Proposed changes to nursing facility payment under consideration by CMS would reduce financial incentives to provide therapy, and would do so with such force – providing higher reimbursement to skilled nursing facilities (SNFs) that provide fewer types of therapy to residents over a shorter period of time or no therapy at all – that it would actually encourage facilities not to provide therapy. Further, the Jimmo v. Sebelius mandate to cover maintenance therapy would be completely ignored.
Ever since the new prospective payment system for Medicare coverage of skilled nursing facilities (SNFs) was first implemented in 1998, the system has faced ongoing criticism. Critics, including the Medicare Payment Advisory Commission[1] and the Department of Health and Human Services’s Office of Inspector General,[2] report that the reimbursement system encourages over-utilization of therapy services, insufficient payment for nursing services, and inaccurate payment for non-therapy ancillary services (chiefly prescription drugs). The Centers for Medicare & Medicaid Services (CMS) contracted with Acumen to develop a new reimbursement system to replace the current system. Four Technical Expert Panels considered draft recommendations. In an Advance Notice of Proposed Rulemaking (ANPRM), CMS solicits comments on “options we may consider” for revising the reimbursement system, based on Acumen’s work.[3] CMS sets out a proposed framework for a new Medicare payment system for SNFs, called Resident Classification System, Version I (RCS-I).
CMS describes three goals for the new reimbursement system: (1) more accurately compensating SNFs; (2) reducing incentives for SNFs to deliver therapy based on financial considerations, rather than resident need; and (3) maintaining simplicity, to the extent possible.[4]
As described in detail in the ANPRM (and illustrated in the chart below), RCS-I dramatically changes the financial incentives for SNFs.[5] Under RCS-I, SNFs would receive higher reimbursement if they provided 15 or fewer days of Medicare coverage and only one form of therapy (not three). Medicare reimbursement would also be higher if 50-75% of a SNF’s Medicare days were billed as non-rehabilitation. In contrast, Medicare reimbursement would be lower for SNFs providing care to the oldest residents (age 90+), to residents receiving three types of therapy, or to residents having 31 or more days of care paid by Medicare.
Comments on the ANPRM are due June 26, 2017. If CMS goes forward with replacing the Medicare SNF reimbursement system with a new system, it will publish a Notice of Proposed Rulemaking a year from now.
Current Medicare Reimbursement System for SNFs
The current system, called Resource Utilization Groups (RUGs), uses a case-mix component and a non case-mix component (which reflects room and board and various capital costs). The case-mix component uses resident assessment information to determine a resident’s classification for payment purposes. RUG-IV has two case-mix categories – nursing (which includes non-therapy ancillary services) and therapy (which includes physical, occupational, and speech therapy). A resident’s RUG classification is based on the higher of the two case-mix categories. Payment for residents in therapy groups reflects the amount of therapy that a SNF reports providing. There are now 66 RUG-IV classifications based on resident assessment information. At this time, more than 90% of residents are assigned to a rehabilitation-based RUG.[6]
Resident Classification System, Version 1
Instead of the RUGs’ two components (nursing and therapy) for case-mix adjustment, RCS-I creates four distinct case-mix categories – physical/occupational therapy, speech language pathology, nursing, and non-therapy ancillaries – with separate case mix adjustments for each.
Area | Case-mix adjustments | Number of case-mix categories |
Physical therapy/occupational therapy | *Clinical reasons for hospital stay (5 categories) | 30 case-mix groups |
Speech language pathology | *Clinical reasons for hospital stay | 18 case-mix groups |
Nursing case mix | CMS is considering using the non-rehabilitation RUG for the nursing component, which is based on the staff-time motion study called Staff Time and Resource Intensity Verification (STRIVE) that was used to develop the nursing case-mix categories for RUG-IV | 43 case-mix groups |
Non therapy ancillaries (NTA) | *Resident comorbidities | 6 case-mix groups |
For speech language pathology and NTA, CMS calculates how much of the variation in current per day costs are predicted by the case-mix adjustments that it proposes. For speech language pathology, the 18 case-mix groups account for only 14.5% of the variation in per day costs; for NTA, only 11.7%.
