Rule: Medicare: Home Health Prospective Payment System 2008 CY Refinement and Rate Update

Wednesday,
August 29, 2007
Part II
Department of
Health and Human
Services
Centers for Medicare & Medicaid Services
42 CFR Part 484

Medicare Program; Home Health
Prospective Payment System Refinement
and Rate Update for Calendar Year 2008;
Final Rule
mstockstill on PROD1PC66 with RULES2
VerDate Aug<31>2005 16:35 Aug 28, 2007 Jkt 211001 PO 00000 Frm 00001 Fmt 4717 Sfmt 4717 E:\FR\FM\29AUR2.SGM 29AUR2
49762 Federal Register / Vol. 72, No. 167 / Wednesday, August 29, 2007 / Rules and Regulations
DEPARTMENT OF HEALTH AND comments on CMS regulations with an Sharon Ventura, (410) 786–1985 and
HUMAN SERVICES open comment period.’’ (Attachments Katie Lucas, (410) 786–7723 (for general
should be in Microsoft Word, issues). Kathy Walch, (410) 786–7970
Centers for Medicare & Medicaid WordPerfect, or Excel; however, we (for clinical OASIS issues). Doug Brown,
Services prefer Microsoft Word.) (410) 786–0028 (for quality issues).
2. By regular mail. You may mail Mollie Knight, (410) 786–7948; and
42 CFR Part 484 written comments (one original and two Heidi Oumarou, (410) 786–7942 (for
[CMS–1541–FC] copies) to the following address ONLY: market basket issues).
Centers for Medicare & Medicaid SUPPLEMENTARY INFORMATION:
RIN 0938–AO32 Services, Department of Health and Submitting Comments: We welcome
Human Services, Attention: CMS–1541– comments from the public on the 2.71
Medicare Program; Home Health FC, P.O. Box 8012, Baltimore, MD percent reduction to the Home Health
Prospective Payment System 21244–8012. Prospective Payment System (HH PPS)
Refinement and Rate Update for Please allow sufficient time for mailed rates for 2011, as set forth in this final
Calendar Year 2008 comments to be received before the rule with comment period, to assist us
AGENCY: Centers for Medicare & close of the comment period. in fully considering this issue and
Medicaid Services (CMS), HHS. 3. By express or overnight mail. You developing policies.
ACTION: Final rule with comment period. may send written comments (one Inspection of Public Comments: All
original and two copies) to the following comments received before the close of
SUMMARY: This final rule with comment address ONLY: Centers for Medicare & the comment period will be available for
period sets forth an update to the 60-day Medicaid Services, Department of viewing by the public, including any
national episode rates and the national Health and Human Services, Attention: personally identifiable or confidential
per-visit amounts under the Medicare CMS–1541–FC, Mail Stop C4–26–05, business information that is included in
prospective payment system for home 7500 Security Boulevard, Baltimore, MD the comment. We post all comments
health services, effective on January 1, 21244–1850. received before the close of the
2008. As part of this final rule with 4. By hand or courier. If you prefer, comment period on the following Web
comment period, we are also rebasing you may deliver (by hand or courier) site as soon as possible after they have
and revising the home health market your written comments (one original been received: http://www.cms.hhs.gov/
basket to ensure it continues to and two copies) before the close of the eRulemaking. Click on the link
adequately reflect the price changes of comment period to one of the following ‘‘Electronic Comments on CMS
efficiently providing home health addresses. If you intend to deliver your Regulations’’ on that Web site to view
services. This final rule with comment comments to the Baltimore address, public comments.
period also sets forth the refinements to please call telephone number (410) 786– Comments received timely will also
the payment system. In addition, this 7195 in advance to schedule your be available for public inspection as
final rule with comment period arrival with one of our staff members. they are received, generally beginning
establishes new quality of care data Room 445–G, Hubert H. Humphrey approximately 3 weeks after publication
collection requirements. Building, 200 Independence Avenue, of a document, at the headquarters of
Finally, this final rule with comment SW., Washington, DC 20201; or 7500 the Centers for Medicare and Medicaid
period allows for further public Security Boulevard, Baltimore, MD Services, 7500 Security Boulevard,
comment on the 2.71 percent reduction 21244–1850. Baltimore, Maryland 21244, Monday
to the home health prospective payment (Because access to the interior of the through Friday of each week from 8:30
system payment rates that are scheduled HHH Building is not readily available to a.m. to 4 p.m. To schedule an
to occur in 2011, to account for changes persons without Federal Government appointment to view public comments,
in coding that were not related to an identification, commenters are phone 1–800–743–3951.
underlying change in patient health encouraged to leave their comments in Table of Contents
status (section III.B.6). the CMS drop slots located in the main
I. Background
DATES: Effective date: These regulations lobby of the building. A stamp-in clock A. Requirements of the Balanced Budget
are effective on January 1, 2008. is available for persons wishing to retain Act of 1997 for Establishing the
Comment date: We will consider a proof of filing by stamping in and Prospective Payment System for Home
public comments on the provisions in retaining an extra copy of the comments Health Services
section III.B.6 that deal with the 2.71 being filed.) B. Deficit Reduction Act of 2005
percent reduction to payment rates in Comments mailed to the addresses C. Updates to the HH PPS
2011. To be assured consideration, indicated as appropriate for hand or D. System for Payment of Home Health
comments must be received at one of courier delivery may be delayed and Services
the addresses provided below, no later received after the comment period. II. Summary of the Provisions of the CY 2008
Proposed Rule
than 5 p.m. on October 29, 2007. Submission of comments on
III. Analysis of and Response to Public
ADDRESSES: In commenting, please refer paperwork requirements. You may Comments on the CY 2008 Proposed
to file code CMS–1541–FC. Because of submit comments on this document’s Rule
staff and resource limitations, we cannot paperwork requirements by mailing A. General Comments on the CY 2008 HH
accept comments by facsimile (FAX) your comments to the addresses PPS Proposed Rule
transmission. provided at the end of the ‘‘Collection 1. Operational Issues
You may submit comments in one of of Information Requirements’’ section in 2. The Schedule for Implementation of the
this document. CY 2008 Refinements
four ways (no duplicates, please):

For information on viewing public 3. Complexity of the System
1. Electronically. You may submit B. Case-Mix Model Refinements
electronic comments on specific issues comments, see the beginning of the
1. General Comments
in this regulation to http:// SUPPLEMENTARY INFORMATION section.
2. Later Episodes
www.cms.hhs.gov/eRulemaking. Click FOR FURTHER INFORMATION CONTACT: 3. Addition of Variables
on the link ‘‘Submit electronic Randy Throndset, (410) 786–0131. 4. Addition of Therapy Thresholds
Federal Register / Vol. 72, No. 167 / Wednesday, August 29, 2007 / Rules and Regulations 49763
5. Determination of Case-Mix Weights reasonable cost basis and be initially reasonable cost-based system that was
6. Case-Mix Change Under the HH PPS based on the most recent audited cost used by Medicare for the payment of
7. Case-Mix Groups report data available to the Secretary, home health services under Part A and
8. OASIS Reporting and Coding Practices and (2) the prospective payment Part B.
C. Payment Adjustments
amounts be standardized to eliminate For a complete and full description of
1. The Partial Episode Payment (PEP)
Adjustment the effects of case-mix and wage levels the HH PPS as required by the BBA, see
2. The Low Utilization Payment among HHAs. the July 2000 HH PPS final rule.
Adjustment (LUPA) Section 1895(b)(3)(B) of the Act
B. Deficit Reduction Act of 2005
3. The Significant Change in Condition addresses the annual update to the
(SCIC) Adjustment standard prospective payment amounts On February 8, 2006, the Deficit
4. Non-Routine Medical Supplies (NRS) by the home health applicable increase Reduction Act (DRA) of 2005 (Pub. L.
D. The Outlier Policy percentage as specified in the statute. 109–171) was enacted. This legislation
E. The Update of the HH PPS Rates Section 1895(b)(4) of the Act governs affected updates to HH payment rates
1. The Home Health Market Basket Update the payment computation. Sections for calendar year (CY) 2006. The DRA
2. The Rebasing and Revising of the Home also required HHAs to submit home
Health Market Basket
1895(b)(4)(A)(i) and (b)(4)(A)(ii) of the
3. Wage Index Act require the standard prospective health care quality data and created a
4. Home Health Care Quality Improvement payment amount be adjusted for case- linkage between that data and payment
5. CY 2008 Payment Updates mix and geographic differences in wage beginning in CY 2007.
IV. Provisions of the Final Rule With levels. Section 1895(b)(4)(B) of the Act Specifically, section 5201 of the DRA
Comment Period requires the establishment of an changed the CY 2006 update from the
V. Collection of Information Requirements appropriate case-mix adjustment factor applicable home health market basket
VI. Regulatory Impact Analysis that adjusts for significant variation in percentage increase minus 0.8
A. Overall Impact costs among different units of services. percentage points to a 0 percent update.
B. Anticipated Effects
Similarly, section 1895(b)(4)(C) of the In addition, section 5201 of the DRA
C. Accounting Statement
Addendum A. CY 2008 Wage Index for Rural Act requires the establishment of wage amends section 421(a) of the Medicare
Areas by CBSA; Applicable Pre-floor and adjustment factors that reflect the Prescription Drug, Improvement, and
Pre-reclassified Hospital Wage Index relative level of wages, and wage-related Modernization Act of 2003 (MMA) (Pub.
Addendum B. CY 2008 Wage Index for Urban costs applicable to home health services L. 108–173, enacted on December 8,
Areas by CBSA; Applicable Pre-floor and furnished in a geographic area 2003). The amended section 421(a) of
Pre-reclassified Hospital Wage Index compared to the applicable national the MMA requires that for home health
Addendum C. Comparison of the CY 2007 average level. These wage-adjustment services furnished in a rural area (as
HH PPS Wage Index and the CY 2008 factors may be used by the Secretary for defined in section 1886(d)(2)(D) of the
HH PPS Wage Index
the different geographic wage levels for Act) on or after January 1, 2006 and
I. Background purposes of section 1886(d)(3)(E) of the before January 1, 2007, that the
Act. Secretary increase the payment amount
A. Requirements of the Balanced Budget Section 1895(b)(5) of the Act gives the otherwise made under section 1895 of
Act of 1997 for Establishing the Secretary the option to make additions the Act for home health services by 5
Prospective Payment System for Home or adjustments to the payment amount percent. The statute waives budget
Health Services otherwise made in the case of outliers neutrality for purposes of this increase
The Balanced Budget Act of 1997 because of unusual variations in the since it specifically states that the
(BBA) (Pub. L. 105–33) enacted on type or amount of medically necessary Secretary must not reduce the standard
August 5, 1997, significantly changed care. Total outlier payments in a given prospective payment amount (or
the way Medicare pays for Medicare fiscal year (FY) may not exceed 5 amounts) under section 1895 of the Act
home health services. Section 4603 of percent of total payments projected or applicable to home health services
the BBA governed the development of estimated. furnished during a period to offset the
the home health prospective payment In accordance with the statute, we increase in payments resulting in the
system (HH PPS). Until the published a final rule (65 FR 41128) in application of this section of the statute.
implementation of a HH PPS on October the Federal Register on July 3, 2000 to The 0 percent update to the payment
1, 2000, home health agencies (HHAs) implement the HH PPS legislation. The rates and the rural add-on provisions of
received payment under a cost-based July 2000 final rule established the DRA were implemented through
reimbursement system. requirements for the new HH PPS for Pub. 100–20, One Time Notification,
Section 4603(a) of the BBA provides home health services as required by Transmittal 211 issued on February 10,
the authority for the development of a section 4603 of the BBA, as 2006.
HH PPS for all Medicare-covered home subsequently amended by section 5101 In addition, section 5201 of the DRA
health services provided under a plan of of the Omnibus Consolidated and requires HHAs to submit data for
care that were paid on a reasonable cost Emergency Supplemental purposes of measuring health care
basis by adding section 1895 of the Appropriations Act (OCESAA) for Fiscal quality, and links the quality data
Social Security Act (the Act), entitled Year 1999, (Pub. L. 105–277), enacted submission to payment. This
‘‘Prospective Payment For Home Health on October 21, 1998; and by sections requirement is applicable for CY 2007
Services,’’ to the Act. 302, 305, and 306 of the Medicare, and each subsequent year. If an HHA
Section 1895(b)(1) of the Act requires Medicaid, and SCHIP Balanced Budget does not submit quality data, the home
the Secretary to establish a HH PPS for Refinement Act (BBRA) of 1999, (Pub. L. health market basket percentage
all costs of home health services paid 106–113), enacted on November 29, increase will be reduced 2 percentage
under Medicare. 1999. The requirements include the points.

Section 1895(b)(3)(A) of the Act implementation of a HH PPS for home
requires that (1) the computation of a health services, consolidated billing C. Updates to the HH PPS
standard prospective payment amount requirements, and a number of other As required by section 1895(b)(3)(B)
include all costs for home health related changes. The HH PPS described of the Act, we have historically updated
services covered and paid for on a in that rule replaced the retrospective the HH PPS rates annually in a separate
Federal Register document. In those in condition adjustment (SCIC market basket was primarily due to the
documents, we also incorporated the adjustment). For certain cases that increase in the benefit cost weight.
legislative changes to the system exceed a specific cost threshold, an The CY 2008 proposed rule (72 FR
required by the statute after the BBA, outlier adjustment may also be 25358) also proposed refinements to the
specifically the MMA. On November 9, available. payment system. Extensive research was
2006, we published a final rule titled conducted to investigate ways to
‘‘Medicare Program; Home Health II. Summary of the Provisions of the CY improve the performance of the case-
Prospective Payment System Rate 2008 Proposed Rule mix model. This research was the basis
Update for Calendar Year 2007 and We published a proposed rule in the for our proposals to refine the case-mix
Deficit Reduction Act of 2005 Changes Federal Register on May 4, 2007 (72 FR model. We proposed to refine the case-
to Medicare Payment for Oxygen 25356) that set forth a proposed update mix model to reflect different resource
Equipment and Capped Rental Durable to the 60-day national episode rates and costs for early home health episodes
Medical Equipment; Final Rule’’ (CMS– the national per-visit amounts under the versus later home health episodes and
1304–F) (71 FR 65884) in the Federal Medicare prospective payment system to expand the case-mix variables
Register that updated the 60-day for home health services. In accordance included in the payment model. For
national episode rates and the national with section 1895(b)(3)(B) of the Act, 2008, we proposed a 4-equation case-
per-visit amounts under the Medicare the standard prospective payment mix model that recognizes and
HH PPS for home health services for CY amounts are to be increased by a factor differentiates payment for episodes of
2007. In addition, the November 2006 equal to the applicable home health care based on whether a patient is in
final rule ended the 1-year transition market basket update for those HHAs what is considered to be an early (1st or
period that consisted of a blend of 50 that submit quality data as required by 2nd episode in a sequence of adjacent
percent of the new area labor market the Secretary. The proposed home episodes) or later (the 3rd episode and
designations’ wage index and 50 percent health market basket update for CY 2008 beyond in a sequence of adjacent
of the previous area labor market was 2.9 percent. For HHAs that fail to episodes) episode of care as well as
designations’ wage index. We also submit the required quality data, the recognizing whether a patient was a
revised the fixed dollar loss ratio, which home health market basket update high therapy (14 or more therapy visits)
is used in the calculation of outlier would be reduced by 2 percentage or low therapy (13 or fewer therapy
payments. According to section points. visits) case. We defined episodes as
5201(c)(2) of the DRA, this final rule adjacent if they were separated by no
Sections 1895(b)(4)(A)(ii) and (b)(4)(C)
also reduced, by 2 percentage points, more than a 60-day period between
of the Act require the Secretary to
the home health market basket claims. Analysis of the performance of
establish area wage adjustment factors
percentage increase to HHAs that did the case-mix model for later episodes
that reflect the relative level of wages
not submit required quality data, as revealed two important differences for
and wage-related costs applicable to the
determined by the Secretary. episodes occurring later in the home
furnishing of home health services and
health treatment compared to earlier
D. System for Payment of Home Health to provide appropriate adjustments to
episodes: higher resource use per
Services the episode payment amounts under the
episode and a different relationship
Generally, Medicare makes payment HH PPS to account for area wage
between clinical conditions and
under the HH PPS on the basis of a differences. As set forth in the July 3, resource use. We also proposed that
national standardized 60-day episode 2000 final rule (65 FR 41128), the additional variables include scores for
payment rate that is adjusted for case- statute provides that the wage certain wound and skin conditions;
mix and wage index. The national adjustment factors may be the factors more diagnosis groups such as
standardized 60-day episode payment used by the Secretary for the purposes pulmonary, cardiac, and cancer
rate includes the six home health of section 1886(d)(3)(E) of the Act for diagnoses; and certain secondary
disciplines (skilled nursing, home hospital wage adjustment factors. In the diagnoses. The proposed 4-equation
health aide, physical therapy, speech- CY 2008 proposed rule (72 FR 25449), model resulted in 153 case-mix groups.
language pathology, occupational we proposed to use the 2008 pre-floor In addition, we proposed to replace
therapy, and medical social services) and pre-reclassified hospital wage index the current single therapy threshold of
and medical supplies. Durable medical (not including any reclassification 10 visits with three therapy thresholds
equipment covered under home health under section 1886(d)(8)(B) of the Act) at 6, 14, and 20 visits. We proposed that
is paid for outside the HH PPS payment. to adjust rates for CY 2008 and would payment for additional therapy visits
To adjust for case-mix, the HH PPS uses publish those final wage index values in between the three thresholds would
an 80-category case-mix classification to the final rule. increase gradually, incorporating a
assign patients to a home health As part of the CY 2008 proposed rule declining, rather than a constant,
resource group (HHRG). Clinical needs, (72 FR 25435), we also proposed to amount per added therapy visit. The
functional status, and service utilization rebase and revise the home health proposed approach would not reduce
are computed from responses to selected market basket to reflect FY 2003 total payments to home health providers
data elements in the OASIS assessment Medicare cost report data, the latest because the payment model would still
instrument. available and most complete data on the predict total resource cost. We noted
For episodes with four or fewer visits, structure of HHA costs. In the proposed that the combined effect of the new
Medicare pays on the basis of a national rebased and revised home health market therapy thresholds and payment
per-visit amount by discipline, referred basket, the labor-related share was gradations was expected to reduce the
to as a low utilization payment 77.082 (an increase from the current undesirable emphasis in treatment
adjustment (LUPA). Medicare also labor-related share of 76.775). The planning on a single therapy visit
adjusts the national standardized 60-day proposed non-labor-related share was threshold, and to restore the primacy of
episode payment rate for certain 22.918 (a decrease from the current non- clinical considerations in treatment
intervening events that are subject to a labor-related share of 23.225). The planning for rehabilitation patients.
partial episode payment adjustment increase in the proposed labor-related In the May 4, 2007 proposed rule (72
(PEP adjustment) or a significant change share using the FY 2003 home health FR 25395), we further proposed to make
an adjustment for case-mix that was not to measure nominal change in case-mix appropriate alternative payment policy.
due to a change in the underlying health under the HH PPS, we identified an 8.7 However, we solicited the public for
status of the home health users. Section percent change (increase) in the average suggestions and comments on this
1895(b)(3)(B) of the Act requires that in CMI that would not be due to a change aspect of the HH PPS for ways to
compensating for case-mix change, a in the patient health status (1.233, 2003 improve the PEP adjustment policy.
payment reduction must be applied to rate ¥1.134, September 2000 baseline =
Section 1895(b)(5) of the Act also
the standardized payment amount. At 0.099; 0.099/1.134 = 0.087).
allows for the provision of an addition
the time of publication of the proposed Consequently, we proposed to account
or adjustment to account for outlier
rule, the most recent available data, for that 8.7 percent in case-mix change,
that we considered to be nominal by episodes, which are those episodes that
from which to compute an average case- incur unusually large costs due to
mix weight, or case-mix index, under reducing the national 60-day episode
rate by 2.75 percent, per year, for 3 years patient care needs. Under the HH PPS,
the HH PPS rule, was from 2003. Using outlier payments are made for episodes
the 2003 data, the average case-mix (subject to change upon analysis of
newer, 2005 data for the final rule), for which the estimated cost exceeds a
weight per episode for initial episodes threshold amount. The wage adjusted
was 1.233. Analysis of a 1-percent beginning in CY 2008.
Additionally, we proposed to modify fixed dollar loss (FDL) amount
sample of initial episodes from the
a number of existing HH PPS payment represents the amount of loss that an
1999–2000 data under the HH IPS
adjustments. Specifically, we proposed agency must bear before an episode
revealed an average case-mix weight of
modifying the LUPA by increasing the becomes eligible for outlier payments.
1.125. Standardized to the distribution
payment, by $92.63, for LUPA episodes Section 1895(b)(5) of the Act requires
of agency type (freestanding proprietary,
that occur as the only episode or the that the estimated total outlier payments
freestanding not-for-profit, hospital-
initial episode during a sequence of may not exceed 5 percent of total
based, government, and skilled nursing
adjacent episodes. It has been suggested, estimated HH PPS payments. With
facility (SNF)-based) that existed in
by the industry, that LUPA payment outlier payments having increased in
2003 under the HH PPS, the average
rates do not adequately account for the recent years, and given the unknown
weight was 1.134. We noted this time
front-loading of costs in an episode. Our effects that the proposed refinements
period is likely not free from
analysis showed that these types of may have on outliers, we proposed to
anticipatory response to the HH PPS,
LUPAs require longer visits, on average, maintain the FDL ratio of 0.67. We
because we published our initial HH than non-LUPA episodes, and that the
PPS proposal on October 28, 1999. The stated, in the proposed rule (72 FR
longer average visit length is due to the 25434), that we believed this would
increase in the average case-mix using start of care visit, when the case is
this time period as the baseline resulted continue to meet the statutory
opened and the initial assessment takes requirement of having an outlier
in an 8.7 percent increase (from 1.134 to place. Consequently, these analyses
1.233; 1.233–1.134=0.099; 0.099/ payment outlay that does not exceed 5
indicate that payments for such percent of total HH PPS payments,
1.134=0.087; 0.087×100=8.7 percent). episodes may not offset the full cost of
We proposed that the 8.7 percent of while still providing for an adequate
initial visits. We also proposed number of episodes to qualify for outlier
case-mix change that occurred between eliminating the significant change in
the 12 months ending September 30, payments. We further stated in the
condition (SCIC) payment adjustment.
2000 and the most recent available data proposed rule (72 FR 25434) that we
The current SCIC policy allows an HHA
at the time from 2003 be considered would rely on the latest data and best
to adjust payment when a beneficiary
case-mix change unrelated to change in analysis available at the time to estimate
experiences a SCIC during the 60-day
health status, also referred to as episode that was not envisioned in the outlier payments and update the FDL
‘‘nominal case-mix change.’’ We original plan of care. Because of the ratio in the final rule if appropriate.
proposed to apply this reduction over 3 apparent difficulty HHAs have in Finally for CY 2007, we specified 10
years at 2.75 percent per year. Our interpreting the SCIC policy, their OASIS quality measures as appropriate
analysis on the average case-mix under negative margins, the decline in the for measurements of health care quality.
the HH PPS using an Abt Associates’ occurrence of SCICs, and the estimated These measures were to be submitted by
case-mix study sample from October little impact on outlays in eliminating HHAs to meet their statutory
1997 to April of 1998 as the baseline the SCIC policy, we proposed to requirements to submit data for a full
revealed an increase in the average case- eliminate the SCIC policy. increase in their home health market
mix of 23.3 percent (from 1.0 during In the development of the HH PPS, basket percentage increase amount. For
October 1997 to April 1998 to 1.233 in non-routine medical supplies (NRS) CY 2008, we proposed to expand the set
2003). Because we believed the HHAs were accounted for by attributing $49.62 of 10 measures by adding up to 2
response to BBA provisions, such as the to the standardized episode payment. In National Quality Forum (NQF)-endorsed
home health interim payment system the CY 2008 proposed rule (72 FR measures. The proposed additional
(HH IPS) during this period, could have 25427), we proposed to apply a severity measures for 2008 were as follows:
produced data from this sample that adjustment to the NRS portion of the
reflected a case-mix in flux, we were not HH PPS standardized episode payment. • Emergent Care for Wound Infection,
confident that the trend in the case-mix Specifically, we proposed a five-severity Deteriorating Wound Status
index (CMI) between the time of the Abt group level approach that we believe • Improvement in the Status of Surgical
Associates case-mix study sample and would account for NRS costs based on Wounds
2003 data, used in the analysis for the measurable conditions, would be
proposed rule, reflected only changes in feasible to administer, and offered Accordingly, for CY 2008, we
proposed to consider the 12 OASIS
nominal coding practices. Conversely, HHAs some protection against episodes

