19188 Federal Register /Vol. 65, No. 69/Monday, April 10, 2000/Proposed … · 2010-09-25 · 19188 Federal Register/Vol. 65, No. 69/Monday, April 10, 2000/Proposed Rules DEPARTMENT

19188 Federal Register / Vol. 65, No. 69 / Monday, April 10, 2000 / Proposed Rules

DEPARTMENT OF HEALTH ANDHUMAN SERVICES

Health Care Financing Administration

42 CFR Parts 411 and 489

[HCFA–1112–P]

RIN 0938–AJ93

Medicare Program; ProspectivePayment System and ConsolidatedBilling for Skilled Nursing Facilities—Update

AGENCY: Health Care FinancingAdministration (HCFA), HHS.ACTION: Notice of proposed rulemaking.

SUMMARY: This proposed rule sets forthupdates to the payment rates used underthe prospective payment system (PPS)for skilled nursing facilities (SNFs), forfiscal year 2001. Furthermore, itspecifically proposes changes to theSNF PPS case-mix methodology.Annual updates to the PPS rates arerequired by section 1888(e) of the SocialSecurity Act, as amended by theMedicare, Medicaid and State ChildHealth Insurance Program BalancedBudget Refinement Act of 1999, relatedto Medicare payments and consolidatedbilling for SNFs. In addition, thisproposed rule sets forth certainconforming revisions to the regulationsthat are necessary in order to implementamendments made to the Act by section103 of the Medicare, Medicaid and StateChild Health Insurance ProgramBalanced Budget Refinement Act of1999.

DATES: We will consider comments ifwe receive them at the appropriateaddress, as provided below, no laterthan 5 p.m. on June 9, 2000.ADDRESSES: Mail written comments (1original and 3 copies) to the followingaddress: Health Care FinancingAdministration, Department of Healthand Human Services, Attention: HCFA–1112–P, P.O. Box 8013, Baltimore, MD21244–8013.

If you prefer, you may deliver yourwritten comments (1 original and 3copies) to one of the followingaddresses:Room 443–G, Hubert H. Humphrey

Building, 200 Independence Avenue,SW., Washington, DC 20201, or

Room C5–15–03, 7500 SecurityBoulevard, Baltimore, MD 21244–8150.Because of staffing and resource

limitations, we cannot accept commentsby facsimile (FAX) transmission. Incommenting, please refer to file codeHCFA–1112–P. Comments received

timely will be available for publicinspection as they are received,generally beginning approximately 3weeks after publication of a document,in Room 443–G of the Department’soffice at 200 Independence Avenue,SW., Washington, DC, on Mondaythrough Friday of each week from 8:30to 5 p.m. (phone: (202) 690–7061).FOR FURTHER INFORMATION CONTACT:Dana Burley, (410) 786–4547 or Sheila

Lambowitz, (410) 786–7605 (forinformation related to the case-mixclassification methodology).

John Davis, (410) 786–0008 (forinformation related to the WageIndex).

Bill Ullman, (410) 786–5667 (forinformation related to consolidatedbilling).

Steve Raitzyk, (410) 786–4599 (forinformation related to the facility-specific transition rates).

Bill Ullman, (410) 786–5667 and SusanBurris (410) 786–6655 (for generalinformation).

SUPPLEMENTARY INFORMATION: Copies: Toorder copies of the Federal Registercontaining this document, send yourrequest to: New Orders, Superintendentof Documents, P.O. Box 371954,Pittsburgh, PA 15250–7954. Pleasespecify the date of the issue requestedand enclose a check or money orderpayable to the Superintendent ofDocuments, or enclose your Visa orMaster Card number and expirationdate. Credit card orders can also beplaced by calling the order desk at (202)512–1800 (or toll free at 1–888–293–6498) or by faxing to (202) 512–2250.The cost for each copy is $8. As analternative, you can view andphotocopy the Federal Registerdocument at most libraries designatedas Federal Depository Libraries and atmany other public and academiclibraries throughout the country thatreceive the Federal Register.

To assist readers in referencingsections contained in this document, weare providing the following table ofcontents.

Table of ContentsI. Background

A. Current System for Payment of SkilledNursing Facility Services Under Part Aof the Medicare Program

B. Requirements of the Balanced BudgetAct of 1997 for Updating the ProspectivePayment System for Skilled NursingFacilities

C. The Medicare, Medicaid and State ChildHealth Insurance Program (SCHIP)Balanced Budget Refinement Act of 1999

D. Skilled Nursing Facility ProspectivePayment—General Overview

1. Payment Provisions—Federal Rates2. Payment Provisions—Transition Period

3. Payment Provisions—Facility-SpecificRateII. Update of Payment Rates Under theProspective Payment System for SkilledNursing Facilities

A. Federal Prospective Payment System1. Cost and Services covered by the Federal

Rates2. Methodology Used for the Calculation of

the Federal RatesB. Case-Mix Adjustment and OptionsC. Wage Index Adjustment to Federal RatesD. Updates to the Federal RatesE. Relationship of RUG–III Classification

System to Existing Skilled NursingFacility Level-of-Care Criteria

III. Three-Year Transition PeriodIV. The Skilled Nursing Facility Market

Basket IndexA. Facility-Specific Rate Update FactorB. Federal Rate Update Factor

V. Consolidated BillingVI. Provisions of the Proposed RuleVII. Collection of Information RequirementsVIII. Response to CommentsIX. Regulatory Impact Analysis

A. BackgroundB. Impact of this Proposed Rule

X. FederalismRegulations TextTechnical Appendix A

A. Creation of the Analytic SampleB. Characteristics of the SampleC. Test and Validation SamplesD. Creation of Measure of Non-Therapy

Ancillary Charges from SNF Claims1. Cost-to-Charge MultiplierE. Analysis and Findings—RUG–III

Refinements1. Costs for Beneficiaries Who Qualify for

Both Extensive Services andRehabilitation

2. Non-Therapy Ancillary Index ModelsF. Model Performance1. RUG–III CMI Adjustment2. RUG–III (proposed, version 2001)3. Weighted Index Model (WIM1)4. Weighted Index Model 2 (WIM2)5. Unweighted Index Model (UWIM)G. RUG–III Medications Data1. Creation of MDS-Based Cost Measures2. RUG–Based Imputation Method3. State and Year-Based Imputation

Method

In addition, because of the manyterms to which we refer by abbreviationin this rule, we are listing theseabbreviations and their correspondingterms in alphabetical order below:ADL—Activity of Daily LivingBBA—Balanced Budget Act of 1997BBRA—Balanced Budget Refinement Act of

1999BLS—(U.S.) Bureau of Labor StatisticsCPI—Consumer Price IndexHCFA— Health Care Financing

AdministrationHCPCS—HCFA Common Procedure Coding

SystemIFC—Interim Final Rule with CommentsMDS—Minimum Data SetMSA—Metropolitan Statistical AreaPPI—Producer Price IndexPPS—Prospective Payment SystemPRM—Provider Reimbursement Manual

VerDate 20<MAR>2000 22:49 Apr 07, 2000 Jkt 190000 PO 00000 Frm 00002 Fmt 4701 Sfmt 4702 E:\FR\FM\10APP4.SGM pfrm03 PsN: 10APP4

19189Federal Register / Vol. 65, No. 69 / Monday, April 10, 2000 / Proposed Rules

RUG—Resource Utilization GroupSCHIP—State Child Health Insurance

ProgramSNF—Skilled Nursing Facility

I. Background

A. Current System for Payment ofSkilled Nursing Facility Services UnderPart A of the Medicare Program

Section 4432 of the Balanced BudgetAct of 1997 (BBA) (Pub. L. 105–33)mandated the implementation of a perdiem prospective payment system (PPS)for skilled nursing facilities (SNFs),covering all costs (routine, ancillary,and capital) of covered SNF servicesfurnished to beneficiaries under Part Aof the Medicare program, effective forcost reporting periods beginning on orafter July 1, 1998. The SNF PPSpayment methodology features a case-mix adjustment that utilizes data fromthe comprehensive assessment processrequired for every SNF beneficiary inorder to group them clinically in termsof their degree of resource intensity. Thecase-mix adjustment is designed toensure that the amount of the PPS perdiem payment is appropriate to theindividual beneficiary’s actualcondition, and is sufficient to purchasethe full range of care and services thata beneficiary with a particular clinicalprofile would typically be expected torequire. We are setting forth thisproposed rule in accordance withsection 1888(e)(4)(H)(ii) of the SocialSecurity Act (the Act), which requiresus to publish each year in the FederalRegister any changes in the case-mixclassification system that we use tomake the case-mix adjustment.Although we are not proposing anyother changes in the overall PPSpayment methodology at present, we arenonetheless including a detaileddiscussion of the overall paymentmethodology in section I.C. below, inorder to provide a context for theproposed changes to the case-mixclassification system. In addition, weare incorporating revisions based on theMedicare, Medicaid and State ChildHealth Insurance Program (SCHIP)Balanced Budget Refinement Act of1999 (BBRA). Major elements of thesystem were implemented in an interimfinal rule that was published in theFederal Register on May 12, 1998 (63FR 26252), and in a final rule that waspublished in the Federal Register onJuly 30, 1999 (64 FR 41644). Theseelements are discussed in greater detailin section I.C. below, and include:

• Rates: Per diem Federal rates wereestablished for urban and rural areasusing allowable costs from fiscal year(FY) 1995 cost reports. These rates alsoincluded an estimate of the cost of

services that, before July 1, 1998, hadbeen paid under Part B but furnished toMedicare beneficiaries in a SNF duringa Part A covered stay. Rates are case-mixadjusted using a refined classificationsystem (Resource Utilization Groups,version III (RUG–III)) based onbeneficiary assessments (using theMinimum Data Set (MDS) 2.0). Theproposed refinement to the RUGclassification system is based on criticalanalysis which examined variousoptions to account more precisely forthe variation in non-therapy ancillaryservices in our payments and the careneeds of medically complex patients.The proposed RUG refinement includesthe addition of new categories andincorporation of an ancillary index, asdiscussed in further detail in sectionII.B. In addition, the Federal rates areadjusted by the hospital wage index toaccount for geographic variation inwages. At this time, data for the FY 2001hospital wage index is not yet available;therefore, the index applied in thisproposed rule is the same index used inthe July 30, 1999 update notice. We willbe updating the wage index in the finalrule using the latest hospital wage data.Further, the rates are adjusted annuallyusing an SNF market basket index.Lastly, as a result of section 101 of theBBRA, for SNF services furnished on orafter April 1, 2000, and before the laterof October 1, 2000, or implementationby the Secretary of Health and HumanServices of a refined RUG system, perdiem adjusted payments are increasedby 20 percent for 15 RUGs falling undercategories for Extensive Services,Special Care, Clinically Complex, HighRehabilitation and MediumRehabilitation. This 20 percent increaseserves solely as a temporary, interimadjustment to the payment rates andRUG–III classification system aspublished in the final rule of July 30,1999, until we have had the opportunityto implement the case-mix refinementsproposed in this rule. At that point, thetemporary adjustment afforded by the20 percent increase will no longer beapplicable, as payment will be made inaccordance with the newly-refinedRUGs. The RUG–III groups to which thisadjustment applies are: SE3, SE2, SE1,SSC, SSB, SSA, CC2, CC1, CB2, CB1,CA2, CA1, RHC, RMC and RMB. Inaddition, for FY 2001 and FY 2002, theadjusted Federal per diem payment to afacility is increased by 4 percent in eachyear, calculated exclusive of the 20percent RUG rate increase.

• Transition: The SNF PPS includes a3-year, phased transition that blends afacility-specific payment rate with theFederal case-mix adjusted rate. The

blend used changes for each costreporting period after a facility migratesto the new system. For most facilities,the facility-specific rate is based onallowable costs from FY 1995. As aresult of section 102 of the BBRA of1999, SNFs may elect immediatetransition to the Federal rate on or afterDecember 15, 1999 for cost reportingperiods beginning on or after January 1,2000. There is no such election for costreporting periods beginning beforeJanuary 1, 2000. SNFs may electimmediate transition up to 30 days afterthe start of their cost reporting period.

• Coverage: The PPS statute did notchange Medicare’s fundamentalrequirements for SNF coverage.However, because RUG–III classificationis based, in part, on the beneficiary’sneed for skilled nursing care andtherapy, we have attempted wherepossible to coordinate claims reviewprocedures with the outputs ofbeneficiary assessment and RUG–IIIclassifying activities. For example, webelieve that when an initial Medicarerequired (5-day) assessment, properlycompleted, places the beneficiary in oneof the upper RUG–III classifications thatwe designate as representing a coveredlevel of SNF care (see section II.E. ofthis preamble), this provides the basisfor us to assume that the beneficiaryneeded such care upon admission andat least up until the assessmentreference date for the initial Medicare-required assessment. We will, however,continue to make individual reviewdeterminations for claims of thoseindividuals who classify in one of thelower RUG–III categories.

• Consolidated Billing: The statuteincludes a billing provision thatrequires a SNF to submit consolidatedMedicare bills for its beneficiaries forvirtually all services that are coveredunder either Part A or Part B. Thestatute excludes a small list of services(primarily those of physicians andcertain other types of practitioners). Asdiscussed later in this preamble, section103 of the BBRA has identified certainadditional services for exclusion,effective April 1, 2000.

As noted above, an interim final ruleimplementing the SNF PPS waspublished in the Federal Register onMay 12, 1998, for which the commentperiod was initially scheduled to closeon July 13, 1998. A subsequent noticeextended the public comment period foran additional 60 days (July 13, 1998, (63FR 37498)), and a second noticereopened the comment period foranother 30 days (November 27, 1998 (63FR 65561)). In addition, a correctionnotice was published October 5, 1998(63 FR 53301) that made a number of



minor technical and editorialcorrections to the interim final rule. Inthe July 30, 1999, final rule weresponded to the public commentsreceived on the interim final rule andmade a number of modifications in theregulation. This final rule was followedby a correction notice published onNovember 4, 1999 (64 FR 60122), whichmade a technical correction to the finalrule’s preamble. Also on July 30, 1999,we issued an update notice (64 FR41684), followed by a correction noticepublished on October 5, 1999 (64 FR54031). We have also issued severalProgram Memoranda on claimsprocessing and billing under the SNFPPS that are available on the SNF PPShome page at the HCFA website on theInternet, at the following location:<www.hcfa.gov/Medicare/snfpps.htm>

B. Requirements of the Balanced BudgetAct of 1997 for Updating the ProspectivePayment System for Skilled NursingFacilities

As described above, section1888(e)(4)(H) of the Act requires that wepublish in the Federal Register:

1. The unadjusted Federal per diemrates to be applied to days of coveredSNF services furnished during the FY.

2. The case-mix classification systemto be applied with respect to theseservices during the FY.

3. The factors to be applied in makingthe area wage adjustment with respectto these services.

In addition, in the July 30, 1999 finalrule, we indicated that we wouldannounce any changes to the guidelinesfor Medicare level of caredeterminations related to Part A SNFservices or to the RUG–IIIclassifications.

This proposed rule updates the ratesas mandated by the Medicare statute.

C. The Medicare, Medicaid and StateChild Health Insurance Program(SCHIP) Balanced Budget RefinementAct of 1999

As a result of enactment of the BBRA,there are several new provisions thatresult in adjustments to the PPS forSNFs. The following highlights themajor provisions involving the PPS forSNFs:

Temporary Increase in Payment forCertain High Cost Residents

As noted previously, section 101 ofthe BBRA provides for a temporary, 20percent increase in the per diemadjusted payment rates for 15 specifiedRUGs, falling under categories forExtensive Services, Special Care,Clinically Complex, High Rehabilitationand Medium Rehabilitation. The

specific RUG–III groups to which thisadjustment applies are: SE3, SE2, SE1,SSC, SSB, SSA, CC2, CC1, CB2, CB1,CA2, CA1, RHC, RMC, and RMB. Thestatute provides that the 20 percentincrease takes effect with SNF servicesthat are furnished on or after April 1,2000, and continues until the later ofOctober 1, 2000, or implementation bythe Secretary of a refined RUG system.Thus, the 20 percent increase servessolely as a temporary, interimadjustment to the payment rates andRUG–III classification system aspublished in the final rule of July 30,1999, until we have implemented thecase-mix refinements that we nowpropose elsewhere in this document,which we expect to accomplish byOctober 1, 2000. Once we haveimplemented the case-mix refinements,the temporary adjustment afforded bythe 20 percent increase will no longer beapplicable, as we will then makepayment in accordance with the newly-refined RUGs.

For FY 2001 and FY 2002, section 101of the BBRA also provides for an across-the-board increase in the adjustedFederal per diem payment rates by 4percent in each year, calculatedexclusive of the 20 percent RUG rateincrease discussed above. Unlike the 20percent increase, which is targeted atcertain particular RUG–III groups, this 4percent increase will apply equally toall RUG groups.

Election For Immediate Transition toFederal Rate

As noted earlier, under section 102 ofthe BBRA, all SNFs may now elect tobypass the transition and be paid basedupon 100 percent of the Federal rate.This election applies to cost reportingperiods beginning on or after January 1,2000. There is no such election for costreporting periods beginning beforeJanuary 1, 2000. SNFs may make thiselection beginning on or after December15, 1999 and up to 30 days after the startof their cost reporting periods. Anelection to bypass the transition iseffective for all subsequent periods andcannot be rescinded once it is effective.Further information can be found inProgram Memorandum A–99–53.

Special Payment Adjustment for CertainSNFs

Section 155 of the BBRA provides thatPPS payments to certain SNF providerslocated in Baldwin or Mobile County,Alabama, will be based on 100 percentof their facility specific rates for costreporting periods that begin in FY 2000or FY 2001. In addition, it requires thatthe facility specific portion of theirpayment rate be calculated using data

from their cost reporting periodbeginning in FY 1998. In order to beeligible for this special payment, a SNFmust meet the following criteria: beganparticipation in the Medicare programbefore January 1, 1995; have at least 80percent of the total inpatient days of thefacility in the cost reporting periodbeginning in FY 1998 comprised ofpersons entitled to Medicare; and, belocated in Baldwin or Mobile County,Alabama.

Special SNF PPS Payment Provisionsfor SNFs with Certain Types of PatientPopulations

Section 105 of the BBRA addsparagraph (12) to section 1888(e) of theAct and permits certain SNFs to receive50 percent of the facility specific rateand 50 percent of the Federal per diemrate, effective from November 29, 1999,until September 30, 2001. In order to beeligible, a SNF must: have been certifiedas an SNF under Medicare prior to July1, 1992; be a hospital-based facility;and, in the cost reporting periodbeginning in FY 1998, have had apatient population, eligible for Part Abenefits, of which at least 60 percentwere ‘‘immuno-compromised secondaryto an infectious disease,’’ with ‘‘specificdiagnoses specified by the Secretary.’’The statute gives the Secretary theauthority to specify the diagnosisassociated with this provision, and webelieve the legislative history providessome guidance concerning theapplication of this provision. The HouseWays and Means Committee report (H.Rep. 106–436, Part 1 at 47) indicatesthat this provision is directed atfacilities that serve ‘‘* * * veryspecialized patients * * * whosemedical conditions are not well-accounted for in the RUG classificationsystem.’’ The Senate Finance CommitteeReport (S. Rep. 106–199 at 8) indicatesthe need to study ‘‘* * * alternativepayment methods for skilled nursingfacilities that specialize in providingcare to extremely high cost, chronicallyill populations * * *’’ such as ‘‘afacility that exclusively specializes incaring for AIDS patients * * *’’ In lightof this general Congressional intent, webelieve that the scope of this provisionshould be limited and propose that thisprovision be applied to humanimmunodeficiency virus (HIV) as codedin ICD–9–CM with the following code:042.

