statistical tests of significance account for the effect of weighting.

The margin of sampling error including the design effect for the full sample is plus or minus 3 percentage points. Numbers of respondents and margins of sampling error for key subgroups are shown in the table below. For results based on other subgroups, the margin of sampling error may be higher. Sample sizes and margins of sampling error for other subgroups are available by request. Note that sampling error is only one of many potential sources of error in this or any other public opinion poll. Kaiser Family Foundation public opinion and survey research is a charter member of the Transparency Initiative of the American Association for Public Opinion Research.

1. C Cox, G Claxton, L Levitt, How Affordable Care Act Repeal and Replace Plans Might Shift Health Insurance Tax Credits. https://modern.kff.org/health-reform/issue-brief/how-affordable-care-act-repeal-and-replace-plans-might-shift-health-insurance-tax-credits/   ↩︎
2. The ACA actually includes a provision allowing nongroup insurance plans to be offered in more than one state under interstate compacts. For more information, see Levitt, L, JAMA Forum: Those Pesky Lines Around States, October 2016. https://newsatjama.jama.com/2016/10/19/jama-forum-those-pesky-lines-around-states/   ↩︎

For Medicare patients, hospitalizations can be stressful; even more so when they result in subsequent readmissions. While many readmissions are unavoidable, researchers have found wide variation in hospitals’ readmission rates, suggesting that patients admitted to certain hospitals are more likely to experience readmissions compared to other hospitals.¹  A number of studies show that hospitals can engage in several activities to lower their rate of readmissions, such as clarifying patient discharge instructions, coordinating with post-acute care providers and patients’ primary care physicians, and reducing medical complications during patients’ initial hospital stays.² 

Through Congressional direction and previous Administration initiatives, Medicare has begun implementing incentives to reduce hospital readmissions. One example, and the focus of this Issue Brief, is the Hospital Readmission Reduction Program (HRRP), which financially penalizes hospitals with relatively high rates of Medicare readmissions.³  This Issue Brief, updated with new estimates for 2017, describes the HRRP, analyzes the program’s impact on Medicare patients and hospitals, examines the decline in national readmission rates, and discusses implementation issues that policymakers have raised.

Main Findings

• Total Medicare penalties assessed on hospitals for readmissions will increase to $528 million in 2017, $108 million more than in 2016. The increase is due mostly to more medical conditions being measured. Hospital fines will average less than 1 percent of their Medicare inpatient payments.
• For 2017, 78 percent of Medicare patient admissions are projected to be in hospitals receiving either no readmission penalty or penalties of less than 1 percent of the hospital’s Medicare inpatient payments. Fewer than 2 percent of Medicare patient admissions will be in hospitals receiving the maximum financial penalty.
• Nationally, beneficiary readmission rates started to fall in 2012, and have continued to drop since then, suggesting that hospitals and clinicians may have adopted new, system-wide interventions soon after the HRRP was enacted by the ACA.
• Across all five years of the HRRP, certain types of hospitals are more likely than others to incur penalties. These include hospitals with relatively higher shares of low-income beneficiaries and major teaching hospitals. Congress recently enacted legislation to incorporate a socioeconomic adjustment in how hospital performance is measured, based on each hospital’s share of inpatients who are dually qualified for Medicare and full Medicaid. The implementation of this methodology will need to consider several key policy issues.

Background: Defining a hospital readmission

Generally speaking, a hospital readmission occurs when a patient is admitted to a hospital within a specified time period after being discharged from an earlier (initial) hospitalization. For Medicare, this time period is defined as 30 days, and includes hospital readmissions to any hospital, not just the hospital at which the patient was originally hospitalized. Medicare uses an “all-cause” definition of readmission, meaning that hospital stays within 30 days of a discharge from an initial hospitalization are considered readmissions, regardless of the reason for the readmission. This all-cause definition is used in calculating both the national average readmission rate and each hospital’s specific readmission rate. Starting in 2014, CMS began making an exception for planned hospitalizations (such as a scheduled coronary angioplasty) within the 30-day window; these are no longer counted as readmissions.