the average case-mix for a sample data with extremely high NRS costs. Finally, quality measures submitted by HHAs to
set for 12 months ending September 30, we did not propose to modify the CMS for episodes beginning on or after
2000 (HH IPS baseline) was found to be existing Partial Episode Payment (PEP) July 1, 2006 and before July 1, 2007 as
1.125, standardized to 1.134. Using this Adjustment. At the time of the proposed meeting the reporting requirement for
time period as the base-line from which rule, our analysis did not suggest a more CY 2008.
III. Analysis of and Responses to Public degree of accuracy based on the original implementation of the HH PPS
Comments on the CY 2008 Proposed information received to date. In all in response to concerns from the home
Rule instances where we foresee submission health industry that requiring RAPs for
In response to the publication of the or processing lags affecting the accuracy brief LUPA episodes presented an
CY 2008 HH PPS proposed rule, we of claim payments under the refined administrative burden. Absent
received approximately 150 items of system, we are designing processes to consistent feedback throughout the
correspondence from the public. We retrospectively adjust paid claims at the home health industry that the benefits
received numerous comments from point when the delayed information is of removing this billing mechanism
various trade associations and major received. For example, the CWF will would outweigh the costs, we plan to
organizations. Comments also originated automatically adjust claims up or down retain the no-RAP LUPA process.
from HHAs, hospitals, other providers, to correct for episode timing (early or However, we note this billing
suppliers, practitioners, advocacy later, from M0110) and for therapy need mechanism is an operational issue and
groups, consulting firms, and private (M0826) when submitted information is we have not received many comments
citizens. The following discussion, found to be incorrect. on this issue. It should be further noted
arranged by subject area, includes our No cancelling and resubmission on that requiring the submission of RAPs
responses to the comments and, where the part of HHAs will be required in for all episodes will not necessarily
appropriate, a brief summary as to these instances. Additionally, as the speed the submission of those RAPs in
whether or not we are implementing the proposed rule noted, providers have the all cases. RAPs, like no-RAP LUPAs, can
proposed provision or some variation option of using a default answer also be submitted at any point in the
thereof. reflecting an early episode in M0110 in timely filing period.
cases where information about episode Comment: One commenter asked
A. General Comments on the CY 2008 sequence is not readily available. whether home health services received
HH PPS Proposed Rule Comment: Most commenters when a beneficiary is enrolled in a
supported the elimination of OASIS Medicare Advantage (MA) Plan will be
1. Operational Issues
item M0175 from the case-mix model, as considered in determining the sequence
Overall, commenters were pleased they sometimes found it difficult to of adjacent episodes in cases where the
with the proposed changes to the HH code accurately. Some commenters beneficiary has disenrolled from the MA
PPS. However, commenters did express thought that we were eliminating Plan and resumes his or her coverage
concerns over the burden they M0175 from the OASIS entirely, and under the Medicare fee-for-service
perceived that would be placed on supported that. Several recommended program.
HHAs to accomplish a number of the that we also stop retrospective M0175 Response: Medicare does not typically
proposed changes. audits. One asked that we keep M0175 receive claim-by-claim or individual
Comment: Commenters generally as a case-mix variable, and apply the service data on beneficiaries enrolled in
appreciated CMS’s plan to automatically points to patients who have been MA Plans. As a result, the information
adjust claims to reflect the actual admitted directly from a hospital. is not available to determine whether a
amount of therapy provided versus that Response: We appreciate the support beneficiary has been receiving home
initially reported in OASIS item M0826, of our decision to eliminate M0175 as a health services under the plan or for
Therapy Need, but two commenters case-mix variable. We are not how long. Medicare systems will
noted that for payment adjustments to eliminating M0175 from the OASIS, as determine sequences of adjacent
be made accurately, Medicare’s is explained in section III.E.4, but only episodes based on the fee-for-service
Common Working File (CWF) system removing it from the case-mix model. episode information currently housed in
must contain timely, accurate The M0175 item’s results across the four the CWF and accessible to Medicare
information. Numerous commenters equations were difficult to interpret, and providers through eligibility inquiry
were concerned that the creation of the item’s explanatory power (with transactions.
M0110 (Episode Timing) would be respect to contribution to the R-squared Comment: A commenter believed that
burdensome, as agencies do not have statistic) was small. Therefore, M0175 the addition of multiple payment tiers
the information to complete them. The was not included as a case-mix variable based on therapy usage would create a
commenters did not want to be in our final case-mix model. problem concerning beneficiary
penalized if M0110 was answered The M0175 item is part of the original notification of their financial obligation
incorrectly, and wanted to avoid HH PPS case-mix model and was to pay for home health services. Many
administrative burden from having to reflected in the determination of beneficiaries are now enrolled in
cancel and resubmit final claims and payments under that system. The Medicare replacement plans that require
Request for Anticipated Payments retrospective M0175 audits are still a co-pay on the episodic rate. The
(RAPs). necessary to correct payments that were Medicare Conditions of Participation
Response: CMS has made efforts over made inappropriately under the original (CoPs) at 42 CFR 484.10 require that the
the last several years to reduce internal HH PPS. These payment corrections HHA notify the patient in advance of his
processing delays and ensure that the have been repeatedly recommended to or her liability for payment. The
CWF is updated with claim receipts CMS by HHS’s Office of Inspector commenter believed some consideration
more quickly overall. While new errors General. needs to be made about the obligations
may arise that delay processing, we will Comment: One commenter proposed of HHAs to meet this requirement as it
seek to correct them as swiftly as that the timeliness of information on is virtually impossible to calculate the
possible in light of all the competing Medicare systems would be increased rate and provide notices of the changing
demands on our systems. by the removal of the option to submit rate prior to providing service.
The factor that most affects the no-RAP LUPA claims. The commenter Response: The provisions of this rule
timeliness and accuracy of the CWF is believes that requiring RAPs for all apply to Medicare’s fee-for-service HH
how promptly within the 15 to 27 episodes will speed submission of PPS and do not apply to Medicare
month timely filing period each episodes to Medicare. Advantage/Medicare Choice plans
provider submits its claims. Medicare Response: The no-RAP LUPA billing where co-pays for home health services
systems can only process to the greatest mechanism was created as part of the provided under the plan may exist. As
long as the patient meets the Medicare relative case-mix weights, the pricing Response: We do not agree that
fee-for-service eligibility requirements, procedure applied after accounting for agencies are unable to plan
and the HHA provides covered services standardized resource costs adjusts for operationally and financially for these
that are reasonable and necessary based geographic differences in cost levels. We changes. We worked with a large, 20-
on the patient’s plan of care, there have no data to effectively evaluate the percent sample of 2005 claims, which
would be no financial obligation on the comments on the disadvantages would not permit us to produce
part of the patient. However, if the attributed to rurally residing accurate summaries at the agency level
patient asks the HHA for services beneficiaries. for many agencies, which would be
outside the scope of the Medicare home Comment: A commenter suggested required for a file of the type mentioned
health benefit, or the HHA provides raising the RAP to 75 percent of the base by the commenter. Our proposed rule
non-covered services, the HHA would rate. Another commenter noted that the impact table provided average case-mix
be required to provide the patient with proposed rule is silent on the need to weights for agencies to use as estimates,
financial liability information via the increase the RAP, even though program according to the detailed subgroup to
Advanced Beneficiary Notification abuse of the RAP has not materialized. which they belong. Consistent with
(ABN). The multiple payment tiers (that This commenter proposed that the RAP resources available, we opted to provide
is, multiple therapy thresholds) would be increased to 80/20 for all providers a simple preliminary grouper to assist
not affect the determination of the who have participated in the HH PPS agencies in understanding the impacts.
patient’s financial liability. That since its inception, and noted that CMS We also provided preliminary grouper
liability would be outside the scope of would retain the right to reduce this logic (‘‘pseudocode’’) for software
the Medicare home health benefit, and level for abuse of the RAP. The developers assisting some agencies to
would be determined between the HHA commenter further proposed that less evaluate the impacts.
and the patient. This comment is established providers could operate Comment: A number of commenters
beyond the scope of this final rule with under current RAP rules until they had noted that home health agencies provide
comment period, which deals with a 5 year record of responsible Medicare quality care that saves Medicare money
payment under HH PPS to fee-for- performance. in hospital or other inpatient facility
service HHAs. Response: Before HH PPS benefits. Several commenters expressed
Comment: Several commenters wrote implementation, HHAs were concern that the proposed changes do
that smaller, rural agencies are accustomed to billing Medicare on a 30- not consider today’s health picture, with
particularly disadvantaged by the day cycle or receiving periodic interim an aging population, a wave of baby
changes in the proposed rule. They were payments. The change to a 60-day boomers entering retirement, a shortage
concerned that the proposed changes episode of care under HH PPS, of nurses, high fuel costs, and the cost
will limit the ability of agencies to combined with concerns over delays of technological advances such as
survive or compete, which could limit due to claims processing times, telehealth and physician’s portal.
access for patients. This may impact documentation requirements, and Response: The goal of the refinements
rural patients more than urban patients. medical review, led us to address in this regulation is to pay as accurately
Another commenter noted that CMS agency cash flow concerns in our 1999 as possible given the case-mix of
derives resource costs by weighting each HH PPS proposed rule. At that time, we patients in home health agencies today.
minute reported on the claim by the proposed a split percentage payment to We appreciate the broad context
national average labor market hourly ensure that agencies have adequate cash referenced in this comment, and will
rate for the discipline, and summing the flow to maintain quality services to continue to work with the home health
total. The commenter believed that it is beneficiaries. In 2000, we implemented industry and the public to understand
not realistic to attribute the same the RAP which paid 60 percent up front and anticipate changes that affect proper
resource cost to rural beneficiaries as to for an initial episode, as we recognized pricing of home health services.
urban beneficiaries, who have more that some administrative costs were Comment: A commenter suggested
social programs available to them. front-loaded; the remaining 40 percent that we revise the regulation requiring
Additionally, this method does not would be paid after submission of the that orders and plans of care for home
account for the significant travel costs final claim. We allowed a RAP of 50 health patients be signed by a physician.
associated with rural beneficiaries. The percent for a subsequent episode, with Another commenter asked that the CoPs
commenter added that this is why there the remaining 50 percent paid upon be changed to allow therapists, in
has periodically been a rural add-on. receipt of the final claim. addition to nurses, to open a case, as it
Response: Our impact tables show We expect agencies to follow normal could improve the ability to accurately
that rural agencies, on average, will business practices with regard to project therapy requirements for
experience a modest reduction in total financing their operations. The current patients.
payments between 2007 and 2008—less RAP percentage splits are reasonable Response: We appreciate these
than 2 percent. Factors in the reduction given the RAP’s purpose, therefore, we comments, but note that this regulation
are discussed in section VI.B. These do not see a need to increase them. updates the HH PPS payment rates and
include the small reduction in the Moreover, we believe our current does not change any of the CoPs.
average case-mix weight in 2008 among process protects against abuse, as an Sections 1814(a)(2)(c) and
rural agencies, the impact of the wage agency’s RAP may be reduced or 1835(a)(2)(A)(ii) of the Act require that
index, and several other factors withheld when protecting Medicare orders and plans of care be established
discussed in that section. The offsetting program integrity warrants this action. and periodically reviewed by a
positive effect of the annual payment Comment: Two commenters wrote physician. The CoP dictating the
update offsets most of the total negative that they are unable to make meaningful physician signature requirements on the
effect of the changes. public comment because CMS has not plan of care is detailed in 42 CFR
Medicare prices are adjusted for the released the impact file that would 484.18(b) and (c).
cost differences among different enable modeling of the proposed Moreover, in 42 CFR 484.55(a)(1),
locations. Although we use changes. Agencies are unable to plan agencies are required to have a
standardized national average resource operationally and financially for these registered nurse conduct an initial
cost estimates for developing the changes. assessment. We note, however in 42
CFR 484.55(a)(2), the home health CoP with a corrected version. The correct incapacitation and need for medically
regulations state that ‘‘when version was promptly posted on the necessary nursing or rehabilitative or
rehabilitation therapy service * * * is CMS Web site. assistive services, while they continue
the only service ordered by the Comment: Regarding dual eligibles, a to meet the homebound requirement.
physician, and if the need for that commenter suggested that CMS improve Agencies are expected to apply the
service establishes program eligibility, the alignment of HHRGs and Medicare statutory eligibility and coverage
the initial assessment visit may be made coverage guidelines for homebound criteria.
by the appropriate rehabilitation skilled status and medical necessity, Comment: A commenter questioned
professional.’’ particularly for cases that receive whether the increase seen in costs of
Comment: A commenter noted that coverage under ‘‘Assessment and late episodes is due to end-of-life care
CMS currently uses salary information Observation’’ or ‘‘Management and given to patients who did not want
to estimate the costs of a visit, and does Evaluation of the Care Plan’’ guidelines. hospice care.
not include overhead costs. This Improved alignment of the payment Response: We appreciate the
method assumes indirect costs are system and coverage rules is critical to comment. We note, however, our
proportional to direct costs. The addressing ongoing disputes between analysis did not focus on whether or not
commenter believes this assumption state Medicaid agencies and the the patient had a terminal illness.
may be incorrect, and suggested Medicare program regarding Third Party 2. The Schedule for Implementation of
examining cost report data to see if Liability. the CY 2008 Refinements
further review provides better data on Response: These comments are
overhead costs. This information could outside the scope of this regulation; In the May 4, 2007 proposed rule, we
be combined with claims information however, we will take them under proposed to implement the finalized
about home health charges to better consideration when evaluating the need updates and refinements on January 1,
assess labor costs. These two sources of for additional guidance on Medicare 2008. However, we did recognize that
information could be used to compute coverage guidelines. there may be operational considerations,
the per-visit discipline costs for Comment: A commenter is concerned affecting CMS or the industry, which
different types of episodes. that the proposed HH PPS refinements could necessitate an implementation
Response: CMS’ methodology does place emphasis on therapy and would schedule that results in certain
assume that overhead costs are support a system that provides for the refinements becoming effective on
proportional to direct labor costs. We utilization of restorative nursing as a different dates (a split-implementation).
will continue to consider the substitution for therapist visits. The We solicited the public for suggestions
appropriate role of cost reports in expansion of this type of service and comments on this matter.
understanding potential improvements utilization will ultimately provide better Comment: Several commenters
to our methodology. At this time, we patient outcomes and address the expressed concern about the amount of
believe the role is limited, as growing demand for restorative services. time available for providers to make any
demonstrated by the limitations on cost Response: The proposed refinements necessary changes to their billing
report reliability pertaining to the were developed within the disciplines systems and administrative processes
derivation of cost-to-charge ratios for the covered by the home health benefit. A between the publication of this rule and
analysis of NRS payments. We urge specialty of restorative nursing is not the implementation date of episodes
agencies to put more resources into recognized within those disciplines. beginning on January 1, 2008. They
accurately completing the cost reports Moreover, we do not have evidence were concerned about the
for future use in payment refinements. about effects on patient outcomes from administrative burden, and that CMS
Comment: A commenter suggested implementing the commenter’s does not have a contingency plan to
that the recommendations from the two proposal. facilitate interim payments to HHAs that
Technical Expert Panel (TEP) meetings Comment: A commenter believed it is are unable to bill Medicare under the
be shared with the industry, and that important for CMS to align regulatory revised HH PPS. A contingency
the industry be allowed to provide and reimbursement decisions so that payment arrangement would ensure that
feedback, as these affected the they reflect the needs of patients as no provider is presented with a
development of the proposed rule. outlined by the Institute of Medicine. significant cash flow problem because of
Response: The TEP was administered The commenter stated that the proposed the tight timeframe involved. Several
by Abt Associates. The panel was not regulation signals a change in which the commenters suggested we convene an
asked for, nor did it produce, consensus home health industry would be asked to ongoing series of implementation
recommendations. Abt Associates used move from its current focus on acute meetings including Medicare
TEP participants as a sounding board and rehabilitative services to the contractors, the home health
about differing aspects of the research provisions of more long-term care community, and the vendors who
approach and the refinements emerging services of the type offered prior to HH support the home health industry to
from it at the time of the TEP meeting. PPS implementation. The commenter reduce the likelihood of delays and
Comment: A commenter asked that asked CMS to clarify whether it prefers errors. One commenter asks for
we provide detailed technical Medicare home health services to additional resources to help providers
specifications and grouper software emphasize more sophisticated cope with this major change. Another
with issuance of the final rule. treatments or whether it expects home asked that we not follow a split-
Response: We intend to issue detailed health services to be used solely for implementation plan.
specifications and a grouper software long-term care and/or custodial services, Response: While the changes
package as soon as possible after the which have traditionally been the described by this rule are significant,
issuance of this rule. purview of Medicaid. their overall impact on provider billing
Comment: A commenter noted that Response: We disagree that the practices are far less extensive than
there was an error in Table 5 posted to proposals signal a shift away from acute those required for the initial
CMS’ Web Site. and rehabilitative services. The implementation of HH PPS. We also
Response: Table 5 was originally proposals recognize that a minority of anticipate the time period between the
posted with an error, but was replaced patients have an extended period of issuance of this final rule with comment
period and the implementation date will time between the issuance of this rule Response: We recognize that there
be longer than the period that was and the effective date (January 1, 2008) was an inadvertent technical error in the
available between publication of the than there was for the initial May 4, 2007 proposed rule in that July
final rule on July 3, 2000, and initial implementation of the HH PPS. 3, 2007 was incorrectly noted as the
implementation of the HH PPS on Consequently, we believe that there will close of the comment period.
October 1, 2000. CMS expects to issue be sufficient time for agencies and their Subsequent to that publication, a
final implementing instructions and vendors to make the changes necessary correction notice was published on May
educational materials about the case- to implement the system on January 1, 11, 2007 (72 FR 26867), noting that error
mix refinement changes as soon as it is 2008. Regarding the home health CoPs, and correctly stating that the end of the
feasible after finalization of the these are on a separate track from our comment period for the HH PPS
proposals contained in this final rule home health payment regulations, and proposed rule was June 26, 2007 and
with comment period. We also plan to will be implemented through a separate not July 3, 2007.
conduct outreach through industry rule-making process. We believe we made reasonable
associations and representatives of While we recognize that efforts to quickly alert the public to the
software companies that serve home implementing the updates and error such that adequate time to
health agencies to facilitate this refinements of this rule is an ambitious comment on the proposed rule was
transition. task, we believe that it is in the best provided.
CMS plans to conduct calls with interest of the industry, CMS, and home 3. Complexity of the System
vendors, hold OASIS training, and health recipients to implement a
continue the use of the home health In general, our goal for the proposed
finalized set of refinements without
Open Door Forums (ODFs) as refinements was to ensure that the home
further delay and without a split-
mechanisms to provide information to health payment system continues to
implementation. The refinements will
HHAs regarding implementation. produce appropriate compensation for
work together to improve the accuracy
Regarding cash flow issues and providers while creating opportunities
and appropriateness of the HH PPS,
contingency plans, CMS is taking steps, for home health agencies to manage
which has not undergone major
internally, to test systems changes home health care efficiently. We also
refinements since its inception in believe it is important in any refinement
before implementation. We do not feel October of 2000. Updates to the HH PPS
that the vulnerabilities that existed to maintain an appropriate degree of
are not linked, specifically, to coding operational efficiency.
when we moved from a cost-based manuals, and thus there would be no
system to a prospective payment system Comment: Several commenters stated
advantage to delaying implementation that the goal of ‘‘operational simplicity’’
exist today in moving to a refined HH to any future coding manual update.
PPS system. Consequently, we do not is not achieved by the proposed
CMS will make every effort to refinements. One commenter stated that
feel it is necessary to create an elaborate communicate the instructions necessary
contingency plan as was needed for the the proposed system is twice as
for HHAs to implement all of the complex as the current system, thus
implementation of the HH PPS.
changes to the HH PPS, in a timely making it more difficult for providers to
Comment: Several commenters
manner so that implementation of these understand how it works. Moreover, the
expressed that an implementation date
of January 1, 2008 be delayed because changes occurs as smoothly as possible. commenter stated it will make it more
the HH PPS reform changes are Comment: Several commenters difficult for providers to manage the
significant, and providers will have to expressed that the comment period was level of services provided for each
educate all of their employees on the too brief to afford providers enough time HHRG with the payment for that HHRG.
changes in addition to working closely to understand the proposed changes and Response: We acknowledge the
with the vendors to initiate complex IT assess the impact that the changes will proposed refined system is more
changes. Because as providers, they have on their businesses. complex than the current system. The
must also implement the changes Response: We provided the 60-day proposed refinements to the current
throughout the organization, to both comment period from the date of system represent an attempt to pay more
clinical and financial staff, the display, with the 60-day period for accurately for the range and intensity of
commenters suggested that CMS delay comments ending on June 26, 2007. We home health services that are provided
the implementation date to October 1, acknowledge that in the publication of to our beneficiaries.
2008 to allow ample time for providers the May 4, 2007 proposed rule, the The proposed refinements are derived
to make all the necessary adjustments. comment period was incorrectly listed from the concepts that form the basis of
The commenters also requested that as closing on July 3, 2007. The correct the current payment approach. We agree
CMS release of the home health CoPs date for the close of the comment period that any refinements to the system will
coincide with the implementation of HH was June 26, 2007. Recognizing the take time and training to learn. CMS has
PPS refinement requirements to ease the implication of this incorrect date, CMS conducted extensive outreach regarding
burden of staff training. It was also alerted the public to the correct date the proposed refinements. We have
suggested that the implementation be through listserves, open door forums, posted a Fact Sheet which summarizes
linked to future ICD–9–CM coding and the publication of a correction the proposed changes on our home
manuals. notice on May 11, 2007 (72 FR 26867). health Web site to assist agencies in
Response: We recognize that the We believe the comment period, as understanding the differences between
changes described in this rule are corrected, provided adequate time for the current system and the proposed
significant. However, the overall impact commenters to review the proposals and refinements. We have developed and
on provider billing practices is far less assess their options. posted an Excel toy grouper, which
significant than the impact resulting Comment: Several commenters allows agencies to see the effect of the
from the initial implementation of the questioned the listing of an earlier new proposal on their payments (see
HH PPS when we were moving from a deadline on the internet for submission ‘‘Toy Grouper’’ on the CMS Home
reasonable cost-based system to that of of public comments, June 26, 2007, Health Web site at: http://
a prospective payment system. And as rather than the deadline published in www.cms.hhs.gov/center/hha.asp). We
mentioned previously, there is more the Federal Register, July 3, 2007. have posted the draft pseudocode for
the HHRG grouper software at the same for therapy episodes would increase based upon the most recent data
Web site address. We also continue to gradually between the first and third available, and reflects any philosophical
plan for additional training and therapy thresholds. For a complete or diagnosis changes that the industry
outreach. description of the proposed case-mix has experienced.
We have also developed claims refinements model and the underlying Comment: A commenter suggested
processing procedures to reduce the research, we refer readers to the CY that the case-mix refinement model was
amount of administrative burden 2008 HH PPS proposed rule (72 FR too complex, and suggested that we
associated with using a more complex 25358–25420) published on May 4, simplify it so that the assessment can
case-mix model. For example, providers 2007. drive clinical and functional dimension
do not have to determine whether an scores that are the same regardless of the
episode is early (the initial episode in a 1. General Comments number of therapy visits or timing of the
sequence of adjacent episodes or the Comment: A commenter wrote that an episode. Subsequent factors could be
next adjacent episode, if any) or later industry analysis of 2006 HH PPS data added into the case-mix for the
(all adjacent episodes beyond the using the proposed case-mix model sequential number of the episode and
second episode) if they choose not to. showed a decline in reimbursement for for the number of visits.
Information from Medicare systems will specific populations with congestive Response: Based on our data analysis,
be used during claims processing to heart failure (CHF), chronic obstructive implementing the commenter’s
automatically address this issue. We pulmonary disease (COPD), ulcers, suggestion would ignore patterns in the
will also relieve providers of the diabetes, orthopedic diagnoses, and data that we think reflect differences
responsibility for resubmitting a claim if neurological diagnoses. Given these between patients and would thereby
the number of therapy visits delivered findings, the commenter asked how the reduce accuracy. We have tried to strike
during an episode is more than or less proposed case-mix refinement could a balance between simplicity and
than the number originally forecasted improve reimbursement. The complexity. The new system is more
on the OASIS. commenter suggested that CMS use complex than the old system but this is
Comment: A commenter stated that more current diagnosis data so as not to a natural outgrowth of our attempt to
the Excel toy grouper did not allow for skew the results, and score secondary pay more accurately for the range and
enough digits in the ICD–9 codes to diagnoses. Other commenters echoed intensity of home health services that
effectively capture the degree of change the concern that the refinement was can be provided to our beneficiaries.
needed. The commenter also noted that based on ‘‘old’’ data. A couple of As noted in the discussion of
each case had to be added individually, commenters noted that there has been a complexity in section III.A.3, a system
which resulted in increased entering philosophical change to front-load visits may seem initially overly complex
time; the results were confusing to the in home health which has not been when it is new. We believe the proposed
commenter. captured by the data. refinements are clearly focused, and are
Response: We believe that the Response: We are unable to a logical outgrowth of the original
requirement that the ICD–9 codes be specifically address the industry payment system. We detail our attempts
entered exactly as they appear in the analysis mentioned above without more to make the proposed refinements easier
proposed rule and the current grouper detailed information on their analysis. to understand and implement in a
documentation does not negate the We note the proposed case-mix model previous comment in section III.A.3.
usefulness of the Excel toy grouper. The pays for more diagnoses than under the Comment: One commenter noted that
instructions imbedded in the Excel toy current HH PPS model, including the proposed diagnosis changes may
grouper specify the requirements for recognition of point-bearing diagnoses negatively impact providers who are
entering the ICD–9 codes. We provided for heart disease and COPD. Agencies currently providing care to those in
the Excel toy grouper as a courtesy to will continue to receive points to the early episodes with less than 14 therapy
allow users to more easily calculate the extent that patients have certain visits. Those providers have worked
proposed new CY 2008 HHRGs and conditions or diagnoses (for example, hard to help patients become
resulting payments rather than having ulcers, diabetes, orthopedic diagnoses, independent and rehabilitated as soon
only the grouper pseudocode for and neurological diagnoses). Agencies as possible.
analysis. Moreover, the majority of can also receive points for secondary Response: Our proposal was intended
feedback from commenters regarding diagnoses, thereby accounting for to refine and to better fit costs incurred
the Excel toy grouper indicated that the multiple co-morbidities. Also, the by agencies for patients with differing
tool is helpful and easy to use. proposed case-mix model allows points characteristics and needs under the
for some resource intensive interactions. prospective payment system. The
B. Case-Mix Model Refinements Furthermore, agencies will be receiving resource cost estimates are derived from
In the proposed rule, we proposed to improved reimbursement for supplies, minutes spent on visits in the home
refine the case-mix model to reflect particularly those related to ulcers or during a 60-day period. The source of
different resource costs for early home wounds. We believed the model as the minutes data is a very large,
health episodes versus later home proposed would better align agency representative sample of Medicare
health episodes and to expand the case- costs with payments. claims. Therefore, we expect that the
mix variables included in the payment We further note that the proposed proposal does reflect agencies’ average
model. We proposed additional refinement research was based upon costs for patients with characteristics
variables including scores for certain data files created from a 20-percent measured on the OASIS and used in
wound and skin conditions; more sample of claims data collected between defining payment groups.
diagnosis groups such as pulmonary, 2001 and 2004. OASIS data was further Comment: While supporting the
cardiac, and cancer diagnoses; and linked to claims and cost reports. concept behind the new case-mix
certain secondary diagnoses. We also However for this final rule with system, a commenter is concerned about
proposed to replace the current single comment period, we used more recent any payment system that ties payments
therapy threshold of 10 visits with three data, claims processed from 2005, with explicitly to the level of services
therapy thresholds (6, 14, and 20 visits). the associated OASIS data. Therefore, provided. Under the proposed system,
In addition, we proposed that payment this final rule with comment period is HHAs could seek higher payments by
providing more therapy or providing patients, and reduce the provider’s 484.18(a). All diagnoses listed in OASIS
later episodes of home care. The preference for some patients. M0230/240 and M0246 should be
commenter notes that HHA margins will Response: We appreciate the pertinent and are expected to be listed
increase with the number of therapy commenter’s assessment of the in the patient’s POC.
visits. proposed changes to the case-mix
Response: We are attuned to concerns system, and agree that the proposed 2. Later Episodes
about payment incentives that could refinements improve the performance In the proposed rule, for 2008 we
drive up therapy visits unnecessarily. and payment accuracy of the HH PPS. proposed a 4-equation case-mix model
We implemented a gradual increase in We agree that these changes will reduce that recognizes and differentiates
payments between the proposed first incentives to select patients based upon payment for episodes of care based on
and third therapy thresholds to achieve perceived financial advantages. whether a patient is in what is
two goals: (1) To better match costs to Comment: A commenter noted that an considered to be an early (1st or 2nd
payments; and (2) to avoid incentives analysis of the coefficient of variation episode in a sequence of adjacent
for providers to distort patterns of good (CV) of the proposed HHRGs found it to episodes) or later (the 3rd episode and
care created by the increase in payment be more internally homogeneous. The beyond in a sequence of adjacent
that would occur at each proposed average CV has dropped from 0.81 in episodes) episode of care as well as
therapy threshold. As a disincentive for the current system to 0.75 for the recognizing whether a patient was a
agencies to deliver more than the proposed HHRGs. The reduction in high therapy (14 or more therapy visits)
appropriate, clinically determined variation means that the new resource or low therapy (13 or fewer therapy
number of therapy visits, we also groups are better at identifying episodes visits) case. Early episodes are defined
proposed that any per-visit increase with similar resource use than under the as to include not only the initial episode
incorporate a declining, rather than a current system. Further, the reduction in a sequence of adjacent episodes, but
constant, amount per added therapy in within-group variation reduces the also the next adjacent episode, if any,
visit. We will monitor the impact of the potential for providers to select the least that followed the initial episode. Later
changes implemented, including on costly patients in a resource group and episodes are defined as all adjacent
home health agency margins, and will makes a modest improvement in the episodes beyond the second episode.
propose further refinements to the accuracy of the system. Episodes are considered to be adjacent
therapy threshold, as well as other Response: We agree with the if they are separated by no more than a
aspects of the HH PPS, if warranted. commenter, and believe that the 60-day period between claims. The
Comment: Several commenters were proposed payment system better analysis of the performance of the case-
concerned that paying more for later matches payments to costs. We also mix model for later episodes revealed
episodes would lead to gaming, with believe that the payments will be more two important differences for episodes
patients on service longer than is accurate, and will benefit patients as occurring later in the home health
appropriate. One commenter noted the well as agencies. treatment compared to earlier episodes:
growth in HHAs in her area had led to Comment: Since this is the first time (1) Higher resource use per episode and
more competition for patients; providers the case-mix index has been updated (2) a different relationship between
may not be discharging patients when since the inception of HH PPS, and clinical conditions and resource use.
they should. Additionally, this considering the rapid pace of change Comment: We received a question
commenter felt the fiscal intermediaries that can occur in health care delivery, about the case-mix weights for early
(FIs) concentrate review activities on a commenter suggested CMS update the versus later episodes when the service
larger agencies where there is the case-mix index with greater frequency utilization is for 16 to 17 therapy visits
greatest potential for risk of harm to to ensure that payments reflect agency (S2; see table 3, III.B.5). In all other
beneficiaries or where the dollars costs. gradients except this one, the case-mix
recovered are greater. The commenter Response: We will continue to weight is greater for later episodes than
encouraged discussion and investigation monitor the performance of any for early episodes. The commenter
of these issues. Another commenter was finalized case-mix model, and will make asked why in this case the later episodes
concerned that the proposed case-mix changes to it as necessary. Future were not associated with a higher case-
refinements created incentives for less refinements may occur at more frequent mix weight.
efficient and less effective care if intervals, depending on the research Response: The model results in Table
agencies provided unneeded care just to outcomes. We recognize that changes in 4 of the proposed rule (72 FR 25388)
extend the length of stay. A third health care delivery may also affect the indicated that the higher cost for later
commenter felt that the proposal would model, and will monitor those as well. episodes was associated with clinical
lead to unwarranted recertification of Comment: A commenter asked CMS and functional severity levels above the
episodes. to accept all pertinent diagnoses. The base levels C1 and F1, and not at or
Response: We appreciate the concerns commenter believed that without a below the base levels C1 and F1. The
and will monitor the use of home health complete clinical picture, the ability to amount isolated in the payment
visits. Additionally, we will share these accurately assess patient severity, regression associated with 16 to 17
concerns with the Regional Home evaluate outcomes, and make policy therapy visits was simply not higher for
Health Intermediaries (RHHIs). decisions is seriously jeopardized. later episodes.
Comment: A commenter’s analysis of Response: We agree that a complete Comment: Several commenters asked
the proposed changes to the case-mix clinical picture of the patient is for clarification of the definition of early
system found that it would result in a necessary to accurately assess patient and later episodes and adjacent
more even distribution of payments severity and evaluate outcomes. To episodes.
relative to costs. The commenter’s qualify for Medicare coverage of home Response: Early episodes are defined
analysis resulted in a more uniform health services, a beneficiary must be as the initial episode or the next episode
payment to cost ratio. The commenter under the care of a physician who in a sequence of adjacent episodes.
noted the proposed refinement would establishes the plan of care (POC). The Therefore an early episode can be the
reduce the differences in financial POC must contain all pertinent first or second episode in a series of
returns among different types of diagnoses as stipulated in 42 CFR adjacent episodes, or even the first and
only episode that a patient has. Later February 13, 2008 visit would be adjacent episodes. Under this scenario,
episodes are defined as all subsequent considered an adjacent episode. that original second episode is now
adjacent episodes beyond the second Intervening stays in inpatient considered to be the third episode in the
episode. Episodes are considered to be facilities will not create any special sequence of adjacent episodes, thus
adjacent if they are contiguous, meaning considerations in counting the 60-day changing its status from that of an early
that they are separated by no more than gap. If an inpatient stay occurred within episode to that of a later episode.
a 60-day period between episodes. This an episode, it would not be a part of the Comment: A commenter noted that
means any gaps are less than or equal gap, as counting would begin at ‘‘day CMS determined its four equation
to 60 days in length. In determining a 60’’ which in this case would be later model based on information collected
gap, we only consider whether the than the inpatient discharge date. If an from the OASIS data set. The data
beneficiary was receiving home health inpatient stay occurred within the collection is required for both Medicare
care from traditional fee-for-service period after the end of HH episode and and Medicaid patients. The commenter
Medicare. If the beneficiary transfers before the beginning of the next one, stated that the analysis by CMS
from a managed care plan, that time those days would be counted as part of included a period of time when
under managed care is considered part the gap just as any other days would. instructions dictated collection of all
of the gap. If episodes are received after a information from payer sources. The
For example, if the beneficiary has not particular claim is paid that change the data is inclusive of the Medicaid
received home health care through sequence initially assigned to the paid patients, who under Medicare
traditional Medicare for at least 60 days, episode (for example, by service dates regulations, would not be eligible for the
and then receives home health care from falling earlier than those of the paid third or additional episodes of care. The
agency A, that is an early episode. If that episode, or by falling within a gap commenter questioned the type of
episode receives a PEP adjustment and between paid episodes), Medicare patients served in third or later
agency B recertifies the beneficiary for systems will initiate automatic episodes, noting that the CMS data
a second episode, that second episode is adjustments to correct the payment of suggest that few patients fall into the
also an early episode. However, the any necessary episodes. new equations. The commenter believed
beneficiary could have received home Upon receipt of a HH episode coded that one group of patients includes
health care from other traditional to represent the early episode in a those with severely infected wounds,
Medicare providers within 60 days sequence, Medicare systems will search Parkinson’s disease, Amyotrophic
before coming to agency A. The the episode history records that are Lateral Sclerosis (ALS), stroke, or
designation of early or later would maintained for each beneficiary. If two similar conditions, while another group
depend upon how many adjacent or more adjacent episodes are found on includes those receiving B–12 injections
episodes of care were received prior to that history, the claim for the new and catheter care, or Medicaid patients.
coming to agency A. The CWF will episode will be recoded to represent its Response: We used data from
examine claims upon receipt in sequence correctly and paid according Medicare episodes only, linked to the
comparison to all previously processed to the changed code. In addition, when OASIS assessment that generated the
episodes to make sure the episode is any new episode is added to those HHRG. Medicare episodes include
correctly designated as early or later. history records for each beneficiary, the episodes of some patients who are
The 60-day period to determine a gap coding representing episode sequence dually eligible for Medicare and
that will begin a new sequence of on previously paid episodes will be Medicaid. Later episodes include both
