Medicare Program; Prospective Payment System and Consolidated Billing for Skilled Nursing Facilities

SUMMARY: This final rule responds to comments submitted by the public

on our May 12, 1998 interim final rule, that implemented provisions in

section 4432 of the Balanced Budget Act of 1997 regarding Medicare

payment for skilled nursing facility services. This legislation

established a prospective payment system, a consolidated billing

provision, and a number of related changes.

EFFECTIVE DATE: These regulations are effective on September 28, 1999.

FOR FURTHER INFORMATION CONTACT:

Dana Burley, (410) 786-4547 (for information related to the case-mix

classification methodology).

John Davis, (410) 786-0008 (for information related to the Federal

rates).

Jackie Gordon, (410) 786-4517 (for information related to consolidated

billing).

Steve Raitzyk, (410) 786-4599 (for information related to the facility-

specific transition payment rates).

Bill Ullman, (410) 786-5667 (for information related to coverage and

level of care determinations).

Laurence Wilson, (410) 786-4603 (for general information).

SUPPLEMENTARY INFORMATION: To assist readers in referencing sections

contained in this preamble, we are providing the following table of

contents.

Table of Contents

I. Background

A. Payment Provisions--Federal Rate

B. Payment Provisions--Transition Period

C. Payment Provisions--Facility-Specific Rate

D. Consolidated Billing for Skilled Nursing Facilities

II. Provisions of the Interim Final Rule

III. Analysis of and Responses to Public Comments

A. Federal Rates--Outliers/Non-therapy ancillaries (NTAs)

B. Federal Rate Calculation

C. Federal Rates--Part B Add-on

D. Facility-specific Rates-Transition

E. Minimum Data Set (MDS) Assessments

1. Billing Issues

2. Corrections

3. Other Medicare Required Assessment (OMRA)

F. Certification and Recertification

G. MDS Scheduling Requirements

1. Grace Days

2. Completion and Locking

3. Discharge and Leave of Absence

H. Other Medicare MDS Requirements

I. Medical Review

J. Rehabilitation Therapy Services and PPS

K. RUG-III Groups

L. Nurse Staffing and the Staff Time Measurement Studies

M. SNF Coverage and Level of Care Determinations

N. SNF Consolidated Billing

O. Scope of Extended Care Benefits

P. Impact Analysis

IV. Provisions of the Final Regulations

V. Collection of Information Requirements

VI. Impact Analysis

A. Background

B. Impact of this Final Rule

C. Rural Hospital Impact Statement

D. Unfunded Mandates

In addition, because of the many terms to which we refer by acronym

in this rule, we are listing these acronyms and their corresponding

terms in alphabetical order below:

ADLs Activities of daily living

ASC Ambulatory Surgical Center

BBA Balanced Budget Act of 1997

CAH Critical access hospital

CBO Congressional Budget Office

CFR Code of Federal Regulations

CORF Comprehensive Outpatient Rehabilitation Facility

CPI Consumer Price Index

CPI-U Consumer Price Index for All Urban Consumers

CPT [Physicians'] Current Procedural Terminology

DME Durable medical equipment

ESRD End stage renal disease

FI Fiscal intermediary

GAO General Accounting Office

HCFA Health Care Financing Administration

HCPCS HCFA Common Procedure Coding System

HIPPS Health Insurance Prospective Payment System

ICD-9-CM International Classification of Diseases, Ninth Edition,

Clinical Modification

MDS Minimum Data Set

MEDPAR Medicare Provider Analysis and Review File

MGCRB Medicare Geographic Classification Review Board

MIM-3 Medicare Intermediary Manual, Part 3

MRI Magnetic Resonance Imaging

MSA Metropolitan Statistical Area

NHCMQD [Multistate] Nursing Home Case-mix and Quality Demonstration

OBRA 87 Omnibus Budget Reconciliation Act of 1987

OIG Office of the Inspector General

OMRA Other Medicare Required Assessment

PM Program Memorandum

PPS Prospective payment system

PRM Provider Reimbursement Manual

PRO Peer Review Organization

RAI Resident Assessment Instrument

RAPs Resident Assessment Protocols

RUG-III Resource Utilization Groups, version III

SNF Skilled nursing facility

SOM State Operations Manual

STM Staff time measure

I. Background

Section 4432 of the Balanced Budget Act of 1997 (BBA) (Public Law

105-33) mandated the implementation of a per diem prospective payment

system (PPS) for skilled nursing facilities (SNFs), covering all costs

(routine, ancillary, and capital) of covered SNF services furnished to

beneficiaries under Part A of the Medicare program, effective for cost

reporting periods beginning on or after July 1, 1998. Major elements of

the system include:

Rates: Per diem Federal rates are established for urban

and rural areas using allowable costs from fiscal year (FY) 1995 cost

reports. These rates also include an estimate of the cost of services

that, before July 1, 1998, had been paid under Part B but furnished to

SNF residents during a Part A covered stay. Rates are case-mix adjusted

using a resident classification system (Resource Utilization Groups,

version III (RUG-III)) based on resident assessments (using the Minimum

Data Set (MDS) 2.0). In addition, the Federal rates are adjusted by a

wage index to account for geographic variation in wages. Finally, the

rates will be adjusted annually using an SNF market basket index.

Transition: The SNF PPS includes a 3-year transition that

blends a facility-specific payment rate with the Federal case-mix

adjusted rate. The blend that is used changes each cost reporting

period after a facility migrates to the new system. For most

facilities, the facility-specific rate is based on allowable costs from

FY 1995.

Coverage: The PPS legislation did not change Medicare's

fundamental statutory requirements for SNF coverage. However, because

RUG-III classification is based, in part, on the resident's need for

skilled nursing care and therapy, we have attempted where possible to

adapt the existing claims review procedures to coordinate them with the

outputs of resident assessment and RUG-III classifying activities, as

discussed later in this preamble.

[[Page 41645]]

Consolidated Billing: The statute includes a billing

provision that requires an SNF to submit consolidated Medicare bills

for its residents for virtually all services that are covered under

either Part A or Part B. The statute excludes a small list of services

(primarily those of physicians and certain other types of

practitioners). A related statutory provision requires SNFs to use HCFA

Common Procedure Coding System (HCPCS) coding on all Part B bills, and

specifies that they are to be paid an amount determined in accordance

with the otherwise applicable Part B fee schedule for the particular

item or service.

Effective Date: The SNF PPS is effective for cost

reporting periods beginning on or after July 1, 1998. The law provides

that the consolidated billing and coding requirements are effective for

services and items furnished on or after July 1, 1998.

An interim final rule implementing the SNF PPS was published in the

Federal Register on May 12, 1998 (63 FR 26252), and the comment period

was initially scheduled to close on July 13, 1998. A follow-up notice

(63 FR 37498, July 13, 1998) extended the public comment period for an

additional 60 days, and a second notice (63 FR 65561, November 27,

1998) reopened the comment period for another 30 days. In addition, a

correction notice (63 FR 53301, October 5, 1998) made a number of minor

technical and editorial corrections to the interim final rule. We have

also issued several Program Memorandums (PMs) on claims processing and

billing under the SNF PPS that are available on the SNF PPS home page

at the HCFA website on the Internet, at the following location:

www.hcfa.gov/medicare/snfpps.htm>.

As described in the interim final rule, the BBA requires

implementation of a Medicare SNF PPS for cost reporting periods

beginning on or after July 1, 1998. Under the PPS, SNFs are no longer

paid under the previous, reasonable cost-based system, but rather

through per diem prospective case-mix adjusted payment rates applicable

to all covered SNF services. These payment rates cover all the costs of

furnishing covered skilled nursing services (that is, routine,

ancillary, and capital-related costs) other than costs associated with

approved educational activities. Covered SNF services include

posthospital SNF services for which benefits are provided under Part A

and all items and services that, prior to July 1, 1998, had been paid

under Part B (other than physician and certain other services

specifically excluded under the BBA), but furnished to SNF residents

during a Part A covered stay.

A. Payment Provisions--Federal Rate

The statute sets forth a fairly prescriptive methodology for

calculating the amount of payments under the SNF PPS. The PPS uses per

diem Federal payment rates based on mean SNF costs in a base year

updated for inflation to the first effective period of the system. We

developed the Federal payment rates using allowable costs from

hospital-based and freestanding SNF cost reports during the base year

(that is, for reporting periods that began in FY 1995). The data used

in developing the Federal rates also incorporate an estimate of the

amounts that were paid separately under Part B for covered SNF services

furnished during the base year to individuals who were residents of a

facility and receiving Part A covered services.

In developing the rates, we update costs to the first effective

year of the PPS (15-month period beginning July 1, 1998) using an SNF

market basket index, and standardize for facility differences in case-

mix and for geographic variations in wages. Providers that received

``new provider'' exemptions from the routine cost limits are excluded

from the data base used to compute the Federal payment rates. In

addition, costs related to payments for exceptions to the routine cost

limits are excluded from the data base used to compute the Federal

payment rates. In accordance with the formula prescribed in the BBA, we

set the Federal rates at a level equal to a weighted mean of

freestanding costs plus 50 percent of the difference between the

freestanding mean and a weighted mean of all SNF costs (hospital-based

and freestanding) combined. We compute and apply separately payment

rates for facilities located in urban and rural areas.

The Federal rate also incorporates adjustments to account for

facility case-mix using a resident classification system that accounts

for the relative resource utilization of different patient types. This

classification system, RUG-III, uses resident assessment data (from the

MDS) completed by SNFs to assign residents into one of 44 groups. SNFs

complete these assessments according to an assessment schedule

specifically designed for Medicare payment (that is, on the 5th, 14th,

30th, 60th, and 90th days after admission to the SNF).

For Medicare billing purposes, there are specific codes associated

with each of the 44 RUG-III groups, and each assessment applies to

specific days within a resident's SNF stay. SNFs that fail to perform

assessments timely are paid a default payment for the days of a

patient's care for which they are not in compliance with this schedule.

In addition, we adjust the portion of the Federal rate attributable to

wage-related costs by a wage index.

For the initial period of the PPS, beginning on July 1, 1998, and

ending on September 30, 1999, the payment rates were contained in the

interim final rule. For each succeeding fiscal year, we will publish

the rates in the Federal Register before August 1 of the year preceding

the affected Federal fiscal year. Pursuant to section 1888(e)(4)(E)(ii)

of the Social Security Act (the Act), for FY 2000 through 2002, we will

increase the rates each year by a factor equal to the SNF market basket

change minus one percentage point. For subsequent fiscal years, we will

increase the rates by the applicable SNF market basket change.

B. Payment Provisions--Transition Period

Beginning with a provider's first cost reporting period beginning

on or after July 1, 1998, there is a transition period covering three

cost reporting periods. During this transition phase, SNFs receive a

payment rate comprising a blend between the Federal rate and a

facility-specific rate based on each facility's FY 1995 cost report.

Under section 1888(e)(2)(E)(ii) of the Act, SNFs that received their

first payment from Medicare on or after October 1, 1995, receive

payment according to the Federal rates only.

For SNFs subject to the transition, the composition of the blended

rate varies depending on the year of the transition. For the first cost

reporting period beginning on or after July 1, 1998, we make payment

based on 75 percent of the facility-specific rate and 25 percent of the

Federal rate. In the next cost reporting period, the rate consists of

50 percent of the facility-specific rate and 50 percent of the Federal

rate. In the following cost reporting period, the rate consists of 25

percent of the facility-specific rate and 75 percent of the Federal

rate. For all subsequent cost reporting periods, we base payment

entirely on the Federal rate.

C. Payment Provisions--Facility-Specific Rate

For most facilities, we compute the facility-specific payment rate

used for the transition using the allowable costs of SNF services for

cost reporting periods that began in FY 1995 (cost reporting periods

beginning on or after October 1, 1994, and before October 1, 1995).

Included in the facility-specific per diem rate for most facilities is

an

[[Page 41646]]

estimate of the amount that was paid separately under Part B for

covered SNF services furnished during the base year to individuals who

were residents of the facility and receiving Part A covered services.

Under section 1888(e)(3)(A) of the Act, the facility-specific rate (in

contrast to the Federal rates) includes amounts paid to SNFs for

exceptions to the routine cost limits. In addition, we also take into

account ``new provider'' exemptions from the routine cost limits, but

only to the extent that routine costs do not exceed 150 percent of the

routine cost limit.

We update the facility-specific rate for each cost reporting period

after FY 1995 to the first cost reporting period beginning on or after

July 1, 1998 (the initial period of the PPS) by a factor equal to the

SNF market basket percentage increase minus 1 percentage point. For the

FYs 1998 and 1999, we update this rate by a factor equal to the SNF

market basket increase minus 1 percentage point, and, for each

subsequent year, we update it by the applicable SNF market basket

increase.

D. Consolidated Billing for Skilled Nursing Facilities

Section 4432(b) of the BBA sets forth a consolidated billing

requirement applicable to all SNFs providing Medicare services. SNF

consolidated billing is a comprehensive billing requirement (similar to

the one that has been in effect for inpatient hospital services for

well over a decade), under which the SNF itself is responsible for

billing Medicare for virtually all of the services that its residents

receive. As with hospital bundling, the SNF consolidated billing

requirement does not apply to the services of physicians and certain

other types of medical practitioners. In a related provision, section

4432(b)(3) of the BBA requires the use of fee schedules and uniform

coding specified by the Secretary of Health and Human Services (the

Secretary) for SNF Part B bills. The law provides that these

requirements are effective for services furnished on or after July 1,

1998.

II. Provisions of the Interim Final Rule

In the interim final rule that was published on May 12, 1998, we

made a number of revisions in the regulations in order to implement

both the PPS and the SNF consolidated billing provision and its

conforming statutory changes:

With regard to payment, we revised the regulations in 42

CFR part 413, subpart A (that deal with Medicare payment to providers

of services) to reflect the replacement of the existing reasonable cost

reimbursement methodology for SNFs by the new SNF PPS.

We revised the regulations to provide that for SNF

residents who are in a covered Part A stay, Medicare makes payment

under the PPS described in new subpart J of part 413, effective with

cost reporting periods beginning on or after July 1, 1998.

For SNF residents who are not in a covered Part A stay, we

revised the regulations to provide that Medicare makes payment on the

basis of the otherwise applicable Part B fee schedule amounts,

effective for services furnished on or after July 1, 1998.

We made a conforming change in subpart B of part 483

(requirements for long term care facilities) to indicate that the

frequency of resident assessments is subject to the timeframes

prescribed under the SNF PPS in the new subpart J of part 413.

We made a number of revisions to implement the

consolidated billing provision, under which the SNF itself has the

Medicare billing responsibility for virtually all of the services that

its residents receive.

We revised the regulations in part 410 (payment of

benefits under Part B) to provide that Part B makes payment for these

services to the SNF rather than to the beneficiary. We also made

conforming changes with regard to Part B coverage of certain individual

medical and other health services.

We revised part 411 (exclusions from coverage) to exclude

from coverage any service furnished to an SNF resident (other than

certain specified service categories) when billed to Medicare by an

entity other than the SNF itself, and we added a definition of an SNF

``resident'' for purposes of this provision.

We revised the regulations in subpart B of part 489

(Medicare provider agreements) to add compliance with the consolidated

billing provision to the specific terms of an SNF's provider agreement.

We revised subpart C of part 424 (claims for payment) to

require the inclusion of an SNF's Medicare provider number on claims

for physician services furnished to an SNF resident, and the inclusion

of HCPCS coding on an SNF's Part B claims.

We made a number of conforming changes in subparts C, D,

and F of part 409 of the regulations which describe, respectively, the

scope of covered SNF benefits under Part A, the criteria for

determining a covered SNF level of care, and benefit period

determinations.

