HEALTH POLICY CENTER Post-Acute Care and Medicare Solvency Reducing Excessive PAC Payments Can Promote Financial Sustainability Bowen Garrett October 2023 Medicare provides health insurance coverage for 66 million elderly and disabled Americans but faces serious short- and long-term financial pressures. To address these pressures, policymakers need to consider options that involve raising additional revenues, finding ways to generate program savings, or likely both. To the extent program savings are required, policymakers will generally want to avoid program changes that would reduce program benefits or could have adverse effects on patient care. Two areas where current Medicare payment rates have been deemed excessive and warrant payment reductions are Medicare Advantage and post-acute care (PAC). We considered overpayment to Medicare Advantage plans in an earlier brief (Berenson, Garrett, and Shartzer 2022; see also MedPAC 2023a). Here, we consider payments to PAC providers for services to enrollees in traditional Medicare (TM). We examine the spending of four types of PAC providers, their payments in relation to cost, and proposals to reduce Medicare spending for PAC. To provide more context for weighing these proposals, we examine which TM enrollees use PAC (by age and income) and how their total program spending is allocated across payers (Medicare, out-of-pocket, Medicaid, or supplemental plan). Medicare enrollees who need recuperation and rehabilitation services after an acute inpatient hospital stay can receive PAC in skilled nursing facilities (SNF), inpatient rehabilitation facilities (IRF), long-term care hospitals (LTCH), or at home through the home health care benefit (HH). Under rules that vary by PAC setting, enrollees may also be admitted into PAC directly from the community rather than discharged directly into PAC after a hospital stay. This approach is common for home health. About half of TM beneficiaries who used PAC services had a prior hospital stay (MedPAC 2019). There is a different Medicare payment system for each of these four settings of PAC, with costs and payments varying considerably depending on setting of care. Because the types of cases treated in the different PAC settings overlap, payments for patients with similar care needs can differ substantially by setting (MedPAC 2015; Wissoker and Garrett 2014). This issue is exacerbated by geographic variability in use patterns, the supply of PAC providers, and a lack of evidence-based criteria for which types of patients should receive PAC and what setting is most appropriate for a given patient. 1 The Medicare Payment Advisory Commission (MedPAC), the independent congressional agency that advises Congress on Medicare payment policy and other program issues, has long determined that payments to providers in three PAC settings-SNF, HH, and IRF-are too high relative to the costs of providing the services and could be reduced without harming patient care (Linehan 2012; MedPAC 2019). Both to manage the use of PAC services and to address the issue of overpayment, MedPAC has previously recommended implementation of a unified PAC payment system that spans the four PAC settings and bases payment on patient characteristics rather than the setting of PAC care, paired with an overall payment rate reduction to generate Medicare program savings (MedPAC 2016, 2017). Though payment rate reduction is not an inherent feature of a unified payment system, implementing a unified PAC PPS presents an opportunity to lower payments if Congress has not already done so. CMS has been developing its approach to a unified PAC payment system and submitted a congressionally mandated report in July 2022.2 MedPAC submitted its second mandated report to Congress in June 2023 (MedPAC 2023b). Further Congressional action would be needed to implement a unified PAC payment system. PAC is mostly paid under the Hospital Insurance (HI) component (Part A) of the Medicare program, which helps pay for hospital and most institutional services, including SNF, IRF, LTCH, and HH care directly preceded by a hospital or SNF stay. Payments for Part A services are made from the HI trust fund, which receives its revenue from dedicated sources, the most important being a payroll tax imposed on workers' earnings. HH care provided without a preceding hospital or SNF stay is paid under Medicare Part B, which also helps pay for physicians' outpatient services, laboratory tests, physician- administered drugs, and durable medical equipment. Part B is financed by beneficiary premiums and federal general revenues deposited into the Supplemental Medical Insurance trust fund, as is Part D, which helps pay for self-administered prescription drugs. Medicare spending has increased from just above 2 percent of GDP in 2000 to about 4 percent in 2020, partly for demographic reasons that will eventually subside (the aging of the baby boomer generation). The latest Medicare trustees' report projects Medicare spending will increase to 6 percent of GDP by 2040. In the nearer term, the Medicare trustees' report projects that the HI trust fund will be insolvent sometime around 2031 (Medicare Trustees 2023). At that point, accumulated trust fund balances would no longer be able to fill the running gap between HI spending and its dedicated sources 2 POST-ACUTE CARE AND MEDICARE SOLVENCY of revenue. The total financing gap (projected spending in excess of projected revenues) for HI over the next 10 years (from 2023 to 2032) is $333 billion.3 The Congressional Budget Office (CBO) more recently updated its expected date of HI insolvency to after 2033, pushing the date out from its estimate of 2030 a year ago.4 Therefore, while the expected date of HI insolvency may be later than previously estimated, there is uncertainty in the timing. Estimates in recent years have shifted, given volatility in the use of health care