episodes will be counted in most checked to see if the presence of the Medicare-only and dually eligible
instances from the calculated 60-day newly added episode causes the need patients with a variety of conditions and
end date of the episode. That is, in most for changes to those episodes. If the needs.
cases CWF will count from ‘‘day 60’’ of need for changes is found, Medicare To summarize, we are implementing
an episode without regard to an earlier systems will initiate automatic the proposed aspect of the case-mix
discharge date in the episode. The adjustments to those previously paid model that recognizes and differentiates
exception to this is for episodes that episodes. payment for episodes of care based on
were subject to PEP adjustment. In PEP For example, a given episode is whether a patient is in what is
cases, CWF will count 60 days from the initially determined to be, and paid as considered to be an early or later
date of the last billable home health the second episode (early) in a sequence episode of care as we believe that it
visit provided in the PEP episode. of episodes. After some period of time, better accounts for the higher resource
Regarding PEP adjustments, consider a claim is submitted by another HHA use per episode and the different
the following example: An episode is that occurs before the previously relationship between clinical conditions
opened on January 1, 2008 which would designated first episode in the sequence and resource use that exists in later
normally span until February 29, 2008. of adjacent episodes and is less than 60 episodes.
If this episode were not subject to a PEP days before the beginning of that
adjustment, any episode within 60 days previously designated first episode. In 3. Addition of Variables
following February 29, 2008 would be such a case, the episode corresponding In the proposed rule, for 2008 we
considered an adjacent episode. In the to the newly submitted claim becomes proposed to expand the case-mix
case of a PEP adjustment, the the first episode of this sequence of variables to include scores for
determination of an adjacent episode adjacent episodes and thus is conditions such as infected surgical
would no longer be based on day 60, but considered to be an early episode. The wounds, abscesses, chronic ulcers, and
would instead be based on the latest episode previously designated as the gangrene; more diagnosis groups such as
billable visit in the episode. Assume in first episode in the sequence of episodes pulmonary, cardiac, and cancer
the example, the patient is transferred to now becomes the second episode in the diagnoses; and certain secondary
another HHA (triggering the PEP sequence of adjacent episodes and is diagnoses.
adjustment) on February 15, 2008 but thus still considered to be an early Comment: Several commenters were
the last billable visit is provided on episode. The real change occurs with concerned that we had not included a
February 13, 2008. In this case, any the episode previously described as the variable for informal caregivers. One
episode within 60 days following the second episode in the sequence of commented that higher costs for these
patients are not captured because of the morbidities, patient non-compliance, or Response: In considering variables for
unmeasured effects of multiple co- living alone. inclusion in the model, we analyzed the
morbidities, patient non-compliance, Comment: Several commenters were relationship between resource use and
and the tendency to live alone. Several concerned that a variable for Medicare/ patient characteristics. We were able to
commenters felt that CMS’ policy Medicaid dual eligibles was not measure resource use directly from the
position on caregivers placed the fear of included in the payment model. One claims sample and patient
negative incentives above the needs of commenter noted that the increased characteristics from the OASIS
the beneficiary. Commenters were costs associated with dual eligibles have assessments. Variables were assessed for
concerned that payment incentives been confirmed by MedPAC in hospital statistical performance and for policy
might limit access for patients without DSH studies, and it is unlikely that appropriateness. Diabetes is taken into
caregivers or result in institutional care. these costs disappear once the patient is account as a point-bearing case-mix
Others suggested that we refine OASIS in home health. Another noted that diagnosis under the current HH PPS,
items related to caregiver access to these patients have longer lengths of and under this final rule with comment
produce more reliable information about stay and multiple co-morbidities. period continues to receive points as
the actual roles caregivers play in Several commenters noted that either a primary or a secondary
meeting the day-to-day needs of home Medicaid numbers are not consistently diagnosis (see Table 2A for the points
health patients, and the time they are reported in OASIS because Medicaid is given).
available. Some commenters expressed not the primary payer. Others suggested Our research did not find the
concern that these patients would have that CMS compare the impact of proportion of home health beneficiaries
difficulty accessing care due to their Medicaid eligibility by studying 85 or older to be increasing. The
high costs. We were asked to conduct resource use of a sample of home health literature reports that those 85 or older
further research into the role of patients enrolled in a Medicaid program were actually less likely to be admitted
caregivers and their affect on costs. from Medicaid files against home health to home health agencies (McCall et al.,
Response: OASIS item M0350 asks patients without Medicaid. 2003). Additionally, we tested an age
whether there are assisting persons in Response: HHAs are required to variable and found it was not associated
the home, other than the home care complete OASIS item M0065, which with greater resource use after
agency staff. We recognize that the data asks for the patient’s Medicaid number, controlling for other factors. As such,
collected by this item is limited in the whether or not Medicaid is the we did not include it in our case-mix
information it collects regarding reimbursement source for the home care model. Accordingly, we did not propose
caregivers. However, in the absence of episode. CMS has sought to improve the to include a variable for those 85 and
other data, we used this item in our accuracy of the OASIS data through older in the refinements.
analysis. We found that on average, extensive training and guidance on Comment: A commenter stated that
episodes without caregivers would be proper use of OASIS. Additionally, the the proposed rule refers to unnamed
underpaid. However the score to be OASIS guidelines provide clear variables which while correlated with
gained by adding this variable was not instructions to complete M0065. higher home health cost, were not
large, and the overall ability of the four- Therefore we believe it is appropriate to considered in the case-mix because of
equation model to explain resource use M0065 in an analysis of resource negative treatment incentives they could
costs is minimally improved by adding use in patients with Medicaid. After create. The commenter believed CMS
this variable. As we noted in the accounting for a broad range of clinical should specify these alternatives which
proposed rule, we believe this variable and functional factors which predict were not adopted along with the reason
raises significant policy concerns. We resource use, M0065 was found to have for dismissing them.
maintain that a case-mix adjustment a low score, suggesting that having Response: As in our original HH PPS
should not discourage assistance from Medicaid is not a strong predictor of proposal, we avoided including a score
family members of home care patients, resource use. Accordingly, we did not for catheter-using patients in the case-
nor should it make patients feel that propose to include a Medicaid variable mix system, out of concern that this
there is some financial stake in how in the case-mix model. Using our final would work against catheter removal at
they report their familial supports analytic data set, we rechecked the the appropriate time. However, for the
during convalescence. We believe that contribution of this variable to explain proposed refinement approach, we did
adjusting payment in response to the home health resource use. We found no include a score in the non-routine
absence of a caregiver would introduce change from what was described for this supplies model out of concern that
negative incentives with adverse affects variable in the proposed rule. Consistent agencies would fail to admit patients
on home health Medicare beneficiaries. with our original policy on this item, we with supplies costs.
We will continue to study the effects of did not include this variable in the final Comment: A commenter objected to
caregivers on the case-mix model. four-equation model of this rule. We the proposal to eliminate M0610
Using our final analytic data set, we will continue to study the effect of dual (behavioral problems) as a case-mix
rechecked the contribution of this eligibles on the case-mix model, and we variable. The commenter noted that
variable to explain home health encourage HHAs to complete M0065 as patients with behavioral problems,
resource use. We found no change from required. including those without formal
what was described for this variable in Comment: A commenter asked that psychiatric diagnoses, consume large
the proposed rule. Consistent with our we evaluate the impact of adding a case- amounts of resources. The commenter
original policy on this item, we did not mix variable for patients aged 85 or asked for further data to support
include this variable in the final four- older, who have greater care needs, and removal of M0610.
equation model of this rule. We will for diabetics. The commenter also Response: We have added case-mix
continue to explore additional expressed concern that providers in scores to the system for psychiatric
refinements to the OASIS instrument to Southern states would be more affected conditions, as they are better markers
gather more information regarding the by proposed policies noted in the for increased resource use related to
roles caregivers play in home health proposed rule, as these parts of the behavioral problems than M0610. When
care and to better quantify any country serve larger populations of two the psychiatric conditions were
unmeasured effects of multiple co- groups at high risk for diabetes. included in the model, M0610 does not
add further predictive power (that is, it M0246. When a fracture code is chronic ischemic heart disease’’ codes
was not statistically significant). assigned to M0246 it will be expected to the case-mix model, with one
Comment: Several commenters asked that the appropriate aftercare V-code exception. We are not including code
that V-codes be included in the case- from V54.1 through V54.8 will be 414.9, ‘‘Chronic ischemic heart disease,
mix diagnosis list as they are assigned to M0230. We note, however, unspecified’’, because this is a
appropriately prevalent in home care that assigning of V54.01, V54.02 and nonspecific code and there are
due to ICD–9 coding guidelines. One V54.09 is considered generally numerous specific codes that we would
commenter suggested V-codes be added inappropriate in the post-acute care expect to be used for this condition. As
as interactions. A number of setting. noted previously, we believe the
commenters also asked for more Comment: The proposed rule implementation of the refined HH PPS
guidance regarding coding, especially in designates the dementia codes 290.0 will better reflect more accurate
the use of V-codes. Several commenters series as manifestation codes in the payments, and we are taking steps to
noted that they have had to hire Psych 2 diagnosis group. A commenter ensure the least amount of burden for
certified coders. stated those codes can only be placed as HHAs.
Response: We have included selected secondary diagnoses, but the proposed Comment: Several commenters noted
codes from the V44 and V55 code rule only offers points when Psych 2 that the neuro 3 code list included ICD–
categories in Tables 2B and 10B. The conditions are primary diagnoses. 9 diagnosis 436, which is an outdated
major use of V-codes in the home health Patients with these diagnoses require code. They asked that it be replaced
setting occurs when a person with a considerable resources even when the with 434.91.
current or resolving disease or injury primary focus of the plan of care is Response: We are aware of the ICD–
encounters the health care system for another diagnosis. Commenters 9–CM changes effective October 1, 2004
specific aftercare of that disease or suggested allowing case-mix points to the classification of unspecified
injury. V-codes are less specific to the when Psych 2 diagnoses are in the cerebrovascular accident (CVA). Before
clinical condition of the patient than are secondary position. this change these conditions were
numeric diagnosis codes. A single V- Response: The ICD–9–CM code indexed to 436, Acute but ill-defined
code could substitute for various category 290, Dementia, codes are listed cerebrovascular disease. In order to
numeric codes each of which describes in the ‘‘Psych 2—Degenerative and other comply with the ‘‘ICD–9–CM Official
a specific different clinical condition. organic psychiatric disorders’’. The Guidelines for Coding and Reporting’’,
For more guidance regarding coding ICD–9–CM code category 290 codes are effective November 15, 2006, we have
especially in the use of V-codes please point bearing regardless of whether the deleted codes in categories 430–437
see the CDC Web site noted below to codes are primary or secondary listed in the ‘‘Neuro 3-Stroke’’
obtain a copy of the ICD–9–CM Official diagnoses. We have removed the diagnostic category of Table 2B of the
Coding Guidelines effective November manifestation designation for these proposed rule. The conditions in
15, 2005. (http://www.cdc.gov/nchs/ codes. categories 430–437 identify the cause of
datawh/ftpserv/ftpicd9/ftpicd9.htm.) Comment: Commenters noted that key the initial onset of an acute stroke and
Comment: CMS currently allows surgical complication codes (996 and must not be assigned in the home health
points for bowel ostomies, but 997 series) have been omitted from the setting.
reimbursement points should be case-mix. These series include joint Agencies should use ICD–9–CM code
allocated to all ostomies. A commenter prosthesis complications, amputation category 438, Late Effects of
suggested we add V55.0–V55.9 to the complications, skin graft complications, Cerebrovascular disease, for conditions
non-routine supply list to capture transplanted organ complications, etc. occurring at any time after the onset of
patients needing supplies for non-bowel They believed these codes should be an acute stroke. The coding guidelines
ostomies. added to the case-mix diagnoses. indicate that these ‘‘late effects’’ include
Response: It is important to note that Response: We disagree. It is not neurologic deficits that persist after the
all ostomies were not included in the appropriate to add these codes to the initial onset of conditions classifiable to
original HH PPS payment because the case-mix because these codes represent 430 through 437. The neurologic deficits
OASIS instrument does not capture all complications that are typically treated caused by cerebrovascular disease may
ostomies, for example, the tracheostomy initially in the inpatient setting. be present from the onset or may arise
is not included in the OASIS Comment: One commenter asked that at any time after the onset of the
instrument. Therefore, we do not have we add 728.87 and 781.3 back to the condition classifiable to 430 through
data for all ostomies. However, we have table of point-bearing diagnosis codes. 437.
tested the non-routine supplies for This commenter also asked that we add To summarize, we deleted diagnosis
stoma conditions for which we have the 414 series of diagnosis codes. codes from Table 2B in the following
added appropriate ‘‘status (V44) V- Response: We disagree with the situations:
codes’’ and ‘‘attention (V55) V-codes’’ to suggestion that 728.87, Muscle • The code was assigned to a minor
the model. weakness (generalized) and 781.3, Lack condition or mild symptom that may be
Comment: A commenter asked that of Coordination, should be added to found in the elderly population;
we include fracture aftercare codes and Table 2B. The conditions assigned to the • The code was a non-specific code or
orthopedic correction codes (V54.01– 781.3 and 728.87 diagnosis codes are • The code could not be assigned
V54.9) as point bearing codes. identified as nonspecific conditions that within the home health setting.
Response: The HH PPS does not rely represent general symptomatic We believe the deletion of these codes
on V-codes, except as mentioned above. complaints in the elderly population as directly correlates with the goals
Therefore we are continuing to require such. We believe inclusion of these stipulated in the proposed rule.
agencies to list the underlying problem codes would threaten to move the case- Specifically, the proposed rule
that led to the V-codes in M0246 of the mix model away from a foundation of stipulated that the case-mix system
OASIS assessment. The numeric reliable and meaningful diagnosis codes avoid, to the fullest extent possible,
fracture codes are listed in Table 2B and that are appropriate for home care. nonspecific or ambiguous ICD–9–CM
are expected to be assigned when We agree with the addition of the codes, codes that represent general
indicated to our optional payment item diagnostic category 414, ‘‘Other forms of symptomatic complaints in the elderly
population, and codes that lack begin including all patients with any time after the onset of the condition
consensus for clear diagnostic criteria constipation symptoms, not just those classifiable to 430–437. Table 2C
within the medical community. The who are more severely affected. includes these codes as deletions from
diagnosis codes listed in Table 2C at the Furthermore, the ICD–9–CM category Table 2B of the proposed rule.
end of section III.B.5 are identified as 564 (Functional Digestive Disorders Not
minor conditions or mild symptoms that Elsewhere Classified) specifically Acute Myocardial Infarction
may be found in the elderly population excludes those clinical conditions that We have also revised code category
or identified as non-specific conditions are more accurately identified by other 410, Acute Myocardial Infarction, in the
and as noted above, have been deleted more specific ICD–9–CM diagnostic
‘‘Heart Disease’’ category of Table 2B of
as point-bearing diagnosis codes. The codes. Therefore, codes 564.00, 564.01,
the proposed rule, to comply with ICD–
following discussion provides further 564.02 and 564.09 have been deleted
9–CM coding instruction (see Table 2C
explanation of the specific changes to from the Gastrointestinal Disorders
the diagnoses occurring in Table 2B diagnostic category in Table 2A (found at the end of section III.B.5 for the list
(also found at the end of section III.B.5): at the end of section III.B.5). Most of the 410 codes to be included). The
• Deletion of constipation and mild, patients with significant constipation code category 410 has been replaced in
unspecified burns; symptoms can be captured with other Table 2B with specific codes from
• Deletion of acute stroke codes ICD–9–CM diagnostic codes that are category 410, (410.x2 ). The specific
(categories 430–437); more specific than the codes for codes designate an episode of care
• Revision of code category 410, constipation. following the initial episode of care. The
Acute Myocardial Infarction and fifth-digit sub-classification of 2 is for
• Addition of code category 414, First Degree Burns use with code category 410 to designate
Other forms of chronic ischemic heart A first degree burn is a minor self- an episode of care following the initial
disease. limited condition that usually requires episode when the patient is admitted for
Constipation no professional medical attention. The further observation, evaluation, or
skin typically displays mild redness treatment for a myocardial infarction
The clinical condition of constipation without blisters. The most common that has received initial treatment but is
(ICD–9–CM codes 564.00, 564.01, example of a first degree burn is mild still less than 8 weeks old.
564.02, and 564.09) was originally sunburn. Neither bandages nor medical
included in the GI group. Occurrences We have also revised code category
supplies are required for first degree
of constipation as a primary diagnosis 045, Acute Poliomyelitis, in the Neuro
burns. This condition is often not coded
were extremely rare. Therefore, the as a diagnosis for medical billing 2-Peripheral Neurological disorders
analysis was conducted with because it rarely requires any section of Table 2B to correlate with
constipation as a secondary diagnosis professional medical treatment. ICD–9–CM coding instructions by
separate from the rest of the diagnoses Therefore the actual frequency of first replacing this code with code 138, Late
in the GI group. The results of this degree burns is underreported in effects of acute poliomyelitis(see Table
analysis show 2, 5, 1, and 5 points from medical claims databases. Because the 2C at the end of section III.B.5).
leg 1 to leg 4, respectively, of the four- severity of this condition is so minimal,
equation model (please see Table 2A at Chronic Ischemic Heart Disease
we do not think it is appropriate to
the end of section III.B.5). However, this include it in the four-equation case-mix We also evaluated the appropriateness
likely reflects selective coding by model. In addition, no medical supplies of code suggestions from commenters,
providers of only those patients with are required for treatment of this and we have inserted codes from ICD–
more severe forms of this condition condition so it would be inappropriate 9–CM code category 414, other forms of
without inclusion of the many patients to include it in Table 10B for Non- chronic ischemic heart disease to Table
with mild constipation symptoms. Routine Supplies. 2B. The only code from category 414
Constipation is both a clinical symptom that was not included is 414.9, ‘‘Chronic
and a medical diagnosis (ICD–9–CM Late Effects of Cerebrovascular Disease
ischemic heart disease, unspecified’’
564). It is relatively common in the To comply with the ‘‘ICD–9–CM due to the non-specificity of the code
elderly population with a prevalence Official Guidelines for Coding and and the fact that we would expect that
ranging from 15 to 20 percent in the Reporting’’, Effective November 15, other codes from this category would be
community setting. The clinical acuity 2006 we have deleted codes in used if appropriate.
of patients with constipation can range categories 430–437 listed in the ‘‘Neuro
from asymptomatic to extreme distress 3-Stroke’’ diagnostic category from Table 2C lists those codes noted above
(including abdominal pain and Table 2B of the proposed rule. The that have been deleted or added to Table
impending bowel obstruction). The conditions in categories 430–437 2B in the proposed rule. Tables 2A, 2B,
ICD–9–CM codes, however, do not identify the cause of the initial onset of and 2C are found at the end of section
distinguish the severity levels of these an acute stroke and must not be II.B.5. We recognize that some HHAs
patients. Since there are no specific assigned in the home health setting. have used ICD–9–CM coding in the past
diagnostic clinical criteria for The ICD–9–CM coding guidelines which will no longer meet future coding
constipation that are widely accepted stipulate the assignment of code standards, as discussed above. For
throughout the medical community, category 438, Late Effects of example, some acute stroke codes were
clinicians are free to assign this Cerebrovascular disease, for conditions recognized in the original case-mix
diagnosis to all patients with even occurring at any time after the onset of system, and we included them in the
minimal symptoms of constipation an acute stroke. The coding guidelines modeling of the refined system finalized
regardless of severity. If additional indicate that these ‘‘late effects’’ include in this rule to capture the effects on the
points were allowed for constipation neurologic deficits that persist after the diagnosis group score. However, we
under the HH PPS, we would expect to initial onset of conditions classifiable to assume that these acute stroke codes
find a large increase in the number of 430–437. The neurologic deficits caused will not be used in the future, and these
patients with this diagnosis simply by cerebrovascular disease may be changes are reflected in the codes listed
because HHAs would be allowed to present from the onset or may arise at in Table 2B.
4. Addition of Therapy Thresholds professional judgment of certified provide a natural disincentive to game
In the proposed rule, for 2008, we agencies and their clinicians to select the system, and that imposing on the
proposed to discontinue the use of a appropriate courses of treatment for regression model a mildly decelerating
single 10-therapy threshold, for the their patients. trend in the resources per added therapy
purpose of payment, and proposed to Comment: Many commenters visit between 6 and 20 therapy visits
implement three therapy thresholds at supported our proposal to have multiple will further mitigate against gaming. We
6, 14, and 20 visits. We proposed using therapy thresholds. However, several detail the resource cost values that
graduated steps (groupings of 1 to 4 questioned the point allocation for impose a decelerating trend in the four-
visits) between these three thresholds to functional variables in relation to equation model in Table 1. We have
provide an equitable increase in therapy. One commenter was concerned updated this table using 2005 data. If a
that this could lead to gaming, where potential problem is detected through
payment that would not otherwise occur
agencies prescribe 14 visits instead of 10 data analysis processes with our RHHIs,
between the three threshold levels. As a
visits, noting that almost all patients then the RHHIs may conduct Medical
disincentive for agencies to attempt to
who need 10 physical therapy or rehab Review of claims identified as potential
reach a therapy level higher than the
visits could benefit from 14 visits. The problems to determine if the services
appropriate, clinically determined
commenter was concerned that the cost rendered were reasonable and
number of therapy visits, we proposed
to agencies would be prohibitive, and necessary.
to decelerate the increase in payment
would force them to replace physical Comment: While supporting the
with each grouping of additional
therapists with physical therapy concept of a graduated therapy
therapy visits between the therapy
assistants, to drop therapy services threshold, several commenters were
thresholds. altogether, or gaming to receive concerned that the reimbursement
For example, if the current proposed
reasonable reimbursement. Another decrease was so substantial. One
model produces an average value for
commenter noted that the dollar commenter noted that his calculations
each additional grouping of therapy
increments between 6 and 14 visits were showed that it would require 17 therapy
visits above 6 and below 14 visits, we
so modest that they may create payment visits under the proposed system to
would incrementally decrease the
deficits. receive the same therapy adjustment as
marginal payment for each grouping of Response: We appreciate the under the current system, when the 10-
therapy visits as the number of therapy comments supporting our multiple therapy threshold is met. The
visits grow. At this time, no study has therapy thresholds. We disagree with commenter noted the resource intensity
been performed to study the clinically the commenter’s concern that our of therapy services, and asked that we
appropriate number of visits primarily increased therapy thresholds will be consider a greater payment allocation
because of the resources required to cost prohibitive and will force providers for visits from 10 to 14. Another
perform such a study. Under fee-for- to replace physical therapists with commenter noted that the new therapy
service Medicare, beneficiaries can physical therapy assistants or to drop thresholds will minimize payment for
select clinicians to treat and act on their therapy services altogether. The goal of orthopedic cases. This commenter
behalf so long as the clinicians meet the the case-mix refinements is to better recommended that the therapy
CoPs, such as licensing (qualified nurses align payment with actual agency costs. threshold be changed to 6, 12, and 20
and therapists), and other forms of Changing to multiple therapy thresholds to allow adequate compensation for
credentialing (CoPs). In the research with a gradual increase in payment therapy visits.
vacuum that exists, the Medicare better aligns costs and payments and Response: The original 10-visit
program relies upon the providers to avoids incentives for providers to distort therapy threshold supported treatment
determine the clinically appropriate patterns of good care. plans involving 10 therapy visits and
number of visits. However, we found Specifically, because we used higher, so one should not expect that
that a payment system with an incentive multiple regression to derive the point weights under the original system for 10
such as the 10-visit-therapy threshold values, with indicator variables for visits would be comparable to weights
indicated that such reliance was therapy visits (for example, 7 to 9 under the new system for 10 therapy
perhaps misplaced. Our revised system therapy visits) included in the visits. Compared to the original system,
of multiple thresholds and smoothing regression model, the point allocations weights under the new system are more
(that is, graduated per-visit payments for functional variables take into precise with respect to the cost of a
between the thresholds) is an attempt to account the range of visits into which given range of therapy (for example, a
reduce the financial incentive that we the treatment plan falls. The point range of 16 to 17 therapy visits). It is
saw as distorting clinically appropriate allocations therefore serve to define important to understand that the
decision making. MedPAC has stated more precisely the average resources regression method modeled the addition
repeatedly that the home health benefit used by a patient given that a certain to total resource cost for treatment plans
would be enhanced by a better range of therapy visits is to be delivered. with each range of therapy visits in
understanding and definition of We are aware that the new threshold of Table 4 of the proposed rule—not just
appropriate clinical standards (e.g., 14 therapy visits may be misperceived the addition to cost from therapy visits.
Report to the Congress: Medicare as a new target for treatment. We do, Therefore, the services utilization
Payment Policy, MedPac, March 2006, however, intend to monitor severity levels cannot be noted strictly
p. 195). We believe it would take years administrative data for indications of as direct costs for added ranges of
of research to determine with sufficient gaming, which could include shorter therapy visits, though the cost of added
precision for payment purposes and lengths for prior therapy visits and therapy visits is certainly very
claims processing what is clinically increased frequencies of episodes with important in producing the values noted
appropriate. We will continue to rely on 14 or more visits without evidence that in Table 4 of the proposed rule and thus
the RHHIs during normal medical an increase in the number of therapy the proposed relative case-mix weights.
review operations to consider therapy visits was appropriate for the patients. The proposal was not intended to
treatment plan appropriateness on a We believe that the need to spend on propose minimized payment for
case-by-case basis. Of course, we also therapy visits, in order to get paid for orthopedic cases, but to reflect to the
continue to rely in good faith on the high therapy treatment plans, will best of our ability the treatment
practices extant in the data for different commenter noted that the Excel toy when developing recommendations for
types of patients and costs experienced grouper produced an increased payment refinements to the HH PPS.
by a wide range of patients in the data of $402 for the seventh visit. Comment: A commenter strongly
analyzed. Response: We cited the source for the disagreed that patients with a high risk
Comment: A commenter stated that starting value of $36 in the proposed of falls should be used as an example of
the variations in payment introduced by rule (72 FR 25364). It was the addition patients with a treatment plan
multiple therapy thresholds were not to total resource cost from comparing commonly requiring 6 therapy visits (72
consistent with a regression model. This episodes with 7 therapy visits to FR 25363). The comment did not
commenter’s initial analysis indicated episodes with 6 therapy visits, based on include an alternate illustration or
that agencies can obtain significant a variant of the four-equation model that example of a common treatment plan
additional payments when they provide allowed for a separate marginal addition requiring 6 therapy visits, however, the
14 therapy visits as opposed to 13 to cost associated with each separate, commenter did agree with us that there
therapy visits when all other OASIS individual number of therapy visits. are therapy treatment plans within the
answers remain constant, even though Thus, this value was entirely data 6 visit range.
the scoring in the 3rd and 4th equations driven, given the entire set of clinical, The commenter stated that ‘‘clinical
is different from the scoring in the 1st functional, and therapy indicator experience with homebound Medicare
and 3rd equations. The commenter variables used in the four-equation patients at high risk for falls indicates
stated that the inconsistencies found in model. In the final version, the updated that these patients typically have
this review make it difficult to analysis yielded a starting value of $42 significant problems with balance and
understand how CMS arrived at the instead of $36. The declining trend was gait. They may also be receiving
proposed increments between HHRGs. modeled by decrements of 1.5 units treatments that elevate their risk,
The commenter asks for additional instead of 1 unit. Please see Table 1 at including the use of diuretics.’’ The
information on how CMS arrived at the the end of this section for details. It commenter is concerned that payment
increments in payment between the should be understood that the resource contractors will apply this example to
various levels of therapy services cost measure is not equivalent to the the medical review process and deny
proposed. average cost of a therapy visit, as it is needed visits to patients at risk for falls
Response: For an early episode, Table derived from national Bureau of Labor who have extensive therapy needs.
4 in the proposed rule indicated that Statistics survey data on the direct
agencies would receive an additional Response: We used the example of
hourly wage and benefit cost of therapy-
$2,191.76¥$1,771.84=$419.42 before patients with a risk of falls as typically
related clinical disciplines in home
wage adjustment for treatment plans receiving six therapy visits based on
care. We convert minutes per episode
involving 14 or 15 therapy visits. For input from Abt Associates, using
reported on claims into resource cost
later episodes, agencies would receive information from their TEP. According
dollars using the national wage and
an additional $2,198.69- to the TEP, physicians may deliberately
benefit data. Table 4 of the proposed
$1,907.93=$290.76. In the final version rule indicated that the therapy order short term plans of care for
of Table 4, which is based on CY2005 increment for services utilization patients because they want the patient
data, agencies would receive an severity S3 encompasses treatment to proceed to outpatient therapy as soon
additional $366.03 for early episodes plans that include 7, 8, or 9 therapy as possible. A short-term plan of care of
and $504.44 for later episodes. These visits. We intend to monitor payments six visits will typically involve
values result from using indicator under the system in the future for evaluation, safety/falls assessment and
variables in the regression for differing evidence that agencies are failing to prevention intervention, with the
ranges of therapy visits (ranges provide the full range of visits included possibility of more than one therapy
indicated in Tables 3 and 4 of the in each S-level. discipline being involved.
proposed rule) and from reintroducing Comment: Several commenters We disagree with the commenter that
the decelerated payments per added questioned our assumption that most the RHHIs will apply the example of
therapy visit at the stage of the payment patients would require 6 to 13 visits and patients with a high risk of falls as a
regression. Our technique for that 14 or more therapy visits would not basis for their decision on the
reintroducing the decelerated payments be normal. They note that therapy determination of coverage. Section
was to estimate a variant of the four- services are resource intensive. A 20.1.2 in Chapter Seven of the Medicare
equation model that did not incorporate commenter disagreed with our Benefit Policy Manual explains the
deceleration. From this, we were able to statement that several common following: ‘‘The intermediary’s decision
compare the added payments for the treatment plans only require about 6 on whether care is reasonable and
proposed ranges of therapy visits with visits, using the example of falls. necessary is based on information
and without deceleration in order to Response: Abt Associates conducted reflected in the home health plan of
adjust the services utilization (S-level) TEP meetings on December 15, 2005 care, the OASIS as required by 42 CFR
marginal resource cost estimates of the and March 14, 2006. These TEP 484.55 or a medical record of the
payment regression appropriately. meetings provided an opportunity for individual patient. Medicare does not
Comment: Several commenters experts, industry representatives, and deny coverage solely on the basis of the
questioned the $36 estimated marginal practitioners in the field of home health reviewer’s general inferences about
cost of adding a seventh therapy visit to care to provide feedback on Abt’s patients with similar diagnoses or on
an episode with 6 therapy visits and the research examining the HH PPS and data related to utilization generally, but
deceleration of payments, as the source exploration of payment policy bases it upon objective clinical evidence
for this information was not cited, and alternatives. Abt received input from regarding the patient’s individual need
the dollars appear to be significantly TEP members as to what the appropriate for care.’’ It is at the discretion of the
below agency costs. One commenter levels for the therapy threshold would contractor to determine the use of its
asks for additional information be based on clinical conditions of home resources. If a potential problem is
regarding how CMS identified an health patients. Different sets of therapy detected through their data analysis
incremental cost of $36 between the 6th thresholds were discussed at TEP processes, then they may conduct
and 7th therapy visits. Another meetings. Abt considered this feedback Medical Review of claims to determine
if the services rendered were reasonable proposed system. The commenter asked As explained in the proposed rule (72
and necessary. that analysis of changes in therapy FR 25388), we collapsed all episodes
Comment: A commenter was under the new system be a key priority with visits over 19 when we saw the
concerned that CMS planned to conduct for future research. The commenter also results of the four-equation model.
automatic medical reviews of every noted that higher payments for third These episodes are grouped in the
episode requiring 20 or more therapy and later episodes appear reasonable, payment regression, and severity
visits. While this commenter agreed that but suggested further research into the distinctions are made using the
such cases are unusual, there was nature of third and subsequent episodes. breakpoints described in that last
concern that the threat of automatic Response: We agree that financial column (20+ therapy visits) of Table 3,
medical review could provide an incentives can affect care provided, and Severity Group Definitions: Four-
incentive for providers to restrict the we will monitor the effects of the equation model (72 FR 25387).
number of visits to individuals who refined payment system. We will be We note the labeling of Table 3 in the
need a higher level of intervention. analyzing changes in therapy under the proposed rule left the impression among
Another commenter asked if HHAs refined system and will conduct further some readers that there was a fifth
should anticipate an increase in therapy refinement research as appropriate. equation. The commenter may have
Additional Documentation Requests Comment: A commenter noted that been confused because Table 3 in the
(ADRs) from the RHHIs, at least adding therapy thresholds in the revised proposed rule shows a separate column
initially, as we validate the case-mix regression model improved the for all episodes with 20 or more visits,
appropriateness of the new therapy ability of the model to predict resource which can give the appearance of a five-
thresholds and the accuracy of provider use, with substantially increased R- equation model rather than a four-
coding. The commenter noted that squared for both early and later equation model. However, there are
increases in ADRs lead to unfunded episodes, as compared to the R-squared only four equations from which to draw
increases in administrative costs, even if values for a single therapy threshold case-mix points. Table 2A of the
they result in no adjustments. model (72 FR 25365, May 4, 2007). The proposed rule gives a description of
Response: The intermediary’s commenter asked what the improved R- each diagnosis group, followed by four
decision on whether care is reasonable squared values were, and if they were columns with the four ‘‘legs’’ of the
and necessary is based on information statistically significant. Further, the four-equation model. If an episode has
reflected in the home health plan of commenter asked if there were concerns 20 or more visits, the case-mix points
care, the OASIS as required by 42 CFR that the randomness being measured would come from the second leg if it is
484.55 or a medical record of the was truly not random, which would an early episode, and from the fourth leg
individual patient. Medicare does not raise questions about the if it is a later episode. The table column
deny coverage solely on the basis of the appropriateness of a linear regression headers indicate that these two legs are
reviewer’s general inferences about model and its associated R-squared. for 14 or more therapy visits. As
patients with similar diagnoses or on Response: Abt Associates estimated explained in the proposed rule, we
data related to utilization generally, but models without therapy thresholds found strong similarities in the case-
bases it upon objective clinical evidence using the basic four-equation structure. mix-adjusted costs for early and later
regarding the patient’s individual need The basic four-equation structure episodes with 20 or more therapy visits.
for care. As mentioned above, it is at the incorporates a threshold at 14 therapy In other words, the results of the four-
discretion of the contractor to determine visits. After adding thresholds to this equation model indicated that predicted
the use of its resources. If a potential model at 6 and 20 visits, and adding costs for the same clinical and
problem is detected through their data per-visit therapy variables, the R- functional severity levels across the two
analysis processes, then they may squared statistic increased by equations (equations 2 and 4) were
conduct Medical Review of claims to approximately 0.10. We subsequently highly similar. Therefore, to reduce the
determine if the services rendered were modified the approach to the per-visit number of groups and thereby simplify
reasonable and necessary. therapy variables, as described in the the system at the payment regression
Medical review targets problem areas proposed rule. We believe the linear stage, we treated episodes with 20 or
which demonstrate significant risk to model is appropriate based on results of more therapy visits the same (that is, we
the Medicare program as a result of experimentation with nonlinear used the same indicator variables for
inappropriate payments, over- specifications during the research. This clinical and functional severity,
utilization, abusive billing and technical topic is treated in the Abt regardless of whether the episode was
unnecessary services. Here, the Associates Final Technical Report. from the early or later equation for 14
Medicare Contractors (RHHIs) use Comment: A commenter noted that plus therapy visits).
different parameters to target their the four-equation model actually In summary, upon examining the CY
review of home health claims. The contains a fifth equation for 20 or more 2005 data on the resource cost trends by
decision regarding which claim to therapy visits and asked for clarification number of therapy visits, we changed
review depends on the information regarding how to code as early or later the starting value for the marginal cost
obtained from data analysis which episodes in this case. of going from six therapy visits to seven
includes all providers submitting claims Response: The OASIS item for early therapy visits from $36 to $42,
for payment. A provider’s claims may be or later episodes (M0110) needs to be consistent with the observed value in
subject to review if they do not meet the completed for all episodes, regardless of the data. The declining trend was
coverage, coding, and billing guidelines the number of therapy visits. The modeled by decrements of 1.5 units, as
contained in the statute, regulations, estimated number of therapy visits must shown in Table 1, because the marginal
coverage guidance, CMS manuals, and also be entered into OASIS (M0826). value observed in the data was no
contractor policies. The episode will then be assigned an higher than $30 when going from 14 to
Comment: A commenter noted that appropriate HHRG by the grouper, and 15 therapy visits. Had we used
providers are sensitive to financial priced out correctly by the Pricer. The decrements of 1.0 units, as in the
incentives associated with therapy system will automatically verify the proposed rule, the imposed values
visits, but that it is difficult to anticipate accuracy of the early/later designation, would have descended to a value of $34,
how utilization may change under the and correct the payment if necessary. which is less consistent with the
observation when going from 14 to 15 We are implementing the three of modeling with the newer, most
therapy visits. Using 1.5-unit therapy thresholds of 6, 14, and 20. The current 2005 data. The deceleration of
increments, the imposed values groups of visits in final Table 1, used to the increase in payment with each
descended to a value of $29, which is achieve graduated steps of increased individual visit between the therapy
more consistent with the actual data. payment between the therapy thresholds is being implemented as in
thresholds, have not changed as a result the final Table 1 (see below).
TABLE 1.—RESOURCE COST VALUES IMPOSING DECELERATION TREND IN FOUR-EQUATION MODEL
Resource cost
Number of values imposed
Equation and services utilization severity level therapy visits in in regression
severity level procedure
1st and 2nd Episodes, 6–13—Therapy Visits:

S3 ............................................................................................................................................................. 7, 8, 9 42, 40.50, 39
S4 ............................................................................................................................................................. 10 37.50
S5 ............................................................................................................................................................. 11, 12,13 36, 34.50, 33
1st and 2nd Episodes, 14–19—Therapy Visits:
S1* ............................................................................................................................................................ 14*, 15 *, 29
S2 ............................................................................................................................................................. 16, 17 27.50, 26
S3 ............................................................................................................................................................. 18, 19 24.50, 23
3rd+ Episodes, 6–13—Therapy Visits:
S3 ............................................................................................................................................................. 7, 8, 9 42, 40.50, 39
S4 ............................................................................................................................................................. 10 37.50
S5 ............................................................................................................................................................. 11, 12, 13 36, 34.50, 33
3rd+ Episodes, 14–19—Therapy Visits:
S1* ............................................................................................................................................................ 14*, 15 *, 29
S2 ............................................................................................................................................................. 16, 17 27.50, 26
S3 ............................................................................................................................................................. 18, 19 24.50, 23
* No value was imposed in the regression procedure for a 14th therapy visit (because the regression intercept estimate for the grouping step
automatically includes the resource cost impact of this visit).
5. Determination of Case-Mix Weights weights amongst all episodes. They stay patients is not made at the expense
claimed that this would simplify the of shorter-stay patients, as our data
In the proposed rule, we revised the model and eliminate the difficulties of analysis showed a modest difference in
case-mix weights, as noted in the determining early or later status of resource cost over the 60-day
previous sections of this final rule with patients using the CWF. One commenter certification period. That some patients
comment period, describing the proposed that we use a two-equation at the start of care need frequent visits
refinements. In this section, we describe model that excludes reference to is accounted for in our data by the
the final revisions to the case-mix model enhanced reimbursement for the third resource cost measure for the entire 60-
and the determination of the final case- and fourth episodes. The commenter day period. We agree that agencies
mix weights. For specifics, see the tables suggested that not having increased should follow best practice guidelines
at the end of this section. reimbursement for later episodes would that are intended to bring about early
Comment: A number of commenters more accurately reflect the way the independence and avoid hospital
supported the higher case-mix weights majority of patients are receiving care readmissions by front-loading visits
for third and subsequent episodes of and reduce the incentive to drive up when appropriate. Further, we do not
care. However, two commenters were costs and possibly reduce patient believe the payment incentives
concerned that the analysis weighted independence. associated with the long-stay equations
third and subsequent episodes more Response: The later episodes reflect are so strong as to that they distort the
highly because Medicaid data is patients who tend on average to have fundamental goals of returning patients
included in the OASIS (M0150), and higher resource needs and extended to health and independence as soon as
Medicaid patients account for 85 stays in home health care. The later possible.
percent of all third and subsequent episode distinction resulted from our Comment: A commenter asked if the
episodes. They noted that most agencies attempts to differentiate the resources M0230/240/246 case-mix scores can
have fewer than two episodes per needed by long-stay patients. Many now be combined or should only the
patient, and would be adversely affected observers in the past indicated it would highest case-mix score be considered in
by the proposed weights. Another noted be appropriate for the case-mix system evaluating the clinical dimension. The
that patients new to home health often to recognize that the Medicare home commenter asked that we clarify Table
have a high degree of anxiety, and health benefit serves a minority who are 2A of the proposed rule, and asked how
therefore need more frequent contact. experiencing an extended period of to handle episodes with 20 or more
Additionally, ‘‘best practice’’ guidelines illness and incapacitation. It is not visits. Another commenter asked if only
recommend a higher level of care during possible to always identify all these those co-morbidities that are actually
the first few weeks of a home health cases upon admission, and an being addressed in the care plan are to
episode. This commenter asked CMS to administratively feasible way to address be included.
reconsider a payment adjustment based this situation is to create a provision Response: Case-mix scores from
on early rather than later episodes. specifically for these cases when they different diagnosis groups in Table 2A
Several commenters suggested reach a milestone indicative of an are additive; a diagnosis group is a line
eliminating the early or later episode extended stay in home care. The item in the table. Points cannot be given
distinction and redistributing the provision for separate groups for long- more than once for diagnoses in the
same group. For example, a patient with F5 based on having 0 to 30 or more OASIS assessments, a period before V-
both heart disease and hypertension points. Under the proposed four- codes were allowed on OASIS. For
would not get points twice for item 11 equation model, an episode receives a validation, we used a random 20%
in Table 2A. However, a patient with a functional score of severity level F1, F2, sample of 2005 claims linked to OASIS
Neuro 3 diagnosis who meets criteria for or F3 based on having 0 to 10 or more assessments to create an analytic file for
points for Items 16 and 17 in Table 2A points, and depending on the episode modeling case-mix. We examined the
would be eligible for points from both timing and number of therapy visits. diagnoses fields on the OASIS
items. A summary of the guidelines Because the models are not directly assessments (M0230/M0240/M0245) for
used in scoring is posted at the CMS comparable, it cannot be assumed that indications that some diagnoses groups
home health Web site and entitled ‘‘Toy fewer points under the proposed model in the proposed model might be
Grouper Logic Guidelines’’ (Web site results in a negative payment reported at differing rates in 2005 than
address: http://www.cms.hhs.gov/ adjustment. in 2003, and we did find some changes.
center/hha.asp). In the footnote to the The points given in Table 2A of the For example, we observed lower rates of
final Table 2A, we have clarified that proposed rule were derived from reporting primary diagnoses for the
scores are additive. modeling actual claims data, and neurological diagnosis groups,
In addition, the commenter may have represent prior experience in home orthopedic groups other than gait
been confused because Table 3 shows a health care. The score is the value of the abnormality, cardiac group, and some of
separate column for all episodes with 20 regression coefficient for the variable, the cancer diagnosis codes. We observed
or more visits, which can give the and measures the impact of the data somewhat higher primary diagnosis
appearance of a five-equation model element on total resource cost of the rates for the diabetes, hypertension, and
rather than a four-equation model. episode. For this final rule with degenerative and other organic
However, there are only four equations comment period, we updated the psychiatric groups. Secondary diagnosis
from which to draw case-mix points. dataset using 2005 data in the regression reporting typically decreased only by
Table 2A gives a description of each analysis, and this resulted in some about 1 percentage point for each of the
diagnosis group, followed by four changes in the scores presented in Table proposed diagnosis groups. Moreover, a
columns with the four ‘‘legs’’ of the 2A of this rule. We will also continue preliminary validation of the model on
four-equation model. If an episode has to study the case-mix model, and will FY 2005 data indicated that the results
20 or more visits, the case-mix points make additional refinements as needed. were substantially the same as the
would come from the second leg if it is Comment: A commenter noted that it results of modeling resources in the
an early episode, and from the fourth leg appears that some individual items in four-equation structure using FY 2003
if it is a later episode. The table column Table 2A of the proposed rule have the data. We concluded that the proposed
headers indicate that these two legs are potential to move the clinical dimension four equation model in the proposed
for 14 or more therapy visits. from the lowest (C1) to the highest (C3).
rule was reliable notwithstanding
Comment: A number of commenters Response: This is correct. We
reporting changes expected from the
expressed concern about the impact of determined the points based on our
research. One example would be an introduction of V-codes on OASIS. We
changes made to the point allocation for
early episode with a primary diagnosis made a number of refinements based on
OASIS functional variables in
in the skin 1 group (item 25 in Table the validation model we estimated using
relationship to therapy. The current
2A); diagnoses in this category are the FY 2005 analytic file. We
case-mix system allocates 6 to 9 points
resource intensive. subsequently updated the data to CY
for M0700 (ambulation) deficits.
Comment: Several commenters asked 2005 and made some further
However, the proposed case-mix
that we clarify the reason for linking the refinements. The final results are shown
refinement system allocates zero points
for ambulation deficits in two of the case-mix adjustment for 781.2 (gait in Tables 1, 2a, and 3. The R-square
three equations, including both abnormality) with pressure ulcers. statistic for the final case-mix model is
equations for 14 or more therapy visits. Persons receiving therapy for gait 0.45.
Two commenters also noted that the training are not typically bed or chair Major differences in the 2005 data
point allocation for M0690 (transfers) bound and therefore it is unlikely that compared to the 2003 data concerned a
were affected unless the patient they would have pressure ulcers. small number of the primary and
required 13 or more therapy visits. They Additionally, points are not allocated secondary diagnosis groups we
were concerned that the proposed new for the gait disorder diagnosis in the 14 identified for the case-mix model in the
case-mix methodology was not plus therapy visit equations. proposed rule: Cancer and psychiatric
capturing the appropriate points to Response: The regression model conditions [affective and other
allow for necessary resources for indicated that patients with pressure psychoses, depression (Psych 1 Group)
functionally impaired patients. The ulcers are overall more clinically and degenerative and other organic
commenters proposed that CMS study compromised if they also have the psychiatric disorders (Psych 2 Group)].
this further before imposing a negative diagnosis of 781.2 than pressure ulcer When we examined the model’s
adjustment. patients without the diagnosis of 781.2. estimates of cancer-related marginal
Response: The proposed four- As to the points allocated for this type resources and marginal resources of the
equation model cannot be compared on of patient, because we are adopting a Psych 1 group, we found that a
a point-by-point basis with the current graduated payment for therapy in the 14 distinction between primary and
case-mix model. The models are based plus visit category, the gait disorder secondary diagnoses was not needed, as
upon different data sets, and the model diagnosis does not add any additional scores were generally similar across the
structures are different (for example, a explanatory power to the model and is equations. For Psych 2, only primary
single equation model versus a four- not statistically significant. diagnoses contributed to this group in
equation model; a single therapy In summary, in the proposed rule, we the proposed rule model. However, the
threshold versus multiple therapy stated our intention to update the data updated estimates indicated secondary
thresholds). Under the current model, used for the four-equation model and diagnoses should be recognized in the
an episode receives a functional score validate the model. We based our model, so we combined secondary with
severity level of F0, F1, F2, F3, F4, or proposal on FY 2003 claims and linked primary diagnoses into a new group for
these psychiatric conditions. Because Interactions involving the other three changes for specific scores. For
these changes eliminated distinctions neurological groups also reflected some example, a primary diagnosis of
between primary and secondary changes. For example, we found that diabetes incurred an increase of one
diagnosis positioning on OASIS M0230/ separating the interactions of functional point in three of the four equations,
M0240, we welcomed them as a limitations with multiple sclerosis while the interaction of stroke and
simplification of the case-mix model. (Neuro 4) into two line items in the dysphagia incurred a loss of one point
We also believe there are advantages proposed table 2A did not work well in in the third equation and a gain of one
from moving away from separate scores the new data, despite results obtained point in the first equation.
for primary and secondary diagnosis with the data used for the proposed We tested a suggestion from a
reporting. Specifically, it reduces rule. However, combining all four commenter to include V-codes from
potential incentives to alter the functional limitation interactions ICD–9–CM for stoma. We defined
placement of codes based on financial recognized in the proposed model variables using selected V-codes to serve
considerations. The final model produced useful results. Based on as markers for patients with stoma other
includes two diagnosis groups with estimates from the new data, we also than colostomies and gastrostomies,
differing scores for primary and modified the interaction of toileting which were already measured or
secondary diagnoses: Diabetes and with the remaining neurological groups, proxied in our variable set. This change
certain skin conditions [specifically, brain disorders and paralysis (Neuro 1) resulted in the addition of two major
traumatic wounds, burns, and post- and peripheral neurological disorders types of stoma. Specifically, we added
operative complications (Skin 1)]. (Neuro 2). The data revealed that appropriate variables in both the case-
peripheral neurological disorders mix model and the NRS model to
In addition, we added stroke (‘‘Neuro (Neuro 2) in this interaction were no capture patients with resource needs or
3’’ diagnosis group) as a primary longer statistically significant, so this supplies cost needs due to tracheostomy
diagnosis, irrespective of any group was removed from the and urostomy/cystostomy. We are
interactions. The final result in the interaction. implementing as final the case-mix
updated data of using this re-defined In the 2005 data, a cost-increasing weights and scoring resulting from the
stroke variable was an added score in effect from incontinence was not four-equation model with therapy
equation 2 of the model (early episodes, observed, so it was deleted from the thresholds at 6, 14, and 20 therapy visits
14 or more therapy visits). Along with four-equation model. An interaction in and with an early or later episode
this change, the data revealed some the proposed model involving distinction. We have updated our
differences in the cost-increasing incontinence and certain neurological modeling to use 2005 data, which
interactions with stroke, which are conditions [brain disorders and resulted in some changes in case-mix
reflected in the final model. The final paralysis (Neuro 1) was no longer weights and item scoring. We are
model indicates added points when statistically significant, so this variable implementing as final the versions of
stroke is accompanied by dressing and/ was removed as well. Tables 2A, 2B, 2C, 3, 4, and 5 that are
or ambulation functional limitations, as Other differences in the four-equation shown below.
well as dysphagia. model generally were small point BILLING CODE 4120–01–P
ER29AU07.000</GPH>
ER29AU07.001</GPH>
ER29AU07.002</GPH>
ER29AU07.003</GPH>
ER29AU07.004</GPH>
ER29AU07.005</GPH>
ER29AU07.006</GPH>
ER29AU07.007</GPH>
ER29AU07.008</GPH>
ER29AU07.009</GPH>
ER29AU07.010</GPH>
ER29AU07.011</GPH>
ER29AU07.012</GPH>
ER29AU07.013</GPH>
ER29AU07.014</GPH>
ER29AU07.015</GPH>
ER29AU07.016</GPH>
ER29AU07.017</GPH>
ER29AU07.018</GPH>
ER29AU07.019</GPH>
ER29AU07.020</GPH>
ER29AU07.021</GPH>
ER29AU07.022</GPH>
ER29AU07.023</GPH>
ER29AU07.024</GPH>
ER29AU07.025</GPH>
ER29AU07.026</GPH>
ER29AU07.027</GPH>
ER29AU07.028</GPH>
ER29AU07.029</GPH>
ER29AU07.030</GPH>
ER29AU07.031</GPH>
ER29AU07.032</GPH>
ER29AU07.033</GPH>
ER29AU07.034</GPH>
ER29AU07.035</GPH>
ER29AU07.036</GPH>
ER29AU07.037</GPH>
ER29AU07.038</GPH>
ER29AU07.039</GPH>
ER29AU07.040</GPH>
ER29AU07.041</GPH>
ER29AU07.042</GPH>
ER29AU07.043</GPH>
ER29AU07.044</GPH>
ER29AU07.045</GPH>
ER29AU07.046</GPH>
ER29AU07.047</GPH>
ER29AU07.048</GPH>
ER29AU07.049</GPH>
BILLING CODE 4120–01–C baseline from which we measured the As a result of various studies, analysis
6. Case-Mix Change Under the HH PPS increase in case-mix. The time period of OASIS data, and changes to the home
was free from any anticipatory response health benefit as due to the BBA, we
Section 1895(b)(3)(B)(iv) of the Act to the HH PPS, and data from this time stated our belief that change in case-mix
specifically provides the Secretary with period were used to develop the original of 13.4 percent between the time of the
the authority to adjust the standard HH PPS model. Also, this is the only Abt Associates case-mix study and the
payment amount (or amounts) if the nationally representative dataset from end of the HH IPS period reflected
Secretary determines that the case-mix the 1997 to 1998 time period that substantial change in the real case-mix.
adjustments resulted (or would likely measured patient characteristics using
result) in a change in aggregate In contrast to that 13.4 percent, we
an OASIS assessment form comparable considered that the 8.7 percent increase
payments that is the result of changes in to the one currently adopted for the HH
the coding or classification of different in the national case-mix index between
PPS. However, agencies included in this the HH IPS baseline and the CY 2003
units of services that do not reflect real sample were volunteers for the study
changes in case-mix. The Secretary may could not be considered a real increase
and could not be considered a perfectly in case-mix. Trend data on visits from
then adjust the payment amount to representative, unbiased sample.
eliminate the effect of the coding or the proposed rule (72 FR 25393),
Furthermore, the response to Balanced resource data presented in the proposed
classification changes that do not reflect Budget Act of 1997 provisions such as
real changes in case-mix. rule (72 FR 25394), and our analysis of
the home health interim payment
In the proposed rule, in order to changes in rates of health characteristics
system (HH IPS) during this period
identify whether the adjustment factor on OASIS assessments and changes in
might produce data from this sample
was needed, we first determined the reporting practices all led to our
that reflect a case-mix in flux; for
current average case-mix weight per conclusion that the underlying case-mix
example, venipuncture patients were
paid episode. The most recent available of the population of home health users
suddenly no longer eligible, and long-
data from which to compute an average term care patients were less likely to be was essentially stable between the HH
case-mix weight, or case-mix index admitted. Therefore, we were not IPS baseline and CY 2003. Our research
(CMI), under the HH PPS was from confident the trend in the CMI between showed that HHAs have reduced
2003. Using the most current available the time of the Abt Associates study and services while the CMI continued to
data from 2003, the average case-mix 2003 reflected only changes in coding rise. In addition to the trend analysis,
weight per episode for initial episodes practices due to real change in case-mix. we conducted several additional kinds
is 1.233. To proceed with the CMI We then looked to the HH IPS of analyses of data and documentary
adjustment, next we determined the baseline period, the 12 month period materials related to home health case-
baseline year needed to evaluate the ending 9/30/2000. Analysis of a 1- mix coding change. The results
trend in the average case-mix per percent sample of initial episodes from supported our view that the change in
episode. the 1999 through 2000 data under the the CMI since the HH IPS baseline
There were two different baseline HH IPS revealed an average case-mix mostly reflected provider responses to
years that were considered from which weight of 1.125. Standardized to the the changes that accompanied the HH
to measure the increase in case-mix: 1) distribution of agency type (freestanding PPS, including particulars of the
A cohort that used home care from proprietary, freestanding not-for-profit, payment system itself and changes to
October 1997 to April 1998 (the Abt hospital-based, government, and SNF- OASIS reporting requirements. Our
case-mix study sample which was used based) that existed in 2003 under the analyses indicated generally modest
to develop the current case-mix model) HH PPS, the average weight was 1.134. changes in overall OASIS health
and 2) the cohort that used home care We noted this time period was likely characteristics between the two periods
during the 12 month period ending not free from anticipatory response to noted above, a specific pattern of
September 30, 2000 (HH IPS Baseline). the HH PPS, because we published our changes in scaled OASIS responses that
The increase in the average case-mix initial HH PPS proposal on October 28, was not indicative of material
using the Abt Associates case-mix study 1999. The increase in the average case- worsening of presenting health status,
sample as the baseline was 23.3 percent mix using this time period as the various changes in the OASIS reporting
(from 1.0 to 1.233). There were several baseline was 8.7 percent (from 1.134 to instructions that helped account for
advantages to using data from Abt 1.233; 1.233–1.134=0.099; 0.099/ numerous coding changes we observed,
ER29AU07.050</GPH>
Associates case-mix study as the 1.134=0.087; 0.087*100=8.7 percent). and a large increase in post-surgical
patients with their traditionally lower of nominal case-mix change and adjust they would expect OASIS data shown in
case-mix index. the national standardized 60-day Table 10 of the proposed rule to change
Therefore, based upon our trend episode payment rate accordingly for accordingly.
analysis we believed the change in the any nominal change in case-mix that Response: Our identification of case-
case-mix index between the Abt case- might occur. We proposed to implement mix change was based on a number of
mix sample (a cohort admitted between a 3-year phase-in of the total downward factors that revealed coding changes to
October 1997 and April 1998) and the adjustment for nominal changes in case- higher clinical, functional, or utilization
HH IPS period (the 12 month period mix by reducing the national severity without an actual change in the
ending September 30, 2000) is due to standardized 60-day episode payment status of home health patients. These
real case-mix change. We took this view, rate by 2.75 percent each year up to and are described in detail in the HH PPS
even though we understood that there including CY 2010. That annual proposed rule (72 FR 25392–25422).
could be some issue as to whether this reduction percent was based on the new Since publication of the proposed
period was affected by case-mix change current estimate of the nominal change rule, we updated our analysis to use 100
due to providers anticipating, in the last in case-mix that occurred between the percent of the HH IPS file for our
year of HH IPS, the forthcoming case- HH IPS baseline (+0.099) and 2003. baseline and a 20-percent sample of
mix system, with its incentives to However, we also stated that, if, at the 2005 claims data. We used all episodes
intensify rehabilitation services. The time of publication of the final CY 2008 rather than just initial episodes. This
change from these two periods is from HH PPS rule, updates of the national change in our sample selection
1.00 to 1.134, an increase of 13.4 claims data to 2005 indicated that the approach does not materially change the
percent. However, we did not propose nominal change in case-mix between estimate of case-mix change, whether
to adjust for case-mix change based on the HH IPS baseline and 2005 was not comparing the baseline to HH PPS 2003
this change in values, as some of that +0.099, we would revise the percentage or HH PPS 2005. The 2005 data yielded
change reflected real change in case- reduction in the next year’s update. The an average CMI of 1.2361, as compared
mix. However, we did propose that the revision would be determined by the to the average CMI of 1.0960 from the
8.7 percent of case-mix change that ratio of the updated 3-year annual 100 percent HH IPS sample. Therefore,
occurred between the 12 months ending reduction factor to the previous year’s the updated change measurement is
September 30, 2000 (HH IPS baseline, annual reduction factor. For example, (1.2361 ¥1.0960)/ 1.0960 = 12.78
CMI=1.134), and the most recent the scheduled annual reduction factor percent. As explained in the summary at
available data from 2003 (CMI=1.233), was estimated to be 0.9725 (equivalent the end of this section, where we
be considered a change in the CMI that to a 2.75 percent reduction); for CY 2008 describe the results of the Abt
does not reflect a ‘‘real’’ change in case- we would multiply this reduction factor Associates model we used to identify
mix, but rather is a ‘‘nominal’’ change by the ratio of the updated reduction real case-mix change, we adjusted this
in case-mix. We proposed a reduction in factor to 0.9725. Therefore, for the CY result downward by 8.03 percent to get
HH PPS national standardized 60-Day 2010 rule, which would govern the third a final case-mix change measure of
episode payment rate to offset the and final year of the proposed case-mix 11.75 percent (0.1278 * (1¥0.0803) =
change in coding practice that has change adjustment transition period, we 0.1175). To account for the 11.75
resulted in significant growth in the would obtain the CY 2007 national percent increase in case-mix which is
national case-mix index since the average CMI to compute the updated not due to a change in the underlying
inception of the HH PPS that is not value for nominal case-mix change health status of Medicare home health
related to ‘‘real’’ change in case-mix. adjustment. Again, we would form the patients, we are finalizing the proposed
Our past experience establishing other ratio of the updated adjustment factor to 2.75 percent reduction of the national
prospective payment systems also led us the previous year’s effective adjustment standardized 60-day episode payment
to believe a proposal to make this factor. The annual updating procedure rate for 3 years beginning in 2008 and
adjustment for nominal change in case- avoids a large reduction for the final extending that adjustment period to a
mix was warranted. In other systems, year of the phase-in, in the event that fourth year via a 2.71 percent reduction
Medicare payments were almost the CY 2007 national average case-mix for 2011. We are seeking comment on
invariably found to be affected by index reflects continued growth since the 2.71 percent case-mix change
nominal case-mix change. We CY 2005. adjustment for 2011.
considered several options for We stated our plan to continue to We have conducted several analyses
implementing this case-mix change monitor changes in the national average to determine if any portion of the above
adjustment. Those options included CMI to determine if any adjustment for case-mix change measurement could be
accounting for the entire ¥8.7 percent nominal change in case-mix is considered real versus nominal, i.e. not
increase in case-mix with an 8.0% warranted in the future. related to real change in the essential
adjustment in CY 2008, incorporating an Comment: A number of commenters underlying health status of the home
adjustment of ¥5.0 percent in CY 2008 asked that we eliminate the 2.75 percent health user population. First, Abt
and an adjustment of ¥2.7 percent in case-mix change adjustment. They Associates developed a model to predict
CY 2009, or incorporating an adjustment argued that the acuity of home care the case-mix weights on large samples
of ¥4.35 percent in CY 2008 and an patients is rising, citing earlier which is described at the end of this
adjustment of ¥4.35 percent in CY discharges from hospitals or skilled section. The model accounted for
2009. However, because of the potential nursing facilities. A number of changes in the age structure of the home
impact our proposed adjustment might commenters argued that patient health user population, and changes in
have on providers, we proposed and characteristics have changed, with more the types of patients being admitted to
requested comment on whether to patients 85 and older receiving home home health. To account for changes in
adjust for the nominal increase in health care, along with more patients the types of patients, we used four main
national average CMI by gradually with resource intensive diagnoses. classes of variables: Variables describing
reducing the national standardized 60- Several commenters noted the increase (1) the utilization of Medicare Part A
day episode payment rate over 3 years. in patients with knee or hip services in the 120 days leading up to
During that period we stated that we replacements. Another noted that if home health, (2) the type of
would continue to update our estimate providers were inflating the case-mix, preadmission acute care stay when the
patient last had such a stay, (3) variables percent. The share of total patients home health caseload due to these
describing living situation, and (4) admitted with CHF acute discharges groups is not large enough to drive the
variables summarizing Part A declined from 3.31 percent to 2.62 national case-mix nominal average to
expenditures in the 120 days leading to percent, a decline of 21 percent. The the CMI levels reached in our follow-up
home health. The variables for changes share of total patients admitted with year, 2005. Further, we have taken the
in the type of acute care stay classified CVA acute discharges declined steadily, contribution of this effect into account
stays into APR DRG case-mix groups, a from 1.52 percent to .97 percent, a one- in the Abt Associates case-mix model
classification system that incorporates a third decrease. Admissions for hip described above and at the end of this
severity classification for each case-mix replacements exhibited no clear trend; section.
group, basic type of stay (procedure the range of rates during the period is While we have seen an increase in the
versus medical) indicator, and risk of between 1.36 percent and 1.64 percent. proportion of patients with diabetes
mortality indicators during the stay. We For these conditions, the results are not according to OASIS diagnosis coding
also incorporated a set of variables clearly indicative of more severe case- information, our research showed that
describing agency ownership and mix. HHAs have reduced services while the
organizational form, to adjust for the We note that admissions for knee case-mix index continued to rise. We
large effect on measured case-mix from replacements are rising, from 1.89 identified a dramatic decline in the
the change in the types of agencies that percent to 2.75 percent in the years from number of home health visits per 60 day
occurred since the HH IPS baseline. The HH IPS to 2005. However, the overall episode (Table 6). The average number
model is described in detail at the end percent of knee replacement patients in of visits per episode in 2005 was 20.53,
of this section. the national home health caseload is not compared to 26.88 during HH IPS.
The results of the analysis indicated large, at less than 3 percent at any given After adjusting for wage and benefits
that a small amount of measured case- time. We accounted for the change in growth (by holding wage and benefit
mix change is real, but that most of it the share of caseload due to knee estimates constant at FY 2000 levels),
is unrelated to the underlying health replacement patients in the Abt we find that average resource costs have
status of home health users. Associates case-mix model using the declined slightly from 1999 to 2005,
Second, some commenters suggested APR DRG classifications, described from $451.39 to $447.41 (see Table 7).
that HHA patients have more resource above and at the end of this section. The For most of the calendar quarters
intensive diagnoses. We conducted results from the model indicated that displayed in Table 7, average resource
analyses using FY 2000 through CY this change, in combination with other costs after adjusting for wage growth
2006 data for several conditions changes that were offsetting, was not were substantially below the HH IPS
emblematic of home health patients. enough to move the real case-mix index baseline. At the same time, the case-mix
The analyses indicated that admissions more than a small amount beyond the indexes at admission and for total
to home health agencies were down baseline. episodes have increased (see Table 7).
slightly for persons with hip fractures, Third, we examined the length of time Resource costs are based on visit time
congestive heart failure, and between discharge and the home health reported on claims, and thus are labor-
cerebrovascular accidents. These results episode start, to develop evidence that, related. If the CMI is increasing,
are shown in Table 8, ‘‘Percent Share of on average, patients enter home care in suggesting that patients are more
Home Health Admissions and Mean a more sickly condition than was the clinically severe, have more functional
Time Prior to Entering a Home Health case in FY 2000. Table 8 shows the impairments, and require more visits,
Episode, for Five Conditions, FY 2000– average number of days between acute we would have expected resource costs
CY 2005’’. Estimates are based on a 10 care discharge and the first day of the to increase as well. However, by 2005
percent random sample (n=388,684 to home health episode for patients with average resources per episode were still
522,973, depending on the calendar acute discharges due to the same five below HH IPS levels, after adjusting for
year; statistically these are considered conditions: Hip fracture, congestive wage growth. Notably, it is not until
large samples). The data for CY2006 heart failure, cerebrovascular accident, 2005 (when, according to Bureau of
come from the first quarter of the year hip replacement, and knee replacement Labor Statistics wage survey data, wages
only. We used total episodes, both surgery. The results show no change in rose significantly), that unadjusted
initial and recertification episodes, for the mean time prior to entering a home resources are significantly higher than
this analysis. As our previous analysis health episode for the first three the HH IPS baseline level.
on the 1 percent HH IPS sample and the conditions. We believe this result partly Comment: Several commenters noted
20 percent CY 2003 sample indicated no reflects increased use of institutional that the growth in Medicare Advantage
significant shift in the balance between post-acute care among the home health (formerly known as Medicare + Choice)
initial and non-initial episodes, we population. Specifically, there was an programs has shifted low acuity patients
believe that the annual rates and means increased use of SNFs and LTCHs out of traditional Medicare, leaving
in the table are appropriately measured, between the HH IPS baseline and CY those patients with higher needs in
and account for the complete mix of 2000. SNF stays grew by 2.8 percent, traditional Medicare. They felt this
patients seen by agencies. For defining and SNF days of stay grew by 8.5 contributed to an increase in the average
the type of acute discharge, we used the percent. LTCH hospital days grew by 38 case-mix index.
same definitions that were used in a percent. IRF stays and days did not Response: Medicare Advantage
CMS study cited by one commenter who grow, but IRF use is only one-third that programs provide managed care benefits
noted that increases in knee of SNF use among home health patients. which are different from the traditional
replacement patients have occurred As shown in Table 8, days prior to Medicare benefit. For further
(CMS, ‘‘Medicare Beneficiary Access to entering home health declined for hip information on these managed care
Rehabilitation Care,’’ June 8, 2007). replacement and knee replacement benefits, please refer to the Internet only
According to Table 8, the share of total patients. As commenters have manual 100–01, ‘‘Medicare General
patients admitted to HHAs with hip suggested, these statistics may reflect Information, Eligibility, and
fracture acute discharges in the 14 days less use of post-acute institutional care Entitlement’’, chapter 5, subsection 80.
leading up to home health declined over on average for these two groups. This manual is available on CMS’ home
the period, from .82 percent to .59 However, the increasing share of the health Web Site at http://
www.cms.hhs.gov/center/hha.asp. Our analyses in the proposed rule to coding improvement, so there should
These managed care programs were not reviewed information pertaining to be a prospective adjustment as well. The
considered in our analysis of the case- changes in OASIS guidance and commenter suggested CMS consider a
mix change adjustment as they are potential coding improvements that may combined (retrospective and
separate benefits from traditional have resulted. In August 2000 official prospective) case-mix change
Medicare. We cannot make comparisons guidance on OASIS coding affected a adjustment for this rule that would be
or draw conclusions based upon any number of case-mix items. Functional taken over a longer period of time.
benefit other than traditional Medicare. items began to emphasize the patient’s Furthermore, the commenter suggested
Comment: Many commenters felt that ability to perform the item safely. This CMS should also continue to evaluate
the 2.75 percent case-mix change may have caused several ADL statistics coding changes in future years to
adjustment failed to account for OASIS to shift away from the completely determine if additional coding
training on accurate assessment and on independent level. Another August improvement is occurring.
OASIS use. The commenters felt this led 2000 change in OASIS instructions Response: While we agree it would be
to OASIS scores which reflect a more affected the pain item, M0420. beneficial to have a more systematic
accurate picture of the home health Additional strategies for assessing pain approach to measuring changes in
patient rather than case-mix up-coding. were offered, and guidance on whether OASIS coding practices, to do so in a
Two commenters noted that there was the pain was well controlled took into manner suggested by the commenter
systematic undercoding prior to training account patient adherence to pain would require significant new
and guidance on OASIS and diagnosis medication. Many patients trade off resources, especially since the methods
coding. Some commenters argued that pain control for diminution of involve primary data collection. We will
CMS has benefited from agency medication-related side-effects. These explore methods to examine agency
undercoding, resulting in agencies changes likely increased the number of coding practices. To make the best use
underpaying themselves. patients assessed with pain. The OASIS of administrative data, rather than
Response: We agree that some of the instructions regarding assessment of expensive-to-collect primary data, we
changes seen in OASIS characteristics urinary incontinence were also intend to analyze changes in
are partly due to emphasis on proper expanded to consider mobility and relationships among types of resources
application of OASIS guidelines. We cognition, which may have led to used in the episode, by case-mix group
also believe that there were incentives increased rates of reporting of this item. and type of patient, controlling for the
driven by payment and quality program Furthermore, in August 2000 there most reliable measures of patient
changes that interacted with the were two changes to the OASIS manual condition available. This may provide
subjective aspects of the assessment that could have increased the number of evidence to supplement our monitoring
process to cause nominal coding patients with surgical wounds. First, the of resources presented in the proposed
changes. Diagnosis coding entails some definition of a surgical wound was rule and this regulation. We will
discretion by the Agency: In some cases expanded to include medi-port sites and continue to monitor average minutes per
more than one diagnosis could other implanted infusion devices or visit reported on claims. We will also
reasonably be called primary. Thus, we venous access devices. Therefore more monitor changes in the comorbidities
believe the significant growth, for skin openings could be assessed as reported alongside primary diagnoses,
example, in orthopedic diagnoses partly wounds under M0488, a case-mix item, to assess changes in relationships
reflects the financial incentives that provided the site is the most among the diagnoses reported on
colored the diagnosis selection process. problematic. The second change OASIS. We will examine diagnosis
Our examination of National Claims allowed a muscle flap performed to coding and OASIS item coding for
History data revealed an increase in surgically replace a pressure ulcer to be coding improvements as well as abuses.
Medicare knee replacement patients. considered a surgical wound, and not a We agree that the refinements will
However, these patients account for pressure ulcer. This again would have present another opportunity for case-
only about 2.75 percent of the national added to the number of surgical mix change due to coding
home health caseload at any given time. wounds. improvements. We did not pursue a
With such a small share of the caseload, All the above we believe indicates prospective adjustment for nominal
they do not drive the case-mix index by that the increased reporting rates seen in case-mix change because we believe it is
themselves. Hip replacement patients some OASIS items do not represent a subject to error. We believe our proposal
did not increase as a share of episodes change in underlying health status of to phase in adjustments based on
by 2006, although their share appeared HH PPS patients. Numerous retrospective analysis is an appropriate
to increase slightly between HH IPS and commenters noted that they had response. Phasing in adjustments limits
CY 2003 (see Table 8). However, changed OASIS coding as a result of the demands for operational
Medicare hip replacement patients also training. This is consistent with adjustments by agencies. Our
are not a large factor in the overall home nominal versus real change in patient retrospective approach is consistent
health caseload, accounting for only characteristics. with this regulation’s request for further
between 1.36 percent and 1.64 percent Comment: A commenter wrote that in comment from the public on the fourth
of episodes in the years 2000 to 2006. future, it would be beneficial to have a year of case-mix change adjustment,
Further, ADL functioning can be more systematic approach to measuring which is based on results of our
difficult to assess due to variability changes in OASIS coding practices. For empirical analysis since the proposed
within patients and the multiple example, CMS should consider efforts rule was issued.
dimensions of functional limitations. such as the collection of OASIS from Comment: A commenter noted that
Quality measures and financial independent entities for comparison to the proportional increase in therapy
incentives may combine to bias agencies agency assessments or on-site visits to services is due to both a decrease in
towards assessing a patient with a more- check agency coding practices. The other services and the underutilization
severe rating at the start of care. commenter noted that the need for of therapy services in past episodes of
Incentives apparently led to high- better data is particularly acute because care prior to HH PPS. Additionally, the
therapy treatment plans, aided by the this rule will present another use of therapists in collaboration with
10-therapy threshold. opportunity for case-mix increases due nurses has helped ensure more accurate
coding of the OASIS, particularly in the incentives. Implications of the analysis similar resource use than under the
functional component area. of case-mix change performed by Abt current system.
Response: We agree that there has Associates suggest the shift to more Comment: Several commenters wrote
been a shift toward rehabilitative intensive therapy plans cannot be that the average annual per patient
services, which increased the explained by changes in patient health expenditures for home health services
proportion of therapy services relative status. dropped from 2001 to 2003, and
to skilled nursing or home health aide We recognize and appreciate the therefore do not suggest that case-mix
services. This suggests there may have contribution of therapists in weights are increasing.
been some substitution of therapy collaboration with nurses in ensuring Response: Data from the annual
services for nursing services and OASIS coding accuracy. As noted Medicare & Medicaid Statistical
perhaps for home health aide services. previously, increases in coding accuracy Supplement indicate that annual
We have not identified any studies contribute to nominal case-mix change. payments per user of home health
substantiating the idea that therapy was Improvement in coding accuracy has services have actually increased from
underutilized, nor have we identified also occurred with the introduction of $2,936 in the year 2000 to $4,314 in
studies indicating that the dramatic other prospective payment systems. 2005. Our analysis clearly shows that
drop in aide services undoubtedly Comment: Several commenters felt the average case-mix weights have
means that aides were overutilized. One the 2.75 percent case-mix change increased. Generally, payments per user
unpublished study of the service adjustment was based upon a flawed are affected by increases in the billed
reductions during HH IPS suggests that analysis, with an insufficient sample case-mix weights and by annual rate
beneficiaries who were financially size. They cited the reduction in the updates.
better off did increase their use of Comment: From 2000 to 2003, HHAs
model’s R-squared along with
privately paid care services as a result altered care practices to achieve
MedPAC’s report that the coefficient of
of the reduction in services which came improved patient outcomes, shifting
variation was greater than 1 for 60 of the
about during the HH IPS period. from dependency-oriented care to care
80 case-mix groups. designed to achieve self-sufficiency and
Whether this indicates that services
Response: Based on the updated independence. The increased use of
were previously overprovided is unclear
analysis, the final case-mix change therapy services and decreased use of
(McKnight, Robin, ‘‘Home Care
measurement was based upon 100 home health aides are indicative of this
Reimbursement, Long-term Care
percent of HH IPS claims and a 20- change. Changing to multiple therapy
Utilization, and Health Outcomes,’’
percent sample of 2005 HH PPS claims, thresholds to align payment incentives
NBER Working Paper Series, Working
a greater number of HH IPS claims than with care and the use of a case-mix
Paper #10414, National Bureau of
used in the proposed rule. Both absolute change adjustment that primarily
Economic Research, Cambridge, MA
sample sizes are considered quite large reflects growth in therapy utilization is
April 2004). Accordingly, review of the
in statistical terms. Therefore sample an unnecessary adjustment that
studies does not enable us to draw a
firm conclusion about which types of size can no longer be considered an ‘‘double-dips’’ on rate adjustments.
services could be characterized as issue in the case-mix change adjustment Response: One goal of the case-mix
under- or overutilized before HH PPS. calculation. We did not use the refinements is to better match payments
However, the implications of the results regression model cited by the with agency costs. Changing to multiple
of the Abt Associates model of case-mix commenter to determine the amount of therapy thresholds with a gradual
change (described at the end of this the case-mix change adjustment; increase in payment better aligns costs
section) are that during HH PPS however we used regression analysis to and payments and avoids incentives for
agencies provided more therapy to model the case-mix index, relying on a providers to distort patterns of good care
patients than they did under HH IPS, set of variables that were independent of that would occur at each proposed
and that most of this increase cannot be agency coding incentives (see the therapy threshold. As a disincentive for
explained by changes in patient health analysis description at the end of this agencies to provide more care than is
status. section). appropriate, we proposed that any per-