The current focus in the HRRP is on readmissions occurring after initial hospitalizations for selected conditions—namely, heart attack, heart failure, pneumonia, chronic obstructive pulmonary disease (COPD), elective hip or knee replacement, and coronary artery bypass graft (CABG). CMS also collects hospitals’ overall readmission rates (regardless of initial diagnoses), but these overall rates are not currently used in the HRRP to calculate readmissions penalties.

The first five years of the Hospital Readmission Reduction Program

The HRRP was established by a provision in the Affordable Care Act (ACA) requiring Medicare to reduce payments to hospitals with relatively high readmission rates for patients in traditional Medicare. Starting in 2013 as a permanent component of Medicare’s inpatient hospital payment system (i.e., not a temporary demonstration project), the HRRP applies to most acute care hospitals. Exempt hospitals include psychiatric, rehabilitation, long term care, children’s, cancer, and critical access hospitals, as well as all hospitals in Maryland.⁴ 

Under the HRRP, hospitals with readmission rates that exceed the national average are penalized by a reduction in payments across all of their Medicare admissions—not just those which resulted in readmissions. Before comparing a hospital’s readmission rate to the national average, CMS adjusts for certain demographic characteristics of both the patients being readmitted and each hospital’s patient population (such as age and illness severity). After these adjustments, CMS calculates a rate of “excess” readmissions, which links directly to the hospital’s readmission penalty—the greater each hospital’s rate of excess readmissions, the higher its penalty.⁵  Each year, CMS releases each hospital’s penalty for the upcoming year in the Federal Register and posts this information on its Medicare website.

For the first year, fiscal year 2013, the maximum penalty was 1 percent of the hospital’s base Medicare inpatient payments, increasing to 2 percent for 2014, and was fully phased-in at 3 percent starting in 2015 (Table 1).⁶  When calculating each hospital’s readmission rate, CMS uses three full years of hospital data. Accordingly, the upcoming 2017 penalties were based on hospital readmissions that occurred from July 2012 through June 2015.

For penalties levied in 2013 and 2014, CMS focused on readmissions after initial hospitalizations for three selected conditions: heart attack, heart failure, and pneumonia. For penalties levied in 2015, CMS added COPD, and elective hip or knee replacement.  For 2017 penalties, CMS expanded the types of pneumonia cases that were assessed and calculated readmission rates following CABG surgery.

total fines will increase in 2017 to over $500 million, due to higher average penalties resulting from more medical conditions being measured

Across all hospitals, the average 2017 penalty will be a 0.58-percent reduction in base Medicare payments for all inpatient admissions.  Among only penalized hospitals, the average fine will be a 0.74-percent reduction in Medicare inpatient payments, a 13-percentage point increase from the prior year (Table 1). With these increases, CMS estimates that total penalties across all hospitals will total $528 million, $108 million more than in 2016. The share of hospitals receiving a penalty has remained relatively flat (78 to 79 percent) over the past three years.

The increase in average and total penalties for 2017 is due mostly to a larger number of medical conditions included in the calculations for the 2017 readmission penalty.  Specifically, as described above, CMS added CABG and expanded the cohort of pneumonia diagnoses to the list of initial diagnoses eligible for assessing hospital performance on readmissions.  A similar increase in average penalties occurred in 2015 when CMS added two diagnoses (COPD and hip or knee replacement) to the list of initial diagnoses being assessed for readmission, which was found to play a larger role in penalty increases than phasing in the maximum penalty.

A second reason that average penalties have not fallen stems from the statutory requirement that penalty assessments be based on average hospital performance, rather than on fixed targets.  In other words, CMS assesses hospital penalties based on a curve, resulting in a certain percentage of hospitals always be penalized, regardless of improvements in national readmission rates.  Advantages and disadvantages of changing this formula are discussed later in this brief.