As noted previously, the PPS legislation did not change the basic

statutory definition of an SNF level of care. However, because RUG-III

classification is based, in part, on the resident's need for skilled

nursing care and therapy, our revisions in the level of care criteria

reflected an attempt where possible to coordinate claims review

procedures with the outputs of resident assessment and RUG-III

classifying activities. For example, we believe that an initial 5-day

assessment, properly completed, that places the resident in one of the

upper 26 RUG-III classifications provides the basis for us to assume

that the resident needed a covered level of SNF care upon admission and

at least up until the assessment reference date of the initial

Medicare-required 5-day assessment. We will, however, continue to make

individual review determinations for claims of individuals who classify

in the lower 18 RUG-III categories.

III. Analysis of and Responses to Public Comments

We received almost 500 comments on the SNF PPS interim final rule

published on May 12, 1998 (63 FR 26302). Comments were submitted by

nursing homes and other providers, suppliers and practitioners (both

individually, and through their respective trade associations), State

agencies, nursing home resident advocacy groups, elected officials,

health care consulting firms, and private citizens.

The comments basically fell into three broad areas. The first

involved the payment rates, including treatment of ``outlier''

situations and non-therapy ancillaries, calculation of the Federal

rates themselves and of the Part B add-on, and the transition from

facility-specific rates to the Federal rates. The second area concerned

the clinical aspects of the SNF PPS, including MDS assessment and

scheduling requirements, certification and recertification procedures,

medical review criteria, treatment of rehabilitation therapy under the

RUG-III classification system, nurse staffing and staff time

measurement studies, and coverage and level of care determinations. The

third broad area involved the consolidated billing requirement and the

scope of the extended care benefit.

As noted in the interim final rule, because of the large number of

items of correspondence we normally receive on Federal Register

documents published for comment, we are unable to acknowledge or

respond to them individually. In particular, a number of commenters on

the interim final rule raised extremely technical and detailed

questions regarding the MDS and the

[[Page 41647]]

billing process. These questions are of a nature that would more

appropriately be addressed through manual instructions and other

issuances than in these regulations. In this final rule, we are

addressing the general concerns raised by the commenters. A summary of

the major issues and our responses follows:

A. Federal Rates--Outliers/Non-therapy Ancillaries (NTAs)

Comment: We received a number of comments expressing concern over

the ability of the PPS to provide adequate payment for certain outlier

or extraordinary cases. Several of the comments noted specific examples

of these cases, such as HIV-infected patients with significant drug

therapy needs, patients receiving intravenous (IV) drug therapy for

antibiotic-resistant infections, ventilator-dependent patients, or

simply patients with generally high costs. A number of commenters

recommended the adoption of an outlier payment process or exceptions

process to provide higher payments for these cases.

Other comments suggested use of a later base year (for example, FY

1997) or add-on to the rates in order to recognize changes made by

facilities after 1995, the year on which the rates are based. These

commenters argued that many facilities increased the scope of services

provided to beneficiaries and served a higher acuity resident

population after 1995 and, therefore, the costs associated with

providing this higher level of care were not reflected in the

calculation of the Federal rate.

Response: Section 1888(e)(4) of the Act provides specific

requirements related to the formula and cost data to be used in

computing the Federal rates. The statute provides that ``the amount of

the payment for all costs * * * of covered skilled nursing facility

services'' during the transition period is ``equal to'' a prescribed

blended payment, and after the transition period is ``equal to'' the

applicable adjusted Federal per diem rate. The statute does not provide

for additional payments over and above these prescribed amounts. While

the Act includes specific statutory authority for the application of

outlier policies in relation to the acute care hospital PPS (section

1886(d) through (f) of the Act), home health PPS (section 1895 of the

Act), and inpatient rehabilitation PPS (section 1886(j) of the Act), it

does not provide such explicit authority with regard to the SNF PPS.

However, we are concerned about this matter and are pursuing the basic

issue of the accuracy of payments through an examination of the case-

mix classification system.

In addition, the statute mandates use of the FY 1995 cost data in

the development of the payment rates. It should be noted that when the

rates were computed, the FY 1995 data were the latest available to

compute the rates. We believe the Congress took this into consideration

when developing the statutory language related to the computation of

the Federal rates as well as the specific impact of using the 1995 data

on the accumulation of Medicare savings, a key goal of the BBA.

We also note that while the Congress provided for Medicare

budgetary savings through the SNF PPS (which had an obvious downward

effect on the rates), there are numerous reports by the U.S. General

Accounting Office (GAO) and Office of the Inspector General (OIG)

suggesting Medicare payment for SNF ancillary services under cost

reimbursement was inappropriately inflated in the past. If correct,

this would mitigate the impact of the budgetary savings. The OIG

includes an expanded discussion of this concept in a 1998 report on the

SNF PPS titled ``Review of the Health Care Financing Administration's

Development of a Prospective Payment System for Skilled Nursing

Facilities'' (Number A-14-98-00350).

We understand the concerns expressed in the comments related to

this issue. As discussed in the impact analysis accompanying the

interim final rule, the SNF PPS will have a varying impact on

providers. Because ``prices'' are based on averages, SNFs should expect

that certain patients cost more than payments and others less. The

extent to which certain facilities can provide quality care, while

incorporating efficiencies in their purchasing of services and

operations, will affect how well they manage under this payment system,

which uses mean-based prices rather than reasonable costs. Financial

performance should, therefore, be determined by looking across each

facility's Medicare population, not on a patient specific comparison of

costs and the payment rate under which the rate would become

essentially a limit.

We will focus our efforts on ensuring that these prices are as

accurate as possible with respect to the resources used by Medicare

beneficiaries. The SNF PPS, through case-mix classification and

adjustment, currently reflects a full range of SNF patient types with

varying characteristics and degrees of resource intensity. Through

research and refinements to the PPS, we will try to ensure that the PPS

not only continues to account for a high level of resource intensity,

but improves in terms of its sensitivity to less common conditions or

patient types. This aspect of our plan is discussed later in the

context of the comments on payment for certain ancillary services.

Comment: There were a number of comments expressing concern with

the adequacy of the PPS rates to cover the costs of ancillary services

other than occupational, physical, and speech therapy (non-therapy

ancillaries), including such things as drugs, laboratory services,

respiratory therapy, and medical supplies. Prescription drugs or

medication therapy were frequently noted areas of concern due to their

potentially high cost for particular residents. Some commenters

suggested that the RUG-III case-mix classification methodology does not

adequately provide for payments that account for the variation in, or

the real costs of, these services provided to their residents. A number

of commenters stated their belief that the payment rates do not

generally reflect the costs of certain of these services (for example,

drugs or respiratory therapy).

Recommendations from commenters included removing all or some of

these services from the PPS rates and continuing to pay for them on a

cost basis, and making changes to the case-mix system and indices to

account for these services more accurately.

Response: We are aware of the challenges certain providers have

faced as they transition from a payment system based on reasonable

costs to one that uses mean-based prices such as the SNF PPS. In fact,

many of the same concerns raised in the comments to the interim final

rule were voiced by hospitals when we implemented the hospital PPS

system in the early 1980s. However, we believe this is an important

issue that calls for a broader discussion of the PPS itself, and

requires the clarification of certain technical issues related to the

PPS and to the statute.

Section 1888(e)(1) of the Act requires that the PPS provide payment

for ``all costs'' (including routine, ancillary, and capital related

costs) of covered SNF services. Consistent with the statute, the PPS

rates are based on 1995 allowable costs calculated from Medicare Part A

cost report data and applicable Part B allowable charges. Thus, a

facility's historical costs (from FY 1995) of drugs, laboratory

services, respiratory therapy, and other non-therapy ancillary services

were captured in these cost reports and reflected in both the Federal

and facility-specific transition rates.

In addition, many of these non-therapy ancillary services (for

example, respiratory therapy, IV medications, and

[[Page 41648]]

IV feedings) are captured both directly and indirectly in the case-mix

methodology and result in higher payments for SNFs. The issue of

whether the nursing case-mix index adequately reflects the relative

costs of non-therapy ancillary services was one that was studied in the

development of the interim final rule and the associated payment rates.

As indicated in the preamble of that rule, using MDS assessments to

classify patients into RUG-III groups, we compared the relative charges

for non-therapy ancillaries to the nursing case-mix indices for each

RUG-III group. We found that the pattern of the two relative amounts

was similar across the RUG-III groups. That is, RUG-III groups with

high nursing weights also tended to have relatively high charges per

stay for non-therapy ancillaries.

Based on this comparison, we concluded that it was reasonable to

include non-therapy ancillary costs in the nursing component of the

rate. Accordingly, the idea that the PPS rates do not reflect the cost

of respiratory therapy, drugs, and other non-therapy ancillaries is

simply not accurate. Whether the accuracy of the rates can be enhanced

in this regard is a subject for research and development that we

discuss below.

The recommendation to remove or ``carve out'' these services as a

class from the PPS rates and continue to pay for them on a cost basis

raises some fundamental concerns related to both the statutory and

conceptual framework of the PPS. As discussed above, section

1888(e)(A)(1) of the Act requires that the PPS provide payment for

``all costs'' (including routine, ancillary, and capital related costs)

of covered SNF services. The conference report associated with section

4432 of the BBA explicitly states that under the SNF PPS, ``services

and supplies provided to residents will be included in pre-determined

per diem payment rates.''

Beyond the threshold issue of statutory language, the issue of

whether specific services should be identified and paid separately

appears to conflict with certain fundamental concepts embodied in a

PPS. Carried to its logical conclusion, this approach is antithetical

to the very concept of the SNF PPS itself, which is based on bundling

services for similar patients and paying an average, prospectively

determined amount for all services included in the bundle. The PPS rate

already recognizes differences between nursing, rehabilitation therapy,

and ancillary services, as well as non-case-mix components.

It is important to consider the budgetary impact of the commenters'

proposal to remove certain services from the PPS rates and to continue

paying for them on a cost basis. The budgetary impact would be

significant and would reduce the savings to Medicare associated with

the SNF PPS provisions of the BBA. Implementing the provision in a way

that would have a budget neutral impact on savings (for example, a

downward adjustment to the Federal rates) would penalize providers that

have made changes to their operations in order to provide services more

efficiently, and would benefit those that have not. Therefore, we

believe that further disaggregation of the payment rate would not be

consistent with the objectives of prospective payment from a

conceptual, statutory, or budgetary perspective.

Finally, we agree with the commenters' recommendation that we

explore the potential for refinements to the PPS and, more

specifically, the case-mix classification system (RUG-III) to ensure

that it continues to account more accurately for the services provided

to SNF residents. We consider the continuing adequacy of the PPS rates,

and the case-mix methodology in particular, to be a high priority. We

believe very strongly that the case-mix methodology should be

periodically evaluated to determine the appropriateness of the RUG-III

groups in relation to changes in patient care practices and the

Medicare population.

In addition, the conference report language associated with section

4432 of the BBA specifically recommended examining payment for

medication therapy in the context of the SNF PPS. Accordingly, we are

funding substantial research to examine the potential for refinements

to the case-mix methodology, including an examination of medication

therapy, medically complex patients, and other non-therapy ancillary

services.

We are currently funding two research contracts to determine the

potential for refinements to the RUG-III model. The first contract was

awarded in FY 1997 and provides preliminary analysis and alternatives

for refinements using a limited database. The next phase of the

research focuses on fully developing these options using more extensive

data. Completion of the research is targeted for January 1, 2000.

Potential refinements to the case-mix model may include the division of

the current 44 groups or the addition of new ones based on items

currently on the MDS 2.0 (for example, new extensive care groups

combining both medical ancillaries and rehabilitation).

In addition, a new payment index (or set of relative weights) based

on ancillary charges, rather than the current staff-time based indices,

is being explored for the non-therapy ancillary component of the PPS

rates. Any refinements to the RUG-III model and case-mix indices that

result from this research would have a distributional effect on

payments resulting in a new set of payment weights across the various

groups. If the research supports refinements, we anticipate their

implementation in conjunction with the October 1, 2000, update to the

PPS rates. This time line is dictated by the complexity of the research

and by operational and regulatory requirements, including publication

of a proposed rule.

It should be noted that the BBA provisions establishing the SNF PPS

provided for over $9 billion in savings to Medicare (in fee for

service) as a result of the statutory formula used for developing the

rates. Accordingly, an SNF's current costs may well exceed the PPS

rates if the SNF does not revise the historical purchasing and charging

practices that it followed under the preexisting cost-based payment

system.

B. Federal Rate Calculation

Section 4432(a) of the BBA amended section 1888 of the Act by

adding a new paragraph (e) that provides for the establishment of per

diem Federal payment rates under the SNF PPS. These rates encompass all

costs of furnishing covered skilled nursing services (that is, routine,

ancillary, and capital-related costs), other than costs associated with

approved educational activities. In the interim final rule, we

established a new subpart J in the regulations at 42 CFR part 413, that

describes this new payment methodology. In this section of the

preamble, we are providing responses to comments on a number of

important issues related to the Federal rates. These include payment

for non-rehabilitation ancillary services, outlier cases, and a variety

of issues related to the data and design of the Federal payment rates.

In addition, we are providing for a minor increase in the unadjusted

rates effective October 1, 1999, based on the recommendation of one

commenter.

Comment: We received a number of comments recommending that we

periodically recompute the PPS rates using the most recent data.

Reasons commonly mentioned include that rebasing would allow the PPS to

recognize changes over time in the intensity and scope of services

provided in SNFs, and that it would provide an opportunity for re-

standardization of the

[[Page 41649]]

payment rates using actual resident assessment (MDS) data.

Conversely, we received comments that recommended against rebasing

payment rates periodically. These commenters were concerned that

because the PPS provides incentives for SNFs to provide services more

efficiently and eliminate distinct parts (that would tend to lower

average SNF costs, as determined from Medicare cost reports), the

impact of rebasing the rates would be unfair, since it would tend to

penalize providers for being efficient.

Response: While we are not able to predict the absolute impact on

SNF costs of the incentive for SNFs to provide services more

efficiently or their continued desire to maintain distinct parts under

PPS, we have no doubt that the PPS will result in some downward

pressure on costs. Anecdotal evidence up to this point certainly

supports this conclusion.

Section 1888(e)(4)(A) of the Act requires a 1995 base year. Section

1888(e)(5)(A) of the Act specifically provides for the establishment of

an SNF market basket index, while section 1888(e)(4)(E) of the Act

requires that the SNF PPS rates be updated annually using that index.

As discussed in response to earlier comments, we believe that it is

appropriate to recognize changes over time in the Medicare population

or care delivery practices in SNFs in the context of case-mix

adjustments. Our periodic evaluation of the case-mix classification and

indices will provide an opportunity for making refinements to the PPS

that recognize changes in the intensity and scope of services provided

in SNFs.

Comment: We received several comments regarding certain costs that

were not included in the computation of the Federal rate. Specifically,

the commenters expressed concern that all SNFs receiving ``new''

provider exemptions from the routine cost limits and all allowable

costs associated with atypical services exceptions to the cost limits

have not been included in the data used for computation of the Federal

rates.

The commenters suggested that it is unfair to exclude the cost

associated with those providers that are providing atypical levels of

care. Further, they noted that these are the same providers that would

have a high case-mix in the new payment rates and, therefore, should be

included. Virtually all of these commenters suggested that the rates

are distorted due to the exclusion of many providers and costs of

furnishing atypical services.