and revenues tied to economic performance through the pandemic and its aftermath. If HI becomes insolvent, Medicare will only be able to pay an estimated $0.90 on the dollar for hospital services, which likely would disrupt patient care. Many consider it inevitable that Congress will be forced to act by cutting Medicare spending, raising additional revenue for the HI trust fund, or both (Garrett, Shartzer, and Arnos 2021). Unlike in the HI trust fund, balances in the Supplemental Medical Insurance trust fund are automatically replenished with general revenues (and federal deficit financing) when they run low, and beneficiary premiums for Parts B and D are increased. Key takeaways emerged from our review of PAC spending trends, policy proposals, and analysis of users of PAC services: ▪ Medicare payments for SNF, HH, and IRF have long been considered excessive. Reducing these payments may be an easy choice to help put Medicare's financing on a more sustainable path, though more comprehensive solutions will eventually require harder choices. ▪ Moving payments by Part A for home health services to Part B would be an expedient way to deal with the HI deficit, but it would do nothing to address Medicare spending as a whole and could lead to increased Part B premiums. ▪ TM enrollees who use SNF and home health are older and need more costly health services; spending for enrollees with SNF stays and home health care is 10 times and 7 times higher than spending for enrollees who use neither PAC service. Thus, managing the overall care of PAC users in TM, not only the PAC spending, offers substantial opportunities for savings and quality improvement. Bundled payment policies that encourage coordination between the acute care hospital and PAC providers are one approach that could generate savings and improvement in care for the patient. ▪ Policies that would aim to generate program savings by increasing cost-sharing obligations for users of PAC would add to already high financial burdens for beneficiaries lacking supplemental or Medicaid coverage unless increased cost sharing for PAC were accompanied by policies that would reduce beneficiaries' overall exposure to high out-of-pocket costs. ▪ The exact date of Medicare HI insolvency is uncertain-the expected date being pushed to around 2031 provides more opportunity to spread necessary spending reductions or revenue increases over time. Making reasonable payment adjustments now could further postpone the date of HI insolvency. POST-ACUTE CARE AND MEDICARE SOLVENCY 3 Spending Varies by PAC Setting Table 1 provides an overview of the use of PAC and spending across PAC settings within TM. It also reports the latest available projected margins (payments relative to costs) for each setting, payment changes recommended by MedPAC, and actual (or proposed) payment rate changes for each setting for 2023 and 2024. In 2021, 1.2 million Medicare enrollees used SNF, and 3.0 million enrollees used home health. Fewer enrollees used IRF (335,000), and fewer still used LTCH (71,000 in 2020, latest data available). Spending was higher for SNF ($28.5 billion) than home health ($16.9 billion), despite more enrollees using home health, given the higher cost structure of the institutional SNF setting. Medicare spending for IRF was $8.5 billion, similar to SNF in terms of spending per user (about $24,000 per user for SNF as compared with $25,000 for IRF, data not shown in table). LTCH spending was $3.4 billion in 2020, which amounts to about $48,000 per user (not shown in table), making LTCH the most expensive PAC setting on a per-user basis. 4 POST-ACUTE CARE AND MEDICARE SOLVENCY TABLE 1 Post-Acute Care Users, Spending, Margins, and Recommended and Actual Payment Changes in Traditional Medicare by Setting Skilled Inpatient nursing rehabilitation Long-term facilities Home health facilities care hospitals Users 1.2 million 3.0 million 335,000 71,000a Spending ($ billions) $28.5 $16.9 $8.5 $3.4a Medicare aggregate margin (percent) 17.2 24.9 17.0 3.6a Projected margin (percent) 10 17 11 3a Recommended base payment rate −5 −5 −5 2 change for 2023 (percent change)a (estimated) Actual base payment rate change for 5.1b -1.0c 3.7d 3.8e 2023 (percent change) Recommended base payment rate -3 -7 -3 -- change for 2024 (percent change) (see note) Actual or proposed base payment rate 6.4f -1.8g 3.7h 3.6i change for 2024 (percent change) Sources: MedPAC March 2023 Report to Congress (MedPAC 2023a), except where noted. a MedPAC March 2022 Report to the Congress (MedPAC 2022). b "FY 2023 Skilled Nursing Facility Prospective Payment System Final Rule (CMS 1765-F)," CMS.gov Newsroom, July 29, 2022. c "CY 2023 Home Health Prospective Payment System Rate Update Final Rule (CMS-1766-F)," Federal Register 87 (213), CMS, November 4, 2022. d "FY 2023 Inpatient Rehabilitation Facility Prospective Payment System Final Rule (CMS-1767-F)," Federal Register 87 (213), CMS, August 1, 2022. e "FY 2023 Hospital Inpatient Prospective Payment System (IPPS) and Long-Term Care Hospital Prospective Payment System (LTCH PPS) Final Rule-CMS-1771-F," Federal Register 87 (153), CMS, August 10, 2022. f "FY 2024 Skilled Nursing Facility Prospective Payment System Final Rule - CMS-1779-F," CMS.gov Newsroom, July 31, 2023. g "CY 2024 Home Health Prospective Payment System Proposed Rule (CMS-1780-P)," Federal Register 88 (130), CMS, July 10, 2023. h "FY 2024 Inpatient Rehabilitation Facility Prospective Payment System Final Rule (CMS-1781-F)," Federal Register 88 (147), CMS, August 2, 2023. i "FY 2024 Hospital Inpatient Prospective Payment System (IPPS) and Long-Term Care Hospital Prospective Payment System (LTCH PPS) Final Rule - CMS-1785-F and CMS-1788-F," Federal Register 88 (165), CMS, August 28, 2023. Notes: Values for users, spending, and Medicare aggregate margin are for 2021, except for long-term care hospitals, for which 2020 values are