Two additional factors affect daily reimbursement rates. First, based on Acumen’s finding that physical therapy/occupational therapy costs are generally higher and non-therapy ancillary costs are very high at the beginning of a resident’s stay, RCS-I “front loads” payments – that is, it pays higher rates at the beginning of a resident’s stay, rather than a consistent rate for each day in the assessment period. The proposed system then adjusts rates downward on successive days, under what CMS calls the variable per diem adjustment schedule. Thus, on day 24 of a resident’s Part A stay, for example, the daily rate has a 0.96 adjustment factor, which means that the SNF is paid 96% of the Medicare rate for that person. For a resident in a Medicare-covered stay on days 99 and 100, the adjustment factor is 0.71 – i.e., 71% of the daily rate that was assessed at the beginning of the resident’s Medicare-covered stay.[7]
Second, CMS is considering eliminating the multiple assessments that are currently required for Medicare reimbursement – days 5, 14, 30, 60, and 90 – and using only the 5-day assessment and the “significant change” assessment that is otherwise required for all residents, regardless of payment source. This change means that, in the absence of a resident’s significant change, the reimbursement rate remains constant and does not reflect changes in the resident’s actual needs. Downward adjustments in payment during the course of a resident’s stay reflect the variable per diem adjustment schedule, not the resident’s assessed condition.
Calculating a Daily Rate
To calculate the per day rate for a particular resident, RCS-I begins with unadjusted per diem rate for urban and rural facilities.
RCS-1 Unadjusted Federal Rate Per Diem-Urban and Rural[8]
| Nursing | NTA | PT/OT | SLP | Non-case-mix |
Urban Per Diem Amount | $100.91 | $76.12 | $126.76 | $24.14 | $90.35 |
Rural Per Diem Amount | $ 96.40 | $72.72 | $141.47 | $31.06 | $92.02 |
The unadjusted federal per diem rates (urban or rural) are adjusted twice: first, to reflect the four case-mix categories (nursing, non-therapy ancillary services, physical and occupational therapy, and speech language pathology), and second, to account for declining payments under the variable per diem adjusted schedule. These case-mix adjusted rates, as reduced by the adjustment schedule, are added to the non case-mix component to create a single, declining per day rate for each resident.
Winners and Losers Under RCS-I
The ANPRM includes two Tables identifying the impact of RCS-I on reimbursement rates for individual residents and for facilities. Some key changes are highlighted.
Impact Analysis, Resident-Level[9]
Resident characteristics | Higher reimbursement | Percentage change | Lower reimbursement | Percentage change |
Gender | Males | 1.2% | Females | -0.7% |
Age | Residents under 65 | 5.4% | Residents 90+ years | -2.8% |
Medicare/Medicaid dual status | Residents who are dually eligible for Medicare and Medicaid | 2.9% | Residents are not dually eligible for Medicare and Medicaid | -1.9% |
Disability status | Residents who are disabled | 3.9% | Residents who are aged | -1.2% |
Length of SNF stay | Residents with SNF stays of 1-15 days | 15.9% | Residents with stays of 31+ days | -2.5% |
Use of 100-day SNF benefit | Residents not using 100 days | 0.3% | Residents using 100 days | -2.7% |
Length of qualifying acute care stay | Residents with 31+ qualifying inpatient days | 4.6% | Residents with 3 qualifying inpatient days | -2.3% |
Admitted with diagnosis of a stroke | Residents with a stroke | 0.7% | Residents without a stroke | -0.1% |
Presence of cognitive impairment | Residents who are severely cognitively impaired | 6.1% | Residents who are moderately cognitively impaired | -1.8% |
Admitted with, or has diagnosis of, HIV | Residents without HIV | 0.2% | Residents with HIV | -40.0% |
Receipt of IV medications during stay | Residents with IV medication | 22.9% | Residents without IV medication | -2.0% |
Presence of wound infection | Residents with wound infections | 17.9% | Residents without wound infections | -2.8% |
Receipt of therapy services during SNF stay | Residents receiving a single therapy | 37.3% | Residents receiving 3 therapies | -3.9% |
| Residents not receiving any physical therapy | 24.2% | Residents receiving physical therapy | -1.0% |
| Residents not receiving any occupational therapy | 24.8% | Residents receiving occupational therapy | -1.2% |
| Residents receiving only occupational therapy | 41.8% | Residents receiving physical, occupational, and speech therapy | -3.9% |
Non-therapy ancillary costs during SNF stay | Residents with NTA costs of $150 | 19.2% | Residents with NTA costs of $10-$50 | -3.2% |
Use of extensive services | Residents with tracheostomy | 18.1% |
|
|
The ANPRM identifies the impact of RCS-I on reimbursement rates for facilities.