Provision for Part B Add-Ons forFacilities Participating in the NursingHome Case-Mix and Quality (NHCMQ)Demonstration Project

Under prior law, section 1888(e)(3) ofthe Act provided for an add-on to the



payment rates for Part B servicesfurnished during the course of a Part Acovered stay for those facilities that didnot participate in the demonstration thatpreceded SNF PPS. However, the Actdid not provide for a similar add-on forfacilities that did participate in thedemonstration project. Therefore,section 104 of the BBRA amendedsection 1888(e)(3) to provide that SNFsthat had participated in the NursingHome Case Mix and QualityDemonstration (NHCMQ) project areeligible for the inclusion of a Part B add-on amount in their facility specific PPSrates. This provision is effective as ifincluded in the enactment of the BBAand, therefore, applies to all costreporting periods subject to the PPStransition.

For the purpose of computing facilityspecific rates, the base year forproviders participating in the NHCMQdemonstration project is calendar year1997 rather than FY 1995 (which is thebase year for SNFs not participating inthe demonstration project). Therefore,the Part B add-on amounts for thedemonstration SNFs will be calculatedusing data from the appropriate periodsin 1997. Because of the time periodnecessary for us to compute theseamounts, existing Part B data from 1995will be updated for inflation and usedas the bases for payment on an interimbasis until we can develop the finalamounts using the 1997 data, at whichpoint earlier payments will be adjustedto reflect the correct data.

Exclusion of Certain AdditionalServices from the SNF PPS Bundle andConsolidated Billing

The original SNF PPS legislation inthe BBA identified several servicecategories that were excluded from theSNF consolidated billing requirement,as well as from the bundled Part Apayment made under the SNF PPSitself. Effective with services furnishedon or after April 1, 2000, section 103(a)of the BBRA has amended section1888(e)(2)(A) to exclude certainadditional types of services from theconsolidated billing requirement, thusallowing these services to be billedseparately to Part B. Section 103(b) ofthe BBRA has also amended section1888(e)(4)(G) to provide for acorresponding proportional reduction inPart A SNF payments, beginning withFY 2001. We discuss these additionalexcluded service categories in section V.of this preamble, on consolidatedbilling.

D. Skilled Nursing Facility ProspectivePayment—General Overview

The Medicare SNF PPS wasimplemented for cost reporting periodsbeginning on or after July 1, 1998.Under the PPS, SNFs are paid throughper diem prospective case-mix adjustedpayment rates applicable to all coveredSNF services. These payment ratescover all the costs of furnishing coveredskilled nursing services (that is, routine,ancillary, and capital-related costs)other than costs associated withapproved educational activities.Covered SNF services includeposthospital SNF services for whichbenefits are provided under Part A andall items and services that, before July1, 1998, had been paid under Part B(other than physician and certain otherservices specifically excluded under theBBA) but furnished to Medicarebeneficiaries in a SNF during a Part Acovered stay. (For a complete discussionof these provisions, see the May 12,1998 interim final rule (63 FR 26252)).

1. Payment Provisions—Federal Rate

The statute sets forth a fairlyprescriptive methodology for calculatingthe amount of payment under the SNFPPS. The PPS utilizes per diem Federalpayment rates based on mean SNF costsin a base year updated for inflation tothe first effective period of the PPS. Wedeveloped the Federal payment ratesusing allowable costs from hospital-based and freestanding SNF cost reportsfor reporting periods beginning in FY1995. The data used in developing theFederal rates also incorporate anestimate of the amounts that would bepayable under Part B for covered SNFservices to individuals who werereceiving Part A covered services in anSNF. In developing the rates for theinitial period, we updated costs to thefirst effective year of PPS (15-monthperiod beginning July 1, 1998) using aSNF market basket index, andstandardized for facility differences incase-mix and for geographic variationsin wages. Providers that received ‘‘newprovider’’ exemptions from the routinecost limits were excluded from thedatabase used to compute the Federalpayment rates. In addition, costs relatedto payments for exceptions to theroutine cost limits were excluded fromthe database used to compute theFederal rates. In accordance with theformula prescribed in the BBA, we setthe Federal rates at a level equal to theweighted mean of freestanding costsplus 50 percent of the differencebetween the freestanding mean andweighted mean of all SNF costs(hospital-based and freestanding)

combined. We compute and applyseparately the payment rates forfacilities located in urban and ruralareas. In addition, we adjust the portionof the Federal rate attributable to wagerelated costs by a wage index.

The Federal rate also incorporatesadjustments to account for facility case-mix using a classification system thataccounts for the relative resourceutilization of different patient types.This classification system, RUG–III,utilizes beneficiary assessment data(from the Minimum Data Set or MDS)completed by SNFs to assignbeneficiaries into one of 178 groups.The May 12, 1998 interim final rule (63FR 26252) has a complete and detaileddescription of the original (44 group)RUG–III classification system. Adetailed discussion of the proposedchanges to the RUG classificationsystem is found in Section II.B. of thisproposed rule.

The Federal rates reflected in thisnotice update the rates in the July 30,1999 update notice (64 FR 41684) by afactor equal to the SNF market basketindex minus 1 percentage point.According to section 1888(e)(4)(E)(ii) ofthe Act, for FYs 2001 and 2002, we willupdate the rate by adjusting the currentrates by the SNF market basket changeminus 1 percentage point. Forsubsequent FYs, we will adjust the ratesby the applicable SNF market basketchange.

2. Payment Provisions—TransitionPeriod

Beginning with a provider’s first costreporting period beginning on or afterJuly 1, 1998, there is a transition periodcovering three cost reporting periods.During the transition period, SNFsreceive a payment rate comprising ablend between the Federal rate and afacility-specific rate based on eachfacility’s FY 1995 cost report. Undersection 1888(e)(2)(E)(ii) of the Act, SNFsthat received their first payment fromMedicare on or after October 1, 1995receive payment according to theFederal rates only.

For SNFs subject to transition, thecomposition of the blended rate variesdepending on the year of transition. Forthe first cost reporting period beginningon or after July 1, 1998, we makepayment based on 75 percent of thefacility-specific rate and 25 percent ofthe Federal rate. In the next costreporting period, the rate consists of 50percent of the facility-specific rate and50 percent of the Federal rate. In thefollowing cost reporting period, the rateconsists of 25 percent of the facility-specific rate and 75 percent of theFederal rate. For all subsequent cost



reporting periods, we base paymentsentirely on the Federal rates.

As noted earlier, in accordance withsection 102 of the BBRA, SNFs thatwould otherwise be subject to thestatutory three-year, phased transitionfrom facility-specific to Federal rates,may elect to bypass the transition andgo directly to the full Federal rate. Thisamendment applies to elections madeon or after December 15, 1999, exceptthat no election will be effective for acost reporting period beginning beforeJanuary 1, 2000; an election is effectivefor a cost reporting period beginning noearlier than 30 days before the date ofthe election.

3. Payment Provisions—Facility-Specific Rate

For most facilities, we compute thefacility-specific payment rate utilizedfor the transition using the allowablecosts of SNF services for cost reportingperiods beginning in FY 1995 (costreporting periods beginning on or afterOctober 1, 1994 and before October 1,1995). Included in the facility-specificper diem rate is an estimate of theamount that would be payable underPart B for covered SNF servicesfurnished during FY 1995 to individualswho were beneficiaries of the facilityand receiving Part A covered services.The facility-specific rate, in contrast tothe Federal rates, includes amounts paidto SNFs for exceptions to the routinecost limits. In addition, we also takeinto account ‘‘new provider’’exemptions from the routine cost limits,but only to the extent that routine costsdo not exceed 150 percent of the routinecost limit.

We update the facility-specific rate foreach cost reporting period after FY 1995to the first cost reporting periodbeginning on or after July 1, 1998 (theinitial period of the PPS) by a factorequal to the SNF market basketpercentage increase minus 1 percentagepoint. For FYs 1998 and 1999, weupdated this rate by a factor equal to theSNF market basket increase minus 1percentage point, and in eachsubsequent year, we will update it bythe applicable SNF market basketincrease.

Appeals RightsIn enacting SNF PPS, Congress

imposed limitations on the rights ofSNFs to appeal their new payment rates(section 1888(e)(8) of the Social SecurityAct). Similar to the hospital PPS, thenew SNF system begins with atransition period, wherein a portion ofthe payment rates (that is, the facility-specific rate) is based upon thefacilities’ costs in a base period (cost

reporting periods beginning in 1995).The facility-specific portion of the ratephases out over the course of a threeyear cost reporting transition period,after which the SNFs will be paid on afully Federal rate. The statutorylanguage removes the Federal portion ofthe rate from administrative and judicialreview, while allowing for a limitedreview of the facility-specific portion ofthe rate related to an SNFs Part Ahistorical costs from the 1995 base year.The language of the interim final rulewith comment and the MedicareProvider Reimbursement Manual (PRM)contemplate situations whereadjustments are made to thereimbursement amounts allowable inthe base year that are used to set thefacility-specific portion of a provider’sPPS rate. Adjustments may be made inthe cost report settlement process and/or providers may have appealed specificcost report adjustments. Whereadjustments are made to the base yearcosts either through final settlement ofthe cost report or as a result of an appealof the base year Notice of ProgramReimbursement (NPR), suchadjustments may be applied to thefacility-specific portion of the PPS ratefor any cost years that are open or arewithin the time periods subject toreopening under the regulations at 42CFR 405.1885. Additionally, providersmay challenge the facility-specificportion of their rates by appealing thefacility-specific rate notice they receivefrom their fiscal intermediary before thestart of SNF PPS. The fiscalintermediaries will apply anyadjustments resulting from a successfulchallenge to this rate notice to all opentransition years. Providers may alsochallenge their facility-specific rates byappealing their transition year NPRs.Adjustments obtained through a NPRchallenge will only be applied to theyear under appeal. Moreover, inaccordance with the judicial reviewprohibitions contained in section1888(e)(8)(B) of the Act, all reviews offacility-specific rates are limited tochallenges relating to specific MedicarePart A costs in the base year.