In response to this comment, we We also note that the commenter’s visit increase incorporate a declining,
measured the growth in utilization of reliance on the MedPAC comments is rather than constant, amount per added
any therapy services and therapy misplaced as the MedPAC comments therapy visit. The final case-mix change
services above the 10 visit threshold, dealt with a review of the case-mix adjustment addresses nominal case-mix
among total episodes between HH IPS refinements and not of the case-mix change that occurred between the HH
and HH PPS. We found during HH IPS change adjustment. MedPAC’s IPS baseline and 2005, and our adjusted
that 39.90 percent of episodes involved comments, which are publicly available, calculation of that nominal case-mix
therapy services, compared to 50.45 state that MedPAC did not change allows for a real increase in case-
percent of episodes during CY 2005. independently assess the case-mix and mix that reduces the nominal
However, the proportion of episodes patient data in our analysis of case-mix measurement by 8.03 percent. The
using therapy services at a level of 10 change. However, MedPAC analyzed the multiple therapy thresholds and the
visits or more changed from 17.0 refinements in the proposed rule, case-mix change adjustment are
percent to 26.4 percent. Thus, therapy including an analysis of the coefficient unrelated and do not doubly adjust the
utilization at or above the 10 visit of variation (CV). Their CV analysis rate as each adjustment is clearly
threshold grew twice as fast as therapy found that the proposed system yields warranted by the data.
utilization below the 10 visit threshold. more internally homogeneous HHRGs Comment: Some commenters stated
These statistics show that the great bulk with less within-in group variation in their belief that incentives in HH PPS
of the growth in therapy utilization was the number of visits provided. They led many agencies to seek out higher
at or above the ten visit therapy reported that the average CV fell from case-mix cases and avoid lower case-
threshold. 0.81 for the current system to 0.75 for mix cases to maximize reimbursement
We believe the data indicate that the proposed system, and that the drop following HH PPS implementation.
agencies’ therapy treatment plans were in CV meant that the new resource They agreed this would create real case-
strongly influenced by financial groups can better identify episodes with mix change versus nominal change.
Response: In the Abt Associates were not incentivized to create real the case-mix equally with those whose
analysis of changes in the case-mix case-mix change until after HH PPS case-mix was inflated. A more equitable
index, the model controlled for changes implementation. The commenter approach would be to reduce
in health status of home health patients, believed that a review by CMS of its proportionally the proposed cut for
measured independently of the OASIS. data during the HH IPS period would those agencies whose individual case-
From that analysis, we identified a allow it to document the subset of HHAs mix weight was below the mean in the
small amount of real case-mix change whose case-mix was not responsive to study period. Several commenters noted
between the HH IPS baseline and 2005. HH IPS. that their average case-mix remained
An analysis by MedPAC in 2005 Response: CMS has done analysis that stable or declined since HH IPS.
(‘‘Home Health Agency Case-mix and accounts for real case-mix change after Another commenter asked for a ‘‘hold
Financial Performance,’’ MedPAC, HH PPS implementation, and only a harmless’’ provision for the non-profit
Washington, DC, December 2005) small amount of real case-mix change or other efficient HHAs where the case-
addressed the possibility that reductions occurred. The analysis takes the mix index is less than 1.
in total visits per episode along with commenter’s idea into account (see the Response: We believe that it is more
shifts in resources among the case-mix end of this section for details). That is, appropriate to implement a nationwide
groups after HH PPS began gave the case-mix model we used to predict approach to the issue of case-mix
agencies the ability to realize higher real change in case-mix measures the change adjustment. An individual
margins on some case-mix groups national level of real case-mix by CY agency approach would be
(particularly high-therapy case-mix 2005, using CY 2005 data on home administratively burdensome and
groups, with their high weights) more health patients’ characteristics. We difficult to implement. Policies to
than for others. However, while margins compared these results to the national address the identity of agencies in light
may have become advantageous among average from the HH IPS baseline year, of changes to organizational structures
some of the case-mix groups after HH and found that a small increase in real and configurations would need to be
PPS began, we believe, based on the case-mix had occurred. developed. Furthermore, smaller
data, that the real case-mix of those The commenter suggested that some agencies might have difficulty in
groups changed very little. agencies were not incentivized to make providing accurate measures of real
Comment: A commenter argued that case-mix change until the case-mix change because of their small
the underlying premise of the HH PPS implementation of the HH PPS. We
caseloads.
system was to control Medicare home believe that it is more appropriate to
Comment: A commenter wrote that
health utilization through an episodic implement a nationwide approach to
CMS’s findings of coding ‘‘creep’’
payment because CMS was unable to the issue of case-mix change
among other provider types (long term
define appropriate and efficient visit adjustment. As noted previously, an
individual agency approach would be care hospitals, inpatient rehabilitation
levels. Therefore, he believed it is
administratively burdensome and facilities, and acute care hospitals)
inconsistent to recognize the expected
difficult to implement. Policies to discredit the agency’s conclusion that
reduction of visits under HH PPS but
address the identity of agencies in light HHA case-mix change is due to nominal
argued that real case-mix change did not
of changes to organizational structures change rather than real change. Another
occur during that period. He noted that
and configurations would need to be commenter wrote that CMS’ case-mix
such a position demonstrates that the
developed. Furthermore, smaller change findings were consistent with
HH PPS did not increase the efficiency
of care delivery. agencies might have difficulty in the prior experience of other
Response: Our initial analysis in the providing accurate measures of real prospective payment systems.
proposed rule indicated that agency case-mix change because of their small Response: We agree with the
coding practices changed for a variety of caseloads. comment that our case-mix change
reasons, including improved coding, Comment: A commenter noted that findings are similar to those seen in
changes in OASIS instructions, specific CMS asserts that OASIS items not used other prospective payment systems. Our
issues (such as confusion about healing for payment were more stable than those conclusion that case-mix change is
status of surgical wounds and effects of used to increase HH PPS payment. The almost completely due to nominal
education in the proper use of trauma commenter stated that if these items change is based upon multiple analyses
codes in the ICD–9–CM classification reflect patient severity, then these items of health characteristics, of resource
system), as well as financial incentives. should be included in the HH PPS costs, and consideration of other factors
The subsequent Abt Associates analysis payment formula. such as the effects of the Balanced
of real case-mix change reinforced the Response: Our process of selecting the Budget Act of 1997. Regardless of
conclusion that very little of the coding case-mix items was explained in the HH similar findings of nominal change
change reflected real case-mix change. PPS Final Rule, implementing the HH among other provider types, the HH
The trend in resources diverged PPS (65 FR 41193). Essentially, not all specific analyses utilized here show that
dramatically from the trend in the items on the OASIS were equally a case-mix change adjustment in HH
average case-mix weight, particularly important in explaining case-mix, and PPS is appropriate.
through 2004 (see Table 7), without any not all items on the OASIS were equally Comment: Several commenters noted
commensurate link to evidence appropriate to use in a payment system. that the proposed case-mix change
concerning home health cost of care. That does not mean such items are adjustment will cripple home health
Comment: A commenter felt that CMS irrelevant in understanding the health agencies’ ability to survive and compete
assumes that all legitimate change in status of the home health user at a time when home health is the only
case-mix ended with the population. hope for an affordable national health
implementation of HH PPS because the Comment: Several commenters wrote approach. They noted that the nursing
HH IPS created sufficient incentives to that by using the average case-mix shortage and rising fuel costs have
maximize all real case-mix change. This weight, CMS is equally cutting payment driven up agency costs and made it
rationale fails to consider that 20 to both high and low average case-mix difficult for agencies to attract and
percent of HHAs had such high cost agencies. This across-the-board cut retain staff. One commenter believed
limits under HH IPS that these agencies would punish those who did not inflate these costs more than compensate for
any coding ‘‘creep’’ that may have more education and training would help occurred. Implementation of a case-mix
occurred. bring about better coding. He noted change adjustment does not depend on
Response: We share the commenters’ there are differences in FI the effect of the HH PPS refinements
concerns about the nursing shortage and implementation, interpretation, or proposed. We believe that the
rising fuel costs. However, case-mix follow-up related to ICD–9 coding. refinements will better match payments
change is based upon actual patient Response: We recognize that there to costs and have already tested this
characteristics and is not to be used to have been improvements in coding using claims data.
compensate for cost differentials. practices, and we encourage home Comment: Several commenters
Comment: Several commenters noted health agencies to follow ICD–9–CM suggested that the case-mix change
that the shift to high therapy episodes guidelines in coding patient diagnoses. adjustment resulted from the FIs failing
(with 10 or more visits) accounts for Home health coding guidance is to do their jobs. One suggested that the
over 70 percent of the change in case- available on CMS’ Home Health Web appropriate way to resolve upcoding
mix from 1999 to 2003. This occurred Site at http://www.cms.hhs.gov/center/ issues is through medical review. If
because those patients requiring more hha.asp, under ‘‘Billing/Payment’’ and medical review occurred and upcoded
therapy visits are in more clinically and then under ‘‘Home Health Coding and episodes were then adjusted, the case-
functionally severe conditions than Billing’’. ICD–9–CM official coding mix change adjustment is essentially
those who do not. The commenters guidance is available from the Centers ‘‘double-dipping’’, taking back dollars a
recommended that this effect be for Disease Control Web Site at: http:// second time. Another commenter writes
excluded from the case-mix change www.cdc.gov/nchs/datawh/ftpserv/ that there is no medical review data
adjustment calculation and that the ftpicd9/ftpicd9.htm. CMS staff supporting an industry wide pattern of
remaining case-mix change adjustment continues to meet regularly with FI case-mix upcoding. One commenter
be eliminated entirely to recognize the representatives to resolve coding issues suggested we focus on audits and
additional costs to HHAs for training as they arise. recovery of inappropriate payments
staff and making operational Comment: A commenter noted that rather than implement a case-mix
modifications as a result of the CMS assumed relative stability of change adjustment. Another argued that
refinements that are not reimbursed. resource utilization that should have therapy services increases in the case-
Response: Our analysis of OASIS been already matched by a mix weight change has the character of
items in Table 10 of the proposed rule corresponding stability in the case-mix a retroactive claim denial without a
indicated basic stability in the health index. Thus, the commenter believed claim review.
characteristics of HHA patients. Our there is an assumption by CMS that
subsequent analysis of case-mix change Response: Medical review affects such
agencies had perfect understanding and
found a small amount of real change, a small proportion of paid claims that
application of OASIS at the time HH
and therefore, we modified the case-mix we do not believe taking it into account
PPS was implemented.
change adjustment accordingly. Response: CMS did not assume would materially affect the estimate of
Given that more therapy sources were agencies possessed perfect nominal coding change, nor did we rely
provided, the implication of our understanding of OASIS or lesser upon it in performing our case-mix
analysis of real change in case-mix is understanding of OASIS. We based our change adjustment analysis. When we
that more therapy was provided to case-mix change adjustment on the initially reviewed the National Claims
substantially the same patient mix that evidence that patient health status did History files to check for adjustments to
agencies served in the HH IPS period. not change substantially even though HHRGs from medical review, we found
We consider the refinements to be improved understanding of and error in the field containing the
evolutionary, not a paradigm shift in our application of OASIS occurred. information. We decided not to use this
payment methodology. For example, we Comment: A commenter wrote that field in correcting the HHRGs on paid
have added only one new item from the the 2.75 percent case-mix change claims in our research files. However,
OASIS, the item on injectable adjustment rate is really higher because we did correct errors in OASIS item
medication use. In addition, we our calculation is based upon the 2007 M0175 (concerning the patient’s
dropped M0175 from the case-mix base rate after adjusting it for the market preadmission stay history) in our
algorithm, in part due to the challenges basket increase and for outliers. analyses. The statute provides authority
faced by agencies in accurately Response: The case-mix change to take into account and adjust for
ascertaining the information needed for adjustment was correctly applied in the changes in case-mix coding not due to
M0175. Furthermore, we dropped other process of determining the budget changes in the underlying health status
items because they are no longer useful neutral expenditure target in our of home health patients.
in explaining resource use (see payment simulation for the refined HH Comment: One commenter noted that
discussion of changes to the case-mix PPS system. The statute provides that the venipuncture patients who were no
model scoring table, Table 2A, in any case-mix change adjustment be longer eligible for Medicare home health
section III.B.5). Thus, we believe the applied to the national standardized 60- care due to BBA changes had a very low
commenter overstated the impact on day episode payment amount, which case-mix. Their loss from the Medicare
agencies of having to adjust to the includes the market basket update and home health patient population would
refinements. While these case-mix adjustment for outliers. cause the overall average case-mix to
refinements will entail staff training and Comment: Several commenters increase. This could account for some
operational modifications, we believe suggested that we evaluate the impact of portion of the increase in case-mix seen.
the refinements as implemented will the coding changes before implementing Another commenter asked if
result in a better alignment of costs to any case-mix change adjustment or that venipuncture patients were included in
payments, which should benefit the we use claims data to test the impact of the baseline HH IPS sample.
agencies. the coding changes, and make this Response: We accounted for the loss
Comment: One commenter suggested available. of venipuncture patients by using the
that the case-mix change was due to Response: The case-mix change last year of HH IPS as our baseline. At
clinicians determining the ICD–9 coding adjustment is designed to address the such time agencies would have
under the HH PPS, and suggested that case-mix change which has already complied with the changes in patient
eligibility requirements, and this would the time period. Moreover, resources effects of the home health prospective
have been reflected in our claims data. rose steadily throughout most of the payment demonstration used 6 months
Comment: Several commenters noted time period, and these increases are of data on expenditures to control for
that the cost reports do not reflect all compensated through market basket general health status [‘‘The Impact of
agency costs, which included those for updates. Home Health Prospective Payment on
telehealth, that have improved care and Comment: Several commenters were Medicare Service Use and
outcomes. If all agency costs were concerned about the absence of Abt’s Reimbursement’’, Mathematica Policy
included, CMS would see an increase in Technical Report, which made analysis Research, Princeton, N.J., December
resource costs which corresponds to the of the proposed case-mix change 2000]. We chose to use 4 months’ of
increase in the case-mix index. Another adjustment and case-mix refinements data on Part A expenditures in part
commenter wrote that resource costs difficult. because there is no consensus, and our
actually decreased early in HH PPS and Response: We understand the available analysis files captured this
then increased. commenter’s desire for Abt’s Technical measure. We decided to avoid using
Response: The statute does not Report, but note that due to OASIS measures in the model (except
provide payment for Medicare home unanticipated difficulties in completing for reported living situation) in favor of
health services provided via a a useful draft, we were unable to issue measurements external to the home
telecommunications system. Section that report. We intend to issue the final health providers, namely irrefutable
1895(e)(1) of the Act provides that report when it is completed and that the demographic measures, National Claims
telehealth services do not substitute for final draft to be useful to the lay reader. History Part A utilization measures, and
in-person home health services and are We expect that the results will be based hospitalization-related patient
not considered a home health visit for on highly technical analyses that characteristics. As previously noted, we
the purposes of eligibility or payment necessitate careful attention from the lay also adjusted for the change in types of
under the Medicare home health public. We will provide a link to Abt’s Medicare agencies that followed the
benefit. As stated in 42 CFR 409.48(c), report on our Web Site once the report start of HH PPS. We believe that there
a visit is an episode of personal contact is available. is little useful analysis that can be
with the beneficiary by staff of the HHA, Comment: Another commenter garnered from separately measuring
or others under arrangements with the asserted that therapy utilization is the dynamic factors in the Medicare system
HHA for the purposes of providing a most important patient characteristic in that impact the nature of patients served
covered HH service. The provision the case-mix model, but that therapy in home health care. The model we use
clarifies that there is nothing to utilization is discounted in the case-mix measures the actual characteristics of
preclude an HHA from adopting change adjustment analysis. The patients that are in the agency caseload,
telemedicine or other technologies that commenter contended that if therapy and is the best reflection of the case-mix
they believe promote efficiency, but utilization were considered a patient in the HHA.
those technologies will not be characteristic, it would explain most of Comment: A commenter was
specifically recognized or reimbursed by the increase in the average case-mix concerned that because LUPA episodes
Medicare under the home health index, and thus the case-mix change retain their original case-mix, they may
benefit. adjustment could be reduced or be contributing to the increase in the
Our measure of resource costs for eliminated. The commenter suggested average case-mix index.
home health is based upon total minutes that CMS withdraw its proposed case- Response: LUPA episodes were not
of time reported on the claim for each mix change adjustment for 2008, 2009, used in the measurement of case-mix
discipline’s visits. Resource costs result and 2010. Furthermore, CMS should change in either our analysis or in the
from weighting each minute by the design and implement an evaluation Abt Associates model of real case-mix
national average labor market hourly method to analyze changes in case-mix change.
rate for the individual discipline that weight that utilizes proper standards Comment: A commenter wrote that if
provided the minutes of care. Bureau of related to the home health relevant 1.233 actually represented average
Labor Statistics data are used to derive factors in the analysis such as changes Medicare case-mix in 2003, then the
this hourly rate. The sum of the in per patient annual expenditures, average payment, per 60-day episode,
weighted minutes is the total resource patient clinical, functional, and service would have been $2,856. The
cost estimate for the claim. This method utilization data, and dynamic factors in commenter asked that CMS disclose
standardizes the resource cost for all the Medicare system that impact the their average 2003 payment amounts for
episodes in the analysis file. This nature of patients served with home all paid episodes, inclusive of full term
method assumes that the non-labor costs health care. and those experiencing downcode
per episode are proportional to the labor Response: We believe that the Abt adjustments.
costs. Our payment rates with an annual Associates case-mix model was Response: It is not clear how the
market basket updates since the initial developed to measure real changes in commenter got the figure of $2,856. The
HH PPS final rule (July 3, 2000) are case-mix addresses this critique. In standardized national rate per 60-day
designed to reflect the agency’s costs. response to the suggestion in the episode for CY 2003 was $2,159.39. If
Telehealth costs are not part of the comments from the National the commenter multiplies this figure by
home health market basket and thus do Association for Home Care and Hospice, the average case-mix weight for 2003 of
not contribute to the annual updates. we used patient expenditures on Part A 1.233, the result is $2,663 before any
Market basket updates are also intended services in the 4 months leading to the wage adjustment. The $2,663 also does
to account for the changes in wages. home health episode, rather than the not include any adjustments for LUPAs,
Table 7 indicates the trajectory of total of annual expenditures suggested PEPs, or SCICs. The average case-mix
resource costs, with and without in the comment. Studies in the field are weight, of 1.233 from the proposed rule,
adjustment for wage growth. The data not consistent in defining a time period for 2003 is calculated after taking
do indicate that resource costs did for measuring this variable, which is downcoding adjustments but is only
decrease at the beginning of HH PPS. used to serve as a proxy for health calculated from initial episodes.
Adjusted resources remained flat until status. For example, a study by Downcoding adjustments are taken
approximately the last six quarters of Mathematica Policy Research of the when the Request for Anticipated
Payment (RAP) reports a high-therapy population excluded costly long-term were known to be unreliable (M0230,
case-mix group, but the final claim does patients who were embraced by HH PPS M0460). Additionally, the commenter
not. Using a 10 percent sample of 2003 from 2000 to 2003. The commenter noted that the OASIS instrument was a
paid claims data, the average payment noted that the problem with the source of error because it was designed
per initial episode is estimated to be proposed refinements is the case-mix to measure outcomes by asking nurses
$2,614. This figure includes the effects adjuster’s inability to cope with therapy to assess the ability of a patient to do a
of the wage adjustment, as well as the utilization by long term users, not the task, as compared to a performance-
downward effect of adjustments for absence of these patients from the based measure.
SCICs, PEPs, and outliers. system. The commenter cited an April Response: As we have noted, we
Comment: A commenter suggested 2000 GAO report which contends that it refined the case-mix model to better
that CMS re-evaluate the coding of has been difficult to develop a case-mix address some of the concerns expressed
M0488, surgical wounds, as the adjustment method that adequately by the commenter. In the proposed rule,
increased incidence of the early/partial described resource use, particularly for we summarized the case-mix model’s
granulation response is not an example long term users. ability to predict resource use with the
of up-coding only. Rather, it is due to an The commenter noted that by measure of model fit known as the R-
increased understanding of how to statutory directive, HH PPS was crafted squared statistic. We explained that the
appropriately code items per OASIS to ensure quality access to all eligible original HH PPS regulation’s model was
guidelines. beneficiaries; by regulatory design, case- based on initial episodes only. We used
Response: This is an example of mix adjustment was engineered to initial episodes because of sample size
nominal coding change due to improved remove incentives for providers to limitations of the original Abt study
coding practices. As noted in the ostracize expensive patients. The sample of 90 agencies. When we began
proposed rule, we recognized the commenter asserted that CMS’ refinement research using claims from
contribution of such sources of change conclusion that patient characteristics the National Claims History, we added
in determining and assessing the case- remained essentially stable is in direct later episodes to the analysis samples.
mix change adjustment. conflict with the goal of HH PPS to We found that the overall R-squared
Comment: A commenter disputed that create a payment system which would statistic of the original HH PPS case-mix
the average case-mix weight of Abt allow equitable treatment of HH IPS- model after adding the later episodes to
model was 1.0, and argued that the excluded patients and thus create a the HH PPS-period analysis samples
timeframe includes a period in which population that was fundamentally was 0.21. Our data analyses indicate
real case-mix change occurred. different than that which existed in the that the R-squared before adding later
Therefore, the commenter asserted that HH IPS baseline year. episodes to the sample is higher than
the statute does not allow an Response: First, we noted that after 0.21; we reported in the proposed rule
adjustment. the BBA, venipuncture-only patients, that the R-squared statistic on initial
Response: By construction, the who were often the long-term users, episodes was reduced to 0.29 by 2003.
average case-mix weight of the original were no longer eligible for the home The R-squared statistic was originally
Abt model was equal to 1.0. This means health benefit. The exclusion of these 0.34 in the Abt study sample, as noted
that we used the case-mix group patients helped stabilize the in the July 3, 2000 Final Rule (65 FR
assignments in the original Abt case-mix characteristics of the home health 41193). It should be understood that the
study’s sample of episodes, and divided patient population. Second, we are later episodes are a minority of episodes
each group’s average resources by the unclear as to the commenter’s statement (29 percent). Therefore, the model still
overall sample average. Using this that the intent of the HH PPS was to adequately fits approximately 71
approach, the average case-mix weight create a different population group. percent of all episodes.
from this procedure must then be 1.0. High-therapy patients were not absent Furthermore, we disagree with the
The sample was selected to be from the national caseload during the suggestion that the OASIS instrument
representative of home health agencies final year of the HH IPS period. We note was a source of large error. The case-mix
nationally, but we were reliant on here again, as we did in the proposed measure is based on OASIS items, and
volunteers for the study. According to rule, that the utilization of therapy was the scientific reliability of OASIS items
statistical theory, it is highly likely that climbing rapidly during the last year of has been studied. OASIS items used in
another sample of volunteer agencies the HH IPS. Therapy utilization the case-mix model generally have good
selected to be nationally representative continued to climb after HH PPS began. reliability. Item M0460, Stage of most
using the same selection procedure Even if we were to agree that the goal problematic pressure ulcer, and item
would have produced similar estimates of the HH PPS was to redress the M0230/M0240, Diagnoses and severity
of resource cost. It is impossible to possible exclusion of certain high-cost index, have ‘‘substantial’’ reliability,
know how different the 1998 to 2003 patients during the HH IPS, we also note according to a report prepared for CMS
trajectory of the average case-mix weight that our model predicting change in the by the Center for Health Services
might be had other agencies’ data been real case-mix accounts for a possible Research in Denver, Colorado (Volume
available. That is, one reason why we return of HH IPS-excluded patients to 4, OASIS Chronicle and
selected a baseline other than the Abt the system. Recommendations, OASIS and
Associates study sample. Choosing the Comment: A commenter believed that Outcome-based Quality Improvement in
HH IPS baseline allowed us to use a errors built into the original case-mix Home Health Care, Feb. 2002). In this
consistent sample of agencies and one adjuster are so large that it is impossible report, a rating system commonly used
that is nationally representative, to reasonably carve out an 8.7 percent in reliability research was used. A
irrespective of whether any agencies case-mix change adjustment. The ‘‘substantial’’ reliability rating was
would be prepared to volunteer for a commenter noted that service utilization assigned if the weighted Kappa
study. accounted for 62.5 percent of the reliability statistic or percent agreement
Comment: A number of commenters estimated predictive power of the was at least 0.61. For these two items,
felt that HH patient characteristics were original model, the actual R-squared the reliability values were at least 0.70.
not stable. One commenter noted that factor for all episodes was 21.9, and In summary, the performance of the
the baseline 1999 to 2000 HH IPS several significant weighting factors original case-mix model is strong
enough to define a case-mix change a shift to higher-cost visits and a general the model. In effect, the model assumes
adjustment. The measure of model fit reduction in visits. that the population’s real case-mix
comparable to the original one from the Comment: The proposed rule stated would have evolved to the predicted
Abt case-mix study has declined that CMS expected the growth in the levels if HH IPS had continued beyond
somewhat, as might be expected over case-mix index to be accompanied by October 2000, or had HH PPS not been
time. Yet the model fit has remained more consumption of services, but that implemented. The independent
adequate for a strong majority of instead CMS measured slightly lower variables (noted below) used to make
episodes. The OASIS assessment items resource consumption. A commenter the predictions purposely do not come
have acceptable reliability. So we noted that this conclusion does not from OASIS (with one exception, family
disagree with the comment that errors consider that payments to home health situation variables) so that the model is
built into the case-mix adjuster are too agencies during this period were not not based on potentially up-coded
large to be the basis for a case-mix being fully adjusted for inflation, and variables from home health agency
change adjustment. therefore the natural reaction of coding on OASIS. We use demographic
Comment: The proposed rule stated agencies would be to improve efficiency and non-home health Part A claims
that HHAs had no incentive to bring and lower resource consumption when history variables as the predictors. We
about nominal changes in case-mix pre- possible in order to survive. also include agency type and
HH HH PPS. A commenter disputed Response: Margin analysis by organizational form variables which
this, noting that HHAs could have MedPAC, CMS, and the Government help explain the level of case-mix. The
affected the case-mix weight in a Accountability Office has indicated that predictive ability of the full model, as
manner not anticipated or not Medicare margins under HH PPS have indicated by the R-squared statistic, is
responded to by CMS. generally exceeded 10 percent. 0.17.
Response: We based our proposal for Therefore, we find the commenter’s With each successive stage of model
adjusting payments for nominal case- conclusion that agencies responded to development, new sets of variables were
mix change on the observed average ensure survival counterintuitive, added to measure the effect on the
weight from a statistically valid sample because it would appear that in general, average prediction in the sample
representing the last four quarters before the payments made under HH PPS representing the 2005 time period. The
HH PPS began. We believe it is the covered their Medicare costs. We have first phase of the model is based on
appropriate baseline from which to start not studied efficiency outcomes among demographic variables, consisting of a
measuring coding changes that Medicare home health agencies, but large set of age-by-race and age-by-sex
Medicare did not intend to pay for economic theory would suggest that groups. The predicted average case-mix
under HH PPS. We explained the other entities become more efficient under weight did not change appreciably
reasons for using this as the baseline in bundled payment. We also note that when using these variables alone to
the proposed rule (72 FR 25392–25393). experts who study health services have make predictions, although we noted
Comment: A commenter questioned suggested there may be an incentive to that those beneficiaries in the 85-and-
the decision not to use the October 1997 stint on services under prospective older age group grew in prevalence and
through April 1998 study sample data as payment. contributed positively to the case-mix
the baseline. CMS had noted that the To summarize our case-mix analysis, index. This effect was offset by changes
agencies in the sample were volunteers, Abt Associates developed a case-mix in the prevalence of other demographic
and the commenter noted that volunteer prediction model designed to measure groups, to produce only minor change
agencies represented less than 1 percent real change in case-mix. We used two in the average case-mix weight during
of the agencies in existence. The data sets in applying this model. First, this model stage.
commenter also noted that the decrease we estimated the model on an HH IPS The second phase of the model added
in visits does not necessarily result in a sample. The HH IPS sample consisted of 12 variables representing inpatient
decrease in resource costs. He stated 394,479 non-LUPA episodes utilization for acute hospitals, long-term
that if the reduction in visits was representative of total episodes during care hospitals, IRF, and SNF, as
weighted toward lower cost visits (such the last 12 months of HH IPS. The identified in the National Claims
as home health aides), then that would episodes were simulated from claims History. Three variables captured the
imply that a greater portion of the visits using the same methodology that we presence of any hospital, SNF, or IRF
done in subsequent years were higher used to define episodes and link them stays in the 14 days leading up to the
cost visits (nursing, therapy, social to OASIS assessments for our case-mix beginning of the episode. A fourth
worker). The average cost per visit change analysis noted in the proposed variable represented episodes where
would then be higher in those rule. We used the model coefficient there was no acute, IRF, or SNF stay in
subsequent years, and therefore the total estimates to predict case-mix on a HH the 14 days before the home health
resource cost would be higher. The PPS sample. The HH PPS sample episode. An additional 8 variables
commenter gave the elimination of consisted of 876,199 non-LUPA captured the number of inpatient days
venipuncture as a qualifying skill as an episodes representative of total episodes of stay by type of stay during the 14
example. during CY 2005. Both samples were days leading up to the beginning of the
Response: The commenter may have restricted to non-LUPA episodes with a episode, and, before that, the number of
confused an agency which volunteers to matched OASIS assessment from the inpatient days in the period 15 to 120
participate in a study with a voluntary, national OASIS repository. days leading up to the beginning of the
or non-profit, agency. The agencies used The purpose of this case-mix model is episode. The days of stay categories
in the study sample included a mix of to predict the average case-mix weight were: Acute hospital, long-term care
organizational types. in the 2005 HH PPS year, based on a hospital, IRF, and SNF.
We accounted for the use of visits as regression model estimated from the HH The results from adding these
a measure of resource costs by IPS baseline year. Then, only the home variables to the demographic variables
weighting the visit minutes according to health population changes (as were an increase in the average
the labor costs of the discipline represented by the independent prediction of 0.6 percent beyond the
involved. Thus, the resource cost variables for the HH PPS year) affect the average during the HH IPS baseline. The
measure summarizes the effects of both average case-mix weight predicted from proportion of episodes preceded by
hospital stays in the 14 days leading up stay was for a procedure. At the same proposed rule estimate of the HH IPS
to the episode declined between HH IPS time, the proportion of episodes baseline average case-mix weight. The
and HH PPS, 2005, from 38.5 percent to associated with an acute procedure adjustment in the proposed rule
33.4 percent. Since this variable was increased from HH IPS to HH PPS 2005 standardized the HH IPS baseline for the
associated in the model with a 0.09 unit by only one percent, from 19 percent to decline in episodes delivered by
decline in case-mix weight, the lower 20 percent. This meant that the hospital-based agencies. At this stage,
prevalence of acute hospital use was an procedure effect would not be strong in given the contribution of all variables
important factor in the increase in the moving the average prediction between added to this point, the increase in the
average prediction. Another important the HH IPS sample and the HH PPS predicted average case-mix weight
contributor to these results was the sample. compared to the HH IPS baseline was
growth in SNF days, including growth The net effect on the predictions from 0.7 percent.
during the 14 day pre-episode period the model at this stage was to increase Finally, we added expenditure
and the 15- to 120-day pre-episode the level of the case-mix average relative variables for Part A utilization in the
period. These variables were associated to the HH IPS baseline, but the effect 120 days leading up to the home health
with an increase in case-mix weight. was very small. It is notable that the episode. These variables, which were
The average number of IRF days predictive power of the model increased adjusted for price increases, subdivided
declined during the 15- to 120-day pre- by more than three percentage points. In the expenditures by type of stay. The
episode period, from 0.68 during HH addition, the model indicated various expenditures related to long-term care
IPS to 0.52 during HH PPS 2005. (We effects as expected, including hospital stays, SNF stays, and inpatient
again included recertification episodes substantially higher HH PPS case-mix rehabilitation stays were associated with
in the total episodes in this sample.) weight associated with conditions such higher case-mix weights. Because the
While the number of IRF days is as intracranial hemorrhage; model controlled for stay events and
associated in the model with higher cerebrovascular accidents; other days of stay, we believe these variables
case-mix, the decline in total IRF days disorders of the nervous system; may proxy the intensity of care during
between HH IPS and CY 2000 meant respiratory system diagnosis with the inpatient periods. The model
that this factor helped offset the case- ventilator support; respiratory infections estimates using all variables included by
mix increasing effect of the hospital and and inflammations; pneumothorax and this final stage increased the average
SNF days variables on the predictions. pleural effusion; respiratory system case-mix weight compared to the HH
The third phase of the model added signs, symptoms, and other diagnoses; IPS baseline by 0.95 percent.
family situation variables, including major esophageal disorders; hip The unadjusted total measure of case-
whether the patient during the episode fractures; electrolyte disorders except mix change was calculated by taking the
lived alone, with a spouse, with other hypovolemia related; septicemia; difference between the 2005 actual
family members, with paid help or with pneumonia; and complications of average case-mix and the HH IPS actual
others. The results from adding these treatment. The model did not indicate average case-mix (our baseline). This
variables moved the predicted average higher case-mix weights associated with unadjusted measure (12.78 percent)
higher than the baseline by only 0.1 many other hospital case-mix groups, included both real and nominal change.
percent. such as hip and knee replacements, We used our full 6-phase model to
The fourth phase of the model added major and nonmajor respiratory derive the proportion of case-mix
scores of variables representing the procedures, cardiac defibrillator change which was real; the full model
hospital case-mix group assignment for implant, cardiac valve procedures with result yielded a predicted average case-
the last acute hospital stay for the cardiac catheterization, and coronary mix for 2005. When we took the
patient in the National Claims History. artery bypass graft. It should be noted difference between this model result
We used the All-Patient-DRGs (APR again that these effects are estimated and the HH IPS actual average case-mix
DRG) classification algorithm to assign after controlling for whether the stay (our baseline), the result was the real
the case-mix group. We specified was procedure-related. Thus, the case-mix change.
variables for all the APR DRG groups negative coefficient for knee The resulting real case-mix change
that met our sample size standards replacements indicates that the effect of was then divided by the total measure
(minimum of 25 cases). Typically, the having had a knee replacement before of case-mix change (real plus nominal)
stays generating the APR DRG home health reduces the size of the to determine the proportion by which
assignments occurred within six weeks, general positive effect from having had the total measure of case-mix change
and overall three-quarters of the stays a procedure. One of the strongest would need to be reduced in order to
occurred within the previous 8.6 impacts on the predictions came from account for real case-mix change. That
months. The purpose of using these the APR DRG for nonspecific proportion was 8.03 percent. Therefore,
variables was to incorporate more cerebrovascular accident and we reduced the 12.78 percent measure
information about the patient’s precerebral occlusion without of total case-mix change by 8.03 percent
condition, especially some measure of infarction; in the HH IPS sample, about (real case-mix change) to derive the
case severity into the model. The APR 1.2 percent of the episodes were nominal case-mix change adjustment of
DRG algorithm uses comorbidity data on preceded by a stay of this type, but in 11.75 percent (0.1278 * (1 ¥ 0.0803) =
the hospital claim to generate severity the HH PPS 2005 sample the episode 0.1175). This 11.75 percent change in
levels for each case-mix group. As an percentage was down to about 0.4 case-mix is 1.03 percentage points lower
example, the model included four percent. The loss of this type of case than the unadjusted total change in
differing severity levels for knee was one of the important contributors case-mix, which is 12.78 percent.
replacement stays, which are included that offset the case-mix increasing While the total measure of case-mix
in APR DRG group 302. A general effects of some of the other changes. increase is 11.75 percent, it could be
indicator that the stay was procedure- The fifth phase of the model adjusted misinterpreted that the total of the
related was also included. This for the change in the types of home adjustments to be made in each of the
indicator had a large effect in the model, health agencies between HH IPS and CY next four years equals 10.96 percent
suggesting an increase in the HH case- 2005. This adjustment is analogous to (2.75 + 2.75 + 2.75 + 2.71 = 10.96), if
mix weight of about 0.34 if the last acute the adjustment we made in the the adjustment were taken in one year.
This would be an incorrect method of to fully address the 11.75 percent growth in the nominal change in case-
solving for the total adjustment if taken change in case-mix unrelated to real mix since CY 2005. The calculation of
in one year. If we accounted for the full case-mix change. We are seeking the adjusted national prospective 60-day
11.75 percent increase in case-mix in a comment on the 2.71 percent case-mix episode payment rate for case-mix and
single year, that percentage reduction to change adjustment for 2011. We will area wage levels is set forth in 42 CFR
the rates would be 10.51 percent (1/(1 continue to monitor and measure the 484.220. We are revising 42 CFR
+ .1175) = 0.894855; 1 ¥ 0.894855 = nominal change in case-mix. As we 484.220 to address the annual
.1051). Over the 4-year period, we are discussed in the proposed rule, if percentage reductions due to changes in
taking the same 10.51 percent updates of the national claims data case-mix that are not a real change in
adjustment ((1 ¥ 0.0275) * (1 ¥ 0.0275) indicate that the nominal change in case-mix. For this final rule with
* (1 ¥ 0.0275)*(1 ¥ 0.0271) = 0.894823; case-mix between the HH IPS baseline comment period, we are specifically
1 ¥ 0.894823 = 0.105177 = 10.52 and the latest available national claims soliciting comment on the 2.71 percent
percent; a difference of 0.01 percent data show a change, we will revise the adjustment to the HH PPS 60-day
from the single-year total adjustment of percentage reduction in future year’s episode payment rate in the fourth year
10.51 percent is due to rounding). Note update of the annual reduction factor. to account for the change in case-mix
that the percentage reduction is less Similar to how it was described in the that is not considered real, i.e., that is
than the percentage increase; because proposed rule, the revision would be not related to an underlying change in
the new baseline is higher, in determined by the ratio of the updated patient health status.
percentage terms the reduction 4-year annual reduction factor to the
The final versions of tables 6, 7, and
necessary to get back to the original previous year’s annual reduction factor.
8, which are discussed in this section on
baseline will be less than the percentage For the CY 2011 rule, which governs the
case-mix change adjustment, are shown
increase. In determining the yearly fourth and final year of the case-mix
below.
percentage reductions, we first opted to change adjustment transition period, we
keep the 2.75 percent per year reduction would obtain the CY 2008 national
which we had proposed. Accounting for average CMI to compute the updated TABLE 6.—AVERAGE NUMBER OF
the compounding effect of a 2.75 value for the nominal case-mix change HOME HEALTH VISITS PER EPISODE
percent reduction in each of the first 3 adjustment. Again, we would form the
Total home
years, the 4th year reduction necessary ratio of the updated adjustment factor to health visits
to bring about a total reduction of 10.51 the previous year’s effective adjustment Year (excluding
percent is 2.71 percent. Note that the factor. Depending on the growth of the LUPAs)
sum of the 4-year nominal reduction of nominal change in case-mix, measured
10.95 percent is only an approximation in any given subsequent year, in future 1997 ...................................... 36.04
of the 10.51 percent since it does not rulemaking, CMS may adjust the 1998 ...................................... 31.56
HH IPS .................................. 26.88
account for the compounding effect of percentage reduction in the second and/
2001 ...................................... 21.67
the annual reductions. For this final rule or third year, elect to adjust the 2002 ...................................... 21.49
with comment period, we are finalizing percentage reduction in only the fourth 2003 ...................................... 21.01
the proposed 2.75 percent reduction of year, or adjust the percentage reduction 2004 ...................................... 20.66
the national standardized 60-day in any combination of years. The annual 2005 ...................................... 20.53
episode payment rate for 3 years updating procedure avoids a large
Note: Excludes LUPAs, RAPs, episodes
beginning in 2008 and extending that reduction for the final year of the phase- with data problems and no matched OASIS.
adjustment period to a fourth year via a in, in the event that the CY 2008 The HH IPS data is from the 100 percent file
2.71 percent reduction for 2011, in order national average CMI reflects continued for FY 2000.
TABLE 7.—AVERAGE RESOURCE COST AND CMI