Impact Analysis, Facility-Level[10]
Provider characteristics | Higher reimbursement | Percentage change | Lower reimbursement | Percentage change |
Facility size | Small facilities, 0-49 beds | 6.7% | Facilities with 200+ beds | -0.7% |
Ownership status | Non-profit facilities | 3.1% | For-profit facilities | -1.1% |
| Government-owned facilities | 7.6% |
|
|
Institution type | Hospital-based and swing-bed facilities | 15.8% |
|
|
% of SNF stays with 100 day utilization | SNFs with 1-10% of their stays utilizing 100 days | 0.3% | SNFs with 25-100% of their stays utilizing 100 days | -3.9% |
% of SNF stays with Medicare/Medicaid dual enrollment | SNFs with 50-75% of their stays with dual eligible residents | 0.8% | SNFs with 0-10% of their stays with dual eligible residents | -1.7% |
% of SNF utilization days billed as rehabilitation ultra high (RU) | SNFs with 1-10% of the utilization days billed as RU | 28.4% | SNFs with 90-100% of the utilization days billed as RU | -9.9% |
% of SNF utilization days billed as non-rehabilitation | SNFs with 50-75% of the utilization days billed as non-rehabilitation | 45.6% | SNFs with 0-10% of the utilization days billed as non-rehabilitation | -2.2% |
Center for Medicare Advocacy Concerns
As demonstrated by the charts above, the proposed revision to Medicare reimbursement for SNFs dramatically alters the Medicare benefit, encouraging less therapy and shorter Medicare-covered stays, while not necessarily improving nurse staffing levels.
During the meetings and in a follow-up letter and conference call with CMS, TEP members expressed concern with Acumen’s charge, which was, apparently, to identify new payment policies and approaches that would be able to more accurately reflect residents’ current use of nursing, therapy, and non-therapy ancillary services. Whether current use is consistent with requirements of federal law was not within Acumen’s scope of work.
Center attorney Toby S. Edelman, a member of all four TEPs, repeatedly raised the issue of provision of, and payment for, maintenance therapy, as required by Jimmo.[11] Acumen responded that the models were based on existing data. With data documenting the cost of maintenance therapy unavailable and possibly non-existent at this time, maintenance therapy is not reflected in the proposed reimbursement system. Acumen suggested that when data become available, Jimmo costs may be added to later revisions of the reimbursement system. The Center questions how data will become available if Jimmo-mandated maintenance nursing and maintenance therapy are not covered by the reimbursement system.