Response: The statute is very specific regarding the exclusion of

providers that have received ``new'' provider exemptions from the

calculation of the Federal rates. Section 1888(e)(4)(A) of the Act

requires that cost data from SNFs ``that were subject to (and not

exempted from) the per diem limits'' be used in computing the payment

rates. Similarly, the statute specifically requires the exclusion of

allowable costs associated with exceptions granted in the FY 1995 base

year. Section 1888(e)(4)(A)(i) requires the use of the allowable costs

of SNF services ``excluding exceptions payments'' in calculating the

payment rates.

Comment: Several commenters were concerned that we eliminated

certain cost reports from the calculation of the Federal rates on the

basis of their duration. Cost reports in excess of 13 months or less

than 10 months in duration were eliminated from the rate computations.

In addition, concerns were expressed over the use of a geometric

outlier elimination process to remove SNF costs from the data.

Response: As we indicated in the interim final rule, we used only

those cost reports for periods of at least 10 months but not more than

13 months. We excluded those periods that fell outside these parameters

on the basis that those cost reports may not be reflective of a normal

cost reporting period and, therefore, may tend to distort the rate

computation. For example, providers entering or exiting the Medicare

program could have abnormally high or low costs due to fluctuations in

occupancy. This approach does not affect a large number of cost reports

and is consistent with our rate setting methodology in other areas of

Medicare.

Similarly, we believe the application of a geometric outlier

elimination process for the SNF costs used to calculate the payment

rates is an appropriate analytical approach consistent with rate

setting for payment systems in other areas of Medicare. We believe that

three standard deviations from the geometric mean of the log value for

each cost component is a fair level of tolerance that focuses on the

truly aberrant cost values. In addition, this process involved the

removal of both high cost and low cost aberrant values, resulting in a

more equitable and more meaningful computation of the rate components.

We would also add that we used all FY 1995 cost reports that were

available at the time of the development of the interim final rule and

associated payment rates. While some cost reports may not have been

available at that time, we constructed the rates based upon the best

available data and are confident it was more than adequate for

construction of the rates. Finally, a small number of cost reports were

eliminated from the computation of the rates due to faulty or missing

data on critical items.

Comment: Several commenters expressed concern with our methodology

related to the use of a MEDPAR analog in the standardization of the

Federal payment rates. They questioned whether the MEDPAR data were

sufficiently accurate for the purpose of developing payment rates and

referred to the 28 percent difference, reported in the interim final

rule (63 FR 26260), between the therapy index calculated from actual

MDS assessments and the MEDPAR analog-generated index. They noted

additional limitations of the analog, such as the lack of functional

status information and recommended that we use actual MDS data, when

available, to re-standardize the payment rates, possibly in conjunction

with a rebasing of the cost data.

Response: As noted in the interim final rule, an adequate national

sample of MDS data for use in standardizing the Federal payment rates

does not yet exist. In the absence of these data, we believe the MEDPAR

analog, adjusted by the case-mix adjustment factor, provides an

appropriate estimate of case-mix for the purpose of rate

standardization. Based on our comparison of actual MDS and MEDPAR data,

we concluded that limitations of the MEDPAR case-mix analog had no

effect on the nursing component of the rate. Whatever inaccuracy

existed in the MEDPAR analog data, the effect was limited to the

therapy component and tended to increase, not decrease, the payment

rate. The fact that the available MDS data yielded a therapy index

value 29 percent higher than the MEDPAR analog data for the same cases

demonstrates that use of the MEDPAR data alone would have made the

therapy component inappropriately high. That is the reason that the

correction factor was applied to the therapy component.

Comment: A few commenters expressed concern about the adjustment we

made to the cost report data in developing the Federal rates, to

account for providers with cost reports that were not settled. One

commenter indicated that all SNFs should not be penalized by this

adjustment. It was also suggested that the rates be redone in the

future to account for the actual change between the as-submitted cost

reports and the settled ones. In addition, one comment

[[Page 41650]]

addressed the methodological application of the adjustment in the

computation of the rates, suggesting an alternative where the

adjustment is applied to total Medicare routine costs as opposed to

only costs subject to the routine limit.

Response: As we indicated in the interim final rule, the adjustment

made pursuant to section 1888(e)(4)(A)(i) of the Act was applied to

unsettled cost reports and was based on the average ratio for all

providers in 1995, between their as-submitted and settled cost report.

This adjustment is only applied to the cost report data of providers

whose cost report was not settled as of the time we computed the rates.

It is an actuarial adjustment required under the law that affects how

the average SNF costs are determined and does not penalize other

providers with settled reports.

As we indicated in the interim final rule, these adjustment factors

were validated using data from three previous years, that showed this

ratio remains fairly constant. To update and change the rates in the

future based on revised cost reports is impractical. Revisions are

constantly being made to cost reports (for many years) and our

validation exercise indicates the ratios are accurate.

Finally, we have decided to incorporate the methodological

alternative described above and will adjust the unadjusted nursing

case-mix component of the urban and rural Federal rates by +$.32 and

+$.24, respectively. In addition, we will adjust the unadjusted non-

case-mix component of the urban and rural Federal rate by +$.25 and

+$.21, respectively. We believe this refinement in the application of

the adjustment factor may result in a more accurate estimate of the

routine costs of SNFs. This adjustment will be prospective and will be

effective at the next scheduled update of the SNF PPS rates on October

1, 1999. That is the earliest point at which we can implement changes

to the standard claims processing systems.

Comment: We received one comment asking why the issue of payments

for low-volume SNFs was not addressed in the interim final rule.

Response: The new Part A PPS established in section 1888(e) of the

Act applies to all SNFs, and does not include any special treatment for

low-volume SNFs. Section 1888(d) of the Act provided for a separate,

optional payment system for SNFs with less than 1500 days (that is,

low-volume SNFs) in their preceding cost reporting period. However,

according to current law, this special payment system for low-volume

SNFs is only in effect for cost reporting periods beginning before July

1, 1998.

Comment: Numerous comments were received from hospital-based

facilities and their representatives indicating that the rates are too

low and do not recognize the additional overhead incurred in a

hospital-based facility. The commenters pointed out that the Federal

rate uses a mean of the average for all freestanding providers and the

average for all freestanding and hospital-based providers. This

computation double counts freestanding providers, thus lowering the

rates. Some commenters suggested the rates should be redone, or an add-

on or separate rate for freestanding versus hospital-based providers be

established, similar to what was done for routine cost limits.

Response: As many of the commenters have already recognized, the

computation as described above is clearly mandated in the formula set

out in section 1888(e)(4) of the Act.

Comment: We received several comments regarding the wage index that

is used to standardize and adjust the rates. The commenters suggested

that the hospital wage index might not adequately represent wages paid

in SNFs. Many of the commenters pointed out that SNF wages and hours

are excluded from the hospital wage index computation, yet we are

applying it to SNF payments. Most commenters want the wage index

updated periodically and often to reflect the most recent changes in

wages. One commenter suggested that we make other changes to the method

for how the wage index is calculated by including costs that are now

excluded, such as physician salaries, and excluding items like interns'

and residents' salaries. There were also a few commenters who suggested

that any move to a wage index based on SNF wage data be done slowly to

ensure it is done accurately. Most commenters hope to see a wage index

based on SNF data soon. In addition, many commenters want us to use a

later wage index to reflect the recent mandated changes in the minimum

wages rates paid to some employees.

Response: As we indicated in the interim final rule, we are using

the hospital wage data since the SNF wage data have not been completed.

We used the latest completed hospital wage index that was available at

the time of publication. It is our intent to use the latest wage index

data that are complete and available when we publish rates or updates

to the rates in the future.

We have been unable to evaluate a wage index based on SNF wage

data, as not all SNF providers reported data via the worksheet S-3. Now

that we have a full year of wage data for both freestanding and

hospital-based facilities, we will begin to evaluate and analyze the

wage and hourly data from the SNF and hospital-based SNF cost reports.

We will analyze and develop these data to evaluate their accuracy and

validity. It is our intent, if the data are accurate, eventually to use

and publish a wage index based on SNF wage data. However, it has been

our experience in the past that when new wage data are used, they can

result in enormous and erratic shifts in the wage indexes; many

providers could be adversely affected while others experience a

windfall. Therefore, before we use any SNF wage data, we will perform

numerous edits to ensure quality. In addition, we will ask for public

comments once the wage index data are available. Since we have not yet

developed a wage index based on SNF wage data, we do not know the

impact of excluding or including any particular cost centers.

As discussed above and in the interim final rule, until an

appropriate wage index based on SNF data is available, we will use the

latest available hospital wage index data in making annual updates to

the payment rates. We believe that SNFs and hospitals compete in the

same labor market areas and, therefore, absent specific SNF wage data,

we continue to believe that the hospital wage data accurately reflect

the relative wage costs between labor areas. In making these annual

updates, section 1888(e)(4)(G)(ii) of the Act requires that the

application of this wage index be made in a manner that does not result

in aggregate payments which are greater or less than would otherwise be

made in the absence of the wage adjustment. For the initial period of

the SNF PPS, the adjustment required by this section was accounted for

through the standardization of the cost data that formed the basis for

the per diem rate components. By means of standardization, each rate

component was adjusted for wage index and case-mix differences so that

aggregate payments were unaffected by the presence of these payment

adjustors.

Since, for the second PPS year (Federal rates effective October 1,

1999), we plan to update the wage index applicable to SNF payments

using the most recent hospital wage data, it is necessary to ensure

that the aggregate payments in the second year are neither greater nor

less than they would be if we continued to use the wage index from the

initial year. This requirement, established pursuant to section

1888(e)(4)(F)(ii) of the Act, will be met by multiplying each of the

per diem rate components by the ratio of the volume

[[Page 41651]]

weighted mean wage adjustment factor (using the wage index from the

initial year) to the volume weighted mean wage adjustment factor, using

the wage index for the fiscal year beginning October 1, 1999. The same

volume weights are used in both the numerator and denominator and will

be derived from 1997 MedPAR data. The wage adjustment factor used in

this calculation is defined as the labor share of the rate component

multiplied by the wage index plus the non-labor share.

Comment: We received two comments suggesting that the rates should

have an add-on to account for the additional cost of completing

resident assessments and the administrative costs associated with

implementing this new payment system and other unfunded mandates.

Response: We recognize that the increased frequency of assessment

may result in additional costs for SNFs. However, as we indicated in

response to an earlier comment, the Congress mandated both the basic

formula and the fiscal year cost data that we are to use in developing

the rates. To the extent that any of these assessment costs are

included in the base year data, they are reflected in the rates. We

would note that, as we indicated in the interim final rules discussion

of the Paperwork Reduction Act, it was determined that the increased

assessments required and the time to transmit them has a minimal impact

on each individual facility. We recognize that providers will incur

additional costs associated with more frequent assessments but we

believe our current rate scheme is consistent with the law.

Comment: Several commenters suggested that capital should not be

part of the rate, suggesting that it be an add-on or pass-through to

recognize those facilities that were committed to large capital

expenditures incurred after 1995.

Response: In accordance with section 1888(e)(2)(B) of the Act, the

calculation of the Federal rates included the capital costs. We realize

that committed capital expenditures after 1995 may create some hardship

on some providers. However, we believe that the present rate scheme,

which includes capital, is consistent with the language and intent of

the statute. Further, we believe that the capital costs included in the

rates are adequate to cover capital costs that would be incurred for

providers over time.

Comment: We received numerous requests, particularly from rural

hospital-based facilities, suggesting that we allow providers to

reclassify to a nearby adjacent urban area to receive the urban wage

index or the rates applicable to the adjoining urban area, especially

in circumstances where the hospital has been reclassified because it is

in a county that was defined as urban under section 1886(d)(8)(B) of

the Act (sometimes referred to as a ``Lugar'' county) or as a result of

geographic reclassifications based on decisions of the Medicare

Geographical Classification Review Board (MGCRB) or the Secretary under

section 1886(d)(10) of the Act for purposes of the hospital PPS. These

commenters suggested that the SNFs are competing in the same market as

hospitals. One commenter suggested that a board similar to the MGCRB be

established to consider an SNF's request to be reclassified.

Response: While we have broad authority to develop an SNF wage

index, we continue to believe that the reclassifications permitted for

hospitals under sections 1886(d)(8)(B) and 1886(d)(10) of the Act are

specific to hospitals. The Congress could have chosen to extend this

provision to SNFs under section 1888(e) of the Act, but it did not. In

addition, it has been our longstanding policy not to allow or recognize

reclassification for SNFs for payment under the routine cost limits.

Since we hope eventually to develop a wage index specific to SNFs, the

possible effect of reclassification on the wage index is unclear and

might have unintended consequences.

Comment: Two comments were received asking that we consider an

adjustment for the non-labor portion for Alaska and Hawaii providers,

similar to what is done for routine cost limits for SNFs. These

commenters suggested that these areas experience a much higher cost

than those providers in the continental United States and, therefore,

are entitled to this adjustment.

Response: The hospital inpatient PPS does have an adjustment

similar to that requested by these commenters; however, it was mandated

by the statute governing the hospital PPS. By contrast, the Congress

did not provide for such an adjustment in the legislation for the SNF

PPS. Costs incurred by Alaska and Hawaii providers are, of course,

included in the base year computation.

Comment: One comment we received suggested that SNFs that were

subject to the low-volume rates should have been eliminated from the

calculation of the Federal rates. Furthermore, the commenter added that

these providers should be exempt from PPS and continue to be paid under

the low-volume rates.

Response: Section 1888(e)(4)(A) of the Act specifically included

low-volume facilities in the SNF PPS rate calculation.

C. Federal Rates--Part B Add-on

In describing the data to be used in developing the Federal rates,

section 1888(e)(4)(A)(ii) of the Act provides for including an estimate

of the amounts payable under Part B for covered SNF services furnished

during FY 1995 to individuals who were residents of a facility and

receiving Part A covered services. This estimate is also known as the

``Part B add-on.'' In this section of the preamble we are providing an

expanded discussion of the development of the add-on for Part B

services which is included in the Federal rates.

Comment: We received a number of comments questioning the accuracy

of our estimate of Medicare Part B allowable charges associated with

patients in Medicare Part A stays during the FY 1995 base year used for

determining both the Federal and facility-specific payment rates.

Certain commenters cited evidence of missing bills and charges

associated with individual providers for particular types of services

(for example, laboratory services or rehabilitation therapy). In

addition, several commenters suggested that we allow for an appeals

process related to the Part B estimate associated with facility-

specific rates.

Response: We took great care in both the methodological design and

construction of the data sources necessary for the development of this

estimate. We are aware of several independent industry efforts to

review this methodology which found no defects in the design. In this

final rule, we are providing the following, more detailed discussion of

the methodology used for the development of the Part B estimate with

the hope that doing so will clarify our process of determining this

estimate and respond to questions and concerns.

The facility-specific payment rate used for the transition is

computed using the allowable costs of SNF services for cost reporting

periods beginning in FY 1995 (cost reporting periods beginning on or

after October 1, 1994, and before October 1, 1995). Included in the

facility-specific per diem rate is an estimate of the amount payable

under Part B for covered SNF services furnished during cost reporting

periods beginning in FY 1995 to individuals who were residents of the

facility and receiving Part A covered services.

These estimates were developed using allowed charges (including

coinsurance and deductibles) from all Medicare Part

[[Page 41652]]

B claims actually submitted (other than those specifically excluded

from the consolidated billing requirements, such as physician services)

associated with SNF residents in a Part A stay during cost reporting

periods that began in FY 1995. Applying the methodology described

below, we provided the fiscal intermediaries (FIs) in May of 1998 with

the total aggregate amount payable under Part B. In addition, at the

request of the nursing home industry, we included a detailing of

certain components of that amount for informational purposes.