reported (latest available). Projected margins are for 2023, except for long-term care hospitals, for which 2022 projections are reported. For long-term care hospitals, MedPAC recommended the base payment for 2023 be increased by the market basket minus the applicable productivity adjustment, which was estimated to be 2 percent, noting that number could change. In its March 2023 report, MedPAC discontinued making annual payment rate updates for long-term care hospitals, noting that as the number of cases qualifying for payment under that system has declined, it was increasingly concerned about small sample sizes in its analysis of the sector, but that it would continue to monitor the sector and provide periodic status reports. The aggregate Medicare margin for LTCH was 3.6 percent in 2020 (projected to be 3 percent in 2022). Medicare margins for the other three PAC settings in 2021 were high: 17.2 percent for SNF, 24.9 percent for home health, and 17.0 percent for IRF. Projected margins for these three settings for 2023 POST-ACUTE CARE AND MEDICARE SOLVENCY 5 were lower but still at elevated levels: 10 percent for SNF, 17 percent for home health, and 11 percent for IRF. Bolstered by evidence that Medicare enrollees' access to care in these settings is adequate, MedPAC previously and continues to recommend reducing payment rates.5 MedPAC 2022 recommendations would have reduced 2023 payment rates by 5 percent in three settings-SNF, HH, and IRF. Instead, in accordance with current law, CMS increased base payment rates for 2023 by 5.1 percent for SNF and 3.7 percent for IRF. The payment rate for home health was reduced, but only by 1 percent. For LTCH, MedPAC recommended the base payment for 2023 be increased by the usual adjustment (market basket minus the applicable productivity adjustment, projected to be 2 percent), which turned out to be a 3.8 percent increase in payment for 2023. Payment recommendations in MedPAC's March 2023 report to Congress call for 3 percent reductions in SNF and IRF and a 7 percent reduction for home health for 2024. The actual payment rate changes for fiscal year 2024 are a 6.4 percent increase for SNF and a 3.7 percent increase for IRF. A payment reduction of 1.8 percent is proposed for home health for the 2024 calendar year. MedPAC discontinued its annual payment rate updates for LTCH in 2023 (see table 1 note); payments in that system increased by 3.6 percent for fiscal year 2024. It would take an act of Congress to reduce base payment rates to be in line with MedPAC's recommendations. Various Policies Have Been Proposed to Reduce Medicare PAC Spending Table 2 summarizes policy proposals aimed at managing and reducing Medicare PAC spending.6 Reducing PAC Payment Rates The first option to reduce PAC spending would implement 5 percent reductions in payment rates for SNF, IRF, and HH. This policy would have saved an estimated $2.7 billion in 2021 had it been implemented that year. The reduced spending for SNF and IRF would accrue directly to the HI trust fund. However, only a portion of the reduced spending for home health would accrue to Part A because roughly 66 percent of home health spending is paid under Medicare Part B.7 Another proposal, included in the president's fiscal year 2021 budget, would have "address[ed] excessive payment rates for post-acute care providers" and shifted to a unified PAC payment system (HHS 2021).8 The unified PAC payment system would base payments on patients' clinical needs rather than the setting of care (MedPAC 2017). The CBO scored this proposal and estimated it would save $79 billion over 10 years.9 Because the policy was modeled as being implemented in 2021 and phased in, the bulk of the total savings was accrued in the later years of the 10–year window. It is unclear how much of the estimated savings was attributed to efficiency gains under the unified payment system compared with the payment rate reductions. 6 POST-ACUTE CARE AND MEDICARE SOLVENCY Shifting Home Health Spending under Part A to Part B Only about one-third of home health spending is now financed under Part A. There is, therefore, an argument for unifying all home health spending under Part B, moving nearly $6 billion in annual spending away from the HI trust fund.10 This would be more than just an accounting change, as it could raise Part B premiums and increase the general federal revenue needed for Part B. Such a policy could include a provision to limit premium increases, say, for enrollees with lower incomes. Though the shift would help address the near-term issue of HI solvency, it would not address the longer-term issue of Medicare sustainability. POST-ACUTE CARE AND MEDICARE SOLVENCY 7 TABLE 2 Summary of Policy Options to Reform Medicare Payment for Post-Acute Care Policy option Description/rationale Estimate of savings impact Reduce PAC payment This would implement MedPAC's most recent $2.7 billion if implemented in rates for SNF, IRF, and payment rate update recommendations for SNF, IRF, 2021 based on spending levels HH by 5 percent and HH, bringing payments more in line with costs. for SNF, HH, and IRF in table 1 Reduce PAC payment This would implement long-standing $79 billion over 10 years rates and shift to a recommendations to reduce PAC payment levels (2021 to 2030)a unified PAC payment because payments well exceed cost. A unified system payment system would base payments on patients' clinical needs rather than site of care. Shift home health Shifting all home health spending to Part B would Moves $6 billion per year in spending under Part A recognize that most home health use is physician- HI/Part A spending to Part B.b to Part B directed and would ease financial burden on the HI No net savings to Medicare. trust fund. Bundle payment for Make one risk-adjusted lump-sum payment to cover Estimated to save $47 billion PAC with preceding an initial hospital stay and PAC that