Most Medicare patients stay in hospitals with either no penalty or penalties less than 1 percent of their Medicare inpatient payments

For a patient perspective, our analysis finds that for 2017, most beneficiaries will stay in hospitals with low to no penalties.  Specifically, we estimate that 78 percent of beneficiary stays in 2017 will be in hospitals that scored well enough during their previous measurement period to receive either no penalty or penalties equaling less than 1 percent of the hospital’s Medicare inpatient payments. (Figure 1).  Conversely, 22 percent of beneficiary stays will be in hospitals that will receive fines greater than 1 percent of their Medicare payments for inpatient admissions. About 1 percent of Medicare admissions will occur in hospitals that will receive the maximum penalty, a 3-percent reduction in Medicare payments across all inpatient admissions.

Readmission penalties continue to vary by hospital characteristics

Analysis of the variation in penalties by type of hospital suggests that Medicare beneficiaries who go to certain types of hospitals—namely major teaching hospitals and hospitals with relatively greater shares of low-income beneficiaries—are more likely to stay in penalized hospitals and hospitals with higher penalties (Table 2).  This finding is consistent with previous research.⁷  To some degree there is overlap among these two types of hospitals, as major teaching hospitals often serve as safety-net hospitals with higher proportions of low-income patients.  Across all years, hospitals with the smallest share of low-income beneficiaries (quartile 1) are the least likely to be assessed any penalty at all.  For 2017, 66 percent of hospitals in the lowest quartile of low-income patients (as determined by beneficiaries with Social Security Income) will be fined a readmission penalty compared with 86 percent among hospitals with the highest share of low-income beneficiaries.  Rural hospitals also have higher rates of being penalized and higher average penalties. Variations in penalty rates by hospital characteristics have persisted across all five years of the program, although differences were greater in the first year when the maximum penalty was only 1 percent (see Appendix Table for analysis of earlier years).

Readmission rates for Medicare patients started to fall in 2012—when CMS began measuring them for the penalty program—and have declined further since

CMS has been posting individual hospital readmission rates on its Hospital Compare website, in addition to other measures of quality and patient satisfaction, since 2009. Designed for use by Medicare consumers as well as researchers, this website also provides comparisons of each hospital’s Medicare readmission performance to the national average by indicating whether the hospital is “better/worse/no different” than the U.S. National rate. In addition to readmissions following hospitalizations for selected diagnoses, the Hospital Compare website started reporting each hospital’s overall Medicare readmission rates.

Analysis of this database shows that 2012 marks the first measurable declines in readmissions (Figure 2).  Specifically, when the three-year moving average of hospital readmission rates began including data from 2012, the rates fell across all three diagnosis categories, and have continued through subsequent measurement periods.⁸  These notable drops started after Congress enacted the HRRP in the ACA, suggesting that hospitals may have initiated new interventions to lower their readmission rates during the measurement period leading up to the fines, which started in 2013. Additionally, other clinician activities outside the hospitals may have contributed to the decline in readmissions. With the decline in the rate of readmissions, the Department of Health and Human Services estimates 565,000 fewer Medicare patient readmissions from April 2010 through May 2015.⁹ 

Despite the reduction in readmission rates, average HRRP penalties have not declined, as shown earlier in Table 1. A main reason that declines in national readmission rates do not necessarily translate to lower fines is that penalty calculations are based on average hospital performance, rather than on fixed targets, as required by law. In other words, CMS must assess hospital penalties based on a curve, resulting in a certain percentage of hospitals always being penalized, regardless of improvements in national readmission rates.  Advantages and disadvantages of changing this formula are discussed below.

Ongoing policy issues

Concerns raised by researchers and hospital representatives have prompted policymakers to consider refinements to the implementation of the HRRP and look for ways to engage other health care providers and patients. Some of the refinements under discussion include changes in the way hospital measures are risk-adjusted, setting hospital performance targets, and engaging other providers.