At that time, we instructed the FIs that only the item listed as

``Total Part B Add-on Amount'' should be incorporated in the

calculation of the facility-specific rates. We noted that, while the

total Part B amount was an accurate estimate based on the universe of

Part B claims, the assignment of allowed charges into the different

service components was only an approximation due to the level of

specificity of the codes and the variation in supplier billing and

coding practices. The following description details the methodology

used to determine the Part B add-on amounts:

1. Identify Cost Report Period

For each SNF, determined appropriate FY 1995 cost report period.

Used all FY 1995 cost reports on file as of January 30, 1998. If no FY

1995 cost report was available, estimated a FY 1995 period from the

latest cost report available.

2. Create List of Dates for SNF Stays for Each Beneficiary

For each SNF, identified all Part A SNF claims with the discharge

date on the claim falling within the cost report period. For each

beneficiary, identified the dates of each stay during the cost report

period.

3. Identify All Non-Physician Part B Claims

Obtained all Part B physician, supplier, DME claims for 1994, 1995,

and 1996. Omitted all professional services, defined as any service

associated with a physician specialty code. Obtained all Part B

outpatient department facility claims for 1994, 1995, and 1996.

4. Match List of Part A SNF Stays to Part B Claims

By beneficiary, matched list of Part A SNF stays to Part B claims.

Kept all non-physician services or facility claims falling on or

between dates of admit and discharge for each SNF stay.

5. Drop Claims for DME

For non-physician Part B claims, that is, not facility claims,

reviewed all alphanumeric HCPCS and identified and dropped obvious DME

codes, for example, wheelchairs, canes, transcutaneous electrical nerve

stimulation (TENS), glucose monitors, commodes, walkers, bath and

toilet aids, lifts, and oxygen equipment. Because coverage under the

Part B DME benefit is not allowed for beneficiaries in an SNF stay, we

believe that these codes probably occurred on either the day of

admission or the day of discharge or were associated with erroneous

payments.

6. Adjust Outpatient Claims to Reflect Costs

Adjusted total charges on Part B outpatient facility bills to

reflect total Medicare payments using a payment to charge ratio

calculated from FY 95 outpatient cost reports. If no FY 95 cost report

was available, used ratio from FY 94 or, if necessary, FY 93 cost

report. If a FY 93 cost report was not available, used the payment

amount associated with the claim.

7. Drop Outpatient Bills

Removed claims with home health and dialysis provider numbers.

Dropped Part B outpatient facility claims where the SNF provider number

matched the hospital outpatient provider number. Dropped bills with at

least one of the following revenue centers: surgery, emergency room

(ER), ambulatory surgical center (ASC), cardiac catheterization,

computerized axial tomography (CT) scan, and magnetic resonance imaging

(MRI). These outpatient hospital services are excluded from the

consolidated billing requirements.

8. Calculate Totals

Calculated total allowed charges for all non-physician Part B

claims. Calculated total payments for Part B outpatient facility

claims.

9. Create Descriptive Categories Within Totals

At request of certain members of the industry, created general

categories to describe the distribution of dollars among types of

services. Categories are not exact due to the lack of precision in

categories for HCPCS ranges, local codes, and the structure of facility

claims. For example, dollars for laboratory services could appear in

(a) the ``laboratory'' category for non-physician Part B, (b) the

``other'' category for non-physician Part B if the code was local, or

(c) the outpatient department's (OPD) ``other'' category for laboratory

tests conducted by an outpatient facility.

Created categories for non-physician Part B claims using HCPCS and

CPT ranges. Often, broad HCPCS categories capture some unrelated codes.

In addition, temporary local codes had to be placed into the ``other''

category.

The structure of the outpatient facility claims prevents

associating a code with a specific dollar amount. Created outpatient

therapy category by combining all claims from CORF hospitals and any

claim with only one physical therapy (PT), occupational therapy (OT),

or speech-language pathology (SLP) code. Left all remaining bills in

OPD category.

As discussed in the above description of our methodology, a number

of factors prevented us from disaggregating the total Part B allowable

charges precisely into distinct high level categories (for example,

laboratory services). However, we decided to attempt to provide an

approximate breakout by category to provide SNFs some notion of what

their Part B service mix may have looked like in the FY 1995 base year.

While we did note in the listing of Part B add-ons provided to FIs

that the categorization of charges was only an approximation, this

qualification may not have always been understood by providers. We

regret any confusion caused by this breakout. We would note that our

purpose in developing the total estimate of Part B allowable charges

did not go beyond providing an accurate account of the total allowed

charges to be included in the PPS rates, and we believe our estimate

accomplished this. However, even if our purpose had been to map every

charge and HCPCS code precisely to some broad category, once again, the

data and structure of Medicare's billing system would not have

permitted it.

Beyond issues related to the categorization of Part B charges, we

received no comments that contained substantiated evidence of

systematic defects in the methodology or data. We would note that

section 1888(e)(8)(B) of the Act limits administrative review of this

estimate.

Comment: We received numerous comments indicating that we should

publish, or otherwise make available to the public and the industry,

the complete and itemized data that were included in the computation of

the rates. Of particular concern was the percentage of the nursing

case-mix component of the rate that is attributable to nursing services

and non-therapy ancillary costs. Some commenters suggested that they

were

[[Page 41653]]

unable to replicate the rates we published with the data currently

available.

Response: Much of the data necessary to compute the rates have been

available for some time, including the 1995 SNF cost reports and the

MEDPAR files. We have also put data and information related to the

computation of the case-mix indices on our SNF PPS website, at:

www.hcfa.gov/medicare/snfpps.htm>. A public use file containing the

most significant data items relating to the calculation of the

unadjusted Federal rates can also be found on the website. The

standardization and case-mix correction factors are included with the

public use data.

It is our understanding from conversations with a number of users

of the data that the public use file, along with the data that were

already available, has been quite helpful in understanding the

calculation of the rates. In addition, we have honored several requests

under the Freedom of Information Act for data associated with the rate

calculations, and have provided further information through data

release agreements.

Regarding the percentage of the nursing case-mix component of the

rate that is attributable to nursing services and social services and

non-therapy ancillary costs, we agreed with earlier comments to the

interim final rule that the public would benefit by knowing the

percentages for nursing and social services and non-therapy ancillary

services included in the rate. Accordingly, on November 27, 1998, we

published a notice in the Federal Register (63 FR 65561) to reopen

comments to the interim final rule. We also provided the public with a

percentage breakdown of the nursing case-mix component of the rates to

the extent feasible.

Comment: We received a number of comments concerning our discussion

in the interim final rule related to OIG's proposal to adjust the

Federal rates to account for costs in the 1995 base year cost data that

result from medically unnecessary services or improper payments. These

comments strongly recommended that we not proceed with such an

adjustment, citing the already significant downward impact on the

Federal rates of the BBA budgetary savings, the inadequate statistical

basis for pursuing such an adjustment, and insufficient statutory

authority for proceeding with an actuarial adjustment of this type to

the rates.

Response: We are concerned about the application of an adjustment

that would have a downward impact on the Federal rates in light of the

substantial reduction already incorporated into the calculation under

the BBA requirements. According to the impact analysis contained in the

interim final rule, this reduction is 17 percent on average. However,

there is a substantial body of evidence, in the form of OIG and GAO

studies, that at least suggests there were inappropriate services or

improper payments associated with SNF services during the 1995 base

year. Consequently, it could reasonably be argued that exclusion of the

costs of these services from the cost base used to compute the Federal

payment rates is appropriate.

However, we believe that in considering the level of budgetary

savings to incorporate into the statutory formula for establishing the

Federal rates, the Congress took into account the existing cost base

and aggregate SNF payment levels to determine an appropriate level of

budgetary savings. Our policy with regard to this issue will be not to

proceed with such an adjustment in the absence of specific statutory

direction from the Congress.

D. Facility-specific Rates-Transition

Section 1888(e)(2) of the Act provides, for most facilities, a

phased transition from facility-specific payment rates (which reflect

the individual facility's historical cost experience) to the Federal

rates. During such a facility's first three cost reporting periods

under the SNF PPS, it receives a blended payment rate, in which the

Federal portion initially represents 25 percent of the facility's total

payment rate, and then increases by 25 percent increments in each

succeeding period until the facility is paid at the full Federal rate.

In this section of the preamble, we are providing responses to

comments on a number of issues related to the PPS transition period and

the calculation of the facility-specific rates. These include issues

related to the eligibility of certain SNFs for the transition. In

addition, this section includes policy changes related to the

calculation of the Federal rates for certain SNFs with short cost

reporting periods and the eligibility for the transition of SNFs with

cost reporting periods beginning in FY 1994 but including the entire FY

1995 period.

Comment: We received several comments suggesting that we should

define a new SNF as one that first furnished patient care on or after

October 1, 1995, rather than one that first received payment on or

after October 1, 1995, as our present policy dictates.

Response: We understand that there are many concerns regarding the

issue of eligibility for the PPS transition. However, we believe

current policy is consistent with the statute. Section 1888(e)(2)(E) of

the Act specifically refers to the date an SNF first received payment

from Medicare on or after October 1, 1995, as the threshold date.

However, it is important to understand that the threshold for

determining eligibility for the transition period affects providers in

different ways, creating both winners and losers. Thus, while many

providers may want to receive PPS transition payments, many other

providers would rather be paid on the basis of the full Federal rate.

We do not see the benefit of a policy change that creates losers under

the system from winners and vice versa.

Comment: We received a number of comments recommending that we

modify our policy with regard to the PPS transition, to allow existing

SNFs to elect to bypass the transition and be paid 100 percent of the

Federal rate if they had experienced significant shifts in case-mix or

significant capital expenditures after the 1995 base year used for

determining the facility-specific rate. One commenter included a

detailed assessment of this proposed policy, including an estimate of

the aggregate costs to the Medicare program of its adoption.

Response: We understand the concern of SNFs that have operated

under the Medicare program since 1995 or earlier and yet find

themselves disadvantaged by the PPS transition due to changes in their

care delivery model or significant capital expenditures that occurred

after the 1995 base year used for computing the facility-specific rate.

However, we believe our present policy to be reasonable and consistent

with the plain language of the statute. Section 1888(e)(2)(E)(ii) of

the Act sets forth the requirements concerning whether a facility

receives payment under the PPS transition or solely according to the

Federal rates. This section provides that for SNFs that ``first

received Medicare payment for services under this title on or after

October 1, 1995, payment for such services shall be made under this

subsection as if all services were furnished after the transition

period.'' In our view, this language establishes clear criteria related

to provider eligibility for the transition and the appropriate basis

for Medicare payment. Accordingly, we have established a policy which

relies on the date an SNF first received payment (interim or otherwise)

from Medicare to determine the basis of their payment.

Comment: We received one comment asking us to reconsider our policy

regarding eligibility for the transition for providers that do not have

a cost

[[Page 41654]]

reporting period beginning in FY 1995, but whose period contains the

entire 1995 FY. Examples of these cost reporting periods include a 13-

month cost reporting period beginning September 1, 1994, and ending on

September 30, 1995 or reporting periods with a floating beginning date

(that is, tied to a specific day of the week) of September 27, 1994.

Response: In Transmittal 405 of the Provider Reimbursement Manual

(PRM, HCFA Pub. 15-1), we had initially required these providers to be

paid at the Federal rate without a transition period, since these

providers did not have a cost reporting period beginning in FY 1995

(the statutory basis for computing the facility-specific transition

rate). However, we have reconsidered our policy, because these

providers did receive their first payment from Medicare before October

1, 1995. These providers will now be eligible for the transition

period.

In addition, any provider that has been paid the full Federal rate

based on our original policy contained in Transmittal 405 of the

Provider Reimbursement Manual will be held harmless, since they have

already transitioned to the PPS. In short, this means that providers

with a cost reporting period beginning date in 1994 and whose period

contains the full 1995 fiscal year (that is, the 12 months beginning

October 1, 1994, through September 30, 1995), will be able to elect

either a PPS transition based payment or the full Federal rate.

Whichever rate the provider chooses must be used for all the years of

the transition period.

Comment: We received a number of comments regarding our policy on

changes of ownership and mergers as they relate to a provider's

eligibility for the PPS transition.

Response: As discussed earlier in this section, SNFs that first

received payment from Medicare on or after October 1, 1995 receive

payment based on the Federal rate only while SNFs that first received

payment from Medicare prior to October 1, 1995 are paid according to

the transition rate and are precluded from receiving payment solely

based on the Federal rate. In addition, our policy, as stated broadly

in transmittal 405 of the Provider Reimbursement Manual, requires that,

for purposes of determining a provider's eligibility for the

transition, Medicare makes its determination based on the date of first

Medicare payment (interim or otherwise) under the present provider

number.

For example, when an SNF undergoes a change in ownership, such as a

merger or a consolidation, the payment is determined by the payment

history of the surviving entity as indicated by the surviving SNF's

provider number. This conforms with longstanding reimbursement policy

and payment principles as applied under the former reasonable cost

payment system and provides administrative simplicity in addressing

complex transactions among SNFs, hospitals, and other entities.

Comment: We received several comments recommending that we adopt a

policy where SNFs would be allowed to elect to bypass the transition

period and receive payment based on the full Federal rate.

Response: Similar to our response to an earlier comment, we

understand how the transition payment methodology may disadvantage

certain providers. However, section 1888(e)(1) and (2)(E) of the Act

specifically addresses the issue of which providers are paid the full

Federal rate and which ones must receive transition payments. As we

discussed, the statute requires that SNFs that received their first

payment under Medicare before October 1, 1995, are to be paid based on

the transition payment methodology described in the interim final rule.

Comment: We received a number of comments related to the Part B

add-on and the methodology for computing facility-specific rates for

SNFs that participated in the Multistate Nursing Home Case-Mix and

Quality Demonstration (NHCMQD) in 1997. Under the interim final rule,

these facilities did not receive a Part B add-on as part of their

facility-specific rate. The commenters argued that a Part B add-on is

appropriate for these SNFs. Several commenters provided detailed

arguments asserting that a Part B add-on for these providers is legally

supportable under the statute.

Response: It appears to us that a Part B add-on to the facility-

specific rate for providers participating in the NHCMQD in 1997 could

well be an appropriate payment policy in light of the historical

circumstances.

During the NHCMQD, many Medicare Part A patients in these SNFs

received certain ancillary items or services provided by suppliers who

then billed Medicare directly under Part B. However, we find that the

statutory language at section 1888(e)(3)(B) of the Act, that provides

the formula for computing facility-specific rates for NHCMQD providers,

does not support this policy outcome.

Accordingly, we are maintaining the policy, set forth in the

interim final rule, of not including a Part B add-on in the calculation

of facility-specific rates for SNFs participating in the NHCMQD in

1997. We believe this policy is consistent with the statute. The

statute treats NHCMQD providers differently from other facilities. For

most facilities, the statute directs the Secretary to use a 1995 base

year and provides for a Part B ``add-on''; for NHCMQD facilities, the

statute directs the Secretary to use a later base year (1997) and does

not provide for a Part B ``add-on.'' Although a Part B add-on for

NHCMQD facilities might be appropriate as a conceptual matter, the

statute does not provide for a Part B add-on and we do not believe the

lack of a Part B add-on leads to an absurd result.

In our effort to ensure the appropriateness of the payment

methodology set forth in the interim final rule, we have decided to

make a modification to one aspect of the calculation of the facility

specific rates. This change only affects the methodology for

determining the inflation factor applied in the calculation of the

facility specific rates for certain providers with short cost reporting

periods (that is, less than 12 months).

There were three different types of short periods discussed in the

interim final rule:

a. A short period in the base year,

b. A short period in the initial period, and

c. A short period between the base year and the initial period.