follows for 90 from 2014 to 2023, or 1.4 inpatient stay days. Hospitals would have an incentive to reduce percent of nondrug outlays unneeded utilization and spending over the bundled when fully phased inc services. Bundle payment for Bundle payment for SNF, IRF, LTCH, and HH, paying a $8.2 billion (2016 to 2026)d PAC providers single risk-adjusted payment for at least half of total payments. Reduce overall payments by 2.85 percent. Introduce 10 percent Details of who might be exempt not specified. $47 billion from 2010 to coinsurance for home 2019e health use in traditional Medicare Introduce a Establish a per-episode copayment of $150 for home $1 to $5 billion over 5 years copayment for home health episodes not preceded by hospitalization or (estimated in 2011)f health use in post-acute care. traditional Medicare Sources: a "Proposals Affecting Medicare-CBO's Estimate of the President's Fiscal Year 2021 Budget," CBO, March 25, 2020 (item 36). b "Strengthening The Medicare Trust Fund In The Era Of COVID-19," Forefront (blog) Health Affairs, June 10, 2020. c "Bundle Medicare's Payments to Health Care Providers," CBO, November 13, 2013. d "Proposals for Health Care Programs-CBO' Estimate of the President's Fiscal Year 2017 Budget," CBO, March 29, 2016. e "Option 85," in Budget Options, vol. I, Health Care (CBO 2008). f Chapter 2, Recommendation 8-3, in MedPAC March 2011 Report to the Congress (MedPAC 2011). Notes: HH = home health care; HI = Hospital Insurance; IRF = inpatient rehabilitation facility; LTCH = long-term care hospital; PAC = post-acute care; SNF = skilled nursing facility. Using Bundled Payments to Incentivize Efficient Use of PAC Services Bundled payment policies tie together payments for services that are otherwise paid separately to make health care providers more accountable for total patient costs for a larger episode of care. PAC services might be bundled in multiple ways. In 2013, CBO estimated the budgetary impact of a policy 8 POST-ACUTE CARE AND MEDICARE SOLVENCY that would have bundled payment for PAC with the preceding inpatient stay. This policy would make a single lump-sum payment to cover an initial hospital stay and any PAC that follows for 90 days. Under this policy, hospitals would have an incentive to reduce unneeded utilization and spending over the bundled services. CBO estimated the policy would save $47 billion over 2014 to 2023. CBO estimated that savings would amount to 1.4 percent of nondrug outlays if the policy were fully phased in.11 Another approach would bundle payment for the four types of PAC providers, paying a single risk- adjusted payment. This policy had been proposed in the president's fiscal year 2017 budget.12 CBO estimated this would have reduced overall payments by 2.85 percent, saving $8.2 billion from 2016 to 2026.13 This option would be more complicated to implement because it would need to determine which entity would get the payment or how a single payment would be divided. Introducing Beneficiary Cost Sharing for Home Health In TM, home health is one of the few services for which the program imposes no cost-sharing obligation on beneficiaries (whether financed through Part A or B). This has led to concern that home health services may be overconsumed, that is, used even when they provide beneficiaries with low or even no marginal value. Requiring some financial stake in the use of home health on the part of enrollees could also create a check on potentially fraudulent or abusive provision of services by providers, which has been a problem within the home health industry.14 Policy proposals to introduce cost sharing for home health would have a direct savings effect for Medicare by shifting a portion of spending to beneficiaries (or their supplemental plan or Medicaid) and could have a behavioral effect by reducing home health care use of low value to beneficiaries. In 2007, CBO scored a proposal requiring enrollees to pay 10 percent of the cost of home health services. This policy was estimated to save $47 billion from 2010 to 2019 (CBO 2008, 172). Few details were provided on the policy (specifying, e.g., who might be exempt from coinsurance, leading to the conclusion it applied to all TM users of home health). In 2011, MedPAC considered a policy establishing a $150 copayment for home health episodes not preceded by a hospitalization or institutional PAC. This more narrowly targeted policy would discourage overuse of home health episodes from the community. The policy could also exclude episodes with low utilization or dual-eligible beneficiaries. MedPAC estimated this policy would save $1 to $5 billion over five years (estimated in 2011) (MedPAC 2011). A variation of this policy could impose beneficiary cost sharing for home health after, say, the first 60 days. This would be similar to the current policy for SNF care in which beneficiary copayments are imposed after day 20. POST-ACUTE CARE AND MEDICARE SOLVENCY 9 Given Their Health Care Needs, Medicare Enrollees Using PAC Services Are Much More Costly than Other Enrollees To provide additional context for the policy options discussed above, we use the Urban Institute's Medicare policy simulation model, MCARE-SIM, to compare health care spending and other characteristics of TM enrollees by use of PAC services estimated for 2023. MCARE-SIM uses data from the 2015–18 Medicare Current Beneficiary Survey and projects Medicare enrollment and spending estimates to 2023.15 Table 3 reports total spending per enrollee in TM for covered services and breaks down spending by payer. Nearly 34 million of the 37 million TM enrollees did not use SNF or home health services in the past 12 months. This group had an average spending of $9,400 (across all Medicare-covered services). Of this, $7,600 was paid by Medicare, $500 was paid out of pocket, $300 was paid by Medicaid, and $1,100 was paid by a supplemental plan (Medigap