II. Update of Payment Rates Under theProspective Payment System for SkilledNursing Facilities

A. Federal Prospective Payment System

This rule sets forth a proposedschedule of Federal prospectivepayment rates applicable to MedicarePart A SNF services beginning October1, 2000. The schedule incorporates perdiem Federal rates designed to providePart A payment for all costs of services

furnished to a beneficiary of an SNFduring a Medicare-covered stay.

1. Cost and Services Covered by theFederal Rates

The Federal rates apply to all costs(that is, routine, ancillary, and capitalrelated costs) of covered SNF servicesother than costs associated withoperating approved educationalactivities as defined in § 413.85. Undersection 1888(e)(2) of the Act, coveredSNF services include posthospital SNFservices for which benefits are providedunder Part A (the hospital insuranceprogram), as well as all items andservices (other than those servicesexcluded by statute) that, before July 1,1998, were paid under Part B (thesupplementary medical insuranceprogram) but furnished to Medicarebeneficiaries in a SNF during a Part Acovered stay. (These excluded servicecategories are discussed in greater detailin section V.B.2. of the May 12, 1998interim final rule (63 FR 26295–97).Also, as mentioned previously, section103 of the BBRA has identified certainadditional types of services forexclusion from the SNF PPS bundle,and has provided for a correspondingproportional reduction in Part A SNFpayments beginning with FY 2001.).

2. Methodology Used for the Calculationof the Federal Rates

The methodology to compute theunadjusted Federal rates incorporatesseveral changes since we published thefinal rule on July 30, 1999 (64 FR41684). First, to facilitate theincorporation of our proposedrefinement to the case mix classificationsystem, we are creating a newcomponent of the payment rates toaccount for non-therapy ancillaryservices. This component is beingcreated by moving the non-therapyancillary costs used in establishing thenursing case-mix component of thepayment rates to a separate component.For the payment rates associated withurban areas, 43.4 percent of the nursingcase mix component is related to non-therapy ancillary services (includingPart B services). For the payment ratesassociated with rural areas, 42.7 percentof the nursing case mix component isrelated to non-therapy ancillary services(including Part B services). Thesepercentages were previously identifiedin a Federal Register notice datedNovember 27, 1998 (63 FR 65561). Thisnew component of the payment rates ispresented in Tables 1 and 2 of thisproposed rule.

In addition, in accordance withsection 103 of the BBRA, the Federalrates will be adjusted to reflect the



exclusion of certain items and servicesfrom consolidated billing, as explainedpreviously. The complexity and timenecessary for computing the numericadjustment itself does not allow us topresent it in this proposed rule.However, we describe the generalmethodology that we plan to use later inthis preamble (in the discussion of thePPS Rate Tables). As required by thestatute, the rates are updated using thelatest market basket percentage minus 1percentage point. For a completedescription of the multi-step process,

see the May 12, 1998 interim final rule.In addition, based on section 101 of theBBRA, we have provided for a 4 percentincrease in the adjusted Federal rate forFY 2001. This 4 percent adjustment isnot reflected in the rate tables (Tables 1,2, 5, and 6 of this proposed rule). Inaccordance with the statute, it is appliedafter all adjustments (wage and case-mix). See the example in Section III;Table 9, of this proposed rule.

The SNF market basket is used toadjust each per diem component of theFederal rates forward to reflect costincreases occurring between the

midpoint of the Federal FY beginningOctober 1, 1999 and the midpoint of theFederal FY beginning October 1, 2000,and ending September 30, 2001, towhich the payment rates apply. Inaccordance with section 1888(e)(4)(B) ofthe Act, the payment rates are updatedbetween FY 2000 and FY 2001 by afactor equivalent to the annual marketbasket index percentage increase minus1 percentage point. This factor is equalto 1.01833. Tables 1 and 2 below reflectthe updated components of theunadjusted Federal rates.

TABLE 1.—UNADJUSTED FEDERAL RATE PER DIEM: URBAN

Rate component Nursingcase-mix

Medicalancillary

Therapycase-mix

Therapynon-case mix

Non-case-mix

Per Diem Amount ................................................................ $64.49 $49.45 $85.79 $11.32 $58.25

TABLE 2.—UNADJUSTED FEDERAL RATE PER DIEM: RURAL

Rate component Nursingcase-mix

Medicalancillary

Therapycase-mix

Therapynon-case mix

Non-case-mix

Per Diem Amount ................................................................ $62.50 $46.58 $99.11 $12.10 $59.32

B. Case-Mix Adjustment and OptionsAs required by the BBA, HCFA must

publish the SNF PPS case-mixclassification methodology applicablefor the next Federal FY before August 1of each year. This proposed rulediscusses options for refinements to theRUG–III system, describes ongoingresearch and analyses, shares the initialresults that we propose be incorporatedinto the Medicare PPS system effectiveOctober 1, 2000, and solicits commentsfrom all interested parties. During thenext 60 days, comments will bereviewed and considered, additionalanalyses will be conducted, and finaldecisions will be made on the need for,and types of, RUG–III refinements to beimplemented. A final rule will then bepromulgated before August 1, 2000.

Research GoalsWe commissioned a study to review

the RUG–III classification system withparticular emphasis on the care needs ofmedically complex Medicarebeneficiaries and the variation in non-therapy ancillary services within RUG–III categories. This project is a majorpriority for us, the provider industry,and others. The initial researchidentified potential refinements to thesystem that we propose to implementeffective October 1, 2000.

A key part of this research was theexploration of potential refinements tothe Extensive Services category.Previous research showed that the

Extensive category is associated withthe highest per diem non-therapyancillary costs of any of the RUG–IIIcategories. The research also indicatedthat, while the Extensive Servicescategory did capture a disproportionateshare of high cost beneficiaries, therewas considerable variance in costswithin this category as well as withinother categories. In the current project,additional studies were conducted toextend the analysis of non-therapyancillary costs and within-groupvariance to other RUG–III categories.

The researchers focused on thefollowing analyses to identify options,and the results were used to develop theproposed RUG–III refinementsdiscussed in this rule:

1. Evaluate the ability of the currentRUG–III system to predict variance indrug, respiratory or other non-therapyancillary costs.

2. Evaluate the ability of specific MDSitems to predict variance in non-therapyancillary costs, and identify the MDSitems most closely associated withdifferences in non-therapy ancillarycosts.

3. Design/test potential refinements tothe RUG–III methodology.

A detailed description of themethodology used to conduct theseanalyses is included in the TechnicalAppendix A to this proposed rule.

Data SourcesSince ensuring the equity and

accuracy of the SNF PPS has been, andcontinues to be, a major HCFA priority,the studies were initiated shortly afterthe introduction of the new paymentsystem. In fact, the research wasconducted before actual PPS claims andacuity data became available. For thisreason, the analyses described here wereconducted using a large cross-linkedresearch data base that included clinicalassessment data collected from theFederally-mandated MDS, druginformation, our claims data, andorganizational data on nursing homeproviders. The data sets used in theanalyses are described below:

Minimum Data Set (MDS)MDS data were collected from 6

states: Kansas, Maine, Mississippi,Ohio, South Dakota, and Texas. (Asexplained in Technical Appendix A, wewere unable to utilize data from aseventh state, New York, due to thatstate’s use of an all-inclusive paymentrate.) These states were selected becausethe MDS data had been collected andused for rate-setting purposes prior tothe start of the Medicare SNF PPS(either through the HCFA Case-MixDemonstration Project or for stateMedicaid payment systems), andprovided a greater number of MDSrecords over a longer period of timethan available from any other source. Inaddition, previous demonstration



project reliability studies and statevalidation activities indicated agenerally high level of data accuracy.

MDS data used in this study were forcalendar years 1995, 1996 and 1997(except for Texas, where data were onlyavailable for 1997), and includedassessments for Medicare beneficiaries,Medicaid recipients and private paypatients. While some states requiredMDS assessments for all beneficiariesadmitted to the SNF regardless of thelength of stay, most of the assessmentswere prepared following the Federalguidelines in effect at the time; that is,assessments required by day 14 of theSNF admission.

MDS Drug Data

Facilities participating in the HCFACase-Mix Demonstration projectsubmitted medications data as part oftheir MDS assessments. In addition,several of the states, including Maine,South Dakota, and Ohio, required themedications data with every MDS,regardless of payor source. Themedications reported on the MDSs werecollected from seven states, the sixstates used for this study, plus NewYork (see Technical Appendix A fordetails on the use of New York data).

Up to 18 medications administeredduring the assessment reference periodcan be reported on an MDS record. TheMDS drug data were cleansed andverified through a combination ofmanual examination (by either a clinicalpharmacist or physician) andcomputerized reclassification ofNational Drug Codes (NDC). The datawere then ordered into therapeuticgroups for easier analysis.

SNF Claims

All SNF Medicare claims spanningthe years 1995 through 1997 weredownloaded from the HCFA Data Centerand matched to MDS files. The fileswere constructed so that there aremultiple observations per SNF stay ifmultiple MDS assessments wereperformed.