The interim final rule included separate instructions on how to

determine which factor to use for an SNF having a short period. There

was, however, no discussion of how to determine which factor to use if

a SNF had more than one short period. For example, an SNF could have a

short period in the base year and a short period between the base year

and the initial period of the PPS.

We now believe that the instructions for item c should not be

applied to SNFs which have both a short period in the base year and a

short period between the base year and the initial period. If an SNF

has a short period in the base year and a short period between the base

year and the initial period, the instructions in section (a) should be

applied using the short period in the base year.

E. MDS Assessments

Under the SNF PPS, the Federal rate incorporates adjustments to

account for case-mix, using a resident classification system that

accounts for the relative resource utilization of different patient

types. This classification system, RUG-III, assigns beneficiaries into

one of 44

[[Page 41655]]

groups, using assessment data from the MDS that the SNF completes

according to an assessment schedule specifically designed for Medicare

payment.

In the interim final rule, we discussed issues relating to the use

of the RUG-III classification system under the SNF PPS, including

scheduling and other requirements pertaining to the MDS, use of the

RUG-III ``grouper'' software, and the use of an Other Medicare Required

Assessment (OMRA) in certain situations following the discontinuation

of rehabilitation therapy services.

In this section of the preamble, we are providing responses to

comments on a number of issues related to the use of the OMRA, grace

days, and the Health Insurance Prospective Payment System (HIPPS) codes

used to bill Medicare Part A covered SNF stays. We also address

comments and questions about the midnight rule and its effect on the

MDS schedule, and provide clarification regarding counting therapy

minutes on the MDS, as well as the requirements for the therapy plan of

treatment. In addition, we are responding to comments concerning

recognition of respiratory therapy and recreational therapy in the

payment rates and on the MDS.

Comment: We received numerous suggestions of ways to improve the

MDS instrument, the assessment schedule, and the classification system.

These comments included suggestions both to increase and decrease the

frequency of required MDS assessments, to improve the MDS staging of

pressure ulcers, ideas for modifications to individual RUG-III groups,

and commenters' requests that we be more directive in our rules about

how facilities are to spend the payments they receive from Medicare.

Response: We appreciate all of the suggestions and will consider

them in our future work in these areas. The comments were very specific

and too numerous to address in this context. Rather, the subject matter

and degree of specificity of some of these suggested changes would be

more appropriately addressed through manual issuances.

It is also worth noting that at this time, the SNF PPS has been in

effect in most facilities for less than 12 months. In the future, when

providers have achieved greater stability and familiarity with the

system, and we have additional data to guide our decisions, we can

consider making additional refinements such as those suggested by the

commenters.

1. Billing Issues

Comment: There were several questions submitted with the comments

regarding the HIPPS codes used for billing SNF PPS claims. The

questions focused on how to use these codes for billing as

distinguished from MDS coding instructions.

Response: Although these codes were not mentioned in the interim

final rule, we believe that it would be helpful and appropriate to

explain here what the HIPPS codes are as distinct from the MDS

information. The HIPPS codes are 5-character codes used solely for

billing the Medicare FI for the Part A SNF stay. The codes reflect the

RUG-III group into which the beneficiary classified and the reason for

the assessment used for determining the classification. The HIPPS code

does not appear anywhere on the MDS. The reason for assessment

reflected in the HIPPS code is based on information coded in items A8a

and A8b of the MDS, but is not a duplication of the data reported on

the MDS. Rather, a conversion must be made from the information on the

MDS to the reason for assessment identifier that comprises the last two

digits of the HIPPS code.

For instructions for billing on the Unified Billing Form 92 (UB-

92), see Transmittal 405 of the Provider Reimbursement Manual (PRM,

HCFA Pub. 15-1, 7/98) published on our website. These instructions are

sent to our FIs and are also available through them.

Further, in the context of billing procedures, we would also like

to use this opportunity to clarify our policy on Periodic Interim

Payments (PIP). Since the inception of the Medicare program, SNFs

reimbursed on the basis of reasonable costs received interim payments

during their cost reporting year for the cost of Part A services

provided to Medicare beneficiaries. For many years, SNFs have also been

permitted to receive PIP--interim payments paid in equal biweekly

amounts--for these services if they met the requirements in

Sec. 413.64(h) and received intermediary approval. Since July 1987, the

statutory authority for PIP for qualifying SNFs has been in section

1815(e)(2) of the Act. Section 1815(e)(1) of the Act was added to

include certain requirements, in addition to the requirements in

Sec. 413.64(h), specifically applicable to hospitals receiving

prospective payments under section 1886(d) of the Act in order for the

hospitals to receive PIP. Section 1815(e)(2) of the Act clarified that

the additional requirements applicable to those hospitals were not

applicable to other types of providers, including SNFs, entitled to

PIP. Accordingly, the regulations at Sec. 413.64(h) were revised to

provide for the continuing availability of PIP after July 1987 for

these other types of providers, including for Part A services provided

by SNFs.

Interim payments, including PIP, provide cost reimbursed providers

with estimated payments during the cost reporting year pending

submittal and subsequent settlement of a Medicare cost report. A

provider can submit its cost report to the intermediary as late as the

last day of the fifth month after the end of the cost reporting period.

Following submittal, the intermediary's determination of Medicare cost

reimbursement to the provider for services provided to beneficiaries

during the year cannot be made until the cost report is reviewed,

sometimes including audit of the provider's records. Because

determination of Medicare reimbursement takes place after the end of

the cost reporting year, interim payments are needed during the year

until this final payment can be determined.

Because a cost report is not required to calculate prospective

payments, interim payments are not necessary to a provider for services

paid on the basis of prospective payments. Nevertheless, with the

exception of special requirements for hospitals receiving prospective

payments under section 1886(d) of the Act, section 1815(e) currently

provides for the availability of PIP for certain services, including

Part A services provided by SNFs, if the requirements in Sec. 413.64(h)

are met. It does not prohibit PIP for SNFs receiving prospective

payments.

While the BBA eliminated PIP under the provisions mandating a PPS

for home health agencies (HHAs), the Congress made no such requirement

under the statutory provisions related to SNF PPS. This may be because,

like the preceding SNF payment system, the SNF PPS continues to rely on

a daily payment amount, while for the HHA PPS, changes in the unit of

payment were contemplated. However, at this time, we see no reason to

discontinue administratively our existing policy of allowing PIP for

qualified SNFs, though we may choose to evaluate its continuing need in

the future.

Therefore, we are permitting the continued availability of PIP for

services of SNFs paid under the PPS. For those services, PIP is based

on estimated prospective payments for the year rather than on estimated

cost reimbursement. An SNF receiving prospective payments, whether or

not it received PIP prior to receiving prospective payments, may

receive PIP if it meets the requirements in Sec. 413.64(h) and receives

approval by its intermediary. Likewise, if an intermediary determines

that an SNF which received PIP prior to

[[Page 41656]]

receiving prospective payments is no longer entitled to receive PIP, it

will remove the SNF from PIP. As provided in Sec. 413.64(h)(5),

intermediary approval of PIP is conditioned upon the intermediary's

best judgment as to whether payment can be made under the PIP method

without undue risk of its resulting in an overpayment to the provider.

An SNF can receive Medicare payment for the bad debts of Medicare

beneficiaries if it meets the requirements at Sec. 413.80 and

implementing instructions. Payment for these bad debts are not included

in the prospective payments but rather are claimed on the Medicare cost

report. Also, some SNFs may incur costs for an approved medical

education program or may incur other costs that are not included in the

prospective payment. Payment for these costs are determined based on

the completion of a Medicare cost report. Because final payment for

Medicare bad debts and for costs paid outside the prospective payment

system is not determined until the cost report is settled, it is

appropriate that SNFs which receive prospective payments should receive

estimated interim payments during the year for bad debts and for costs

paid outside the prospective payment system. Payments for these costs

are made in equal biweekly payments in the same manner as PIP. There is

no requirement for an SNF to meet in order to receive biweekly payments

for these costs because it is the only type of interim payment made for

them.

The new regulations providing for PIP for SNFs receiving

prospective payments and for biweekly interim payments for costs

outside the prospective payment system closely follow the regulations

at Sec. 412.116 which provide for PIP for hospitals receiving

prospective payments under section 1886(d) of the Act, as adjusted to

remove provisions specifically applicable to those hospitals. As with

Sec. 412.116 for hospitals and Sec. 413.64 for SNFs under the previous

cost-based system, these regulations for SNFs also provide for

accelerated payments in certain situations.

2. Corrections

Comment: We received several comments with questions and

suggestions regarding the policies governing the correction of MDS

errors and billing errors.

Response: The MDS corrections policy is set forth in the State

Operations Manual (SOM, HCFA Pub. 7) by HCFA's Center for Medicaid and

State Operations. The corrections policy applies to all users of the

MDS and, thus, is beyond the scope of this regulation. We address

issues and provide clarification of Medicare policy regarding how to

correct or adjust SNF Part A bills to the Medicare program in the

Provider Reimbursement Manual.

3. Other Medicare Required Assessment (OMRA)

Comment: There were a number of questions about the OMRA. These

included questions about when the OMRA is to be performed and whether

it is a full or comprehensive assessment.

Response: An OMRA is required 8 to 10 days after rehabilitation

therapy is discontinued for Medicare beneficiaries who have been

receiving rehabilitation therapy in the SNF. Specifically, there is

confusion regarding whether or not this assessment type is required in

certain circumstances. For example, when the beneficiary has no further

need for skilled care and has been moved out of the Medicare-certified

portion of the institution before the eighth day following the

cessation of rehabilitation services or when one or two of three

therapy services are discontinued. As stated in our corrections notice

to the interim final rule, published in the Federal Register on October

5, 1998 (63 FR 53301), the OMRA is not required to be a comprehensive

assessment. There are no PPS requirements for comprehensive assessments

(that is, those including Resident Assessment Protocols (RAPs)).

Comprehensive assessments are only required for clinical reasons, as

they have been since implementation of the nursing home reform

requirements enacted in the Omnibus Budget Reconciliation Act of 1987

(OBRA 87, Public Law 100-203).

An SNF must perform an OMRA only for those beneficiaries who

continue to have skilled care requirements after their rehabilitation

therapy services have been discontinued. For those beneficiaries who

are not ready for discharge from the facility, and who continue to

require a Medicare covered skilled level of care, an OMRA must be

performed in order to obtain an accurate classification into one of the

non-therapy RUG-III groups.

The assessment reference date of the OMRA must be set on day 8, 9,

or 10 after the last day any rehabilitation therapy services were

provided. This timing ensures that no therapy minutes will be captured

on the OMRA and that the beneficiary's new classification will be into

one of the non-therapy RUG-III groups. An OMRA will always result in

classification into a non-therapy RUG-III group. For the days between

the cessation of rehabilitation therapy and the assessment reference

date of the OMRA, the beneficiary continues to be covered at the

therapy RUG-III group level to which he or she was classified before

cessation.

We expect that there will be many cases in which the beneficiary

will be discharged from the facility shortly after rehabilitation

therapy services end. Before PPS, beneficiaries were often discharged

from the SNF immediately upon the discontinuation of rehabilitation

therapies. Likewise, many SNF residents who received rehabilitation

therapy services under Medicare Part A were moved to a non-Medicare

level of care following the cessation of therapy services. These same

patterns are expected to continue under the PPS.

In circumstances in which the beneficiary is discharged from the

facility (or from the Medicare-certified portion of a larger,

noncertified institution) before the eighth day following the end of

all rehabilitation therapy, there is no expectation by Medicare that an

OMRA will be performed. If the beneficiary remains in the Medicare-

certified facility through the eighth day following rehabilitation

therapy discontinuation, there must be some clinical reason for his or

her continuing skilled stay that is supported by documentation in the

medical record. We realize that there will be cases in which the

beneficiary stays in the SNF for a number of days after rehabilitation

therapy ends, in order for the facility staff to verify that his or her

status is stable and to assure that the plans for his or her next

destination are appropriate and in the best interests of the

beneficiary.

By contrast, always waiting to perform the OMRA to verify that the

beneficiary is stable and no longer in need of skilled nursing or

therapy services is not appropriate. A pattern of OMRA assessments

immediately preceding discharge from the facility, or from the Medicare

level of care within the facility, would indicate that perhaps the

facility is at times using those 8 to 10 days inappropriately. We

believe it is unfair to the beneficiary to use any of the 100 Medicare

SNF benefit days available in a benefit period unless he or she is

actually in need of skilled services. Likewise, it is an inappropriate

use of Medicare trust fund dollars for Medicare to pay for SNF days

that are not needed by the beneficiary.

The beneficiary should not be kept in a Medicare Part A stay if

skilled services are neither needed, nor being provided. We believe

that nursing homes' clinical staff should know when there are no

[[Page 41657]]

skilled services being provided to a beneficiary. Our guidelines

provided in the PRM (Transmittal 405) reinforce the expectation that

facilities may, and in fact are expected to, act in the best interest

of the beneficiary with regard to use of the beneficiary's limited SNF

benefit days, by ending Medicare Part A coverage appropriately. (See

also the discussion below regarding circumstances that serve to

discontinue a presumption that the SNF level of care requirement is met

by a beneficiary who has classified into one of the upper 26 RUG-III

groups.)

F. Certification and Recertification

Comment: We received a few comments regarding the statutory

requirement for initial certification and periodic recertification as

to level of care, as required under section 1814(a)(2) of the Act.

Response: The comments regarding this particular provision are

addressed later, in the discussion on coverage and level of care

determinations under the SNF PPS. However, we would like to take this

opportunity to clarify that the required certification and

recertification statements are not the same as any requirements

specifically related to the plan of treatment for therapy that is

required for purposes of coverage, or to the overall requirement for

the multidisciplinary plan of care required by the long-term care

facility requirements for participation at section 1819(b)(2) of the

Act.

G. MDS Scheduling Requirements

1. Grace Days

Comment: We received several comments asking about the appropriate

use of the 3-day grace period provided for the Medicare 5-day

assessment. There is some confusion about when use of the grace days

could result in the facility being at a high risk for an audit.

Response: Days six, seven, and eight, of the Medicare covered stay,

were provided as grace days for setting the assessment reference date

for the Medicare 5-day assessment. This assessment is to have an

assessment reference date (MDS 2.0 Item A3a) of any day one through

eight of the Medicare Part A stay. Days one through five are optimal

but days six through eight are also acceptable, and for some residents

may actually be more appropriate; for example, to allow maximum

flexibility for nurses to determine when to set the assessment

reference date for the beneficiary's MDS, and thereby lessen the burden

of the increased frequency of assessments that accompanied the PPS.

Thus, the resident can be assessed using any one of these first eight

days as the assessment reference date for the Medicare-required 5-day

assessment.

However, we discourage the routine use of grace days for assessing

every Medicare admission. We plan to identify patterns of inappropriate

use as we gain a better understanding of what facilities' practice

patterns are. When a facility routinely uses a grace day as the

assessment reference date for the 5-day assessment, it loses the

cushion that these days provide against performing the MDS later than

day eight and, thus, risks being faced with payment at the default

rate.

At this time our main interest is to encourage facilities to

perform assessments timely and to recognize the grace days as a cushion

and to use them as such, rather than as deadlines for setting each

beneficiary's assessment reference date. The grace days are also

provided to offset any incentive that facilities may have to initiate

therapy services before the beneficiary is able to tolerate that level

of activity.

Our discussion in the interim final rule about the possibility of

audits was intended to address the possible practice of routinely using

grace days for Medicare assessments. We were cognizant that the routine

use of a grace day for the 5-day assessment would pose a temptation to

back-date the assessment fraudulently when day eight was missed. We

believed that any facility that routinely used grace days for the

required assessments was liable to have assessments billed at the

default rate; and that the absence of default rate billings in the

facility's claims might indicate that some misrepresentation of the

assessment reference dates had occurred.