or retiree coverage). In stark contrast, for the nearly 3.2 million enrollees who used SNF or home health, average total spending was $62,300, with $55,300 of that paid by Medicare, $1,800 out of pocket, $1,800 by Medicaid, and $3,500 by supplemental plan coverage. TABLE 3 Composition of Medicare Spending for Traditional Medicare Enrollees, by Use of SNF or Home Health Average Spending per Traditional Medicare Enrollee Total spending for Medicare- covered Out of Supplementary Number services Medicare pocket Medicaid plan Did not use SNF or HH in 12 33,861,000 $9,400 $7,600 $500 $300 $1,100 months Used SNF or HH in 12 months 3,191,000 $62,300 $55,300 $1,800 $1,800 $3,500 Used SNF 1,071,000 $87,400 $77,600 $2,600 $2,400 $4,900 Used HH 2,798,000 $59,500 $52,800 $1,600 $1,700 $3,300 Sources: MCARE-SIM; 2023 estimates based on 2015–2018 Medicare Current Beneficiary Survey. Notes: Enrollee groups using SNF and using home health are not mutually exclusive. Supplementary plan spending includes Medigap and retiree coverage. Average spending amounts are across all Medicare-covered services. HH = home health; SNF = skilled nursing facility. Table 3 shows there are nearly 1.1 million SNF users and 2.8 million home health users estimated in 2023. These adjusted survey-based numbers are similar to the analogous figures based on Medicare administrative data for 2021.16 An estimated 678,000 enrollees used both types of PAC and are counted in both groups (not shown). Total spending was higher for SNF users than home health users 10 POST-ACUTE CARE AND MEDICARE SOLVENCY ($87,400 versus $59,500), as were amounts for each payor type. Across all services, out-of-pocket spending was $2,600 on average for SNF users as compared with $1,600 for home health users. SNF and Home Health Users Are Older than Other Medicare Enrollees and Have Lower Incomes The high spending levels for SNF and home health users are partly driven by demographic differences. Table 4 reports distributions of age and income for the same users of SNF and home health described in table 3 and compares their distributions with those for TM enrollees who did not use SNF or home health. Enrollees using SNF and home health are older than TM enrollees, who used neither service. The share age 85 and over is 28.1 percent for SNF users and 26.1 percent for home health users, compared with 8.3 percent for enrollees who used neither type of PAC. SNF and home health users were also more likely to be in the 75 to 84 age group and less likely to be in the 65 to 74 and younger than 65 age groups than enrollees who did not use SNF or home health. TABLE 4 Age and Income Group Distributions for Traditional Medicare Enrollees, by Use of SNF or Home Health SNF users HH users No SNF or HH use Age group (%) <65 6.8 10.2 14.4 65–74 31.9 31.7 53.7 75–84 33.2 32.0 23.6 ≥85 28.1 26.1 8.3 All 100.0 100.0 100.0 Income group (%) <100% of FPL 17.7 21.6 13.8 100–200% of FPL 27.3 26.5 20.5 200–400% of FPL 31.7 29.2 29.0 >400% of FPL 23.3 22.6 36.7 All (%) 100.0 100.0 100.0 Number of enrollees 1,071,000 2,798,000 33,861,000 Sources: MCARE-SIM; 2023 estimates based on 2015–18 Medicare Current Beneficiary Survey. Notes: Enrollee groups using SNF and using HH are not mutually exclusive. FPL = federal poverty level; HH = home health care; SNF = skilled nursing facility. POST-ACUTE CARE AND MEDICARE SOLVENCY 11 By income, SNF and home health users were less likely to be in the top income group (above 400 percent of FPL) and more likely to be in the lower two income groups than TM enrollees who used neither type of PAC. For example, 17.7 percent of SNF users and 21.6 percent of home health users had incomes less than 100 percent of FPL, compared with 13.8 percent of TM enrollees who did not use SNF or home health. While SNF and home health enrollees tend to have lower incomes, the overall pattern shows that these PAC users are well represented in each of the four income categories we report. TM Enrollees Using Home Health Face Substantial Out- of-Pocket Costs When They Lack Other Coverage Table 5 reports shares with different types of coverage and average out-of-pocket spending levels, by income group, for TM enrollees using home health. We focus on home health users because of the policy interest in introducing beneficiary cost sharing for home health (SNF users already face cost-sharing after the 20th day of service). Overall, 51.2 percent of home health users have supplementary insurance (Medigap or retiree coverage), 23.7 percent are Medicaid enrolled, and 25.1 percent are without supplemental coverage or Medicaid. 12 POST-ACUTE CARE AND MEDICARE SOLVENCY TABLE 5 Coverage Shares and Average Out-of-Pocket Spending Levels, by Income Group, for Traditional Medicare Enrollees Using Home Health Average Out-of-Pocket Spending ($) No No Medigap or supplemental supplemental retiree coverage or coverage or Number coverage (%) Medicaid (%) Medicaid (%) All coverage Medicaid Income group <100% of FPL 605,000 17.3 68.5 14.2 $900 $6,600 100–200% of FPL 743,000 49.3 27.6 23.1 $1,600 $7,100 200–400% of FPL 816,000 62.8 4.9 32.4 $2,000 $6,200 >400% of FPL 634,000 71.0 0.5 28.5 $1,700 $6,000 All 2,798,000 51.2 23.7 25.1 $1,600 $6,500 Sources: MCARE-SIM; 2023 estimates based on 2015–18 Medicare Current Beneficiary Survey. Note: Supplemental coverage includes Medigap and retiree plan coverage. The Medicaid group includes Medicare enrollees who report Medicaid enrollment in any month of the year, including both full-benefit Medicare-Medicaid enrollees and partial-benefit Medicare-Medicaid enrollees. FPL = federal poverty level. However, the shares of home health users with coverage vary greatly by income. As we would expect given income-related eligibility rules, few home health users with reported income above 200 percent of FPL are enrolled in Medicaid. The large majority of home health users with incomes less than 100 