In an August 2016 letter to CMS, Edelman expressed additional concerns about the proposed revisions to the Medicare reimbursement system that the TEPs had been reviewing.[12] She wrote that she had thought that the purposes of revising the SNF reimbursement system were to (1) respond to and correct problems that have been identified for many years in the RUGs system; (2) incorporate new understandings of the care and services that are currently coverable under Medicare (for example, Jimmo's recognition of maintenance therapy and nursing); (3) recognize and implement the statutory directive to pay for care and services that SNFs are required to provide (and therefore, to reflect the revised Requirements of Participations for SNFs, when they were published, which happened in October 2016[13]); and (4) recognize the changes in delivery system reforms and ongoing payment changes, such as bundling demonstrations, that affect reimbursement policies going forward.
Does RCS-I meet CMS’s stated goals?
The short answer is no.
First, RCS-I does not appear to more accurately pay SNFs for providing care to residents who are in a Medicare Part A-covered stay. It rearranges payments, but does not necessarily compensate SNFs appropriately for providing the care and services they are required to provide under the federal Nursing Home Reform Law.[14] The revised Requirements of Participation are not reflected in the proposal. The proposed case-mix adjustments for speech language pathology and NTA have minimal correlation with current facility costs.
Second, while RCS-I does reduce financial incentives to provide therapy, it does so with such force – providing higher reimbursement to SNFs that provide fewer types of therapy to residents over a shorter period of time or no therapy at all – that it actually encourages facilities not to provide therapy. Jimmo’s mandate to cover maintenance therapy is completely ignored.
Finally, RCS-I does not maintain simplicity. It is a highly complex system.
Submitting Comments
Comments must be submitted by June 26. When commenting, refer to file code CMS-1686-ANPRM. Comments may be submitted electronically, at http://ww.regulations.gov, by regular mail, by express or overnight mail, or by hand or courier.[15]
[1] MedPAC, Report to the Congress: Medicare Payment Policy, Chapter 8, page 200 (Mar. 2017) (calling for lower rates and a revised reimbursement system. “Under a revised design, payments would increase for medically complex stays and decrease for stays that include intensive therapy that is unrelated to a patient’s care needs.”), http://medpac.gov/docs/default-source/reports/mar17_entirereport.pdf?sfvrsn=0.
[2] OIG. Questionable Billing by Skilled Nursing Facilities (Dec. 2010), https://oig.hhs.gov/oei/reports/oei-02-09-00202.pdf; OIG, Inappropriate Payments to Skilled Nursing Facilities Cost Medicare More Than a Billion Dollars in 2009 (Nov. 2012), https://oig.hhs.gov/oei/reports/oei-02-09-00200.pdf; OIG, The Medicare Payment System for Skilled Nursing Facilities Needs to be Reevaluated (Sep. 2015), https://oig.hhs.gov/oei/reports/oei-02-13-00610.pdf).
[3] 82 Federal Register 20980 (May 4, 2017), https://www.gpo.gov/fdsys/pkg/FR-2017-05-04/pdf/2017-08519.pdf.
[4] 82 Fed. Reg. 20980, 20984.
[5] 82 Fed. Reg. 20980, 21009-21012.
[6] 82 Fed. Reg. 20980, 20981-20982.
[7] 82 Fed. Reg. 20980, 21002, Table 14.
[8] 82 Fed. Reg. 20980, 20987, Tables 1 and 2 combined.
[9] Taken from 82 Fed. Reg. 20980, 21009-21011, Table 18.
[10] Taken from 82 Fed. Reg. 20980, 21011-21012, Table 19.
[11] Jimmo v. Sebelius, Civ. No. 5:11-CV-17 (D. Vt. Settlement approved Jan. 18, 2011) also requires coverage of maintenance nursing. For information about the case, see https://www.medicareadvocacy.org/?s=jimmo&op.x=0&op.y=0.
[12] The letter is available at https://www.medicareadvocacy.org/letter-to-cms-regarding-snf-technical-expert-panel
[13] 81 Fed. Reg. 68688 (Oct. 4, 2016).
[14] 42 U.S.C. §1395 i-3(a)-(h), 1396r(a)-(h), Medicare and Medicaid, respectively.
[15] 82 Fed. Reg. 20980.