Unlike the routine use of grace days described above, we do expect

that many beneficiaries who classify into the rehabilitation category

will have 5-day assessment reference dates that fall on grace days.

There are many cases in which the beneficiary is not physically able to

begin therapy services until he or she has been in the facility for a

few days. Thus, for a beneficiary who does not begin receiving

rehabilitation therapy until the fifth, sixth, or seventh day of his or

her SNF stay, the assessment reference date may be set for one of the

grace days in order to capture an adequate number of days and minutes

in section P of the current version of the MDS to qualify the resident

for classification into one of the rehabilitation therapy RUG-III

groups.

Another reason for the provision of three grace days for the 5-day

assessment was to make it possible for beneficiaries to classify into

the two highest RUG-III rehabilitation sub-categories. Classification

into the Ultra High and Very High Rehabilitation sub-categories is not

possible unless the beneficiary receives the sub-category's minimum

level of services during the first seven days of the stay.

We also intended to minimize the incentive to facilities to provide

too high a level of rehabilitation therapy to newly admitted

beneficiaries. Having these extra few days allows time for those

beneficiaries who need it, to stabilize from the acute care setting and

be prepared for the beginning of rehabilitation in the SNF. We expect

facilities will not compromise any beneficiary's health by beginning

rehabilitation therapy prematurely or at a level that is too rigorous

for the individual's status. In summary, use of grace days is

acceptable and permitted for patients with any condition. However, a

facility that uses grace days routinely may be subject to audit to

determine that assessment reference dates are accurately reflected.

Comment: One commenter requested that we modify the statement at

section II.B.7 of the interim final rule that states SNFs ``must submit

the Resident Assessment Protocols (RAPs) with either the 5-day or the

14-day assessment'' to indicate that the SNFs must submit the completed

RAP Summary Form, section V of the MDS with either the 5-day or 14-day

assessment.

Response: This may be a helpful clarification for providers;

however, we want to be certain that providers fully understand this

requirement. We will take this opportunity to make clear that the RAPs

are not a PPS requirement. The requirements for completion of section V

and the care planning responsibilities of facility clinical staff are

unchanged by the PPS. We included the clinical requirement for RAPs in

the interim final rule in an effort to help providers to understand how

the Medicare required SNF PPS assessments coordinate with the required

clinical assessments.

The requirement for RAPs is entirely outside of the SNF PPS. In

fact, if the clinical initial admission assessment (item AA8a of the

MDS 2.0 = ``01'') was performed before the beneficiary started his

Medicare covered SNF stay, neither the Medicare required 5-day, nor the

Medicare 14-day assessment is required to have a completed section V.

There are no care planning requirements associated with any full MDS

assessment performed solely for the purpose of complying with the

Medicare assessment schedule for a Part

[[Page 41658]]

A Medicare beneficiary's SNF stay. The Medicare PPS requirements are

separate from the clinical requirements. However, we have designed the

Medicare requirements so that an SNF can coordinate the scheduling of

assessments to avoid duplication of effort.

2. Completion and Locking

For Medicare payment, we are requiring that any assessment,

including the 5-day, must be ``completed'' (that is, signed by all

members of the care team) within 14 days of the assessment reference

date (MDS item A3a). That is, the completion date at MDS item R2b, must

be a date that is within 14 days of the date at A3a. Then the

assessment must be ``locked'' within seven days of the date at R2b, and

transmitted to the State in which the SNF operates within 31 days of

the final lock date (State Operations Manual, HCFA Pub. 7).

However, there are other considerations to keep in mind. There is

still the clinical requirement that an Initial Admission Assessment

must be ``completed'' by the 14th day of the nursing home stay. This

means that for a Medicare beneficiary who is newly admitted to the SNF

for a covered Part A stay, the SNF must complete a comprehensive MDS by

day 14, regardless of the assessment reference dates on the Medicare-

required 5 day and 14 day assessments.

As has been the case since the OBRA 1987 requirements were

implemented, a comprehensive assessment (Initial Admission Assessment)

is due to be completed by the 14th day of the SNF stay. In addition,

for Medicare beneficiaries in the SNF for a covered Part A stay, a 5-

day assessment must be performed, with an assessment reference date on

any day one through eight of the Medicare Part A covered stay, and must

be completed within 14 days of the assessment reference date. Also, by

the end of the second week in the Medicare Part A covered stay, the

Medicare 14-day assessment must be performed. This assessment must have

an assessment reference date of any day 11 through 19 (including the 5-

day grace period provided for this assessment).

Given these requirements during the first weeks of the SNF stay,

and considering that Medicare Part A coverage often begins on the day

of admission, we believe that in many cases nursing homes will opt to

complete a single assessment to satisfy the requirements for both the

5-day (or 14-day) assessment and the Initial Admission Assessment. In

this example, the Medicare 5-day assessment, with an assessment

reference date of any day, one through eight of the stay, will be a

comprehensive assessment and will have to be completed within 14 days

of the start of the SNF stay. The day of admission is counted as day

one. The assessment must comply with the requirements for the Initial

Admission Assessment. That is, it must be a comprehensive assessment,

including the RAPs.

When the Medicare 5-day assessment is also used to fulfill the

requirement for the Initial Admission Assessment, the Medicare 14-day

assessment may be performed using any day 11 through 14 of the stay as

the assessment reference date (MDS item A3a) and, in addition, the SNF

may use the five available grace days (through day 19), if necessary.

The Medicare 14-day assessment must then be completed (dated at item

R2b) 14 days after the assessment reference date, locked in seven days,

and so forth. Keep in mind that there are no grace days for completion

of the Initial Admission Assessment. As always, the Initial Admission

Assessment must be completed by day 14. Another factor to consider in

timing completion and locking of assessments is that bills may only be

sent for assessments that have been locked.

3. Discharge and Leave of Absence

Comment: One commenter asked for a definition of ``leave of

absence'' as distinguished from a ``discharge.''

Response: Although this is not a distinction that is specific to

the PPS, we would like to define these terms in the context of

clarifying another somewhat misunderstood aspect of Medicare coverage,

the so-called ``midnight rule'' and the clinical requirements for

Discharge forms and Re-Entry Tracking forms. We received questions from

other commenters on how to handle cases in which the beneficiary is out

of the facility at the time of census-taking, midnight. These

activities are all interrelated and have generated many questions

during the initial phase of PPS implementation. There are a number of

reasons why a beneficiary may leave the SNF for a ``leave of absence.''

These include a temporary home visit, a temporary therapeutic leave, or

a hospital observational stay of less than 24 hours in which the

beneficiary is not formally admitted to the hospital and is not

discharged from the SNF. In each of these situations, there is no

requirement for the SNF to complete a Discharge or a Re-Entry Tracking

form.

When a beneficiary goes to an acute care hospital emergency room

(ER) during his or her SNF stay and is in the ER at midnight, there is

an additional aspect with regard to Medicare payment. According to

Medicare rules, the day preceding the midnight on which the beneficiary

was absent from the facility becomes a day for which the SNF may not

bill Part A of Medicare. This is known as the ``midnight rule.''

However, for clinical purposes, as long as the beneficiary returns to

the facility in less than 24 hours, was not admitted to the hospital,

and was not discharged from the SNF, this time in the ER is considered

a ``leave of absence'' and requires no discharge form.

Likewise, from the perspective of Medicare payment under PPS, there

is no requirement for any additional assessment. The day preceding the

midnight is not a covered Part A day and, therefore, the Medicare

assessment ``clock'' is altered by skipping that day in calculating

when the next Medicare assessment is due. From a clinical standpoint,

the leave of absence does not affect the ``clock'' for the clinical

assessments.

For example, if the beneficiary is due for his 30-day assessment on

March 30 (day 30 of his Medicare covered stay), but he spends midnight

of March 27 in the ER, day 30 of his Medicare Part A covered stay now

falls on March 31, as March 27 does not count as one of the

beneficiary's 100 days of Medicare SNF care. In other words, the count

of days in the Medicare covered stay changes when there is a noncovered

day because the facility cannot count that day as one of the

beneficiary's benefit days. Given the flexibility of the assessment

windows for the Medicare assessments, altering the count of days as

described here should have no more than a negligible effect on

assessment scheduling for facilities.

Of course, a beneficiary who is required to be in the ER at

midnight may well have experienced a significant change in clinical

status. In that case, the facility must comply with the clinical

requirement to complete a Significant Change in Status Assessment when

the beneficiary returns to the SNF. The Medicare payment requirements

and the midnight rule have no bearing on this requirement for

completion of a Significant Change in Status Assessment.

Alternatively, if the beneficiary is in the ER for more than 24

hours, or is actually admitted to the hospital or discharged from the

SNF, a Discharge Tracking form is required. In addition, when the

beneficiary returns to the SNF, a Re-Entry Tracking form is required,

and a Return/Readmission Assessment (MDS 2.0 item A8b=5) must be

performed to restart the Medicare assessment schedule. The Return/

[[Page 41659]]

Readmission Assessment fulfills the requirement for a Medicare 5-day

assessment in this situation, and the next required assessment would be

the Medicare 14-day assessment.

Finally, with regard to MDS scheduling requirements, we are taking

this opportunity to clarify the regulations text at Sec. 413.343(b),

which specifies the assessment schedule required under the SNF PPS. The

current language requires the performance of such assessments on the

5th, 14th, 30th, 60th, and 90th days ``following admission.'' However,

as indicated in the preceding discussion, it is not the admission date

per se that determines the start of the Medicare assessment schedule,

but rather, the commencement of Medicare-covered care in the SNF.

Although Medicare-covered posthospital SNF care often does begin

immediately upon a beneficiary's admission to the SNF, the existing

language fails to address those situations in which such care does not

commence until sometime after the day of admission. The Medicare

required assessment schedule is based only on those days in the

Medicare Part A covered stay and, thus, cannot be scheduled based on

the day of admission per se. Therefore, we are revising the language in

the regulations text to take into account the possibility that a

beneficiary's ``posthospital SNF care'' (that is, SNF care that is

covered under Medicare Part A) may begin subsequent to the day of his

or her actual admission to the facility. The Medicare required

assessments are to be performed so that, using the first day of

posthospital SNF care as day 1, there is a full MDS assessment on the

5th day, the 14th day, the 30th day, the 60th day and the 90th day of

the SNF stay.

H. Other Medicare MDS Requirements

In the interim final rule, we stated that collection of medication

information using a revised version of section U of the MDS would be

required under PPS, beginning October 1, 1999. The criteria we

established for this process anticipated that a refined section U would

be developed to facilitate streamlined data collection, maximize data

accuracy, and minimize burden to facilities. We have, to date, made

considerable progress in our work on the section U refinements.

However, due to systems constraints resulting from the need to achieve

Year 2000 (Y2K) compliance (see the further discussion of the Y2K issue

below in the context of the partial delay in SNF consolidated billing

implementation), we will not be able to implement the refined version

of section U until after the first months of the year 2000 have passed.

Therefore, we have determined that the most straightforward and least

burdensome approach is to defer section U implementation until October

1, 2000.

I. Medical Review

Comment: We received several comments requesting that we publish

the medical review criteria to be used now that PPS is in place. Also,

there were requests that we institute consistent medical review

policies across FIs.

Response: We are currently formalizing the medical review criteria

that will be used in the review of SNF PPS bills. Certainly, one of the

primary goals of the new policy is to provide reviewers with guidelines

that will facilitate consistent national medical review policy, one of

the initial goals of implementing the PPS. We recently published a PM

(PM transmittal No. A-99-20, May 1999) to instruct medical reviewers in

the new process. One aspect of the reviews of SNF PPS bills to be

performed by the FIs focuses on the MDS information and its consistency

with the documentation in the rest of the medical record. In addition,

the review process focuses on identification of instances in which

inappropriate services were provided or in which the beneficiary did

not meet the requirements for Medicare Part A coverage in an SNF.

Comment: There were questions about how the MDS information might

be matched to claims data to facilitate monitoring or auditing of SNF

reporting practices.

Response: The process for matching the bill to the MDS takes place

at HCFA. We use the bill data forwarded to us by the FIs to match to

the appropriate MDS from the HCFA MDS Repository. From these matched or

unmatched files, we generate various reports for use by HCFA and the

FIs in their audit functions.

Comment: We received a comment requesting that we instruct FIs to

give demand bills a high priority within the review process and to

process these submissions no later than 30 days from the date of the

request.

Response: The policy governing how demand bills will be processed

under the SNF PPS will be determined by considering the FIs' overall

workloads, of which the SNF PPS represents only a small portion.

Comment: A commenter requested that we generate and disseminate to

the nursing home industry and to the payers, the full process of

transmission of clinical Medicare Part A information and claims

submission requirements, including documentation requirements needed by

the fiscal intermediary for late assessment reference dates.

Response: The requirements for the transmission of all MDS

assessments can be found in the Federal Register published on December

23, 1997 (62 FR 67174). There are no separate requirements for Medicare

Part A information. The facility must submit the MDS to the State in

which it operates and the State transmits it to us. In contrast, the

SNF submits claims to the FI, as they did before PPS. Each claim is

transmitted to us by the FI after it has been paid, and we match the

claim to the appropriate MDS. The FI may request any information it

deems to be necessary to verify the level of services billed by the

facility.

Comment: We received one comment suggesting that we should exempt

from post-payment review or on-site audit, any 5-day assessment with an

assessment reference date on one of the grace days that results in the

beneficiary's classification into a Low Rehabilitation group.

Response: This comment reflects a misunderstanding of our policy

regarding grace days. As explained above in this final rule, the grace

days are available for use, without penalty. The reference to audits in

the interim final rule was not intended to preclude any appropriate use

of the grace days. Therefore, although the comment indicates that

beneficiaries who classify into one of the low rehabilitation groups

should be exempt from review (presumably because of the requirement for

six days of nursing rehabilitation services in order to qualify for

this RUG-III group), there is no reason for us to consider excluding

any type of Medicare SNF claims from post-payment review.

Comment: Several commenters cited the BBA mandate that we must

implement a quality monitoring system. Section 4432(c) of the BBA

requires the Secretary to establish a medical review process to examine

the effects of the SNF and PPS related provisions on the quality of SNF

services furnished to Medicare beneficiaries, with particular emphasis

on the quality of non-routine covered services and Medicare-covered

physician services.

Response: The quality of care provided to beneficiaries is

paramount in our view. We will use our existing survey and enforcement

activities (along with the new techniques and data that are now

becoming available with the advent of prospective payment) to ensure

the quality of SNF services provided to Medicare beneficiaries.

[[Page 41660]]

In addition to the more traditional medical review process we are

establishing, as described above, we have also begun work toward the

establishment of a quality medical review process that is specifically

designed to fulfill the BBA mandate. We have developed an SNF PPS

Quality Medical Review Pilot project that uses MDS and other data to

monitor and target quality and program integrity problems. This

monitoring will be accomplished by testing a more integrated and

cooperative approach to medical review of SNF services using several

pilot states to partner Peer Review Organizations (PROs), FIs, State

Survey Agencies, and Medicaid agencies to assess, monitor, and improve

the quality of Medicare SNF services under the PPS.