• Risk-adjustment. While Medicare’s readmission measures are adjusted for demographic characteristics associated with higher rates of hospital readmissions (such as age), the initial HRRP statute did not allow adjustments to the penalty calculations based on socioeconomic or community-level factors.  However, subsequent to recommendations by the National Quality Forum (NQF) and the Medicare Payment Advisory Commission (MedPAC) to adjust for socioeconomic status, Congress recently modified the method that CMS will use to assess hospital performance, as outlined in the recently passed 21^st Century Cures Act.¹⁰   Specifically, the new law directs the Secretary of Health and Human Services to divide hospitals into peer groups based on similar shares of inpatients who qualify for both Medicare and full Medicaid and then determine each hospital’s performance on readmissions relative to its peer group.  This change will be effective fiscal year 2019, but the law also allows for the Secretary to consider implementing additional risk adjustments in the future.CMS has previously raised concerns regarding socioeconomic adjustment—namely, the potential that softening penalties for hospitals with lower-income patients could hold those hospitals to a lower standard, and unintentionally weaken incentives to improve health outcomes for disadvantaged patients.¹¹   In part, the peer group methodology may address some of this concern.  However, depending on how the hospital peer groups are divided in the new adjustment methodology, further complicating issues could arise, due to state variation in the criteria for Medicaid eligibility among seniors and people with disabilities.

• Moving goalposts for multiple measures. Hospital performance in the HRRP is essentially graded on a curve because the calculations for determining penalties are based on comparisons to the national average, with future modifications based on hospital peer groups, as described above.  Therefore, for each measure, if the national rate of Medicare readmissions declines—the main goal of the HRRP—then it is possible for hospitals to improve their readmission rates, but still be penalized.  Also, the shifting penalties do not capture hospital performance across all initial hospitalization diagnoses—sometimes referred to as a hospital-wide or all-condition readmission measure.On the one hand, some suggest that if Medicare established fixed target rates for an all-condition readmission measure, hospitals might have an easier time understanding the HRRP, embracing interventions to achieve those targets, and would not risk some readmissions being double-counted (in the case of CABG and heart failure, for example).¹²   On the other hand, others could argue that a fixed target for all conditions might mitigate the level of hospital improvement because it would establish a minimum performance, rather than encourage hospitals to keep up with the average for each of the selected diagnoses, especially relevant when the average improves.

• Other providers play a role. Hospital administrators and policymakers have noted that incentives to reduce readmissions should not rest on hospitals alone because other providers and the patients themselves may play important roles in this effort.  Further, hospitals may have little to no control over the care that patients receive after they are discharged from an inpatient stay.  While researchers readily acknowledge that many readmissions are not preventable, studies show that hospitals can engage in collaborative activities to lower their number of readmissions, such as clarifying patient discharge instructions, coordinating with post-acute care providers, and reducing medical complications during the patients’ initial hospital stays.¹³ 

Discussion

Our analysis finds that most Medicare patient admissions (78%) for 2017 will occur in hospitals receiving either no readmission penalty or penalties of less than 1 percent of their Medicare inpatient revenue. Given the stress and problems that patients experience when readmitted to a hospital, it may be reassuring that most beneficiary stays are in hospitals with relatively lower readmission rates.  While hospitals with higher penalty rates account for a considerably smaller proportion of beneficiary inpatient stays, to the extent that higher readmission rates signal concerns about care quality and discharge planning, patients going to higher-penalty hospitals may be at greater risk for related problems.

Overall, national readmission rates in traditional Medicare, started to decline in 2012 and have continued along that path in subsequent years. This timing suggests that hospitals may have started to implement strategies to lower their readmissions in response to the enactment of the HRRP, with the understanding that the financial penalties (starting in 2013) would be based on performance in prior years.  Further, because the declines are seen for multiple conditions, multipronged and system-based efforts to reduce readmissions are likely, rather than limited improvements in selected clinical treatments.¹⁴   Moreover, the fact that the lower readmission rates have been sustained across several years also suggests that system-wide improvement have played a role.  Simultaneously, other providers in the community may have also started focusing more attention on ways to lower hospital readmissions among their patients.