Resources CMI
Standard-
Period Average ized to CY
resource Admissions All
2000 labor
cost rates
HH IPS
1999Q4 ............................................................................................................................ $451.11 $451.39 1.1165 1.0796

2000Q1 ............................................................................................................................ 468.27 468.27 1.1040 1.0822
2000Q2 ............................................................................................................................ 475.34 475.34 1.1277 1.1026
2000Q3 ............................................................................................................................ 471.64 471.64 1.1448 1.1186
HH PPS
2000Q4 ............................................................................................................................ N/A N/A N/A N/A

2001Q1 ............................................................................................................................ $432.14 $419.60 1.1855 1.1651
2001Q2 ............................................................................................................................ 440.98 428.18 1.1930 1.1801
2001Q3 ............................................................................................................................ 445.96 433.02 1.1980 1.1756
2001Q4 ............................................................................................................................ 446.80 433.84 1.2025 1.1853
2002Q1 ............................................................................................................................ 453.76 426.42 1.2086 1.1843

2002Q2 ............................................................................................................................ 454.65 427.25 1.2027 1.1874
2002Q3 ............................................................................................................................ 457.49 429.92 1.2127 1.1871
2002Q4 ............................................................................................................................ 460.96 433.17 1.2243 1.1996
2003Q1 ............................................................................................................................ 454.77 422.58 1.2182 1.1931
2003Q2 ............................................................................................................................ 461.18 428.53 1.2326 1.2060
TABLE 7.—AVERAGE RESOURCE COST AND CMI—Continued

Resources CMI
Standard-
Period Average ized to CY
resource Admissions All
2000 labor
cost rates
2003Q3 ............................................................................................................................ 460.15 427.58 1.2333 1.2044