Staff Time Measurement (STM) StudyData

This analysis incorporated HCFASTM Study data (combined 1995 and1997). The May 12, 1998 interim finalrule described the STM Study, and themethodology used to incorporate theSTM data into Medicare PPS rate-setting. These data were used to imputestaff time costs for the observations usedin this study.

On-Line Survey Certification andReporting System (OSCAR) Data

The OSCAR data provide facility-levelinformation, such as the results fromannual survey inspections andinformation regarding facility type.OSCAR data from 1991 through 1998were linked serially into a longitudinalfile. The analytic database constructedfor this research has been merged to thislongitudinal OSCAR file through thelinking of facility identifiers, using theOSCAR information from the surveydates closest to the MDS assessmentdata.

Case Mix Research Findings

While maintaining the generalstructure of RUG–III, we found that thetwo most viable ways to refine thesystem are by adding new categoriesand end splits to the system, and bydeveloping a new index system toreflect the variation of non-therapyancillary service costs. Adoption ofthese refinements will add additionalgroups to the case-mix system,somewhat increasing its complexity.This proposed change also mayintroduce some initial uncertainty forproviders, who would have to becomefamiliar with the refined system andmodify existing operational and supportsystems.

In evaluating a particular change, wefirst identified the drawbacks of thatchange (for example, added complexityof the RUG–III model and time andeffort required by providers, contractors,and beneficiaries to assimilate thechange). Then, to evaluate the overalldesirability of the potential change, weweighed these drawbacks against thebenefits, such as the expectedimprovement in payment and clinicalaccuracy. In addition, we evaluatedpotential refinements in terms ofpossible incentives and disincentivesrelated to access, quality and cost-effectiveness of SNF care. Weincorporated this analysis into ourevaluation of potential RUG–IIIrefinements.

After careful review and extensiveanalysis, we then identified severalpossible RUG–III refinements that willimprove the accuracy of SNF PPSpayments. One such refinement is thedevelopment of new categories forbeneficiaries who qualify for both theRUG–III Rehabilitation and ExtensiveServices categories. As expected, ouranalyses indicated that ancillary costswere much higher for Medicarebeneficiaries in the Extensive Servicescategory than for those in othercategories. There are also a significantnumber of beneficiaries who would

classify into the Extensive Servicescategory based on clinical conditionsbut who, because they are also receivingrehabilitation services, classify into oneof the Rehabilitation categories instead(due to the hierarchical logic of theRUG–III classification system). Thesebeneficiaries carry with them the samenon-therapy ancillary costs associatedwith their complex clinical needs eventhough they are classified into a RUG–III Rehabilitation category.

The high costs for beneficiaries in theExtensive Services category suggest thatthe payment rate for Extensive Servicesshould be increased. However,increasing the payment rate withoutfurther adjustments could adverselyaffect provider incentives to providetherapy to beneficiaries requiringExtensive Services. Therefore, weexpanded the scope of the proposedrefinement to include a new category forbeneficiaries who qualify for bothExtensive Services and a RUG–IIIRehabilitation category.

Our research findings showed little orno correlation between the groupswithin the Extensive Services category(that is, SE1, SE2, SE3) and the level ofrehabilitation services used. For thisreason, the structure for the newhierarchy level proposed here wouldmirror that of the existing Rehabilitationcategories. Thus, we would add to thecurrent RUG–III model fourteen (14)new ‘‘Rehabilitation and ExtensiveServices’’ sub-categories that use thesame Rehabilitation sub-category andADL splits as the current system (SeeTable 4 for the proposed RUG–IIIstructure).

The second component of theproposed refinement is the developmentof a separate ‘‘non-therapy ancillary’’index based on clinical variables on theMDS. We tested MDS items to identifyclinical conditions and services that arepredictive of non-therapy ancillarycosts. First, we analyzed each MDSvariable independently, and identifiedall MDS items that had a significantpositive relationship (at the 5 percentlevel) with per diem non-therapyancillary costs. Next, we identifiedcombinations of MDS items that wereassociated with significant costdifferences. We then evaluated variablesfor clinical validity and potentialincentive effects. For example, werejected consideration of indwellingcatheters as case-mix adjustors due tothe potential negative incentive factorsassociated with their use in the index.See Table 3 for a list of MDS items thatwere found to be associated withsignificant differences in ancillary costs.

Once we identified the criticalpredictive variables, we investigated a



number of index model approaches. Wedeveloped weighted and unweightedversions of a non-therapy ancillaryindex. Both versions improved thevariance prediction of the case-mixsystem. The unweighted index modelassigns a non-therapy ancillary levelbased on a count of the variables(selected MDS items) associated withnon-therapy ancillary costs. Under theweighted index model, different weightsare assigned to the selected MDS itemsbased on the difference in costsassociated with the item. In this study,the researchers assigned the weightsbased on quantitative analysis of thedata. With both indices, thresholds weredetermined to form subgroups whichvary logically in cost. However, thesecost variations relate to the researchdata base, and need to be verifiedagainst the national MDS/Medicareclaims data base.

The grouping logic used for therefined RUG–III is very similar to thatcurrently used. The same 108 MDSitems that are used to classifybeneficiaries into the 44 RUG–III groupswill be used to classify beneficiariesinto the refined RUG–III subcategoriesin either the unweighted or weightedindex models. It is only at the last levelof classification that additional MDSitems are considered. The MDS itemsused for the last step of classificationinclude some of the 108 items that areused for the first level of classification

in addition to some others, either aloneor in combinations.

The last step to grouping using theunweighted index model (UWIM) thatwe are proposing is based on a count ofclinical variables, up to a maximum of11. There are 11 ‘‘domains,’’ some ofwhich are comprised of multiple MDSclinical variables. The clinicalconditions and services that define thedomains are shown in Table 3. Withina domain, any one clinical variable, orcombination of variables, satisfies thecriteria for being included in the countfor classification into one of the refinedRUG–III groups. For example, the firstdomain is ‘‘Parenteral/IV feeding withgreater than 76 percent total calories.’’In order for the domain to be countedfor determining the final step in RUG–III classification in the UWIM, the MDSitems K5a and K6a must be coded toreflect the receipt by the beneficiary ofat least 76 percent of total nutritionreceived via parenteral or IV feeding inthe previous 7 days.

Other domains are comprised of manymore MDS items than the parenteral/IVfeeding domain. An example of this isthe domain entitled, ‘‘Oxygen and eitherpneumonia or respiratory infection withfever, or pneumonia or respiratoryinfection, chronic obstructivepulmonary disease, congestive heartfailure, coronary artery disease withshortness of breath.’’ This domain willonly count once toward classificationeven though it is possible for abeneficiary to have values for all of

these clinical conditions. As soon as thegrouper software identifies that onecombination of MDS items’ values ispresent on the MDS that satisfies thisdomain, it will credit the case with acount of 1 in addition to whatever otherdomain criteria are satisfied by theMDS.

The identified clinical variables areused for classification of every MedicareMDS in the Clinically Complex categoryand above, regardless of the otherqualifying conditions and servicesreported on the MDS. This means thata beneficiary who has a count of 2 of therelevant clinical variables, will classifyinto the ‘‘3’’ level of the particularrefined RUG–III subcategory for whichhe or she qualifies. As described above,the ‘‘3’’ level signifies a count of 1 or 2of the clinical variables used fordetermining the non-therapy ancillaryend split.

For example, a beneficiary who haspneumonia, an ADL sum score of 8,dehydration, a fever, and a surgicalwound that requires twice dailydressing changes, will classify to theSpecial Care category. Within theSpecial Care category, the ADL score of8 will classify this beneficiary into the‘‘SC’’ subcategory. The count of theitems that are used to make the finalclassification is 2, as the pneumonia andthe wound care with dressing changesare the two clinical variables that willaffect classification of this beneficiary tothe SC3 group.

TABLE 3.—MDS ITEMS ASSOCIATED WITH DIFFERENCES IN ANCILLARY CHARGES—REFINED VARIABLE LIST FOLLOWINGCLINICAL INPUT

MDS itemsdomains

Percent ofsample

Regressioncoefficient Standard error t–Statistic

Parenteral/IV with >76 percent total calories .................................................. 1 153.97 14.63 10.53Tracheostomy .................................................................................................. 1 109.87 16.57 6.63Suctioning ........................................................................................................ 2 106.76 10.23 10.43IV Medication ................................................................................................... 15 77.33 3.71 20.86Oxygen and either pneumonia or resp. inf. with fever, or pneumonia or

resp. inf., COPD, CHF, CAD with SOB ....................................................... 44 26.42 2.60 10.17Pneumonia ....................................................................................................... 10 25.64 4.06 6.32Tube feeding with >76 percent total calories .................................................. 6 23.21 4.33 5.36Respiratory Infection ........................................................................................ 7 18.81 4.87 3.87Application of dressing with/with-out topical medication and presence of ul-

cers or other skin lesions/ wounds .............................................................. 5 13.38 5.15 2.60Skin wound/ulcer care ..................................................................................... 25 7.01 2.77 2.53Stage 4 Pressure Ulcer ................................................................................... 4 6.87 3.09 2.22

Notes: N = 8,087 (Based on analysis of test sample only—20 percent of observations)Data Source: Medicare MDS and SNF Claims Data 1995–1997, excluding ME, OH, SD.

Using the selected MDS items, wecalculated a non-therapy ancillary indexscore for each MDS and classified themto the appropriate non-therapy ancillarylevel. We are including a more detaileddescription of the non-therapy ancillary

index methodology in TechnicalAppendix A.