percent of FPL are Medicaid enrolled (68.5 percent), and 27.6 percent with incomes from 100 to 200 percent of FPL are Medicaid enrolled. The share with Medigap or retiree coverage increases steeply with income, from 17.3 percent for enrollees in the lowest income group (less than 100 percent of FPL) to 71.0 percent for enrollees in the highest income group. Between these two extremes, those with incomes between 200 and 400 percent of FPL are most likely to be without supplemental coverage or Medicaid (32.4 percent). By paying for all or some of the cost-sharing obligations TM beneficiaries incur for services they use, supplemental coverage and Medicaid can protect beneficiaries from high out-of-pocket costs (Medicare cost-sharing rules differ by supplemental plan and state Medicaid program). Over all coverage types and income groups, average out-of-pocket spending for home health users is $1,600. Across income groups but over coverage types, average out-of-pocket spending ranges from $900 for the lowest income group to $2,000 for those with incomes from 200 to 400 percent of FPL. This variation is largely attributable to the differences in the share with supplemental or Medicaid coverage. POST-ACUTE CARE AND MEDICARE SOLVENCY 13 The last column of table 5 shows average out-of-pocket spending of home health users who do not have supplemental coverage or Medicaid. Over all income levels, average out-of-pocket spending among this group for Medicare-covered services is $6,500. We see little variation across income groups, with out-of-pocket spending ranging from $6,000 in the highest income group (over 400 percent of FPL) to $7,100 for those with incomes 100 to 200 percent of FPL. In sum, users of home health services in TM who lack supplemental coverage or Medicaid face high out-of-pocket costs for the various covered services they use, even though home health does not itself require beneficiary cost sharing in TM. Discussion Post-acute care, including access to home health care without an initial hospital stay, is an important service for Medicare enrollees. Users of PAC are older on average than other Medicare enrollees and have much higher spending for Medicare-covered services, including out-of-pocket spending. Yet in aggregate, Medicare payment for PAC well exceeds its cost for SNF, home health, and IRF providers. MedPAC has long determined that payments to these PAC settings are excessive. High aggregate margins and positive indicators of beneficiary access suggest that reducing payments would not threaten enrollees' access to needed care. Given the impending insolvency of the Medicare HI trust fund, Congress will likely be forced to shore up Medicare finances through payment cuts, additional revenues, or both. Though the expected timing of HI insolvency has shifted and remains uncertain, the sooner Congress acts, the better it can distribute changes to spending and revenue over time and avoid more difficult choices. A rational approach would start by targeting payment reductions to where payments are already deemed excessive. Traditional Medicare payments to SNF, home health, and IRF present such an opportunity that could be part of broader reforms, including implementing a unified PAC payment system, making targeted reductions in excessive payments to Medicare Advantage plans, and other reforms to both Medicare spending and revenue that have been described elsewhere. 17 A unified PAC payment system could better rationalize the use of post-acute care, generate savings over time, and be paired with payment cuts. In its June 2017 report to Congress, MedPAC recommended that Congress implement a unified PAC payment design (MedPAC 2017), and the proposal has subsequently received bipartisan support (Bipartisan Policy Center 2020). MedPAC forwarded its second mandated report on a PAC PPS to Congress in its June 2023 report without voting on whether to recommend implementing a unified payment system. Among the key takeaways, MedPAC noted that, based on its work and that conducted by CMS/ASPE, designing a PAC PPS is feasible and could establish accurate payments, but would necessitate a complex set of companion policies that would be controversial and take many years to implement.18 Another option that could be part of a package is to shift the costs of all home health services to Part B. Although this policy would not solve Medicare's fiscal challenges given the increasingly deficit- financed funding of Part B, it would ease the near-term solvency issue for Part A.19 14 POST-ACUTE CARE AND MEDICARE SOLVENCY Removing home health's special status within TM of having no beneficiary cost-sharing is conceptually appealing and would lower Medicare spending (shifting it to other payers) and potentially reduce use of unnecessary or low-value home health episodes. Nearly 75 percent of home health users in TM have supplemental coverage or Medicaid that could shield them from all or some cost-sharing expenses and reduce any behavioral effect. However, the 25 percent lacking supplemental coverage or Medicaid already face out-of-pocket expenses averaging $6,500, even for those with lower incomes. For these beneficiaries, adding cost sharing for home health would increase already high financial burdens and could lead them to forgo necessary or high-value care, potentially leading to avoidable hospitalization. Unlike TM, some Medicare Advantage plans require beneficiary cost sharing for home health. An analysis of differences in home health use by Medicare Advantage plan enrollees who did and did not have cost sharing for home health found no significant differences in use patterns. It did, however, find increased rates of disenrollment from Medicare advantage among enrollees with greater use of home health (Li et al. 2017). In a comment to this study, Grabowski (2017) provides possible alternative explanations for the lack of an apparent