We are implementing a two-tier strategy using the PRO Special

Project process. This strategy is expected to strengthen program

integrity and quality review in SNFs, promote SNF quality improvement,

deter fraud and abuse, and enhance beneficiary protection. The first

tier is a statistical analysis PRO (StatPRO), that is testing a data

driven approach which analyzes MDS data to flag potential quality of

care and program integrity problems. The MDS data set will be linked

with other HCFA data sets (such as, Medicare Part A and B claims,

OSCAR-Online Survey Certification and Reporting System, HCIS-HCFA

Customer Information System, FI payment, and program integrity data) to

identify patterns and trends in care. The second tier of the project

pilot tests a data based approach using StatPRO and other data to

examine State trends and variations in SNF data and patient care

through the collaboration of quality medical review (QMR) teams

composed of the PRO, FI, and State survey agency in two States (NC and

CO) and in three States (AZ, MA, and MD) the Medicaid Agency is added.

The QMR pilots will field test an integrated model where they will work

together to better understand each other's program integrity and

quality review roles, develop collaborative approaches within their

regulatory authority, test a targeted clinical data driven intervention

strategy, target beneficiary protection, and deterrence of fraud and

abuse. Finally, we will use the vast data resources available from the

national MDS data repository to support our quality initiatives.

J. Rehabilitation Therapy Services and PPS

Comment: Many commenters questioned when rehabilitation therapy may

begin in the SNF stay.

Response: Although rehabilitation therapy may begin as early as day

one of the Medicare Part A SNF stay, we note that all of the

rehabilitation therapy services (PT, OT, and SLP) must meet each of the

following criteria in order to be coded in the MDS as minutes of

rehabilitation therapy:

The service must be ordered by a physician.

The therapy intervention must relate directly and

specifically to an active written treatment regimen established by the

physician after any needed consultation with the qualified

rehabilitation therapy professional and must be reasonable and

necessary to the treatment of the beneficiary's illness or injury

(section 230 of the Medicare Skilled Nursing Facility Manual, HCFA Pub.

12).

An appropriately licensed or certified individual must

provide or directly supervise the therapeutic service and coordinate

the intervention with nursing services.

Even though these three criteria are not new with PPS, the

establishment of a new payment system has heightened interest in

understanding and satisfying these standards. For instance, in addition

to the commenters' question about when rehabilitation therapy services

can begin, we have received many questions during the first year of PPS

implementation regarding standards for supervision of rehabilitation

therapy assistants and aides, and many questions regarding the

physician signature requirements for the rehabilitation therapy plan of

treatment. Accordingly, we will take this opportunity to provide

further clarification of those issues. The rehabilitation therapy

service must be ordered by a physician. The Medicare policy regarding

the requirement for the physician signature on the therapy plan of

treatment has not changed. As is stated in the SNF Manual,

rehabilitation therapy services provided to a beneficiary in a SNF must

be directly and specifically related to an active written treatment

plan established by the physician after any needed consultation with a

qualified therapist. Implementation of the PPS did nothing to alter

this guideline. We will, however, take this opportunity to clarify what

is required for coverage of rehabilitation therapy.

As stated in the language in the SNF Manual cited in the preceding

paragraph, Medicare requires the physician to make decisions regarding

the amount and intensity of rehabilitation therapy services provided to

Medicare beneficiaries in SNFs after consulting with the professional

therapist. This requirement is based on our commitment to ensuring

quality care for Medicare beneficiaries, and also reflects the

requirements for participation (at section 1819(b)(6)(A) of the Act),

which specify that the medical care of every SNF resident must be

provided under the supervision of a physician. Our policy has not

changed, and we are taking this opportunity to clarify that policy. The

physician's responsibility in the development of a rehabilitation

therapy plan of treatment ensures that the services to be provided will

not exceed the beneficiary's abilities as constrained by his clinical

status. In addition, we believe that the physician's clinical judgement

is an important aspect in preventing injuries that can result from the

provision of inappropriate rehabilitation therapy. For example, the

rehabilitation plan of treatment for a beneficiary with a hip fracture

should be developed with an awareness of his or her limitations due to

severe osteoporosis and emphysema. Unless the beneficiary's entire

clinical condition is taken into account, there is a significant risk

of injury and of a compromised medical status.

We expect that the same care will be taken by the physician and SNF

staff to document physician responsibility for developing the therapy

plan of treatment, including precautions, that is reasonably expected

to be taken for any other element of the medical record. We realize,

however, that in the SNF setting there may not be a physician on the

premises every day. Therefore, Medicare allows the professional

therapist to develop a suggested plan of treatment and to begin

providing services based on that plan prior to obtaining the

physician's signature on the plan. We continue to require that the plan

of treatment must be a physician's responsibility after any needed

consultation with a qualified therapist, and that the requirement for

physician verification of the suggested plan of treatment will be

obtained within a reasonable amount of time. However, a physician

signature must be obtained before the facility bills Medicare for

payment for the rehabilitation therapy services provided to the

beneficiary based on the plan of treatment he or she has approved. In

this way, the facility can be sure that the level of therapy for which

it bills Medicare is the level the physician deems to be medically

necessary. We expect that the type and intensity of therapy billed will

always match the type and intensity of therapy on the signed therapy

plan of treatment.

We understand that many physicians use the fax to participate

actively in the review of written plans of care and so

[[Page 41661]]

believe that it is appropriate to accept physicians' faxed signatures

for the plan of treatment. As always, whenever the plan of treatment is

altered in any way, the modification must be made in writing. If the

physician is not the person making the modification, the therapist who

is making the change must notify the physician timely, and the

physician must sign the change within a reasonable amount of time.

In addition to the issues discussed above, we would like to clarify

the requirements for the rehabilitation therapist's initial evaluation

of a Medicare beneficiary in a SNF stay and the requirements for

licensed therapist supervision of therapy assistants and therapy aides

when they provide therapy services to Medicare beneficiaries. The

initial evaluation, performed by the licensed therapist and necessary

for the development of the plan of treatment, must be performed during

the beneficiary's SNF stay. It is not acceptable to use an evaluation

that was performed for instance, in the acute care hospital or the

rehabilitation hospital setting as the evaluation of the beneficiary in

the SNF, because the beneficiary's status must be evaluated as he or

she presents in the SNF setting. The evaluation, and the resultant plan

of treatment, developed in the acute care hospital or rehabilitation

hospital is relevant to the specific type of setting and is not

interchangeable with an evaluation and plan of treatment developed for

the beneficiary in the SNF setting. The time that it takes for the

therapist to perform this evaluation may not be recorded as minutes of

therapy received by the beneficiary.

An appropriately licensed or certified individual must provide or

supervise the therapeutic service and coordinate the intervention with

nursing services. As stated above, Medicare expects that services will

be provided by, or supervised by, appropriately licensed or certified

professionals.

Physical and occupational therapy assistants may provide

rehabilitation therapy services under the supervision of the

professional therapist. A rehabilitation therapy assistant must be

under the general supervision of a professional therapist who is

accessible while the assistant is providing services to the

beneficiary. The therapy assistant cannot supervise a therapy aide. It

is up to the professional therapist to ensure that the assistant is

capable of performing therapy services without the more stringent

``line-of-sight'' level of supervision required by therapy aides.

A therapy aide must be supervised personally by the professional

therapist in such a way that the therapist has visual contact with the

aide at all times. Therapy aides are not to perform any services

without ``line-of-sight'' supervision. Similarly, a therapy aide must

never be responsible for provision of group therapy services, as this

is well beyond the scope of services that they are qualified to

provide.

A therapy student who is participating in field experience must

also be under the ``line-of-sight'' level of supervision of the

professional therapist. Even though these students may become licensed

therapists within months of the field training portion of their school

program, they are not licensed or certified for practice in an

unsupervised status. Further, none of the minutes of therapy services

provided by the students may be recorded on the MDS as minutes of

therapy received by the beneficiary. Medicare recognizes the costs

associated with approved educational activities as a pass-through (see

Sec. 413.85).

Comment: Many commenters had questions about the correct counting

and recording in the MDS of minutes of rehabilitation therapy.

Response: Section P of the current version of the MDS contains the

items that capture the amount of time each nursing home resident spends

receiving rehabilitation therapy. Thus, it is in section P that the

clinician records the number of days and minutes of rehabilitation

therapy (PT, OT, ST) received by the individual beneficiary during the

past seven days, or since admission to the SNF, whichever is shorter.

The directions for completion of section P instruct the assessor to

look back over the ``last 7 calendar days,'' counting only post

admission days and minutes of therapy, when counting the days and

minutes of rehabilitation therapy received by the beneficiary. The

number of minutes recorded here must be the actual number received by

the beneficiary. Seven calendar days are, by definition, consecutive

days.

In the case of a Medicare 5-day assessment, however, the nurse

assessor will choose as the assessment reference date (MDS item A3a),

any day one through eight of the covered stay, and will look back over

the prior seven calendar days (or over the days since admission if

there are fewer than seven days since admission) to count the number of

days upon which more than 15 minutes of therapy were received and the

number of minutes that were received by the beneficiary during those

days. It is irrelevant if there is a break in therapy (for example, for

a weekend or holiday) during that time. For example, if day five of the

stay is chosen as the assessment reference date, the assessor would

look back to admission to count the patient's PT, OT, and ST time. If

the beneficiary received PT for 50 minutes on both the second and fifth

days of the Part A covered stay, that would be recorded as two days of

PT and 100 total minutes of PT. The actual number (not rounded) of

minutes must be recorded on the MDS. Minutes cannot be rounded to

multiples of 10 or 15.

The rehabilitation therapy time reported on the MDS is a record of

the time the beneficiary spent receiving therapy services, not a record

of the therapist's time. As stated in the August 1996 publication, Long

Term Care Resident Assessment Instrument Questions and Answers, Version

2.0, the beneficiary's ``therapy time starts when he begins the first

treatment activity or task and ends when he finishes with the last

apparatus and the treatment is ended.''

Set-up time is included, as is time under the therapist's or

therapy assistant's direct supervision. PT, OT, and ST provided outside

the building may be counted and recorded on the MDS, as long as the

staff who provide therapy are qualified to provide the service. In the

State Operations Manual (SOM, HCFA Pub. 7) Transmittal #272, pp. R64,

``The therapy treatment may occur inside or outside the facility.''

This includes the time it takes for the therapist to take the

beneficiary to his or her home for a home visit before discharge as

long as the therapist uses the time in the car to teach or discuss the

beneficiary's treatment or treatment goals, and for family conferences

when the beneficiary is also present.

Whether the time spent evaluating the beneficiary is counted

depends on whether it is the formal initial evaluation or an evaluation

performed after the course of therapy has begun. The time it takes to

perform the formal initial evaluation and develop the treatment goals

and the plan of treatment may not be counted as minutes of therapy

received by the beneficiary. However, a reevaluation--that is, a hands-

on examination of the beneficiary and not simply an update to the

documentation and revision of the care plan--that is performed once a

therapy regimen is underway (for example, evaluating goal achievement

as part of the therapy session) may be counted as minutes of therapy

received.

This policy was established because we do not wish to provide an

incentive for facilities to perform initial evaluations for therapy

services for patients who have no need of those specialized services.

However, we

[[Page 41662]]

believe that the initial evaluation is an appropriate cost of doing

business. Therefore, the cost of the initial assessment is included in

the payment rates for all Medicare beneficiaries in covered Part A SNF

stays.

For beneficiaries who do not classify into one of the

Rehabilitation RUG-III groups, the therapy non-case-mix component is

part of the daily rate. The amount, $0.91, is reflected in the rate for

all of the non-therapy RUG-III groups.

The Long Term Care Resident Assessment Instrument Questions and

Answers Version 2.0, clarifies how to account for therapy provided to

an individual within a group setting. It states that if the group has

four or fewer participants per supervising therapist (or therapy

assistant under general supervision by the therapist) then it is

appropriate to report the full time as therapy for each patient. The

example used is that of a therapist working with three patients for 45

minutes on training to return to the community. Each patient's MDS

would reflect receipt of 45 minutes of therapy for this session.

Although we recognize that receiving PT, OT, or ST as part of a

group has clinical merit in select situations, we do not believe that

services received within a group setting should account for more than

25 percent of the Medicare resident's therapy regimen during the SNF

stay. For this reason, no more than 25 percent of the minutes reported

in the MDS may be provided within a group setting. This limit is to be

applied for each therapy discipline; that is, only 25 percent of the PT

minutes reported in the MDS may be minutes received in a group setting

and, similarly, only 25 percent of the OT, or the ST minutes reported

may be minutes received in a group setting.

To summarize, the minutes of therapy provided by at least one

supervising therapist (or therapy assistant under general supervision

by the therapist) within a group of four or fewer participants, may be

fully counted, provided that those minutes account for no more than 25

percent of the resident's weekly therapy in that discipline, as

reported in the MDS. The supervising therapist may not be supervising

any individuals other than the four or fewer individuals who are in the

group at the time of the therapy session. Naturally, provision of group

therapy time in excess of the 25 percent threshold is allowable, but

those minutes may not be counted in section P of the MDS for purposes

of RUG-III classification for Medicare Part A beneficiaries.

Under section 1814(a)(2)(B) of the Act, a covered SNF level of care

is defined in terms of those services that necessitate the involvement

of skilled personnel, are needed and received on a daily basis and, as

a practical matter, can be provided only in an SNF on an inpatient

basis. Additionally, the requirements for participation at section

1819(b)(4)(A) of the Act require an SNF to furnish the full range of

nursing and specialized rehabilitative services needed to attain or

maintain each resident's highest practicable state of well-being, in

accordance with the comprehensive plan of care. This means that there

are to be no limits placed on the services to be provided to the

beneficiary due to the facilities interpretations of how many minutes

are ``allowed'' by the given RUG-III group.

The RUG-III classification system uses minimum levels of minutes

per week as qualifiers for classification into the rehabilitation

therapy groups. These minutes are minimums and are not to be used as

upper limits for service provision. Similarly, there are instances in

which beneficiaries in the so-called ``clinical categories,'' Extensive

Services, Special Care and Clinically Complex, will need some limited

amounts of rehabilitation therapy services, which they should receive,

even though they may not require a level that would qualify them for

one of the rehabilitation groups. The SNF PPS is based on averages, and

a facility that continues to provide services as they are needed by its

beneficiaries should receive payments that, in the aggregate, are

adequate to pay for those services. Any policy of holding therapy to

the bare minimum, regardless of beneficiary need, is inconsistent with

the statutory requirements discussed above, and will result in poor

outcomes, longer lengths of stay, and a degradation in the facility's

quality of care.

Section T of the current version of the MDS must be included with

each Medicare PPS assessment, but in the case of a Medicare five day

assessment, the clinician captures minutes of therapy that are

anticipated for the beneficiary during the first two weeks of the

nursing home stay. This makes it possible for the beneficiary to

classify into the appropriate RUG-III rehabilitation group based on the

anticipated receipt of rehabilitation therapy, even though the

assessment is done during the first few days of the SNF stay.

Section T of the current version of the MDS contains three items,

T1b, T1c and T1d, in which the assessor is to record ``ordered

therapies.'' The T1b item asks, ``Has physician ordered any of the

following therapies to begin in FIRST 14 days of stay--PT, OT, or

speech pathology service?'' If the answer to this question is yes, then

the number of expected minutes and days is completed in items T1c and

T1d of the current version of the MDS. If the answer is no, then there

is nothing to report in T1c or T1d.

If the physician orders therapy for 10 days, the projected number

of days in section T will be 10 rather than 14; likewise, if the

physician does not order a limited number of days, the projection will

be based on the entire two weeks, assuming the beneficiary's continued

stay and receipt of services.

The RUG-III grouper takes into consideration both the days and

minutes already received by the beneficiary, as reported in section P

of the current version of the MDS, and the days and minutes expected to

be received in the first two weeks of the stay. The number of days and

minutes expected, as reported in section T, should include those

already received.