An index model can differ withrespect to the RUG-III categories towhich the model is applied. Twooptions that we considered were toapply the index model only to the

Extensive Services category (includingbeneficiaries in rehabilitation who alsoqualify for Extensive Services) or toapply the index option to a broadergroup of RUG-III categories. Theresearch indicated very little differencein ancillary costs for beneficiaries in the



Impaired Cognition, Behavior andPhysical Function categories.Differences in ancillary costs wereidentified within the Rehabilitation,Clinically Complex, Special Care, andExtensive Services groups. For thisreason, we propose to apply the non-therapy ancillary index model to allresidents in the Clinically Complexcategory or above (where over 90percent of Medicare patients fall). Inaddition, we propose to apply a singlenon-therapy ancillary index factor toeach of the lower levels of the RUG-IIImodel (that is, Impaired Cognition,Behavior, and Physical Function).

Index models can also be applieddifferently across RUG-III levels. Themost straightforward method is to applya fixed dollar amount for each level ofthe index. In this case, the add-on for anon-therapy ancillary index score of 3would be the same regardless of thebeneficiary’s RUG-III group. Separateindices can also be calculated for eachlevel of the hierarchy. In this case, thedollar amount of the non-therapyancillary index level of 3 would bedifferent for beneficiaries in differentlevels of the RUG-III hierarchy, forexample, clinically complex, specialcare, rehabilitation, etc. Separate indicesare more appropriate when there issignificant inter-group variance. Usingthe research data base, we foundsignificant variation. In projecting ratesfor both the UWIM (Tables 5 and 6) andWIM 2 (Technical Appendix A, Tables6.1 and 6.2) models, we calculatedseparate index values for each of the 8proposed hierarchy levels. Thisapproach will be analyzed andevaluated using the national PPS/MDSdata base.

Finally, index models can also differwith respect to the number of non-therapy ancillary index groups that areused. Six groups were developed for theweighted index model. Four groupswere used for the unweighted model.The weighted index model performsslightly better than its unweightedcounterpart. However, it adds asignificant level of complexity both interms of the number of additional RUG-III variations and the addition of a newtype of MDS scoring methodology basedon cost instead of clinical criteria. Inaddition, as stated above, the weightedindex model break points are notrepresentative of national ancillarycosts.

On the other hand, the unweightedindex model relies on a count of MDSitems to differentiate among index

levels, an approach similar to that usedcurrently in RUG-III for classificationinto the Extensive Services category. Atthis phase of our analysis, we haveconcluded that the added complexity ofthe weighted model offsets any benefitsgained. Therefore, we are proposing theunweighted non-therapy ancillary indexmodel that will be applied to thecombined Rehabilitation/ExtensiveServices, Rehabilitation, ExtensiveServices, Special Care and ClinicallyComplex categories of the RUG-IIIhierarchy.

Adopting a new Extensive Serviceswith Rehabilitation category and addinga non-therapy ancillary indexcomponent will require modifications tothe naming conventions used to identifyeach RUG-III group. Based on theserecommendations, we have updated theRUG-III structure to incorporate theproposed refinements, as displayed inTable 4. These proposed RUG-III groupsare based upon the existing 3 digit RUG-III coding structure, but will designatethe non-therapy ancillary level as wellas the RUG-III category.

The first letter of the RUG-III codedefines the hierarchy level. First, a newhierarchy level is being added torecognize beneficiaries needing acombination of Extensive andRehabilitation Services. The codes usedto reflect the hierarchy level are alsobeing expanded to identify separatelyeach level of Rehabilitation (that is,Ultra High, Very High, High, Mediumand Low) either in combination withExtensive Services or separately.

RUG CODE—FIRST LETTER

Hierarchy Code

Extensive with Rehabilitation:Ultra High ........................................ JVery High ........................................ KHigh ................................................. LMedium ........................................... MLow .................................................. N

Rehabilitation:Ultra High ........................................ UVery High ........................................ VHigh ................................................. WMedium ........................................... XLow .................................................. Y

Extensive Services ............................. ESpecial Services ................................. SClinically Complex .............................. CImpaired Cognition ............................. IBehavior .............................................. BReduced Physical Function ................ P

The second letter of the proposedRUG–III coding structure is an alphacharacter that indicates the final group

assigned after the RUG–III end-splits(that is, ADLs, depression, restorativenursing) have been calculated.

The third digit of the proposed RUG–III coding structure will indicate thenon-therapy ancillary index level. In theunweighted non-therapy ancillarymodel, there are 4 levels determined bythe number of MDS non-therapyancillary qualifying items (See Table 4for the complete list of qualifiers.)

Indexlevel Number qualifiers met

5 ........... 6 or more.4 ........... 3–5.3 ........... 1–2.2 ........... 0.1 ........... Regular—for impaired cognition

behavior and physical functioncategories.

For example, under the current RUG–III model, a beneficiary whose MDSreflects an ADL sum score of 11, atracheostomy, suctioning, pneumonia,IV medications and receipt of 380minutes per week of physical therapy,would group into the RHB rehabilitationgroup.

In the refined RUG-III model with theunweighted non-therapy ancillaryindex, this beneficiary would group intothe LB4 group with the first digit, L,indicating a combination of ExtensiveServices and High Rehabilitation, thesecond digit, B, indicating the ADL levelof 11, and the third digit, 4, indicatingthe non-therapy ancillary level for abeneficiary with 4 qualifiers. See Table4 for a crosswalk from the current RUG-III groups to the new groups.

In Example 2, we will show theproposed classification for a beneficiarywho receives no rehabilitation services.This beneficiary is a quadriplegic, whohas an ADL sum score of 17, a stage 4pressure ulcer, treatment for thepressure ulcer, pneumonia, and dailyrespiratory therapy. This beneficiarycurrently classifies into the Special Carecategory, into the SSC group. In therefined classification system he or shewill group into the SA4 group, showingthat he or she is in the Special Carecategory, with an ADL sum score of 17–18, and 3–5 of the MDS non-therapyancillary qualifiers.

A naming convention has also beenestablished for the weighted model. Thefirst 2 digits are the same as for theunweighted model. The third digit, thenon-therapy ancillary indicator, usesalpha characters A through F, with ‘‘F’’as the lowest ancillary level.



TABLE 4.—RUG REFINEMENT CROSSWALK

CurrentRUG–IIIgroup

Description of category Non-therapyancillary split

RefinedRUG–IIIgroup

Rehab: At least 720 minutes/week in 1 disciplines, one discipline at least 5 days/week ............Extensive: At least one of the following: IV feeding in last 7 days, IV medications in last 14

days, suctioning in last 14 days, tracheostomy care in last 14 days, ventilator/respirator inlast 14 days

ADL Sum Score: 16–18

6 JA5

3–5 JA41–2 JA30 JA2

Rehabilitation: As above for ultra high rehabilitation .....................................................................Extensive: As aboveADL Sum Score: 9–15

6 JB5

3–5 JB41–2 JB30 JB2

Rehabilitation: As above for ultra high rehabilitation .....................................................................Extensive: As aboveADL Sum Score: 7–8

6 JC5

3–5 JC41–2 JC30 JC2

Rehabilitation: At least 500 minutes/week. At least one discipline 5 days/week ..........................Extensive: As aboveADL Sum Score: 16–18

6 KA5

3–5 KA41–2 KA30 KA2

Rehabilitation: As above for Very High Rehabilitation ...................................................................Extensive: As aboveADL Sum Score: 9–15

6 KB5

3–5 KB41–2 KB30 KB2

Rehabilitation: As above for Very High Rehabilitation ...................................................................Extensive: As aboveADL Sum Score: 7–8

6 KC5

3–5 KC41–2 KC30 KC2

Rehabilitation: High Rehabilitation: At least 325 minutes/week. One discipline at least 5 times/week.

Extensive: As above.ADL Sum Score: 13–18

6 LA5

3–5 LA41–2 LA30 LA2

Rehabilitation: As above for High Rehabilitation ...........................................................................Extensive: As aboveADL Sum Score: 8–12

6 LB5

3–5 LB41–2 LB30 LB2

Rehabilitation: As above for High Rehabilitation ...........................................................................Extensive: As aboveADL Sum Score: 7

6 LC5

3–5 LC41–2 LC30 LC2

Rehabilitation: Medium Rehabilitation: At least 150 minutes/week. Must have therapy on 5days, any discipline combination.

Extensive: As aboveADL Sum Score: 15–18

6 MA5

3–5 MA41–2 MA30 MA2

Rehabilitation: As above for Medium Rehabilitation ......................................................................Extensive: As aboveADL Sum Score: 8–14

6 MB5

3–5 MB41–2 MB30 MB2



TABLE 4.—RUG REFINEMENT CROSSWALK—Continued




Rehabilitation: As above for Medium Rehabilitation ......................................................................Extensive: As aboveADL Sum Score: 7

6 MC5

3–51–2

0

MC4MC3MC2

Rehabilitation: Low Rehabilitation: At least 45 minutes/week on at least 3 days/week. NursingRehabilitation therapy must be provided in two activities, for 15 minutes, 6 days/week.

Extensive: As aboveADL Sum Score: 14–18

6 NA5

3–51–2

0

NA4NA3NA2

Rehabilitation: As above for Low Rehabilitation ............................................................................Extensive: As above.ADL Sum Score: 7–13

6 NB5

3–51–2

0

NB4NB3NB2

ULTRA HIGHRUC.

Rehabilitation: At least 720 minutes/week in at least 2 therapy disciplines. At least one dis-cipline must be provided at least 5 days/week.


6 UA5

3–51–2

0

UA4UA3UA2

RUB ............. Rehabilitation: As above for Ultra High Rehabilitation ..................................................................ADL Sum Score: 9–15

6 UB5

3–51–2

0

UB4UB3UB2

RUA ............. Rehabilitation: As above for Ultra High Rehabilitation ..................................................................ADL Sum Score: 4–8

6 UC5

3–51–2

0

UC4UC3UC2

RVC ............. Rehabilitation: Very High Rehabilitation: At least 500 minutes/week. One discipline at least 5days/week.