Despite these declines, our study shows that readmission rates and penalties continue to be higher among hospitals that have higher proportions of low-income Medicare patients, consistent with research conducted by MedPAC, NQF, and others.¹⁵  Health care providers and policymakers have noted that socioeconomic factors may often play a role in access to home and community support services aiding a patient’s recovery after hospitalizations.  For example, lower-income communities and families may have limited resources for reliable transportation to take patients to follow-up medical appointments, assistance with patient mobility and daily living needs during recovery, and access to foods that meet patients’ special dietary needs.  Further examination of ways to address patient and community needs in lower income areas may offer insights into ways to lower hospital readmissions among patients in hospitals with higher shares of low-income patients.

In light of the socioeconomic issues that have been raised about the HRRP, Congress recently required a new “peer group” method for assessing hospital performance that categorizes hospitals based on their proportion of inpatients who are dually eligible for Medicare and full Medicaid. While this new method aims to identify hospitals serving low-income Medicare patients, the significant variation between states in Medicaid eligibility criteria (such as income thresholds and asset limits) raises several issues.¹⁶  For example, hospitals in states with relatively stricter Medicaid eligibility criteria for seniors and people with disabilities are likely to have lower average incomes among their patients dully covered by Medicare and Medicaid—and potentially higher readmission rates—than hospitals in states with more generous Medicaid eligibility criteria. Therefore, penalties based on readmission rates within the peer groups may not fairly account for patient income if the peer groups span across states.

Similarly, any future changes to the structural financing of the Medicaid program, such as block grant proposals recently discussed in Congress, could have varying effects on states’ ability to provide Medicaid coverage to low-income seniors and people with disabilities.¹⁷   This could further complicate the ability of the HRRP to identify comparable hospital peer groups, if based exclusively on their share of inpatients dully eligible for Medicare and Medicaid.

Finally, because the HRRP affects Medicare payments to hospitals, its effect on Medicare patients is generally indirect.  To the extent that the financial penalties encourage hospitals to implement activities designed to improve care quality and lower their rate of preventable readmissions, the penalty program could be beneficial to Medicare patients and the Medicare program.  Alternatively, some have noted that reducing financial resources to lower-performing hospitals could have a negative impact on their delivery of patient care.  Regarding beneficiaries’ out-of-pocket expenses, a hospital’s penalty status has no direct effect on beneficiaries’ cost sharing during inpatient stays. For readmissions in particular, the inpatient hospital deductible ($1,316 in 2017) is waived, but beneficiaries do face other out-of-pocket liabilities—mostly in the form of coinsurance for separately billed physician services received during their stay—as they would in all inpatient stays.¹⁸ 

In the coming years

With the enactment of the HRRP in the ACA, the aim to reduce preventable hospital readmissions has gained traction among providers and policymakers.  Moreover, key programs, such as Accountable Care Organizations (ACOs), bundled-payment initiatives, and medical home programs, include provider incentives to lower hospital readmissions, either directly or indirectly.  For a description of these models and a summary of the available evidence on spending and quality outcomes, see Payment and Delivery System Reform in Medicare: A Primer on Medical Homes, Accountable Care Organizations, and Bundled Payments.  Additionally, the Community-based Care Transitions Program, also enacted by the ACA, is designed to assess ways that local organizations might partner with hospitals to improve patients’ transitions to other settings, such as skilled nursing facilities or the patients’ home.¹⁹   Also, in traditional Medicare, CMS has recently started allowing physicians to bill Medicare for “transitional care management” after a beneficiary’s discharge from a hospital or other health care facility, in an effort to reimburse physicians for follow-up activities that could reduce readmissions and other complications.