effect of cost sharing on home health utilization in the Medicare Advantage context. Yet, we would still expect those not shielded from cost sharing by supplemental coverage or Medicaid to respond to any cost sharing implemented in TM. Recent policy movement by CMS has been to limit beneficiary cost sharing for home health services in Medicare Advantage.20 Studies have shown that Medicare Advantage enrollees have lower home health and PAC utilization rates than TM enrollees after controlling for observable differences (Skopec et al. 2020). Assuming selection into Medicare Advantage based on unobservable factors is not driving these differences, and overall quality of care is at least as good in Medicare Advantage, such findings suggest that TM can reduce PAC utilization and spending without negative consequences for beneficiaries. Given our finding that spending for enrollees with SNF and home health stays is 10 times and 7 times higher than spending for enrollees who use neither PAC service, management of PAC users' care in TM offers substantial opportunity for savings and quality improvement. Notes 1 Carol Carter, Bowen Garrett, and Douglas Wissoker, "A Unified Medicare Payment System for Post-Acute Care Is Feasible," Forefront (blog), Health Affairs, September 28, 2016, https://www.healthaffairs.org/do/10.1377/forefront.20160928.056838/. 2 "IMPACT Act Spotlights and Announcements," CMS, July 07, 2022, https://www.cms.gov/medicare/quality/initiatives/pac-quality-initiatives/impact-act-2014-data- standardization-cross-setting-measures/impact-act-spotlights-and-announcements. 3 The HI trust fund balance of $202 billion (estimated at the end of 2023) will be spent down each year to fill a portion of that gap, until the accumulated balance is exhausted (2031, according to the trustees). 4 See CBO's May 2022 Medicare baseline projections, https://www.cbo.gov/system/files/2022-05/51302-2022- 05-medicare.pdf, and May 2023 baseline projection, https://www.cbo.gov/system/files/2023-05/51302-2023- 05-medicare.pdf. POST-ACUTE CARE AND MEDICARE SOLVENCY 15 5 In arriving at its payment rate recommendations each year, MedPAC considers a range of measures, including providers' financial situation and measures of beneficiaries' access to care. 6 See also the list of PAC reform policies and discussion produced by the Committee for a Responsible Federal Budget, "Can Post-Acute Care Reforms Save the Medicare Trust Fund?," March 24, 2022, https://www.crfb.org/blogs/can-post-acute-care-reforms-save-medicare-trust-fund. This brief benefited from their presentation of these policy proposals. 7 David M. Cutler, Richard G. Frank, Jonathan Gruber, and Joseph P. Newhouse, "Strengthening the Medicare Trust Fund in the Era of COVID-19," Forefront (blog), Health Affairs, June 10, 2020, https://www.healthaffairs.org/do/10.1377/forefront.20200608.322412/full/. 8 See also "Proposals Affecting Medicare-CBO's Estimate of the President's Fiscal Year 2021 Budget," CBO, March 25, 2020. 9 "Proposals Affecting Medicare-CBO's Estimate of the President's Fiscal Year 2021 Budget." 10 Cutler et al., "Strengthening the Medicare Trust Fund in the Era of COVID-19." 11"Mandatory Spending, Function 570-Medicare: Bundle Medicare's Payments to Health Care Providers," CBO, November 13, 2013, https://www.cbo.gov/budget-options/2013/44898. 12 "HHS FY 2017 Budget in Brief-CMS-Medicare," CMS, January 26, 2017, https://www.hhs.gov/about/budget/fy2017/budget-in-brief/cms/medicare/index.html. Bundled payment for the four types of PAC providers was to be imposed for at least half of total payments, but the method of bundling was not specified. 13 "Proposals for Health Care Programs-CBO's Estimate of the President's Fiscal Year 2017 Budget," CBO, March 29, 2016. 14 See for example, MedPAC 2020. Also see Joyce Famakinwa, "Home Health Sector Remains Area of Focus For Fraud Watchdogs," Home Health Care News, July 13, 2022, https://homehealthcarenews.com/2022/07/home- health-sector-remains-area-of-focus-for-fraud-watchdogs/. 15 For a description of MCARE-SIM's methods, see the methods section and appendix A of Gangopadhyaya et al. (2023). 16 MCARE-SIM and administrative data are not expected to be the same because of sampling variability and because the Medicare Current Beneficiary Survey data weights were not made to align with administrative totals for TM enrollees using SNF or home health care. We make further adjustments to survey weights to account for demographic change and align to administrative totals for 2023, but these adjustments did not aim to align the number of SNF or home health users to administrative data. See Arnos et al. 2020 for the approach to model calibration in MCARE-SIM. 17 "Ten Options to Secure the Medicare Trust Fund," Committee for a Responsible Federal Budget, June 16, 2022, https://www.crfb.org/blogs/ten-options-secure-medicare-trust-fund; "Medicare Solvency," Commonwealth Fund, https://www.commonwealthfund.org/series/medicare-solvency; Cutler et al., "Strengthening the Medicare Trust Fund."; Garrett et al. 2021; Mermin et al. 2023. 18 As described in MedPAC's June 2023 report, implementing a unified PAC-PPS would require companion policies involving changes to Medicare's benefit and coverage rules and establishing uniform Medicare conditions of participation for PAC providers. MedPAC also noted that changes CMS already implemented to SNF, home health, and LTCH PPSs were substantial and addressed some of the shortcomings within those separate payment systems (though not the overlap of cases treated in different settings). 19 Cutler et al., "Strengthening the Medicare Trust Fund." 20 "CMS Limits Medicare Advantage Plan Cost Sharing for Home Health Services," HomeCare News, April 19, 2022, https://www.homecaremag.com/news/cms-limits-medicare-advantage-plan-cost-sharing-home-health- services. 