For example, the beneficiary received an hour of OT on both the

fourth and fifth days (a Monday and Tuesday) of the SNF stay. The

prescribed regimen calls for the beneficiary to receive an hour of OT

daily, Monday through Friday, during the first two weeks in the SNF.

The assessment reference date was set for the fifth day of the stay;

two days and 120 minutes were reported as having been received in

section P of the MDS, and 10 days and 600 minutes were reported as

anticipated in section T. The 10 days and 600 minutes recorded in

section T include the 2 days and 120 minutes already received, in

addition to the upcoming three days and 180 minutes expected to be

received in the first week, and the five days and 300 minutes of

therapy in the second week.

We realize that reporting therapy time that has not yet been

provided is a significant change for providers, but it is in compliance

with the grouper logic and allows the facility to provide the most

accurate representation of the services to be provided to the

beneficiary during the first assessment period.

K. RUG-III Groups

Comment: We received a few comments stating that the ``limits'' on

therapy minutes imposed by the RUG-III groups were too low, and that

more than 720 minutes should be allowed for beneficiaries in the

highest RUG-III groups.

Response: The RUG-III system does not impose limits on the services

a resident may receive; rather, it is used to determine how much

Medicare pays for the services that the resident

[[Page 41663]]

receives. The minutes used to classify beneficiaries into RUG-III

groups are in no way to be taken as upper limits. The 720-minute

threshold for the Ultra High sub-category is a minimum for purposes of

classifying residents. In fact, during the demonstration, there were

beneficiaries who were receiving more that 1,000 minutes per week, and

we expect that there will be similar instances during the national

implementation. All of the groups were created based on a continuum of

minutes being provided, including Ultra High. Just as we expect to see

beneficiaries in the High Rehabilitation sub-category receiving 450

minutes per week, we expect that as many minutes as are needed will be

provided to beneficiaries in the Ultra High groups.

Comment: We received a comment requesting that we explain how the

RUG-III grouper works. The commenter believed that we failed to explain

fully in the interim final rule the grouping logic that restricts

classification into the Rehabilitation Ultra High and Very High sub-

categories to beneficiaries who have a full week of therapy recorded in

section P of the MDS.

Response: The grouper software uses the minutes and days recorded

in sections P and T together to classify beneficiaries into the RUG-III

rehabilitation groups. However, in order for a beneficiary to classify

into the upper two sub-categories, Ultra High and Very High, he or she

must have received at least one full week (five days) of therapy at the

level that would qualify for these groups.

For example, suppose a beneficiary is admitted on Monday, May 1 and

begins PT and OT on May 4. The beneficiary receives 90 minutes of PT

and 60 minutes of OT on the 4th, 5th, 6th, 7th, and 8th of May. The

assessment reference date for the Medicare 5-day assessment is Monday,

May 8. The beneficiary will classify into the Ultra High sub-category

based on having received more than 720 minutes of therapy across at

least two disciplines during the past seven days, as recorded in

section P of the MDS. If, on the other hand, the beneficiary received

this level of therapy on only four of the first seven days in the SNF,

he or she would classify into the High Rehabilitation sub-category

since this is the highest level of classification that is possible when

a minimum of 500 minutes and five days of therapy have not been

provided.

We have posted a tool on our web site that allows the user to

follow the grouper logic manually. It walks through each step of the

grouping logic and we believe it is a useful learning tool. The website

address is:

www.hcfa.gov/medicare/hsqb/mds20/>.

Comment: There were a few comments regarding the use of the

combination of physician visits and order changes to qualify

beneficiaries for the Clinically Complex RUG-III category. One of the

commenters argued that these criteria are unacceptable because an SNF

represents not a medical model but rather a nursing model; as such, the

physician's involvement and participation may be limited, or may result

from consultation sought by the facility's nursing staff due to changes

in a resident's condition or the need for specific services. Another

commenter inquired about the specific definition being used to define a

``physician order change.''

Response: These comments are representative of concerns that have

been expressed during the initial implementation of the SNF PPS. While

we are aware that many facilities operate using a nursing model as

opposed to a medical model of care delivery, the commenter's further

observation on why the nursing staff would consult with the physician

provides the explanation for why physician order changes and visits are

qualifiers for the RUG-III groups.

The RUG-III system uses clinical events, conditions, and services

as indicators of severity. The results of the research that is the

basis of the RUG-III system showed that an increased frequency of

physician visits and order changes are indicators of a beneficiary's

clinical instability. As in the commenter's example, the nursing staff

may consult with the physician due to changes in the beneficiary's

condition that require medical intervention or the need for specific

services that require a physician order.

We would also like to make clear what constitutes an order change.

The specific issues that have been raised include whether an order to

continue a specified treatment is a new order and, therefore, counts as

an order change; whether a sliding scale medication order counts as a

new order every time the clinician administers one of the different

dosages specified in the scale; whether orders written to clarify a

previous order count; and, whether all doctor's visits count in the

number of physician visits item.

A physician's order to continue or renew some specified treatment

or regimen would not be considered to be an order change, nor would an

order written solely to clarify an earlier order. As stated in the Long

Term Care RAI User's Manual, the definition of an order change does not

include admission orders, return admission orders, or renewal orders

without changes. Similarly, a sliding scale dosage schedule that is

written to cover different dosages depending on lab values, does not

count as an order change simply because a different dose is

administered based on the sliding scale guidelines. ``Physician

visits'' are also defined in the Long Term Care RAI User's Manual. The

physician is defined to include an ``MD, osteopath, podiatrist, or

dentist who is either the primary physician or consultant. Also include

an authorized physician assistant or nurse practitioner working in

collaboration with the physician.'' The visit is defined as a partial

or full exam at the facility or in the physician's office.

L. Nurse Staffing and the Staff Time Measurement Studies

Comment: We received a variety of questions related to the staff

time measurement (STM) studies performed in 1995 and 1997 that were

used to set the case-mix indices. These included questions about what

portion of the nurses' time was accounted for in the study, whether all

nursing minutes (resident specific and non-specific) were used, whether

medication aide time was counted, and what nurse staffing mix was used.

Also, the suggestion was made that we should conduct another STM study

after the PPS has been in place for a year.

Response: Before addressing the specific comments, we are taking

this opportunity to provide a brief background explanation of the STM

studies. As stated in the interim final rule, we conducted the STM

studies in 12 States across 154 SNFs and 3,900 residents. The 1997 STM

was performed to supplement the 1995 study to secure additional STM

data from SNFs identified as providing both high quality care and more

than an average level of rehabilitation therapy to patients on their

Medicare-certified nursing units, and to include a broad geographic

distribution of providers.

The STM data collection accounted for all nursing staff time during

the 48-hour collection period. This time included that of the

registered nurses (RNs), licensed practical nurses (LPNs), aides

(certified nursing assistants (CNAs)), and medication aides. The

resident-specific component counted all nursing time of 30 seconds or

more spent in an activity directly attributable to a specific resident.

The non-resident specific component included all time

[[Page 41664]]

not directly related to a specific resident, such as meetings, nursing

unit administration, and staff meal times. Also, if the nursing staff

member worked past the end of the shift, that time was counted as well.

The therapy staff time was collected over a 7-day period. All time

that the therapist, therapy assistant, and therapy aides spent working

in the certified nursing unit was accounted for and was apportioned

between resident specific and non-resident specific, following the same

methodology as was used in the nursing time allocation. All of these

collected time data were used in the development of the indices.

The staffing levels and the nurse staffing mix on the units

selected for the study met the OBRA 87 staffing requirements and

provided more than 110 minutes of daily resident specific nurse staff

time. Both freestanding and hospital-based facilities were used in the

study. Salaries were adjusted based on the American Health Care

Association's 1995 study of national nursing home salaries.

The nurse staffing mix found on the certified nursing units in the

study were determined per unit, based on the mix of residents on the

nursing unit at the time of the data collection. Based on a case-mix of

0.92, the average time across the two staff time studies was: 1.2 hours

per resident, per day of RN time; 0.7 hours per resident, per day of

LPN time; 2.6 hours per resident, per day of CNA time (including

medication aides). This adds up to 4.6 hours per resident, per day of

nursing time.

An important point to understand about the nursing time is how it

affects the rates. The nursing time associated with any one group in

the RUG-III hierarchy does not represent the nursing minutes that must

be provided (and that will be paid for by Medicare) to each of those

beneficiaries. Rather, the minutes are a distributional value--an

average for the RUG-III group--and were an important factor in the

development of the case-mix indices. The weight for each of the 44 RUG-

III groups represents the average resources (including, of course,

nursing) required to care for beneficiaries who classify to that group

relative to the average resources required to care for beneficiaries in

all of the other RUG-III groups. The RUG-III group with a value of 1.0

is identified and the weights for the other groups are calculated once

that has been done. The value of 1.0 indicates that the average

resources required to care for beneficiaries in that group are the

average compared to all of the other groups. Accordingly, the resource

requirements to care for beneficiaries in the other 43 RUG-III groups

are either higher or lower than for the group with the weight of 1.0.

Depending on the distribution of beneficiaries across the RUG-III

groups, the group with the relative value of 1.0 will vary. It is

important to bear this concept in mind, in order to avoid the

misconception that a RUG-III group with a relative weight that changes

from one year to the next has staffing requirements that have changed

from the original staff time measurement study. The RUG-III system does

not impose any new staffing requirements. The data are available from

the HCFA PPS website address:

www.hcfa.gov/medicare/snfpps.htm.>

Comment: One commenter requested that we explain why the Behavior

Category is so low in the RUG-III hierarchy, even though beneficiaries

who classify into that group require intensive amounts of staffing

resources.

Response: The reason for this is that the RUG-III hierarchy is in

large part based on minutes of licensed nursing time and on the

clinical conditions that require the attention of licensed staff. This

is a result of early research findings that indicated that

beneficiaries who have the clinical characteristics that would classify

them into the medically complex categories, like Extensive Care,

generally require much more RN and LPN time than do beneficiaries who

classify into the groups in the lower end of the clinical scale.

Similarly, beneficiaries who classify into groups lower in the

hierarchy generally require less licensed nurse time but, as stated by

the commenter, may require intensive amounts of staffing resources.

Beneficiaries in the Behavior Category may not need much licensed

nurse time, but instead may require a large amount of certified nurse

assistant (CNA) time. Much of the care required in these lower-end RUG-

III groups is of the type provided by CNAs, such as assistance with

activities of daily living (ADLs) and other types of maintenance care.

In general, the need for CNA time is reflected in the beneficiaries'

ADL sum scores, whereas the need for licensed nurse time is predicted

by clinical complexity as reflected by the level in the hierarchy.

Thus, beneficiaries who classify into the Extensive Services category

where the ADL sum score is at least 7, have highly complex clinical

needs and require high levels of both licensed nurse (RN/LPN) and CNA

time. Beneficiaries in the lower-weighted RUG-III groups may also

require skilled nursing care, but generally not as much as required in

the higher groups.

Comment: Several commenters had concerns that the STM was collected

over too short a period, that too few facilities were used and that not

enough of them were hospital-based facilities, and that too few of the

facilities used were located in metropolitan statistical areas (MSAs)

with populations in excess of 500,000 people.

Response: We are confident that the methodology used for the STM

studies was valid and appropriate for the task. Three STM studies were

conducted. The first was in 1990, followed by another in 1995, and the

last in 1997. The staff time studies were conducted in 13 States, in

units of more than 300 nursing homes, representing care provided to

about 12,000 residents. The States included were California, Colorado,

Florida, Kansas, Maine, Maryland, Mississippi, Nebraska, New York,

Ohio, South Dakota, Texas, and Washington. These States are

geographically representative and include rural areas, as well as MSAs

with populations in excess of 500,000.

Within each of these States, the selection of SNFs was guided by

the research design that called for a sample that would adequately

represent units that provide high quality, high-acuity care. The

facilities in the combined 1995 and 1997 study sample were 55 percent

for-profit facilities, 45 percent non-profit, 22 percent hospital-based

facilities; 36 percent of the facilities had a head trauma unit, a

ventilator unit, a special rehabilitation therapy unit, or a dialysis

unit, or had been recommended as a high intensity unit by the Technical

Expert Panel. Although the amount of time spent collecting data on any

particular unit was short, the studies were conducted during different

years and each year's study was performed over a period of months. In

this way, the study was reflective of practice in the facilities in the

aggregate, if not precisely representative of any particular facility

over time.

Comment: One commenter argued that the staffing patterns (using

1995 as the base year) used in developing the rate structure lock SNFs

into historic staffing patterns. Another commenter asserted that since

there is no language in the regulation requiring SNFs to use a certain

proportion of the rate on direct care services, they will not do so,

and suggested that we adopt the staffing recommendations of the

Institute for Geriatric Nursing of the John A. Hartford Foundation.

Response: As indicated earlier, the RUG-III system does not impose

staffing requirements. We do not believe that our use of the 1995 base

year locks facilities into any particular level of nurse staffing,

either directly or indirectly. As stated above, the staffing

[[Page 41665]]

levels in the staff time studies were based on the unit's case-mix at

the time. The study used units that had a mix of payor types.

Nationally, of the 14,000 Medicare certified facilities, fewer than

4,000 have an average daily census of more than 10 Medicare

beneficiaries. We do not believe that it is appropriate to require

staffing standards based on the needs of such a small portion of the

facility's population, but this is an issue that is outside of this

regulation's scope. We will carefully review and consider the findings

of the National Academy of Sciences report (the report cited by the

commenter is part of this larger effort) when it becomes available.

Comment: One commenter suggested that we include speech language

pathology assistant times when we perform our next staff time

measurement study. The commenter recommended that we include the times

for these care providers in our update of the RUG-III case-mix indices.

Response: While we are not prepared to address the issue of when to

conduct another staff time measurement study within the context of this

final rule, we would note that services of speech language pathology

assistants are not recognized for separate coverage under Medicare. In

order for this class of providers to be eligible for Medicare payment,

they must first achieve licensure or some other standard credential

recognized at the national level. To date, these assistants have not

obtained this standing.

Comment: We received a number of comments regarding the treatment

of respiratory therapy services under the RUG-III. Several commenters

expressed concern that facilities would be using inappropriately

trained nurses rather than appropriately trained personnel to provide

respiratory therapy services.

Response: We share the commenters' concern with regard to the

quality of care. As stated in the SNF Manual at section 230.10.B.1,

Medicare requires that respiratory services must be provided by

respiratory therapists or technicians, physical therapists, nurses, or

other qualified personnel. We currently have no evidence that

unqualified personnel are administering respiratory treatments, but

note that the State surveyors monitor long-term care facilities for

such lapses in quality. The rules governing the provision of

respiratory treatment were not altered by the implementation of PPS but

certainly, in light of the PPS and its associated incentives, we are

determined to monitor closely the provision of SNF care, including

respiratory treatments. A key provision in implementing the new payment

system is to safeguard quality of care for nursing home residents, and

this issue warrants particular attention from our quality and

enforcement initiative.

Comment: Several commenters raised concerns with the manner in

which respiratory therapy is recognized in the SNF PPS.

Response: As discussed earlier in this preamble in the context of

the Federal PPS rates, the treatment of respiratory therapy services in

the SNF PPS was the result of careful consideration and extensive

analysis. The RUG-III case-mix classification system, which forms the

basis for the payment rates, does not include respiratory therapy in

the same category as the rehabilitation therapies (occupational,

speech, and physical therapy).

The primary reason for this was the difference in treatment

patterns between respiratory therapy and the rehabilitation therapies.

A secondary reason is that the costs of respiratory therapy services

are not always separately identifiable on SNF cost reports, since

trained nurses are qualified to provide these services, and often do

so. However, we note that all costs from the base year data associated

This text is long and has been trimmed here. Open the source document for the complete record.