As more results become available from Medicare’s payment and delivery system reforms a key question continues to be whether and how care improves for patients with the greatest health care needs—often those who have been hospitalized one or more times during the year.  Reductions in hospital readmission may be a likely outcome on which to focus and may help policymakers identify which models to pursue more broadly.  A continuing decline in preventable readmission rates would help slow the growth in Medicare spending and may also signal improved care for patients during and after their hospitalizations.

1. Jencks, S. F. et al., “Hospitalizations among Patients in the Medicare Fee-for-Service Program,” New England Journal of Medicine Vol. 360, No. 14: 1418-1428, 2009.; Epstein, A. M. et al., “The Relationship between Hospital Admission Rates and Rehospitalizations,” New England Journal of Medicine Vol. 365, No. 24: 2287-2295, 2011. ↩︎
2. Ahmad, F. S. et al., “Identifying Hospital Organizational Strategies to Reduce Readmissions,” American Journal of Medical Quality Vol. 28, No. 4: 278-285, 2013.; Silow-Carroll, S. et al., “Reducing Hospital Readmissions: Lessons from Top-Performing Hospitals,” Commonwealth Fund Synthesis Report, New York: Commonwealth Fund, 2011.; Jack, B. W. et al., “A Reengineered Hospital Discharge Program to Decrease Hospitalization: A Randomized Trial,” Annals of Internal Medicine Vol. 50, No. 3: 178-187, 2009.; and Kanaan, S. B., “Homeward Bound: Nine Patient-Centered Programs Cut Readmissions,” Oakland, CA: California HealthCare Foundation, 2009. ↩︎
3. Other examples include several of the payment delivery system reforms launched by the CMS Innovation Center, including Pioneer Accountable Care Organizations (ACOs), bundled-payment initiatives, and the Independence at Home demonstration, all of which include provider incentives to lower hospital admission and readmission rates, either directly or indirectly.  Additionally, the Community-based Care Transitions Program, also enacted by the ACA, is designed to assess ways that community-based organizations might partner with hospitals to improve patients’ transitions to other settings, such as skilled nursing facilities or the patients’ home. ↩︎
4. Maryland hospitals have a longstanding Medicare waiver that exempt them from the IPPS and allows Maryland to set all-payer rates for its hospital services. ↩︎
5. Hospitals’ excess readmission ratios are also weighted by Medicare’s cost for admissions with the given initial diagnosis. ↩︎
6. Penalties assessed as reductions in base payments on all Medicare inpatient admissions, and do not apply to added payment adjustments, such as graduate medical education payments. ↩︎
7. Sheingold, S., R. Zuckerman, and A. Shartzer, “Understanding Medicare Hospital Readmission Rates and Differing Penalties Between Safety-net and Other Hospitals,” Health Affairs, 35, No.1, January 2016; Kahn, C., T. Ault, L. Potetz, T. Walke, J. Chambers, and S. Burch, “Assessing Medicare’s Hospital Pay-For-Performance Programs And Whether They Are Achieving Their Goals,” Health Affairs, 34, No.8, August 2015; Joynt, K. and A. Jha, “A Path Forward on Medicare Readmissions” New England Journal of Medicine Vol. 368, No. 13, 2013; Medicare Payment Advisory Commission, “Chapter 4: Refining the Hospital Readmissions Reduction Program,” Report to the Congress: Medicare and the Health Care Delivery System, June 2013. ↩︎
8. Research using quarterly data shows declines occurred for some measures earlier. See Zuckerman, R. S. Sheingold, J. Orav, J. Ruhter, and A. Epstein, “Readmissions, Observations, and the Hospital Readmissions Reduction Program,” New England Journal of Medicine, Vol. 374, No.16, 2016; ↩︎