2003Q4 ............................................................................................................................ 464.71 431.81 1.2497 1.2178
2004Q1 ............................................................................................................................ 462.26 427.31 1.2434 1.2117
2004Q2 ............................................................................................................................ 473.42 437.63 1.2572 1.2239
2004Q3 ............................................................................................................................ 476.77 440.72 1.2634 1.2252
2004Q4 ............................................................................................................................ 479.90 443.61 1.2709 1.2314
2005Q1 ............................................................................................................................ 487.19 417.40 1.2680 1.2298
2005Q2 ............................................................................................................................ 509.91 436.87 1.2697 1.2341
2005Q3 ............................................................................................................................ 518.92 444.58 1.2810 1.2358
2005Q4 ............................................................................................................................ 522.22 447.41 1.2882 1.2443
Note: HH IPS data based on 100% National Claims History File. The averages reported in the proposed rule may differ slightly from averages
reported here because of slight changes in methodology and further data cleaning.
TABLE 8.—PERCENT SHARE OF HOME HEALTH EPISODES AND MEAN TIME PRIOR TO ENTERING A HOME HEALTH
EPISODE, FOR FIVE CONDITIONS, FY 2000–CY 2006
Condition FY 2000 CY 2001 CY 2002 CY 2003 CY 2004 CY 2005 CY 2006 *
Hip fracture:
percent share ................................................................ 0.82 0.83 0.75 0.72 0.70 0.62 0.59
days prior to entering .................................................... 7.19 7.12 7.18 7.21 7.30 7.09 7.12
Congestive heart failure:
percent share ................................................................ 3.31 3.05 2.95 2.87 2.71 2.43 2.62
Cerebrovascular accident:
percent share ................................................................ 1.52 1.45 1.40 1.29 1.14 1.03 0.97
Hip replacement:
percent share ................................................................ 1.47 1.64 1.63 1.59 1.64 1.45 1.36
Knee replacement:
percent share ................................................................ 1.89 2.20 2.30 2.43 2.58 2.70 2.75
Note: Time prior to entering is number of days between hospital discharge and beginning of home health episode, for discharges occurring
within 14 days of the start of the home health episode.
For beneficiaries with more than 1 hospital discharge in the 14 day period leading up to the home health episode, time prior to entering is from
the last hospital discharge immediately preceding the home health episode.
* CY 2006 data for first quarter of the year only.
7. Case-Mix Groups to a payment weight will exist in the training of staff. They asked that the
Comment: Two commenters were form of the HIHH PPS code produced by implementation be postponed or be
concerned that the proposed case-mix the Grouper software. We plan that the phased-in.
model results in loss of all identifiable first position of the five position HIHH Response: As we noted previously, we
meaning from a case-mix group or PPS code will represent the payment have tried to strike a balance between
HHRG. The commenters asked for a grouping step that applies to the simplicity and complexity. The refined
mechanism to produce a unique HHRG, episode. The second, third and fourth system is more complex than the old
Health Insurance Prospective Payment positions will represent the clinical, system but this is a natural outgrowth of
System (HIHH PPS) code, or other functional and service domains arrived our attempt to pay more accurately for
designation for each of the 153 case-mix at under the payment equation that the range and intensity of home health
groups and five NRS severity levels. applies for that grouping step. The fifth services that can be provided to our
They believed providers need a unique position will represent the NRS severity beneficiaries.
identifier for each case-mix group to level. The final code structure for these A refined system may seem overly
facilitate communication, analysis, and HIHH PPS codes and the complete list complex just because it is new.
financial comparison. of codes will be published in Medicare However, we believe the proposed
Response: While it is true that the instructions and on our Web site, refinements are clearly focused, and
HHRG code represents the severity shortly after the issuance of this final logically stem from the original case-
rule.
levels in the clinical, functional and mix payment system. We agree that any
service domains, it no longer represents Comment: Several commenters refined system will take time and
a one-to-one match with a case-mix remarked that the increase from 80 to training to learn. As explained in the
weight under the proposed refined 153 HHRGs was complex and would response to a comment in section
payment case-mix system. However, a create an administrative burden. III.A.3, we have taken several measures
code with this one-to-one relationship Additionally, it will require extensive to make the proposed refinements easier
to understand, and we trust that these As we noted in the proposed rule (72 on the condition most related to the
measures will assist HHAs in FR 25361, and 25362), scores were current plan of care. This diagnosis may
implementing this refined system. assigned to certain secondary diagnoses or may not be related to a patient’s
and used to account for the cost- recent hospital stay but must relate to
8. OASIS Reporting and Coding
increasing effects of comorbidities. the services rendered by the HHA.
Practices Comment: A commenter asked that
However, with most diagnosis groups,
Comment: Several commenters we did not make a distinction in the we adopt ICD–10 guidelines, and study
expressed concern that some pressure final case-mix model between primary the impact of coding changes on HH
ulcers are not stageable due to eschar. placement and secondary placement of PPS.
They noted that proper care includes a condition in the reported list of Response: We agree that it is
debridement, which is costly due to diagnoses. We made case-by-case important to have an accurate and
supplies and clinician time. Once decisions on this question based on precise coding system. The Department
debridement occurs, the ulcer would be differences in the impact on resource will continue to study whether or not to
stageable, but the HHA would have no cost between the primary diagnosis and propose ICD–10–CM and ICD–10–PCS
way to note the change in condition secondary diagnosis. If differences were as the new HIPAA standard to replace
since the SCIC adjustment has been small, we combined cases reporting the ICD–9–CM.
eliminated. The commenters conditions, regardless of whether the Comment: A commenter suggests that
recommended allowing staging of these listed position of the diagnosis was M0826 be asked only if the patient is
ulcers in accordance with National primary or secondary. We believe this is expected to be a higher need case.
Pressure Ulcer Advisory Panel an important protection against Response: We disagree. Home health
guidelines. unintended and undesirable incentive providers are expected to assess and
Response: We are aware of recent effects that could arise if agencies document each patient’s need for
revisions issued by the National perceive opportunities to change the therapy. M0826 is required to be coded
Pressure Ulcer Advisory Panel placement of the diagnosis due to non- by providers regardless of the patient’s
(NPUAP). The NPUAP guidance is clinical reasons. expected case-mix assignment. The
essentially permitting the assessment of Concerning the comment suggesting coding of M0826 should be in
a wound for staging when the wound we add more lines for entering compliance with Medicare home health
bed is not completely covered with diagnoses in M0240, we disagree that CoPs 42 CFR 484.55, 42 CFR 484.18,
eschar or slough. If the bed of the ulcer more lines are needed for M0240. and 42 CFR 484.32.
is completely covered with eschar/ However, as noted in the proposed rule, Provider instructions for coding
slough, NPUAP guidance stipulates that we did make changes to the OASIS to M0826 are provided in Chapter 8 of the
the wound cannot be staged until some enable agencies to report secondary OASIS Implementation Manual. Those
of the necrotic tissue is removed. After case-mix diagnosis codes (see 72 FR instructions allow providers to answer
reviewing the NPUAP guidance we have 25362). Specifically, the addition of ‘‘000’’ if no therapy services are needed,
revised the instructions accompanying secondary diagnoses to the proposed or answer with the total number of
the OASIS item to allow a wound to be case-mix system (see Table 2A of the therapy visits indicated or planned for
staged if the bed of the wound is proposed rule, case-mix adjustment the Medicare payment episode for
partially covered by necrotic tissue and variables and scores) requires that the which this assessment will determine
if the presence of eschar does not OASIS allow for reporting of instances the case-mix group. Providers may also
obscure the depth of the tissue loss. in which a V-code is coded in place of answer ‘‘not applicable’’ when this
Comment: We received a number of a case-mix diagnosis other than the assessment will not be used to
comments supporting our decision to primary diagnosis. A case-mix diagnosis determine a Medicare case-mix group.
allow additional case-mix diagnoses for is a diagnosis that determines the HH Comment: A commenter asked that
certain conditions and for allowing PPS case-mix group. Currently, the we expand the wound section of the
points for some comorbidities. One OASIS allows for reporting of instances OASIS to include all wounds, especially
supported the scoring of secondary of displacement involving primary diabetic ulcers and arterial ulcers.
diagnoses to account for the cost- diagnosis only for M0245. Response: The diagnosis codes for
increasing effects of comorbidities. A Consequently, because of the nature and diabetic and arterial ulcers were in the
few commenters suggested more rows significance of the changes needed, as proposed rule for both the case-mix
for entering diagnoses in M0240 noted in the proposed rule, we deleted diagnosis and non-routine supply
(‘‘other’’ diagnoses). They note that to the OASIS item M0245 and replaced it diagnosis tables. As a result of further
follow ICD–9–CM coding guidance with a new OASIS item M0246. research, we are also adding two
based on severity ranking, there will be We disagree with the comments additional arterial ulcer codes to final
many instances where the case-mix suggesting that if ICD–9–CM coding tables 2B and 10B (see ICD–9–CM codes
diagnoses that impact the plan of care guidance is based on severity ranking in 447.2 and 447.8).
and resource utilization will not be the OASIS, there will be many instances However, such review and expansion
captured for patients with multiple co- where the case-mix diagnoses that of OASIS is beyond the scope of this
morbidities, leading to underpayment impact the plan of care and resource rule. OASIS will continue to capture
for the sickest patients if coding rules utilization will not be captured for diabetic and arterial ulcers in both the
are followed. It would also address patients with multiple co-morbidities, diagnosis section and the basic wound-
OASIS diagnosis spaces fields in leading to underpayment for the sickest related section (M0440). OASIS item
preparation for ICD–10, which will patients. It is significant to note that the M0440 measures the presence of a skin
significantly increase the number of logic for determining both the primary lesion or open wound.
required diagnosis codes. and secondary diagnoses remains OASIS items are only part of a
Response: We appreciate the unchanged (see the OASIS comprehensive assessment and include
comments supporting our decision to Implementation Manual, Definition only those items that have proven useful
allow additional case-mix diagnoses and Section of M0230/240 as well as for outcome measurement and risk
for allowing points for comorbidities/ Attachment D to Chapter 8). The factor adjustment. Therefore only the
secondary diagnoses. primary diagnosis is determined based types of wounds that are relevant to
these OASIS purposes or outcome beneficiary elected transfer or a Response: We will share this concern
measurement or risk factor adjustment discharge and return to the same HHA with our fiscal intermediaries and
have been included in OASIS, though within the 60-day period. The PEP suggest that they direct medical review
other types of wounds such as diabetic adjusted episode is paid based on the activities for PEP episodes as
and arterial ulcers are extremely span of days including start of care date appropriate.
important to assess and document in the or first billable service date and Comment: A commenter noted that
patient’s clinical record. including the last billable service date when a PEP occurs due to a transfer to
Comment: A commenter wrote that under the original plan of care before another agency, the first agency is often
changes to the OASIS items M0230/240/ the intervening event. As noted in the surprised. The commenter asks CMS to
246 are complex, and the instructions proposed rule, descriptive analysis was automatically check for proper protocol
need to be clearer for column 4. The conducted to better understand the by the second agency to ensure that the
commenter suggested that the patient characteristics associated with first agency is not caught off guard.
instructions read, ‘‘Complete ONLY IF PEP-adjusted episodes and the Response: We appreciate this
the V-code in Column 2 is reported in circumstances under which PEP- comment. Our analysis of a 20-percent
place of a case-mix diagnosis that is a adjusted episodes occurred. Analysis of sample of 2003 episodes showed that
multiple coding situation.’’ patient characteristics revealed no approximately 3 percent of all episodes
Response: The commenter has appreciable differences between were PEP adjusted. Of those PEP
literally repeated the precise patients in normal episodes (that is, no episodes, approximately 55 percent of
instructions we have issued in Column HH PPS payment adjustments, such as PEP-adjusted episodes involved a
4 of the OASIS, M0230/240/246 as a LUPA, PEPs, or SCICs) and patients in discharge and return to the same HHA,
suggestion for clearer instructions. It is PEP episodes with regard to conditions about 42 percent involved a transfer to
significant to note that Column 4 does or clinical characteristics. The mix of another agency, and approximately 3
stipulate the following: ‘‘Complete visits in PEP episodes was found to be percent involved a move to managed
ONLY if the V-code in Column 2 is similar to that of normal episodes. care.
reported in place of a case-mix Chapter 10 (Section 10.1.13) of the
The descriptive analyses conducted Medicare claims processing manual
diagnosis that is a multiple coding by Abt Associates also looked at the
situation.’’ does provide a process for the initial
different components that make up PEP HHA and the receiving (new) HHA to
In reference to assigning V-codes on episodes. The analysis showed that PEP
the OASIS, a case-mix diagnosis is a follow in when a transfer to another
episodes have significantly shorter HHA results in a PEP situation. In order
diagnosis that gives a patient a score for service periods on average than all
Medicare Home health HH PPS case- for a receiving (new) HHA to accept a
episodes other than LUPA and SCIC beneficiary elected transfer, the
mix group assignment. A case-mix episodes. The number of visits in a PEP
diagnosis may be the primary diagnosis, receiving HHA must document that the
episode, on average, represented 75 beneficiary has been informed that the
‘‘other’’ diagnosis, or a manifestation percent of the average number of visits
associated with a primary or other initial HHA will no longer receive
for normal episodes. We have used the Medicare payment on behalf of the
diagnosis. Diagnoses listed under span of billable visits in the PEP patient and will no longer provide
columns 3 and 4 of OASIS, M0230/240/ payment adjustment because of the Medicare covered services to the patient
246 should be documented on the HHA’s involvement in decisions after the date of the patient’s elected
patient’s Plan of Care in compliance influencing the intervening events for a transfer in accordance with current
with 42 CFR 484.18(a). V-code reporting beneficiary who elected to transfer or patient rights requirements at 42 CFR
on the OASIS became effective in discharge and returned to the same 484.10(e). The receiving HHA must also
October 2003 in compliance with HHA during the same 60-day episode document in its records that it accessed
HIPAA. Providers assigning V-codes on period. Agencies have some flexibility the RHHI inquiry system to determine
the OASIS are expected to comply with in discharge decisions that affect the whether or not the patient was under an
all of the following long-standing home likelihood of incurring a partial episode, established home health plan of care
health diagnosis coding requirements, whether or not a hospital stay and contacted the initial HHA on the
which can be found in the document intervenes. They also have indirect effective date of transfer. In such cases,
entitled ‘‘Medicare Home Health influence on a beneficiary’s decision to the previously open episode will be
Diagnosis Coding’’ on the CMS Home transfer to another home care provider automatically closed in the Medicare
Health Web site at: http:// through the quality of care they provide. claims processing systems as of the date
www.cms.hhs.gov/HomeHealthPPS/ Data suggested that PEP episodes are services began at the HHA the
03_coding&billing.asp. rare and, therefore, the current PEP beneficiary transferred to, as reported in
Comment: Another commenter policy may be serving as a deterrent to the RAP; and the new episode for the
suggested that we revise the instructions premature discharge. Consequently, we ‘‘transfer to’’ agency will begin on that
for M0080 and M0090 to recognize the did not propose to change the PEP same date.
new complexities of completing M0230/ policy. Comment: Several commenters noted
240/246 correctly. Comment: Several commenters raised that PEP episodes are underpaid. Two
Response: Chapter 8 of the OASIS concerns about a specific situation that commenters said that agencies are
Implementation Manual will be updated can arise under the existing PEP policy. especially concerned with PEP
to accommodate changes to the OASIS In the specific situation mentioned, the situations where patients are discharged
items. second provider in the PEP can admit a when the plan of care goals are met but
C. Payment Adjustments beneficiary whose plan of care goals return to the same agency within the 60-
were already met by the first provider. day period, often for a condition that
1. The Partial Episode Payment (PEP) The commenter suggests that the FIs) was not related to the first plan of care.
Adjustment review those admissions to determine if In those cases, agencies can receive a
Currently, HH PPS provides for an the care provided by second agency was significant reduction in payment for the
adjusted proportional payment for 60- medically necessary. A PEP can occur first episode despite provision of all
day episodes interrupted by a because of transfer to another agency. visits authorized under a plan of care.
Similarly, two commenters implemented in such a way that an advice on front-loading visits to avoid
recommended that CMS not apply PEP initial home health agency does not re-hospitalization. The commenter
to cases where the patient is discharged receive appropriate recognition from the suggested that CMS prorate the initial
with the plan of care goals met yet beginning of the episode, recognizing PEP episode based on the ratio of days
returns to the same HHA with a new that currently the PEP begins at the first between the first billable visit and
medical issue. The commenters believed visit rather than the beginning of the discharge to the subsequent agency.
maintenance of the PEP policy in its episode. Response: As stated in the proposed
current form also raises questions Response: We do not believe that it is rule, we believe that HHAs have some
regarding how ‘‘early’’ and ‘‘later’’ appropriate to generate another episode flexibility in discharge decisions that
episodes will be defined in the type based upon a per-visit basis. At the affect the likelihood of incurring a
proposed payment system. inception of the HH PPS, we decided partial episode (72 FR 25423), whether
Response: As discussed in the that paying for LUPA episodes on a per- or not a hospital stay intervenes (72 FR
proposed rule, the PEP adjustment visit basis was appropriate due to the 25423). HHAs also have indirect
provides a simplified approach to the extremely low number of visits influence on a beneficiary’s decision to
episode definition and accounts for key provided in such an episode. One of the transfer to another HHA through the
intervening events in a patient’s care goals of a PPS for home heath was to quality of care they provide.
defined as a beneficiary elected transfer, move away from a system that pays on Additionally, current data suggest that
or a discharge and return to the same a per-visit basis. PEP episodes are rare, and therefore, the
HHA that warrants a new start of care Comment: A commenter suggested current PEP policy may be serving as a
for payment purposes, OASIS, and that CMS eliminate the PEP due to its deterrent to premature discharge. We
physician certification of the new plan adverse clinical, administrative, and believe that the PEP adjustment is
of care (72 FR 25422, 25423). The financial impact. The commenter stated provided in a manner that maintains the
discharge and return to the same HHA PEP adjustments require significant opportunity for Medicare patients to
during the 60-day episode period is only resource utilization for agencies with choose the provider with which they
recognized when a beneficiary reached minimal reimbursement as HHAs front- feel most comfortable. We also note that,
the treatment goals in the original plan load costs. Additionally, the commenter as we did in the proposed rule, in many
of care. The original plan of care must further noted while HHAs have cases an HHA received payment for an
be terminated with no anticipated need developed strategies to minimize additional full episode which it might
for additional home health services for hospitalizations and SNF admissions, not have received had the first episode
the balance of the 60-day period. This the HHAs often cannot affect the not been subject to a PEP adjustment (72
policy ensures that we do not provide patient’s level of acuity or social FR 25423). We do recognize that PEP
full payment for two episodes at any situation, which can result in a PEP episodes provide, on average, 75 percent
time during a given certified 60-day episode. of the average number of visits for
episode. Results from our refinement Response: We disagree with the normal episodes, which parallels the
research provided evidence that there is commenter. We believe the PEP QIO’s advice to HHAs to provide more
some front-loading of visits compared to adjustment is provided in a manner that visits early in an episode of care to
normal episodes, causing PEP episodes maintains the opportunity for Medicare prevent re-hospitalizations.
to have a faster average rate of visits patients to choose the provider with Comment: A commenter asked that
during the span of days used to prorate which they feel most comfortable while we reopen the episode if a patient
the episode payment. ensuring that the Medicare Trust Funds returns to the HHA within 60 days, and
Early episodes are defined to include are protected by a policy that ensures only pay for the time services were
not only the initial episode in a adequate payment levels that reflect the given.
sequence of adjacent episodes, but also care provided by each HHA to a Response: HHAs have some flexibility
the next adjacent episode, if any, that beneficiary in a transfer situation. in discharge decisions that affect the
followed the initial episode as the first Comment: A commenter was likelihood of incurring a partial episode,
two episodes in a sequence of adjacent disappointed that CMS did not make whether or not a hospital stay
episodes. Later episodes are defined as changes in the PEP adjustment to more intervenes. They also have indirect
all adjacent episodes beyond the second accurately allocate costs, believing that influence on a beneficiary’s decision to
episode. Episodes are considered to be the current methodology often transfer to another home care provider
‘‘adjacent’’ if they are separated by no underpays in the case of PEP transfers. through the quality of care they provide.
more than a 60-day period between Specifically, the commenter felt it is Whether or not a given episode remains
episodes. This holds true regardless of particularly troubling when the PEP open is subject to whether or not the
the type of episode. The end of a PEP occurs without the first agency’s goals of the plan of care have been met
episode is denoted as the last billable knowledge as often the patient has had and a particular HHAs’s discharge
visit date. The gap in days between an an intervening hospital stay and is policy. We believe that it would be
episode with a PEP adjustment and the advised by the hospital that it is inappropriate for CMS to dictate
next episode would be calculated using preferable or required that the patient whether or not or when an HHA should
the last billable visit of the PEP and the use a hospital-based HHA upon discharge a patient, as we believe those
from-date of the subsequent episode. discharge, thus generating the PEP. sorts of decisions are best left up to the
Comment: A commenter asked that There are cases where the patient or HHA. Consequently we do not believe
PEPs be considered from the beginning family is confused and seeks care from that a policy to reopen an episode if the
of the episode rather than the first visit a second agency, believing that using patient returns to the HHA within the
due to care coordination activities. The two HHAs is allowable and is better 60 days would be an appropriate policy.
commenter asserted that agencies than having just one. The commenter In addition, we believe that prorating an
should receive at least the LUPA rate if again noted that these visits tend to be episode, as the commenter suggests,
the episodic payment under PEP would front-loaded, and prorating from first to would unnecessarily further complicate
be lower than the LUPA. Moreover, the last billable visit systematically the PEP payment policy.
commenter noted that since the underpays the initiating agency and In summary, there are several
inception of HH PPS, the PEP has been penalizes agencies who follow QIO methods that could be used to refine the
PEP adjustment methodology, as Furthermore, our analysis of NRS that are the first in a series of adjacent
recommended by commenters. Another showed that NRS charges for non-LUPA episodes or the only episode. After
possible approach could involve episodes are almost 3 times higher than updating the payment model using 2005
weighting the payment to reflect the for LUPA episodes. In the proposed data and re-analyzing the characteristics
front-loading of visits, but it is not clear rule, we expressed concerns that adding of all LUPAs, the results continue to
at this time what an appropriate an additional amount to LUPA support providing a revised payment for
approach to refinement of the PEP payments for NRS could promote LUPA episodes, but only for those that
policy would be. We intend to study the increases in medically unnecessary occur as the first episode in a sequence
comments provided, continue public home health episodes, and therefore did of adjacent episodes or the only episode.
discussion on this issue, and look not propose any additional payments for Using the updated 2005 data, the
towards the possible refinement of this NRS costs for LUPA episodes (72 FR additional revised payment for first
adjustment in future rulemaking. 25430.) episode LUPAs or the only episode is
An analysis of a 20-percent sample of $87.93.
2. The Low-Utilization Payment home health episodes covering more Comment: We received universal
Adjustment (LUPA) than 3 years of experience with HH PPS support for the revised LUPA payment,
The low utilization payment revealed that there were approximately but several commenters noted that due
adjustment (LUPA) reduces the 60-day 179,845 LUPA episodes. While some to treatment timing, HHA clinicians
episode payment when minimal LUPA patients were in high severity often must make an additional, non-
services are provided during a 60-day groups, overall LUPA patients had chargeable visit for the sole purpose of
episode. LUPAs are episodes with four somewhat lower clinical and functional completing an OASIS follow-up
or fewer visits and receive a wage- severity. These data indicated that assessment in the required 5-day
adjusted average per visit amount per LUPAs are serving as a low-end outlier window or for a recertification visit.
home health discipline, instead of a full payment for certain episodes that incur These can occur with catheter and
60-day episode payment. The home unexpectedly low costs. Other LUPA vitamin B–12 patients. The commenters
health industry suggests that the LUPA episodes result from expected care claimed the costs for these visits are not
payment rates do not adequately patterns for patients with particular captured in claims data as HHAs are
account for the front-loading of costs in conditions (for example, neurogenic prohibited from billing for assessment-
an episode. In performing our bladder). only visits. Again, this claim often
refinement research, we found that the Section 1861(m)(5) of the Act, occurs with catheter patients. Another
average visit lengths in these initial specifically, includes catheters, catheter commenter noted that CMS only
LUPAs are 16 to 18 percent higher than supplies, and ostomy bags and supplies included an estimate of additional
the average visit lengths in initial non- as a covered home health supply. They minutes of direct service cost for
LUPA episodes. For a complete are considered to be non-routine in assessment in its LUPA cost calculation,
description of the LUPA review, nature, and are bundled into the HH rather than the entire administrative
analysis, and research performed, we PPS payment rates. Catheters and cost the agency bears. Another noted
refer to the CY 2008 HH PPS proposed catheter supplies are on our list of NRS that our analysis may have been
rule (72 FR 25423–27). In the proposed codes subject to consolidated billing influenced by data issues in industry
rule, we proposed to increase payment which is posted on CMS’s home health cost reports. One commenter asked for
by $92.63 for LUPA episodes that occur Web Site at http://www.cms.hhs.gov/ higher reimbursement for acute patients
as the first or only episode in a sequence center/hha.asp (go to ‘‘Billing/ who cannot remain at home and become
of adjacent episodes. Payment’’, and then ‘‘Home Health a LUPA patient through no fault of the
Comment: Several commenters asked Coding and Billing’’). HHA.
that NRS supplies, particularly catheters Comment: While there was Response: We derived a revised final
and ostomy supplies, be reimbursed as widespread support for the revised value for the increase to LUPA episodes
part of the LUPA payment. One LUPA payment, many commenters that occur as the only episode or the
suggested that we develop a NRS add- asked that the additional $92.63 apply initial episode during a sequence of
on using diagnostic categories. Others to all LUPAs and not just to the first and adjacent episodes from a new data base
noted that some LUPAs require wound only LUPA or the initial LUPA in a consisting of visit line items from a
care supplies or chest drains. Several series of adjacent episodes. A number of large, representative sample of claims in
commenters believed that we proposed commenters noted that the 2005. This method enabled us to
to remove the NRS payment from reimbursement still does not cover the measure the entire excess of minutes
LUPAs and asked that we reconsider costs of LUPA episodes and suggested due to both OASIS and administrative
this proposal. One suggested we increasing the payments further. activities of the type cited in the
reimburse HHAs 200 percent of the Response: The proposed additional comment. This database showed that
supply cost to cover overhead or payment of $92.63 was intended to the average excess of minutes for the
establish a fee schedule that lists out cover the front-loading of costs which first visit in episodes that were single
reimbursement rates for medical occurs in an initial assessment in a LUPAs or initial LUPAs in a sequence
supplies. LUPA episode. We analyzed LUPA of episodes was 38.5 for the first visit if
Response: LUPA episodes are paid on episodes and found that the average skilled nursing, 25.1 for the first visit if
a per-visit basis. Currently LUPA visit length for nursing for an initial physical therapy, and 22.6 for the first
payments include NRS paid under a assessment averaged twice as long as the visit if speech therapy. We then
home health plan of care, NRS possibly length of other visits. Similarly, the expressed these excess values as a
unbundled to Part B, and a per-visit initial assessment visit made by a proportion of the average number of
ongoing OASIS reporting adjustment. physical therapist was 25 percent longer minutes for all nonfirst visits in non-
Moreover, contrary to the commenters’ than other physical therapy visits. We LUPA episodes (42.5, 45.6, and 48.6 for
statements, the original 2000 NRS did not find that all visits in LUPA skilled nursing, physical therapy, and
amount of $1.94 included in the LUPA episodes were longer than average, and speech therapy, respectively). We then
per visit rates has been updated as such, we proposed to provide the proportionately inflated the per-visit
annually and has not been removed. additional $92.63 only for those LUPAs payment, using LUPA per-visit payment
rates, in accordance with these excess commenter also wanted it made nursing and physical or speech therapy
values. Finally, using an appropriate set mandatory for all episodes and LUPAs visit rates.
of weights representing the share of to support any request for payment To summarize, additional analysis did
LUPA first visits for skilled nursing based upon severity scores and severity not support that all LUPA episodes are
(77.8 percent), physical therapy (21.7 levels, or such payment will be negated. negatively impacted by the front-loading
percent), and speech therapy (0.5 Another commenter suggested we of assessment costs and administrative
percent), respectively, we calculated the require that supplies be charged on costs. Consequently, for this final rule,
revised increase of $87.93 for LUPA claims in order to receive NRS payment. we are implementing the proposed
episodes that occur as the only episode Response: We will continue to study provision of paying a revised payment
or the initial episode during a sequence supply use, and will make amount to LUPA episodes that occur as
of adjacent episodes. We did not use improvements to our method of the only episode or the first episode in
cost reports in computing the LUPA accounting for NRS costs as the data a sequence of adjacent episodes. That
revised payment amount. We also do warrant. We encourage HHAs to additional amount has been calculated
not take into account the underlying develop in-house mechanisms to to be $87.93, for CY 2008. To account
reasons leading to a LUPA. improve their supply tracking, and to for the additional payment to LUPA
Comment: Several commenters were report supplies used on their claims. In episodes that occur as the first episode
unclear about how we propose to section III.C.4, we address the in a sequence of adjacent episodes or as
identify the timing of a LUPA episode mandatory reporting of supplies. the only episode, and maintain budget
as an only episode or initial episode in Comment: A commenter noted that neutrality, we reduce the national
a series of adjacent episodes. Another CMS has determined that later episodes standardized 60-day episode payment
noted commenter believed that the cost 7 percent more, but has chosen not rate.
LUPA continuing episode will be to differentiate early and later LUPA 3. The Significant Change in Condition
determined from claims data where the episodes. The commenter questioned (SCIC) Adjustment
start-of-care date is the same as the data that increases payment for one
‘‘from’’ date. In the proposed rule, for 2008, we
payment type and does not do the same proposed to eliminate our SCIC policy,
Response: A LUPA episode is 60 days
for another payment type. which allowed an HHA to adjust
long. An initial episode is an episode in
Response: Providing for an additional payment when a beneficiary
which a gap of greater than 60 days
payment for initial and only LUPA experiences a SCIC during the 60-day
exists before the from-date of that LUPA
episodes is actually similar to the episode that was not envisioned in the
episode. A LUPA episode that exists as
an only episode is an episode with a gap concept of early and later episodes original plan of care. The SCIC policy
of greater than 60 days both before the proposed for the full 60-day episode was designed and implemented
beginning and after the end of the LUPA payment. The results of data analysis primarily to protect HHAs from
episode. LUPAs, other than only done on LUPA episodes did not support receiving a lower, inadequate payment
episodes, would be considered as providing a revised payment for LUPA for a beneficiary who unexpectedly got
adjacent episodes to other episodes if no episodes that exist as the second or worse and became more expensive to
more than 60 days occur between the subsequent LUPA episode in a sequence the agency during the course of a 60-day
end of one episode and the beginning of of adjacent episodes, as the case-mix episode. Our margin analysis suggested
the next, except for those episodes that model does for all other types of that, on average, SCIC episodes had
have been PEP-adjusted. episodes. Instead, data do support a negative margins. We proposed to
Comment: A commenter noted that revised payment for initial and only eliminate the SCIC policy based on the
the LUPA payments cover about half the LUPA episodes. findings of our analysis and the
costs of rural agencies, and asked that Comment: While we received apparent difficulty the industry had in
we increase LUPA payment rates, widespread support for the revised interpreting when to apply the SCIC
particularly for rural agencies. LUPA payment, a commenter noted that adjustment policy. For a full description
Response: The per-visit rates used for the analysis focused principally on of the SCIC review and analysis, see CY
payment of LUPA episodes and used in nursing and physical therapy visits for 2008 HH PPS proposed rule (72 FR
the outlier calculation are based on visit LUPAs. The commenter encouraged 25425–25426).
cost data from audited cost reports. We CMS to examine the presence of other Comment: Several commenters were
believe this to be the most appropriate home health service visits (social concerned that with the elimination of
and accurate data on which to base service, occupational or speech therapy) the SCIC, there would be no avenue for
these rates. Currently, there exists no to ensure that the proposed payment reimbursement of supplies that were
rural add-on for home health services amount recognizes all service costs needed as a result of a change in
provided in a rural area. However, incurred with these initial visits. condition. Some commenters used the
LUPA payments are wage adjusted to Response: LUPA episodes average example of a home health patient
account for geographic differences. approximately 2.5 visits. In an initial or admitted with an unobservable pressure
Comment: Several commenters noted only LUPA episode, the first billable ulcer or surgical wound. The ulcer or
that the home health industry had not visit for the episode must be a skilled wound cannot be staged if it is
billed for supplies or kept good records visit. Consequently, the first visits of an unobservable, leaving the HHA with a
of supplies used, and that this initial or only LUPA episode would be minimum HHRG and large supply
contributed to the difficulty in either nursing or physical or speech expenses; the care needs greatly
analyzing NRS use in general and in therapy visits. It is these start of care increase when stageable. One
LUPA episodes. One commenter nursing and physical or speech therapy commenter asked for a simplified
suggested that billing for non-routine visits that occur when the case is supply SCIC to cover unanticipated
medical supplies, specifying the type of opened and the initial assessment takes supply costs that occur when a patient’s
supply and quantity, should be made place, that are longer than the average condition changes.
mandatory for all episodes and LUPAs visit length. Consequently, we believe it Response: As noted in a response to
to gather data for future evaluation of appropriate to base the revised payment a comment in section III.B.8, currently,
diagnosis and rates of payment. The for initial and only LUPA episodes on the OASIS guidelines for M0460 do not
allow a pressure ulcer with any eschar change in condition (SCIC) assessment. the elimination of the SCIC adjustment,
to be staged. We are aware of recent We note our proposal was limited to and to maintain budget neutrality, we
revisions issued by the National eliminating the SCIC payment reduce the national standardized 60-day
Pressure Ulcer Advisory Panel, adjustment from the HH PPS. Currently, episode payment rate. As such, we are
(NPUAP). Essentially, the NPUAP the assessment used in SCIC situations revising 42 CFR 484.205, 484.237, and
guidance permits the assessment of a is used in the quality monitoring aspect 484.240 to remove all references to the
wound for staging when the wound bed of the OASIS. This assessment is a SCIC adjustment.
is not completely covered with eschar or requirement integrated into the CoPs,
4. Non-Routine Medical Supplies (NRS)
slough. If the bed of the ulcer is found at § 484.18(b), and therefore any
completely covered with eschar/slough, change to the CoP requirement is To ensure that the variation in non-
NPUAP guidance stipulates that the beyond the scope of this payment rule. routine supplies is more appropriately
wound cannot be staged until some of Comment: A commenter suggested reflected in HH PPS, we proposed to
the necrotic tissue is removed. After that the adjustment to the national replace the original portion ($43.54) of
reviewing the NPUAP guidance, we standardized 60-day episode payment of the HH PPS base rate that accounted for
have revised the instructions $15.71 for the elimination of the SCIC NRS, with a system that pays for non-
accompanying this OASIS item to allow was incorrect. The commenter suggested routine supplies based on 5 severity
a wound to be staged if the bed of the that since SCICs have little impact on groups. The classification algorithm is
wound is partially covered by necrotic outlays (0.5 percent of total payments based on selected OASIS assessment
tissue and if the presence of eschar does regardless of urban/rural status, items, similar to the way the clinical
not obscure the depth of the tissue loss. ownership, or size) the calculation model was developed. We noted we
We hope this encourages HHAs to should have been $2,521.17 × 0.5 believed the original amount of $43.54
properly treat pressure ulcers and percent = $12.64 rather than the $15.71 (updated through 2008) per episode that
promote their healing. We believe this quoted in the proposed rule and asked accounts for NRS does not accurately
will allow for accurate payment for that the national standardized 60-day reflect the large variation in non-routine
home health patients with wounds that episode payment be adjusted. medical supplies use across patient
are partially covered with eschar/ Response: The adjustments to the type. In general, use of non-routine
slough. national standardized amount reflect medical supplies is unevenly
Comment: A majority of commenters our best estimates of the amount of the distributed across episodes of care in
appreciated the concept behind the budget-neutral target that is allocated in home health. Specifically, we found that
SCIC, but supported our decision to order to account for elimination of the patients with certain conditions, many
eliminate the SCIC, citing complexity SCIC, the LUPA add-on, and other of them related to skin conditions, were
and administrative burden. refinements that are taken as offsets to more likely to require high non-routine
Response: We appreciate the support the national standardized amount. The medical supply utilization. For a
for our proposal to eliminate the SCIC estimates of the cost of these complete description of our analysis
adjustment. adjustments also reflect the interaction and research, we refer readers to the CY
Comment: Several commenters noted of the outlier payments with other 2008 HH PPS proposed rule (72 FR
that if the SCIC is eliminated, payment elements during the 25426–25434).
completion of an ‘‘Other Follow-up’’ simulation. Comment: Several commenters noted
OASIS will not be necessary for Comment: A commenter suggested that conditions that generate high NRS
payment purposes. However, the that the SCIC adjustment not be costs are not accounted for in the NRS
Medicare home health CoPs requires eliminated. Another asked that we weights. They asked that NRS diagnoses
completion of the ‘‘Other Follow-up’’ withdraw our proposal to remove the include catheters, enteral nutrition,
OASIS when there is a SCIC. The SCIC until there had been time to chest drains, gastrointestinal tubes, and
commenters stated that completion of review the other changes resulting from an expanded list of ostomy supplies.
these assessments has been problematic, the refinement. Some commenters noted that wound
inconsistent, and burdensome for Response: The SCIC policy was supply payments are still inadequate.
HHAs, partly because of limited designed and implemented primarily to Commenters asked that the proposed
guidance from CMS regarding the kinds protect HHAs from receiving a lower, case-mix model be changed to allow
of clinical changes that require a new inadequate payment for a beneficiary scoring for these items, and that
comprehensive assessment. Specifically, that unexpectedly got worse and became payment for these items be increased
when a patient does have a change in more expensive to the agency during the beyond what is proposed in the rule.
condition, the plan of care is updated by course of a 60-day episode. Our Response: Section 1861(m)(5) of the
contacting the physician and recording examination of the SCIC adjustment Act defines home health services and
verbal/phone orders. This action by confirmed industry comments that specifically lists catheters, catheter
HHAs is not dependent on completion HHAs have had difficulty applying the supplies, ostomy bags and ostomy
of the OASIS. Additionally, collection SCIC policy, and that margin analysis, supplies as medical supplies.
and submission of OASIS data at this on average, shows that SCIC episodes Accordingly, catheters and catheter
time point often masks improvement have negative margins. We believe that supplies and bowel ostomy supplies are
made in the patient’s condition before it is now appropriate to remove the already included as covered NRS in the
the SCIC. Outcomes measures based on SCIC payment adjustment from HH PPS proposed rule. We also expanded the
the follow-up comprehensive and that the proposed refinement NRS listing of ostomy supplies to
assessment are likely to show less changes would not have had a include those for cystostomy,
improvement than a comparison of the significant impact on the SCIC payment tracheostomy, and urostomy.
patient at start of care and discharge. policy. The proposed rule notes that enteral
The commenters recommended that this In summary, based in part, upon and parenteral nutrition are Part B
Condition of Participation be comments received, as well as our services not covered by the home health
eliminated. continued analysis of this issue, we are benefit and not defined as non-routine
Response: We appreciate the finalizing our proposal to eliminate the supplies. The Medicare coverage
comments regarding the significant SCIC adjustment policy. To account for guidelines for enteral nutrition are
included in the proposed rule, along that HHAs will report NRS costs on their RHHI could not process supply
with a table of ‘‘Enteral Items and their claims. To ensure that NRS costs lines on claims for an unspecified
Services’’ which includes the HCPCS are being reported, claims that do not period of time. Several commenters
codes needed for billing. The table report NRS costs, unless explicitly mentioned high supply costs for
includes codes for tubing and other noted by the HHA that NRS was not particular items, such as chest drains,
supplies needed for administering provided, will be returned to the which can cost $500 to $600 per month.
enteral nutrition. If a home health provider (RTP). For episodes in which Commenters asked that CMS abandon
patient needs enteral nutrition and NRS was provided, the provider will the NRS supply model as proposed as
meets the criteria for coverage, need to resubmit the claim with NRS it would underpay HHAs for supplies
providers may claim reimbursement by reported. For episodes in which NRS used.
using the UB–92 claim form. Payment is was not provided, the HHA will need to Response: In general, we acknowledge
then made by the RHHI under the Part explicitly note that fact on the claim. NRS use is unevenly distributed across
B Medicare Fee Schedule, rather than We will allow a grace period, which episodes of care in home health. While
through the home health benefit. will be determined and communicated most patients do not use NRS, many use
Comment: Most commenters believed in instructions from CMS. This will a small amount, and a small number of
that NRS supplies are underreported; provide stronger incentives to HHAs to patients use a large amount. It is
the industry is grappling with an report NRS, resulting in more accurate important to note that while Durable
efficient mechanism to consistently NRS data for possible future refinements Medical Equipment (DME) is covered
capture the supplies used. While most to this aspect of the HH PPS. We will under the home health benefit, such
commenters appreciated our proposed continue to study supply use, and will items are not included in the HH PPS
increase in our approach to better make improvements to how we account payment and thus can be billed for
account for NRS payments, many noted for and pay for NRS as the data warrant. separately either by the HHA or a DME
that the analysis was based on Comment: A commenter is concerned supplier and are not subject to home
incomplete information that that the bundling of NRS in a budget- health consolidated billing. In
inadequately reflects the providers’ true neutral system will continue to create a developing the proposed approach for
costs. One commenter suggested that growing payment disparity as new and NRS payment, we sought to more
CMS consider requiring agencies to more expensive technologies are accurately match Medicare payments for
report supply costs if they wish to applied to home care. Each year, new NRS to agency costs. The proposed and
receive reimbursement above the first supplies are added to the HH PPS final regression models were developed
severity level. Without such a bundle that did not exist when the after creating additional variables from
requirement, agencies that fail to make baseline was established for HH PPS. OASIS items and targeting certain
the effort to identify and report these The commenter urged CMS to freeze conditions expected to be predictors of
costs will receive the same advantages NRS codes that are currently bundled NRS use based on clinical
as those that do, and would have an and unbundle new NRS technology considerations. The sample only
unfair result. from HH PPS as it emerges. Another included HHAs whose total charges on
CMS was also encouraged to continue commenter asked that NRS be claims matched their total charges on
studying the NRS issue as the reimbursed through the DME fee their cost reports for that same year, and
compensation can fall far short of what schedule. thus, any issues with RHHI processing
agencies expend for their most supply- Response: We appreciate the concern did not impede the analysis.
intensive patients. about supply costs and particularly Since the proposed rule, we updated
Response: We appreciate the about the cost of new technologies. If our data base for the NRS analysis to be
commenter’s concern that without a agencies will report these supplies on representative of episodes from 2004
requirement for HHAs to report NRS on their claims, the costs of supplies, and 2005. This analysis relies on cost
the claim, those agencies that fail to including new technologies, will be reports to derive cost-to-charge ratios for
make the effort to identify and report captured in future data analyses. estimating NRS costs on claims, and the
NRS costs will receive the same Section 1895 of the Act, as added by latest data available incorporated 2004
considerations for payment as those that section 4603(a) of the Balanced Budget cost reports. The results of modeling the
do report NRS. We believe that it is Act of 1997, provided the authority for NRS costs are shown in the scoring
imperative that HHAs report these the development of a HH PPS for all table, Table 10A. Since updating the
supplies on their claims so that we can Medicare-covered home health services data base, we have added several new
improve the accuracy of our system and paid on a reasonable cost basis. Section variables, such as diabetic ulcers, and
better reflect costs when paying for 1895(b)(1) of the Act requires the re-specified the treatment of certain
NRS. Secretary to establish a HH PPS for all wound variables (for example, counts
We have consistently encouraged costs of home health services, including and stages of pressure ulcers) in the
home health agencies to develop in- medical supplies. Therefore, medical final model.
house mechanisms to improve their supplies are bundled into the HH PPS We explored the concern that the
supply tracking, and to report supplies payment, as required by the statute, and proposed 5th severity group level did
used on their claims. Our data for 2003 are subject to consolidated billing. DME, not provide adequate reimbursement for
indicate that the percentages of agencies on the other hand, was explicitly episodes with a high-utilization of NRS.
not reporting supplies on claims to be statutorily excluded from consolidated In response to those comments, and as
similar to percentages that existed billing. a result of further analysis, we are
during the HH IPS baseline. We are Comment: Several commenters were implementing a system that pays for
concerned with the commenter’s concerned that the proposed model for non-routine supplies based on 6 severity
assertion that NRS supplies are reimbursing NRS has poor performance groups. The 6th group is a subset of the
underreported, and the limitations this and a low R-squared of 13.7 percent. previously proposed 5th group. Our
underreporting puts on any future work The commenter cited industry analysis revealed that a small
towards refining payment to HHAs for difficulties in reporting supply costs, percentage of cases in the proposed 5th
providing NRS. To adequately account and high supply costs for particular severity group may not have adequately
for and pay for NRS costs, we expect diagnoses. One commenter noted that reflected the resources required for
providing care in this group. and the current payment system rehabilitation potential, functional
Consequently, in recognizing that a overcompensates these episodes. The limitations, activities permitted,
small percentage of episodes incur final NRS approach better matches NRS nutritional requirements, medications
higher costs than the majority of payments with NRS costs incurred in and treatments, any safety measures to
episodes in the 5th severity group, we the episode. We will continue to look protect against injury, instructions for
split the small percentage of high cost for ways to improve our approach to timely discharge or referral, and any
NRS cases from the 5th severity group account for NRS. other appropriate items.’’ Accordingly,
to form a 6th severity group. Under the Comment: Several commenters noted because the CoPs require that all
final 6 severity NRS approach, the 6th that the NRS analysis was based on pertinent diagnoses are included on the
severity level is associated with a higher 1997 costs rather than more recent data; plan of care, the plan of care should
score and higher payment than any of one suggested using 2005 data. Another include any conditions for which NRS
the severity levels in the proposed rule. suggested that we tie annual increases is necessary for the treatment of those
The R-squared for this final model is in supply costs to a medical supply diagnoses, and NRS should be provided
16.6 percent. The sample was trimmed inflation index. and reported being supplied.
Response: The analysis file used to Comment: Several commenters asked
to eliminate outliers, where outliers
develop the proposed NRS case-mix for additional diagnoses codes to be
were defined to be episodes with NRS
model for the proposed rule was based included in the NRS supply list. A few
costs estimated to be $3,500 or higher.
on 2001 cost reports. The cost reports asked for V44.0–V.44.9 specifically.
The trimming procedure resulted in a
were then linked to claims to determine While they appreciate the attempt to
small loss from the total sample size. A
the cost-to-charge ratios, which were improve NRS payment, several
total of 2,653 episodes were excluded
used to estimate NRS costs for the
(less than 0.09 percent) out of a total commenters noted that the payments are
episodes in the sample. For this final
sample of 2,974,678 episodes. Our still inadequate.
rule, we updated the database upon
sample for the NRS analysis consisted of which our payment proposal for NRS Response: We tested selected stoma
all agencies whose total charges was based to use 2004 and 2005 data. V-codes mentioned by the commenter.
reported on claims matched their total Again, to refine payments for NRS will We selected codes for testing that were
charges reported in the cost reports, but depend on the quality of the data not already represented by other
as these trimming requirements show, available in claims and costs reports for variables in the model. The final NRS
the resulting sample included a relative succeeding years. We note we are model reflects additional conditions for
few questionable sample data points. revising our NRS policy to require scoring, when reported using the
We believe the final regression model HHAs to specifically note on submitted selected V-codes. We also believe under
represents the relationships between claims NRS in any episode in which a our final 6 severity group methodology,
case-mix and NRS cost among a highly NRS is provided. HH PPS will better reflect the NRS costs
representative sample of episodes and Comment: A commenter asked that and usage.
agencies nationally. HHAs only be responsible for providing In summary, we are implementing a 6
While we have not yet developed a NRS for those conditions that are severity group methodology for the
statistical model that has performed included in the plan of care. paying of NRS in the HH PPS, as shown
with a high degree of predictive Response: The plan of care is to be in Table 9 below. We believe that
accuracy, we believe this may due to the established and periodically reviewed adding a 6th severity group better
limited data available to model NRS by the patient’s physician. The CoPs for recognizes episodes with higher NRS
costs, and the likelihood that OASIS HHAs in 42 CFR 484.18 state that ‘‘the costs. To account for paying of NRS
does not have any measures available plan of care developed in consultation through the implementation of a 6-
for some kinds of NRS. Notwithstanding with the agency staff covers all pertinent severity group methodology, and to
these concerns, we are changing the diagnoses, including mental status, maintain budget neutrality, we reduce
payment system because the majority of types of services and equipment the national standardized 60-day
episodes do not incur any NRS costs, required, frequency of visits, prognosis, episode payment rate.
TABLE 9. RELATIVE WEIGHTS FOR NON-ROUTINE MEDICAL SUPPLIES—SIX-GROUP APPROACH