6 VA5

3–51–2

0

VA4VA3VA2

RVB Rehabilitation: As above for Very High Rehabilitation ...................................................................ADL Sum Score: 9–15

6 VB5

3–51–2

0

VB4VB3VB2

...................... Rehabilitation: As above for Very High Rehabilitation ...................................................................ADL Sum Score: 4–8

6 VC5

3–51–2

0

VC4VC3VC2

RHC ............. Rehabilitation: High Rehabilitation: At least 325 minutes/week and at least one discipline 5days/week.


6 WA5

3–51–2

0

WA4WA3WA2

RHB ............. Rehabilitation: As above for High Rehabilitation ...........................................................................ADL Sum Score: 8–12

6 WB5

3–51–2

0

WB4WB3WB2

RHA ............. Rehabilitation: As above for High Rehabilitation ...........................................................................ADL Sum Score: 4–7

6 WC5

3–51–2

0

WC4WC3WC2







RMC ............ Rehabilitation: At least 150 minutes/week and at least 5 days/week in one therapy discipline ... 63–51–2

0

XA5XA4XA3XA2

RMB ............. Rehabilitation: As above for Medium Rehabilitation ......................................................................ADL Sum Score: 8–14

6 XB5

3–51–2

0

XB4XB3XB2

RMA ............. Rehabilitation: As above for Medium Rehabilitation ......................................................................ADL Sum Score: 4–7

6 XC5

3–5 XC41–2 XC3

0 XC2RLB .............. Rehabilitation: Low Rehabilitation: At least 45 minutes/week on at least 3 days/week. Nursing

rehabilitation therapy must be provided in two activities, for 15 minutes, 6 days/week.ADL Sum Score: 14–18

6 YA5

3–5 YA41–2 YA3

0 YA2RLA .............. Rehabilitation: As above for Low Rehabilitation ............................................................................

ADL Sum Score: 4–136 YB5

3–5 YB41–2 YB3

0 YB2SE3 .............. EXTENSIVE SERVICES—(if ADL <7, beneficiary classifies to Special Care) ............................. 6 EA5

IV feeding in the past 7 days (K5a).IV medications in the past 14 days (P1ac).Suctioning in the past 14 days (P1ai).Tracheostomy care in the last 14 days (P1aj).Ventilator/respirator in the last 14 days (P1al).ADL Sum Score: 7–18.

3–5 EA4Qualification for the EA, EB, EC levels is dependent on ADL score and additional clinical quali-

fiers identified in the Special Care and Clinically Complex criteria. No change from the cur-rent RUG–III system.

1–2 EA3

0 EA2SE2 .............. Extensive Services: As above .......................................................................................................

ADL Sum Score: 7–186 EB5

3–5 EB41–2 EB3

0 EB2SE1 .............. Extensive Services: As above .......................................................................................................

ADL Sum Score: 7–186 EC5

3–5 EC41–2 EC3

0 EC2SSC ............. SPECIAL CARE—(if ADL <7 beneficiary classifies to Clinically Complex) ................................... 6 SA5

Multiple Sclerosis (I1w) and an ADL score of 10 or higher ..........................................................Quadriplegia (I1z) and an ADL score of 10 or higher ...................................................................Cerebral Palsy (I1s) and an ADL score of 10 or higher ................................................................Respiratory therapy (P1bdA must = 7 days) ..................................................................................Ulcers, pressure or stasis; 2 or more of any stage (M1a,b,c,d) and treatment (M5a, b,c,d,e,g,h)Ulcers, pressure; any stage 3 or 4 (M2a) and treatment (M5a,b,c,d,e,g,h) ..................................Radiation therapy (P1ah) ...............................................................................................................Surgical, Wounds (M4g) and treatment (M5f,g,h) .........................................................................Open Lesions (M4c) and treatment (M5f,g,h) ...............................................................................Tube Fed (K5b) and Aphasia (I1r) and feeding accounts for at least 51 percent of daily cal-

ories (K6a = 3 or 4) OR at least 26 percent of daily calories and 501cc daily intake(K6b = 2,3,4 or 5).

Fever (J1h) with Dehydration (J1c), Pneumonia (Ie2),Vomiting (J1o) or Weight loss (K 3a) ......Fever (J1h) with Tube Feeding (K5b) and, as above, (K6a = 3 or 4) &/or (K6b = 2,3,4, or 5) ......ADL Sum Score: 17–18 ................................................................................................................. 3–5 SA4







1–20

SA3SA2

SSB ............. Special Care: As above .................................................................................................................ADL Sum Score: 15–16

6 SB5

3–51–2

0

SB4SB3SB2

SSA ............. Special Care: As above .................................................................................................................ADL Sum Score: 7–14

6 SC5

3–51–2

0

SC4SC3SC2

CC2 ............. CLINICALLY COMPLEX— 6 CA5Burns (M4b) ...................................................................................................................................Coma (B1) and Not awake (N1 = d) and completely ADL dependent (G1aa, G1ba, G1ha,

G1ia = 4 or 8).Septicemia (I2g) .............................................................................................................................Pneumonia (I2e) .............................................................................................................................Foot/Wounds (M6b,c) and treatment (M6f) ...................................................................................Internal Bleed (J1j) .........................................................................................................................Dialysis (P1ab) ...............................................................................................................................Tube Fed (K5b) and feeding accounts for: at least 51% of daily calories (K6a = 3 or 4) OR 26

percent of daily calories and 501cc daily intake (K6b = 2, 3, 4 or 5).Dehydration (J1c) ...........................................................................................................................Oxygen therapy (P1ag) ..................................................................................................................Transfusions (P1ak) .......................................................................................................................Hemiplegia (I1v) and an ADL score or 10 or higher .....................................................................Chemotherapy (P1aa) ....................................................................................................................No. Of Days in last 14 there were Physician Visits and order changes: ......................................visits > = 1 days and order changes > = 4 days; or visits > = 2 days and order changes on > = 2

days..........................................................................................................................................................Diabetes mellitus (I1a) and injections on 7 days (O3> = 7).ADL Sum Score: 17–18 .................................................................................................................Positive for Signs of Depression

3–5 CA4

CC1 ............. Clinically Complex: As above ........................................................................................................ADL Sum Score: 17–18 .................................................................................................................

6 CB5

No signs of depression .................................................................................................................. 3–51–2

0

CB4CB3CB2

CB2 .............. Clinically Complex: As above ........................................................................................................ 6 CC5ADL Sum Score: 12–16 .................................................................................................................Positive for Signs for Depression

3–5 CC4

1–20

CC3CC2

CB1 .............. Clinically Complex: As above ........................................................................................................ 6 CD5................. ADL Sum Score: 12–16 ................................................................................................................. 3–5 CD4

No signs of depression .................................................................................................................. 1–20

CD3CD2

CA2 .............. Clinically Complex: As above ........................................................................................................ 6 CE5ADL Sum Score: 4–11 ...................................................................................................................Positive for Signs of Depression

3–51–2

1

CE4CE3CE2

CA1 .............. Clinically Complex: As above ........................................................................................................ 6 CF5ADL Sum Score: 4–11 ...................................................................................................................No Signs of Depression

3–5 CF4

1–20

CF3CF2

IB2 ............... Impaired Cognition: Score on MDS2.0 Cognitive Performance Scale >= 3 .......................... IA1Receiving Nursing rehabilitation therapy in two activities, for 15 minutes, 6 days/week.ADL Sum Score: 6–10.

IB1 ............... Impaired Cognition: Score on MDS2.0 Cognitive Performance Scale >= 3 .................................ADL Sum Score: 6–1

.......................... IB1

IA2 ............... Impaired Cognition: Score on MDS2.0 Cognitive Performance Scale >= 3 ................................. .......................... IC1Receiving Nursing rehabilitation therapy in two activities, for 15 minutes, 6 days/week.ADL Sum Score: 4–5.







IA1 ............... Impaired Cognition: Score on MDS2.0 Cognitive Performance Scale >= 3 .................................ADL Sum Score: 4–5

.......................... ID1

BB2 .............. BEHAVIOR ONLY .......................................................................................................................... .......................... BA1Coded on MDS 2.0 items: 4+ days a week—wandering, physical or verbal abuse, inappro-

priate behavior or resists care; or hallucinations, or delusions checked.Receiving Nursing rehabilitation therapy in two activities, for 15 minutes, 6 days/week.ADL Sum Score: 6–10.

BB1 .............. Behavior: As above ........................................................................................................................No nursing rehabilitation receivedADL Sum Score: 6–10

BB1

BA2 .............. Behavior: As above ........................................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score: 4–5

BC1

BA1 .............. Behavior: As above ........................................................................................................................No nursing rehabilitation receivedADL Sum Score: 4–5

BD1

PE2 .............. Physical Function Impaired ............................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score:16–18

PA1

PE1 .............. Physical Function Impaired ............................................................................................................ADL Sum Score: 16–18

PB1

PD2 .............. Physical Function Impaired ............................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score:11–15

PC1

PD1 .............. Physical Function Impaired ............................................................................................................ADL Sum Score: 11–15

PD1

PC2 .............. Physical Function Impaired ............................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score: 9–10

PE1

PC1 .............. Physical Function Impaired ............................................................................................................ADL Sum Score: 9–10

PF1

PB2 .............. Physical Function Impaired ............................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score: 6–8

PG1

PB1 .............. Physical Function Impaired ............................................................................................................ADL Sum Score: 6–8

PH1

PA2 .............. Physical Function Impaired ............................................................................................................Nursing Rehabilitation received, at level described aboveADL Sum Score: 4–5

PI1

PA1 .............. Physical Function Impaired ............................................................................................................ADL Sum Score: 4–5

PJ1

BC1 .............. .................................................................................................................................................... ((1)) BC1

1Default Code

Additional Research Plans

As noted above, we performed theRUG–III refinement analyses on aresearch data base rather than on PPSMedicare claims and MDS data. Theresearch data base was appropriate andextremely useful in testing hypotheses,and identifying areas where refinementscould be introduced. However, researchdata always have limitations, and HCFAand contractor staff have identifiedseveral areas of concern. Fortunately,since actual PPS claims and MDS dataare now available, we are alreadyconducting additional analyses of theunweighted and weighted models toaddress these concerns and validate theresearch findings.