9. Obama, B., “United States Health Care Reform: Progress to Date and Next Steps,” Journal of American Medicine (July 11 2016). ↩︎
10. Medicare Payment Advisory Commission, “Chapter 4: Refining the Hospital Readmissions Reduction Program,” Report to the Congress: Medicare and the Health Care Delivery System, June 2013; National Quality Forum, “Risk Adjustment for Socioeconomic Status or Other Sociodemographic Factors,” Technical Report, August 15, 2014. ↩︎
11. Final Rule for 42 CFR Parts 405, 412, 413, and 489—scheduled to be published in the Federal Register on August 22, 2016. Federal Register Vol. 79, No. 163: 49659-50536, 2014. ↩︎
12. This issue and other complexities with the penalty formula are discussed in Medicare Payment Advisory Commission, “Chapter 4: Refining the Hospital Readmissions Reduction Program,” Report to the Congress: Medicare and the Health Care Delivery System, June 2013; Lynn, Joanne and Steve Jencks, “A Dangerous Malfunction in the Measure of Readmission Reduction,” MediCaring.org, August 26, 2014. More recently, MedPAC submitted relevant comments to CMS on the IPPS proposed rule http://medpac.gov/-documents-/comment-letters ↩︎
13. See studies listed in endnote 2. ↩︎
14. While hospital readmission rates among Medicare patients has declined, some other types of hospital rates have increased, including outpatient emergency room visits and observation stays.  Medicare claims analysis by Gerhardt et al (2014) suggests that reductions in readmissions in 2012 were largely independent from these other types of hospital uses. (Gerhardt, G., A. Yemane, K. Apostle, A. Oelschlaeger, E. Rollins, and N. Brennan. 2014. “Evaluating Whether Changes in Utilization of Hospital Outpatient Services Contributed to Lower Medicare Readmission Rate.” Medicare & Medicaid Research Review 4(1), E1-E13.) ↩︎
15. Medicare Payment Advisory Commission, “Chapter 4: Refining the Hospital Readmissions Reduction Program,” Report to the Congress: Medicare and the Health Care Delivery System, June 2013; National Quality Forum, “Risk Adjustment for Socioeconomic Status or Other Sociodemographic Factors,” Technical Report, August 15, 2014. Sheingold, S., R. Zuckerman, and A. Shartzer, “Understanding Medicare Hospital Readmission Rates and Differing Penalties Between Safety-net and Other Hospitals,” Health Affairs, 35, No.1, January 2016; Kahn, C., T. Ault, L. Potetz, T. Walke, J. Chambers, and S. Burch, “Assessing Medicare’s Hospital Pay-For-Performance Programs And Whether They Are Achieving Their Goals,” Health Affairs, 34, No.8, August 2015; Joynt, K. and A. Jha, “A Path Forward on Medicare Readmissions” New England Journal of Medicine Vol. 368, No. 13, 2013; ↩︎
16. Watts, M., E. Cornachione, M. Musumeci, “Medicaid Financial Eligibility for Seniors and People with Disabilities in 2015. Kaiser Family Foundation (March 2016)  https://modern.kff.org/medicaid/report/medicaid-financial-eligibility-for-seniors-and-people-with-disabilities-in-2015/ ↩︎
17. For further information on Medicaid reform proposals, such as block grants and per capita caps, see: Rudowitz, Robin, “5 Key Questions: Medicaid Block Grants & Per Capita Caps,” Kaiser Family Foundation, January 31, 2017. https://modern.kff.org/medicaid/issue-brief/5-key-questions-medicaid-block-grants-per-capita-caps/ ↩︎
18. In general, beneficiaries are not responsible for an inpatient deductible if admitted to a hospital within 60 days of a prior hospitalization. ↩︎
19. Based on a subset of participating CBCT organizations, preliminary results find that a small number of organizations were able to achieve reductions in readmissions. Further analysis is expected to highlight both successful and unsuccessful strategies and techniques. (Econometrica, Inc, “Evaluation of the Community-based Care Transitions Program,” submitted to CMS (May 2014). ↩︎