16 POST-ACUTE CARE AND MEDICARE SOLVENCY References Arnos, Diane, Bowen Garrett, Anuj Gangopadhyaya, and Adele Shartzer. 2020. "Calibrating a Microsimulation Model of Medicare Policy Reform to Match Administrative Benchmarks for Enrollment and Spending." Washington, DC: Urban Institute. Berenson, Robert A., Bowen Garrett, and Adele Shartzer. 2022. "Understanding Medicare Advantage Payment: How the Program Allows and Obscures Overspending." Washington, DC: Urban Institute. Bipartisan Policy Center. 2020. "Bipartisan Rx for America's Health Care." Washington, DC: Bipartisan Policy Center. CBO (Congressional Budget Office). 2008. Budget Options, vol. I, Health Care. Washington, DC: CBO. Garrett, Bowen, Adele Shartzer, and Diane Arnos. 2021. "Medicare Solvency Roundtable: Insights from Leading Experts to Keep Medicare on Solid Financial Ground." Washington, DC: Urban Institute. Gangopadhyaya, Anuj, John Holahan, Bowen Garrett, and Adele Shartzer. 2023. "Applying a Premium Cap in Medicare Part B and Part D." Washington, DC: Urban Institute. Grabowski, David C. 2017. "Cost Sharing and Home Health Care." JAMA Internal Medicine 177 (7): 1018–19. HHS (US Department of Health and Human Services). 2021. "Putting America's Health First: FY 2021 President's Budget for HHS," Washington, DC: HHS. Li, Qijuan, Laura M. Keohane, Kali Thomas, Yoojin Lee, and Amal N. Trivedi. 2017. "Association of Cost Sharing with Use of Home Health Services among Medicare Advantage Enrollees." JAMA Internal Medicine 177 (7): 1012–18. Linehan, Kathryn. 2012. "Medicare's Post-Acute Care Payment: A Review of the Issues and Policy Proposals." George Washington University National Health Policy Forum 847: 1–23. Medicare Trustees (Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds). 2023. 2023 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds. Baltimore: Center for Medicare & Medicaid Services. MedPAC (Medicare Payment Advisory Commission). 2015. "Medicare's Post-Acute Care: Trends and Ways to Rationalize Payments." In March 2015 Report to the Congress (chapter 7). Washington, DC: MedPAC. ---. 2016. "Mandated Report: Developing a Unified Payment System for Post-Acute Care." In June 2016 Report to the Congress (chapter 3). Washington, DC: MedPAC. ---. 2017. "Implementing a Unified Payment System for Post-Acute Care." In June 2017 Report to the Congress (chapter 1). Washington, DC: MedPAC. ---. 2019. "Cross-Cutting Issues in Post-Acute Care." In March 2019 Report to the Congress (chapter 7). Washington, DC: MedPAC. ---. 2020. Home Health Care Services. In Report to the Congress: Medicare Payment Policy (chapter 9). Washington, DC: MedPAC. ---. 2021. Home Health Care Services. In Report to the Congress: Medicare Payment Policy (chapter 8). Washington, DC: MedPAC. ---. 2022. March 2022 Report to Congress: Medicare Payment Policy. Washington, DC: MedPAC. ---. 2023a. March 2023 Report to Congress: Medicare Payment Policy. Washington, DC: MedPAC. ---. 2023b. Mandated Report: Evaluation of a Prototype Design for a Post-acute Care Prospective Payment System. In June 2023 Report to the Congress: Medicare and the Health Care Delivery System (chapter 10). Washington, DC: MedPAC. Mermin, Gordon B., Bowen Garrett, Lillian Hunter, and C. Eugene Steuerle. 2023. "Options for Increasing Medicare Revenues." Washington, DC: Urban-Brookings Tax Policy Center. POST-ACUTE CARE AND MEDICARE SOLVENCY 17 Skopec, Laura, Stephen Zuckerman, Joshua Aarons, Douglas Wissoker, Peter J. Huckfeldt, Judith Feder, Robert A. Berenson, Judith Dey, and Iara Oliveira. 2020. "Home Health Use in Medicare Advantage Compared to Use in Traditional Medicare." Health Affairs 39 (6). Wissoker, Doug, and Bowen Garrett. 2014. "Equalizing Medicare Payments for Select Patients in IRFs and SNFs." Washington, DC: MedPAC. About the Author Bowen Garrett is an economist and senior fellow in the Urban Institute Health Policy Center. His research focuses on health reform and health policy, including health insurance and labor markets, Medicare's prospective payment systems, and Medicare financing. He leads the development of the Urban Institute's Medicare policy simulation model (MCARE-SIM). Previously, Garrett was chief economist of the Center for US Health System Reform and McKinsey Advanced Health Analytics at McKinsey and Company. Garrett received his PhD in economics from Columbia University in 1996 and was a postdoctoral research fellow in the Robert Wood Johnson Foundation's Scholars in Health Policy Research Program at the University of California, Berkeley, from 1996 to 1998. Acknowledgments This brief was supported by Arnold Ventures. We are grateful to them and to all our funders, who make it possible for Urban to advance its mission. The views expressed are those of the author and should not be attributed to the Urban Institute, its trustees, or its funders. Funders do not determine research findings or the insights and recommendations of Urban experts. Further information on the Urban Institute's funding principles is available at urban.org/fundingprinciples. The author thanks Anuj Gangopadhyaya for data analysis of the MCBS and MCARE-SIM, Carol Carter, Lee-Lee Ellis, Laura Skopec, Doug Wissoker, and Stephen Zuckerman for helpful feedback, and Sarah LaCorte and Devlan O'Connor for editing. ABOUT THE URBAN INSTITUTE The Urban Institute is a nonprofit research organization that provides data and evidence to help advance upward mobility and equity. We are a trusted source for changemakers who seek to strengthen decisionmaking, create inclusive economic growth, and improve the well-being of families and communities. For more than 50 years, Urban has delivered facts that inspire solutions-and this 500 L'Enfant Plaza SW remains our charge today. Washington, DC 20024 Copyright © October 2023. Urban Institute. Permission is granted for www.urban.org reproduction of this file, with attribution to the Urban Institute. 18 POST-ACUTE CARE AND MEDICARE SOLVENCY