For this proposed rule, we havedeveloped Tables 5 and 6 to illustratethe application of the proposed

refinement to the RUG–III classificationsystem on the FY 2001 Federal per diemrates. In addition, for comparisonpurposes, we have developed rate tablesfor the WIM2 model that are shown inTechnical Appendix A (Tables 6.1 and6.2). However, in reviewing these tables,it is important to recognize thefollowing limitations:

The nursing index is a critical factorin accurately calibrating the system tolink payment to acuity levels. Thenursing indices shown in Tables 5through 6 assume that the distributionof the actual Medicare population is thesame as the distribution of the researchdata base. We are now reworking thesecalculations using national PPS data toensure accurate calibration of thesystem.

Using the actual PPS data base alsoadjusts for a second data limitation: the

extent to which MDS data reflects shortstay patients. The research data baseutilized MDS assessments from 1995through 1997, a period when MDSswere often not completed forbeneficiaries who were in a SNF for lessthan 14 days. By contrast, the PPS database includes short-stay beneficiaries,and we will take any special needs ofthis population into account by usingactual PPS data to validate the initialfindings.

In addition, the methodology used toadjust non-therapy ancillary charges tocost used the older, non-therapyancillary charges and facility cost-to-charge ratios. In developing the PPSdata base, we will use PPS claims dataand the latest available cost-to-chargeratios.

Using the smaller research data base,it was not always possible to obtain a



large number of observations in some ofthe RUG–III groups to fully determineancillary costs with the necessary levelof precision. For that small number ofRUG–III groups, the researchersimputed ancillary costs, and appliedthese imputed costs to the non-therapyancillary index used in the rate-settingprojections. Using the national PPS database will allow better differentiationbetween the non-therapy ancillary indexlevels for the new, combinedRehabilitation and Extensive Servicescategories, particularly in index levels 2and 3 of the unweighted model (and Band C of the weighted model.) (SeeTables 5 and 6 for the UWIM model andTechnical Appendix A Tables 6.1 and6.2 for the WIM2 model.)

Finally, we will continue the processof identifying possible negativeincentives associated with MDS itemsused in the non-therapy ancillary index.We will carefully evaluate each itembefore incorporating it into the finalindex. Then, we will develop methodsto monitor coding practices and toidentify changes in coding patterns foruse in medical review, quality assuranceand program integrity activities. We willissue clarifications, through ProgramMemoranda and other appropriatemeans, of MDS requirements needed tomaintain the integrity of the RUG–IIIsystem.

Using the national PPS data base, wewill recalculate the distribution of thebeneficiary population across RUG–IIIcategories, including the proposedcombined Rehabilitation and ExtensiveServices category. Then, we willperform the necessary analyses andsensitivity tests to compare the resultswith those derived from the researchdata base. We will reevaluate programoptions (for example, unweighted vs.weighted non-therapy ancillary index,etc.) based on the additional analyses,and modify the proposed refinements asneeded. We expect these final analysesto be available in late Spring 2000, andwe plan to incorporate them in the finalrule to be issued before August 1, 2000.

PPS Rate TablesWe are confident that the additional

analyses based on national data willconfirm the need for refinements in theRUG–III model by adding the newcombined Extensive and RehabilitationService groups and by creating a newnon-therapy ancillary index. However,it is very likely the values of some of themodel components (for example,average ancillary cost by RUG–III group,frequency distribution by RUG–IIIgroup, relative weights, etc.) will befurther refined through use of thenational data base. For this reason, it is

important to understand that the valuescontained in these tables will likelychange in the final rule.

While we are confident that theseresearch findings are based on soundmethodology, it is certainly possiblethat additional testing will identify newissues or support variations of themodels to those presented here. Weremain open to suggestions during thecomment period and will carefullyevaluate the validation analyses beforeproceeding to final rulemaking. Toillustrate the impact of these proposedchanges based on the best data currentlyavailable, we have developed rateTables 5 and 6 using the unweightedmodel. (For an additional discussion ofthe weighted model, including aschedule of rates, see TechnicalAppendix A.) These projections shouldnot be viewed as final nursing indices,non-therapy ancillary indices, orpayment rates.

Further, as noted above, we based thenon-therapy ancillary indices on themean adjusted derived cost (that is,charges adjusted by facility ancillarycost-to-charge ratios) of non-therapyancillary services. Mean costs werecalculated separately for each of theeight proposed levels of the RUG–IIIhierarchy. For the research data base,we used the cost-to-charge ratioapplicable to the service date of theclaim. For the follow up analyses usingactual PPS claims data, we are using themost recent available cost-to-chargeratio. We expect that using the newercost-to-charges ratios will enhance theaccuracy of the calculations. However,due to the lag time between SNF PPSclaims submission and cost reportprocessing, it is impossible to match theclaims service dates perfectly with thecost report period used for the cost-to-charge ratios. For the SNF PPS database, we are proposing to useapproximately 9 months of claims datastarting from January 1, 1999, the datealmost all providers became subject toPPS. The cost reports for calendar year1999 are not due until April 2000.

Finally, the research findings in thisproposed rule include the use of‘‘imputed’’ data in situations where thecell size (for example, number ofrecords meeting the criteria for aspecific RUG–III group, etc.) was toosmall for accurate measurement. Whenusing the national data base, we expectthat the relevant data cells will beadequately populated and that allanalyses used in developing the finalrule will be based on actual rather thanimputed data.

These tables reflect two adjustmentsin particular. First, our nursing andtherapy staff time indices (combined

1995 and 1997 staff time data) currentlyused to establish PPS rates have beenadjusted to reflect the new combinedExtensive Services with Rehabilitationcategories. Second, we have adjustedthe nursing case mix component of therate to remove the non-therapy ancillarycomponent that is part of the currentnursing index used in PPS rate-setting.We will need to adjust one or both ofthese components based on theadditional analyses.

We integrated these proposedrefinements into the rate-settingmethodology, and we list the estimatedper diem Federal rates for 178 separateRUG–III classification groups in Tables5 and 6. We list the case-mix adjustedpayment rates separately for urban andrural SNFs (178 each), with thecorresponding case-mix index values.These tables list the rates in total and bycomponent. The application of the wageindex, described later in this section, isthe final adjustment applied to theprojected Federal rates in these tables.

In accordance with section 101 of theBBRA, we will make a four percentupward adjustment to the adjusted perdiem Federal rate for FY 2001. Thisestimated adjustment is shown in Table9.

Finally, these projected rates do notreflect the BBRA requirement (section103) to reduce the Part A SNF paymentrates to account for those services thatare newly excluded from consolidatedbilling and, thus, will be separatelybillable to Part B by the supplier. Asmentioned in section II.A.2. above,because of the complexity of the processand the amount of time needed toimplement this requirement, we areunable at present to adjust the proposedrates to reflect this. However, we willmake these adjustments prospectively inthe final rule establishing payment ratesfor FY 2001, using the methodologydescribed below.

In order to compute the level of thisadjustment, we propose to determinethe per diem amount of allowed chargesassociated with the specific HCPCScodes identified in the statute (and laterin this rule) using the same 1995 dataon Part B services used in establishingthe Federal rates. These data aredescribed in detail in section II.A.2.b ofthe May 12, 1998 interim final rule (63FR 26251) and final rule (64 FR 41644)associated with the implementation ofthe SNF PPS. The per diem amount willbe subtracted from the non-therapyancillary component of the Federal ratesshown in Tables 5 and 6 of this rule. Weexpect this adjustment to be minimal.



Summary of Proposed RUG–IIIRefinements

Based on the research described here,we are proposing the addition of newRUG–III groups to recognize the needsof Medicare beneficiaries with bothheavy medical and rehabilitation needsand the development of an unweightedindex model that would account moreprecisely for the variation in non-therapy ancillary services. Since theresearch shows substantial ancillarycost variation in the Rehabilitation andExtensive Services, Rehabilitation,Extensive Services, Special Care, andClinically Complex categories, we haveproposed four ancillary index levels tocapture variation in ancillary costsaccurately. Since beneficiaries in theImpaired Cognition, Behavior, and

Physical Function categories exhibited amuch smaller ancillary cost variation,we calculated a single ancillary add-onamount. The ancillary add-on amountswere calculated separately for each ofthe eight proposed RUG–III categories.

The refinements will achieveimportant improvements in the PPSmodel, and allow for more accuratepayment rates. In addition, after furtheranalysis and review of publiccomments, we may adjust theseproposed refinements further to reflectactual PPS experience.

Collection of Medication DataIn the interim final rule published in

the Federal Register on May 12, 1998,we stated that we would requirefacilities to complete and include MDSSection U with their Medicare MDS

submissions beginning October 1, 1999.Subsequently, in the final rulepublished in the Federal Register onJuly 30, 1999, we announced a delay ofthat requirement and stated ourintention to require completion ofSection U beginning October 1, 2000.However, we are currently unable toimplement the collection of medicationdata on the MDS beginning October 1,2000. Accordingly, we will not requirecompletion and submission of Section Uof the MDS beginning October 1, 2000,as we had planned. We are currentlyexamining issues related to theimplementation of this requirement andwe plan to address this matter when weimplement the SNF PPS paymentupdate for FY 2001.

BILLING CODE 4120–03–U


