Percentage
Severity Points Relative Payment
of
level (scoring) weight amount
episodes
1 ....................................................................... 63.7 0 .................................................................... 0.2698 $14.12

2 ....................................................................... 20.6 1 to 14 ........................................................... 0.9742 51.00
3 ....................................................................... 6.7 15 to 27 ......................................................... 2.6712 139.84
4 ....................................................................... 5.4 28 to 48 ......................................................... 3.9686 207.76
5 ....................................................................... 3.2 49 to 98 ......................................................... 6.1198 320.37
6 ....................................................................... 0.3 99+ ................................................................ 10.5254 551.00
Note: NRS conversion factor = $52.35. The the HH PPS episode base rate ($49.62), after We have also included the final
NRS conversion factor is the market-basket- adjustment for nominal change in case-mix. versions of Table 10A and Table 10B
updated amount CMS originally included in below.
TABLE 10A.—NRS CASE-MIX ADJUSTMENT VARIABLES AND SCORES

Item Description Score
SELECTED SKIN CONDITIONS
1 ....................................... Primary diagnosis = Anal fissure, fistula and abscess ................................................................................. 15

2 ....................................... Other diagnosis = Anal fissure, fistula and abscess ..................................................................................... 13
3 ....................................... Primary diagnosis = Cellulitis and abscess ................................................................................................... 14
4 ....................................... Other diagnosis = Cellulitis and abscess ...................................................................................................... 8
5 ....................................... Primary or other diagnosis = Diabetic ulcers ................................................................................................ 20
6 ....................................... Primary diagnosis = Gangrene ...................................................................................................................... 11
7 ....................................... Other diagnosis = Gangrene ......................................................................................................................... 8
8 ....................................... Primary diagnosis = Malignant neoplasms of skin ........................................................................................ 15
9 ....................................... Other diagnosis = Malignant neoplasms of skin ........................................................................................... 4
10 ..................................... Primary or Other diagnosis = Non-pressure and non-stasis ulcers .............................................................. 13
11 ..................................... Primary diagnosis = Other infections of skin and subcutaneous tissue ....................................................... 16
12 ..................................... Other diagnosis = Other infections of skin and subcutaneous tissue .......................................................... 7
13 ..................................... Primary diagnosis = Post-operative Complications ....................................................................................... 23
14 ..................................... Other diagnosis = Post-operative Complications .......................................................................................... 15
15 ..................................... Primary diagnosis = Traumatic Wounds and Burns ..................................................................................... 19
16 ..................................... Other diagnosis = Traumatic Wounds and Burns ......................................................................................... 8
17 ..................................... Primary or other diagnosis = V code, Cystostomy care ............................................................................... 16
18 ..................................... Primary or other diagnosis = V code, Tracheostomy care ........................................................................... 23
19 ..................................... Primary or other diagnosis = V code, Urostomy care ................................................................................... 24
20 ..................................... OASIS M0450 = 1 or 2 pressure ulcers, stage 1 ......................................................................................... 4
21 ..................................... OASIS M0450 = 3+ pressure ulcers, stage 1 ............................................................................................... 6
22 ..................................... OASIS M0450 = 1 pressure ulcer, stage 2 ................................................................................................... 14
23 ..................................... OASIS M0450 = 2 pressure ulcers, stage 2 ................................................................................................. 22
33 ..................................... OASIS M0450e = 1 (unobserved pressure ulcer(s)) .................................................................................... 17
34 ..................................... OASIS M0470 = 2 (2 stasis ulcers) .............................................................................................................. 6
35 ..................................... OASIS M0470 = 3 (3 stasis ulcers) .............................................................................................................. 12
36 ..................................... OASIS M0470 = 4 (4+ stasis ulcers) ............................................................................................................ 21
37 ..................................... OASIS M0474 = 1 (unobservable stasis ulcers) ........................................................................................... 9
38 ..................................... OASIS M0476 = 1 (status of most problematic stasis ulcer: fully granulating) ............................................ 6
39 ..................................... OASIS M0476 = 2 (status of most problematic stasis ulcer: early/partial granulation) ................................ 25
40 ..................................... OASIS M0476 = 3 (status of most problematic stasis ulcer: not healing) .................................................... 36
41 ..................................... OASIS M0488 = 2 (status of most problematic surgical wound: early/partial granulation) .......................... 4
42 ..................................... OASIS M0488 = 3 (status of most problematic surgical wound: not healing) .............................................. 14
OTHER CLINICAL FACTORS
43 ..................................... OASIS M0550 = 1 (ostomy not related to inpt stay/no regimen change) .................................................... 27
44 ..................................... OASIS M0550 = 2 (ostomy related to inpt stay/regimen change) ................................................................ 45
45 ..................................... Any ‘Selected Skin Conditions’ (rows 1–42 above) AND M0550 = 1 (ostomy not related to inpt stay/no 14
regimen change).
46 ..................................... Any ‘Selected Skin Conditions’ (rows 1–42 above) AND M0550 = 2 (ostomy related to inpt stay/ regimen 11
change).
47 ..................................... OASIS M0250 (Therapy at home) =1 (IV/Infusion) ....................................................................................... 5
48 ..................................... OASIS M0520 = 2 (patient requires urinary catheter) .................................................................................. 9
49 ..................................... OASIS M0540 = 4 or 5 (bowel incontinence, daily or >daily) ....................................................................... 10
Note: Points are additive, however points from the same diagnosis/condition group. www.cms.hhs.gov/HomeHealthPPS/
may not be given for the same line item in See Table 12b for definitions of diagnosis/ 03_coding&billing.asp for definitions of
the table more than once. Points are not condition groups. primary and secondary diagnoses.
assigned for a secondary diagnosis if points Please see Medicare Home Health
are already assigned for a primary diagnosis Diagnosis Coding guidance at http://
TABLE 10B.—ICD–9–CM DIAGNOSES INCLUDED IN THE DIAGNOSTIC CATEGORIES FOR THE NONROUTINE SUPPLIES
(NRS) CASE-MIX ADJUSTMENT MODEL

ICD–9–CM
Diagnostic Category Manifestation Short Description of ICD–9–CM Code
Code*
Anal fissure, fistula and abscess .. 565 .................... ........................... ANAL FISSURE AND FISTULA.
(NRS) CASE-MIX ADJUSTMENT MODEL—Continued
ICD–9–CM
Code*
566 .................... ........................... ABSCESS OF ANAL AND RECTAL REGIONS.

Cellulitis and abscess .................... 681.00 ............... ........................... FINGER—CELLULITIS AND ABSCESS, UNSPECIFIED.
681.01 ............... ........................... FELON.
681.10 ............... ........................... TOE—CELLULITIS AND ABSCESS, UNSPECIFIED.
681.9 ................. ........................... CELLULITIS AND ABSCESS OF UNSPECIFIED DIGIT.
682 .................... ........................... OTHER CELLULITIS AND ABSCESS.
Diabetic Ulcers .............................. 250.8x & ........................... (PRIMARY OR FIRST OTHER DIAGNOSIS = 250.8x AND PRI-
707.10–707.9. MARY OR FIRST OTHER DIAGNOSIS = 707.10- 707.9).
Gangrene ....................................... 440.24 ............... ........................... ATHERSCLER-ART EXTREM W/GANGRENE.
785.4 ................. M ...................... GANGRENE.
Malignant neoplasms of skin ......... 172 .................... ........................... MALIGNANT MELANOMA OF SKIN.
173 .................... ........................... OTHER MALIGNANT NEOPLASM OF SKIN.
Non-pressure and non-stasis ul- 440.23 ............... ........................... ATHEROSCLER-ART EXTREM W/ULCERATION.
cers (other than diabetic).
447.2 ................. ........................... RUPTURE OF ARTERY.
447.8 ................. ........................... OTHER SPECIFIED DISORDERS OF ARTERIES AND
ARTERIOLES.
707.10 ............... ........................... ULCER OF LOWER LIMB, UNSPECIFIED.
707.11 ............... ........................... ULCER OF THIGH.
707.12 ............... ........................... ULCER OF CALF.
707.13 ............... ........................... ULCER OF ANKLE.
707.14 ............... ........................... ULCER OF HEEL AND MIDFOOT.
707.15 ............... ........................... ULCER OF OTHER PART OF FOOT.
707.19 ............... ........................... ULCER OF OTHER PART OF LOWER LIMB.
707.8 ................. ........................... CHRONIC ULCER OTHER SPECIFIED SITE.
707.9 ................. ........................... CHRONIC ULCER OF UNSPECIFIED SITE.
Other infections of skin and sub- 680 .................... ........................... CARBUNCLE AND FURUNCLE.
cutaneous tissue.
683 .................... ........................... ACUTE LYMPHADENITIS.
685 .................... ........................... PILONIDAL CYST.
686 .................... ........................... OTH LOCAL INF SKIN&SUBCUT TISSUE.
Post-operative Complications ........ 998.11 ............... ........................... HEMORRHAGE COMPLICATING A PROCEDURE.
998.12 ............... ........................... HEMATOMA COMPLICATING A PROCEDURE.
998.13 ............... ........................... SEROMA COMPLICATING A PROCEDURE.
998.2 ................. ........................... ACC PUNCT/LACERATION DURING PROC NEC.
998.4 ................. ........................... FB ACC LEFT DURING PROC NEC.
998.6 ................. ........................... PERSISTENT POSTOPERATIVE FIST NEC.
998.83 ............... ........................... NON-HEALING SURGICAL WOUND NEC.
Traumatic wounds, burns and 870 .................... ........................... OPEN WOUND OF OCULAR ADNEXA.
post-operative complications.
872 .................... ........................... OPEN WOUND OF EAR.
873 .................... ........................... OTHER OPEN WOUND OF HEAD.
874 .................... ........................... OPEN WOUND OF NECK.
875 .................... ........................... OPEN WOUND OF CHEST.
876 .................... ........................... OPEN WOUND OF BACK.
877 .................... ........................... OPEN WOUND OF BUTTOCK.
878 .................... ........................... OPEN WND GNT ORGN INCL TRAUMAT AMP.
879 .................... ........................... OPEN WOUND OTH&UNSPEC SITE NO LIMBS.
880 .................... ........................... OPEN WOUND OF SHOULDER&UPPER ARM.
881 .................... ........................... OPEN WOUND OF ELBOW, FOREARM&WRIST.
882 .................... ........................... OPEN WOUND HAND EXCEPT FINGER ALONE.
883 .................... ........................... OPEN WOUND OF FINGER.
884 .................... ........................... MX&UNSPEC OPEN WOUND UPPER LIMB.
885 .................... ........................... TRAUMATIC AMPUTATION OF THUMB.
886 .................... ........................... TRAUMATIC AMPUTATION OTHER FINGER.
887 .................... ........................... TRAUMATIC AMPUTATION OF ARM&HAND.
890 .................... ........................... OPEN WOUND OF HIP AND THIGH.
891 .................... ........................... OPEN WOUND OF KNEE, LEG, AND ANKLE.
892 .................... ........................... OPEN WOUND OF FOOT EXCEPT TOE ALONE.
893 .................... ........................... OPEN WOUND OF TOE.
894 .................... ........................... MX&UNSPEC OPEN WOUND LOWER LIMB.
895 .................... ........................... TRAUMATIC AMPUTATION OF TOE.
896 .................... ........................... TRAUMATIC AMPUTATION OF FOOT.
897 .................... ........................... TRAUMATIC AMPUTATION OF LEG.
941 except ........................... BURN OF FACE, HEAD, AND NECK.

941.0x and
941.1x.
942 except ........................... BURN OF TRUNK.
942.0x and
942.1x.
(NRS) CASE-MIX ADJUSTMENT MODEL—Continued
ICD–9–CM
Code*
943 except ........................... BURN OF UPPER LIMB, EXCEPT WRIST AND HAND.
943.0x and
943.1x.
944 except ........................... BURN OF WRIST(S) AND HAND(S).
944.0x and
944.1x.
945 except ........................... BURN OF LOWER LIMB(S).
945.0x and
945.1x.
946.2 ................. ........................... BURNS OF MULTIPLE SPECIFIED SITES, BLISTERS, EPI-
DERMAL LOSS [SECOND DEGREE].
946.3 ................. ........................... BURNS OF MULTIPLE SPECIFIED SITES, FULL-THICKNESS SKIN
LOSS [THIRD DEGREE NOS].
946.4 ................. ........................... BURNS OF MULTIPLE SPECIFIED SITES, DEEP NECROSIS OF
UNDERLYING TISSUES [DEEP THIRD DEGREE] WITHOUT
MENTION OF LOSS OF A BODY PART.
946.5 ................. ........................... BURNS OF MULTIPLE SPECIFIED SITES, DEEP NECROSIS OF
UNDERLYING TISSUES [DEEP THIRD DEGREE] WITH LOSS
OF A BODY PART.
998.31 ............... ........................... DISRUPTION OF INTERNAL OPERATION WOUND.
998.32 ............... ........................... DISRUPTION OF EXTERNAL OPERATION WOUND.
998.51 ............... ........................... INFECTED POSTOPERATIVE SEROMA.
998.59 ............... ........................... OTHER POSTOPERATIVE INFECTION.
V-code, Cystostomy Care ............. V55.5 ................. ........................... CYSTOSTOMY—CARE.
V-code, Tracheostomy Care ......... V55.0 ................. ........................... TRACHEOSTOMY—CARE.
V-code, Urostomy Care ................. V55.6 ................. ........................... OTHER ARTIFICIAL OPENING OF URINARY TRACT-
NEPHROSTOMY, URETEROSTOMY, URETHROSTOMY.
To ensure that NRS costs are being outlier payments are no more than 5 target of outlier payments to total HH
reported, claims that do not report NRS percent of total estimated HH PPS PPS payments, and based the change on
costs, unless explicitly noted by the payments. For a full description of our analysis of CY 2005 HH claims.
HHA that NRS was not provided, will outlier policy, we refer to the CY 2008 In the proposed rule (72 FR 25434),
be returned to the provider (RTP). For HH PPS proposed rule (72 FR 25434– we stated that preliminary analysis
episodes in which NRS was provided, 25435). showed that outlier payments, as a
the provider will need to resubmit the The wage adjusted fixed dollar loss percentage of total HH PPS payments,
claim with NRS reported. For episodes (FDL) amount represents the amount of have increased on a yearly basis. With
in which NRS was not provided, the loss that an agency must bear before an outlier payments having increased in
HHA will need to explicitly note that episode becomes eligible for outlier recent years, and given the unknown
fact on the claim. We will allow a grace payments. The loss sharing ratio is 0.80. effects that the proposed refinements
period, which will be determined and As noted in the proposed rule, when the may have on outliers, we proposed to
communicated in instructions from HH PPS system was implemented, we maintain the FDL ratio at 0.67. We
CMS. This will improve data on NRS, in chose a value of 0.80 for the loss-sharing believed that this would continue to
the home health setting, providing us ratio and an FDL ratio of 1.13. In the meet the statutory requirement of
with better data with which to analyze October 2004 final rule, we revised the having an outlier payment outlay that
and evaluate payment to HHAs for NRS FDL ratio to 0.70, based on analysis of does not exceed 5 percent of total HH
in the future. We will monitor the CY 2003 HH PPS data. We believed this PPS payments, while still providing for
accuracy of the 6-severity group updated FDL ratio of 0.70 preserved a an adequate number of episodes to
methodology for payment of NRS. We reasonable degree of cost sharing, qualify for outlier payments. We stated
will continue to monitor the accuracy allowed a greater number of episodes to in the proposed rule that we would rely
and completeness of the reporting of qualify for outlier payments, and yet did on the latest data and best analysis
NRS costs. Finally, we will explore not result in a projected target available at the time to estimate outlier
alternative methods for accounting for percentage of estimated outlier payments and update the FDL ratio in
NRS costs and payments in the future. payments of more than 5 percent. the final rule if appropriate.
Our CY 2006 update to the HH PPS Comment: A commenter supported
D. The Outlier Policy rates, which was based upon CY 2004 our proposed outlier policy but does not
As noted in section II, of this final HH claims data, again revised the FDL understand why it needs to be capped
rule with comment period, outlier ratio from 0.70 to 0.65 to allow even at 5 percent.
payments are made for episodes for more home health episodes to qualify Response: The statute, at section
which the estimated cost exceeds a for outlier payments and to better meet 1895(b)(5) of the Act, limits estimated
threshold amount and are intended to the estimated 5 percent target of outlier outlier payments to no more than 5
address home health episodes that incur payments as a percentage of total HH percent of the total estimated HH PPS
unusually high costs due to patient PPS payments. In our CY 2007 update, payments during a given year.
health care needs. Section 1895(b)(5) of we again changed the FDL ratio from Comment: Commenters stated that the
the Act requires that the estimated total 0.65 to 0.67 to better meet the 5 percent fixed dollar loss (FDL) ratio should be
reduced since the 0.67 FDL ratio will The current model’s estimate of the estimated HH PPS payments are
not result in CMS spending the targeted FDL ratio, using CY 2005 data, is 0.47. estimated to be approximately 5.7
5 percent for outlier payments as a This is higher than the estimate from the percent. We take the 0.7 percent (the
percentage of total estimated HH PPS FY 2003 data, which was 0.42, reflecting percentage amount in excess of the 5
payments. CMS should adjust its growth in the outlier percentage, as percent target) and multiply it by 0.31
technique for calculating the FDL ratio noted earlier. Given current trends, we (the estimated amount of change in the
by using its historical data on actual estimate that we would exceed the 5 FDL ratio for every one percentage point
outlays. percent statutory limit on outlier change in the outlier payment
Response: Given that outlier payments using either the model’s FDL percentage), (0.7 * 0.31), resulting in a
payments as a percentage of total HH ratio of 0.47, or the proposed FDL ratio change in the FDL ratio of 0.22. We add
PPS payments have increased in recent of 0.67. In order to capture the most that 0.22 change in the FDL ratio to the
years and given the unknown effects of recent trends in the increase of outlier FDL ratio in effect in 2007 (0.67),
the proposed refinements, we proposed payments, and to appropriately account arriving at a final FDL ratio of 0.89.
to maintain the FDL ratio at 0.67. At the for seasonal differences that may exist Based on this analysis, we believe that
time of the proposed rule, data in outlier episodes, we compared the setting the FDL ratio at 0.89 would be
indicated that by maintaining the FDL percentage of outlier payments as a the most prudent course given these
ratio at 0.67 we would continue to meet percentage of total HH PPS payments trends and the unknown effects of the
the statutory requirement that estimated from the first quarter of CY 2006 (4.52 refinements on outliers. As previously
outlier payments be no more than 5 percent) and the first quarter of CY 2007 stated, we further believe that a FDL
percent of total estimated HH PPS (4.85 percent). That estimated annual ratio of 0.89 will continue to meet the
payments, yet an adequate number of percentage increase in outlier payments statutory requirement of having an
episodes would qualify for outlier is calculated to be 7.3 percent. We estimated outlier payment outlay that
payments. In the proposed rule, we estimate the percentage of outlier does not exceed the 5 percent of total
indicated that preliminary analysis, payments for CY 2007 by multiplying estimated HH PPS payments, while still
which was based on 2003 data, showed 4.97 percent (the percentage of outlier providing for an adequate number of
the FDL ratio could be as low as 0.42. payments for CY 2006) by 1.073 (the episodes to qualify for outlier payments.
The 2003 data used in Abt’s modeling estimated annual percentage increase in As our best estimate is that an FDL of
of the refined HH PPS for the proposed outlier payments noted above) for an 0.89 is consistent with outlier payments
rule was somewhat limited in that it estimated percentage of outlier of no more than 5.0 percent of total
was not able to take into account more payments as a percent of total estimated estimated HH PPS payments, we will
recent trends in actual outlier HH PPS payments for CY 2007 of 5.33 account for the estimated 5 percent
expenditures. Similarly, Abt’s modeling percent. We multiply the 5.33 percent outlier payments in our updating of the
of the refined HH PPS for this final rule by 1.073, to estimate the percentage of HH PPS rates. We will continue to
is still somewhat limited in that it is not outlier payments as a percent of total monitor the trends in outlier payments
able to take into account the latest estimated HH PPS payments for CY and the effects of the refinements, and
available data on actual outlier 2008. That calculation results in an will adjust the FDL ratio as needed.
expenditures. Consequently, as we estimated percentage of outlier Comment: Several commenters
stated in the proposed rule, in the payments as a percent of total estimated supported eliminating the outlier policy
interest of using the latest data and best HH PPS payments for CY 2008 of 5.7 and redistributing the 5 percent outlier
analysis available, we have performed percent. allocation, which has never been fully
supplemental analysis on more recent We then analyzed the sensitivity of distributed anyway, in order to increase
data in order to best estimate the FDL the percent of outlier payments to total the standardized payment rates. The
ratio. payments to variations in the FDL ratio. commenters believed that the outlier
When we revised the FDL from 1.13 Using simulations of the values of FDLs policy is disadvantageous to efficient
to .70 in CY 2005, we expected to consistent with alternative outlier and effective HHAs. Despite caring for
observe an increase in outlier payments payment percents based on CY 2005 very sick, resource intensive patients,
as a percent of total payments to better data (the latest data available for such some HHAs have never received any
meet our projected target percentage of an analysis), we used linear regression benefit from the outlier policy. The
not more than 5 percent. In addition, for to estimate the change in the FDL ratio commenters suggested that
CY 2006 and CY 2007 (with relatively associated with a 1 percentage point redistributing the outlier allocation to
stable FDLs of .65 and .67), we would change in the percent of outlier the standardized payment rates would
have anticipated that outlier payments payments. That linear regression ensure a more effective use of the
would have remained relatively stable analysis shows that a one percentage budgeted Medicare home health funds.
and not exceed 5 percent of estimated point change in the outlier payment Another commenter suggested we
HH PPS payments for each given year. percentage is associated with a negative reduce the maximum outlier payments
Instead, experience has shown that 0.31 change in the FDL ratio. That is, to as a percentage of total HH PPS payment
outlier payments have been increasing reduce the percent of outlier payments from 5 percent to 1 percent.
as a percent of total payments from 4.1 by one percentage point, it would be Response: We appreciate the
percent in CY 2005 to 4.97 percent in necessary to increase the FDL ratio by comment. However, we continue to
CY 2006 and, we estimate, 5.33 percent 0.31. believe that maintaining an outlier
in CY 2007. These increasing percents Using this analysis we looked to see policy is beneficial to the home health
imply that the cost distribution of what adjustment, to the FDL ratio, community. We have set the loss
episodes is changing and that our would be appropriate in estimating sharing ratio and the fixed dollar loss
estimates of the FDL need to account for outlier payments of up to but not more amount in such a way to preserve a
these changes in order to better match than 5 percent of total estimated HH reasonable degree of cost sharing while
experience and to not exceed the PPS payments in CY 2008. As also allowing an appropriate number of
statutory limit of not more than 5 mentioned above, we have estimated episodes to qualify for outlier payments.
percent as a percentage of total that with an FDL ratio of 0.67, outlier We disagree with the suggestion that
estimated HH PPS payments. payments as a percentage of total we reduce the maximum outlier
percentage from 5 percent of total HH analysis and data available, including market basket. For full description of
PPS payments to 1 percent. We believe trend analysis and linear regression our proposal to revise and rebase the
that the current policy is more analysis described above, we have home health market basket, we refer to
equitable, and that reducing the adjusted the current FDL ratio of 0.67 to the CY 2008 HH PPS proposed rule (72
percentage could result in reducing 0.89. We believe that we have accounted FR 25435–25442). In the proposed
access to home health care by high for the latest observed trends in outlier revised and rebased home health market
needs patients. payments, and incorporated the best basket, the labor-related share would be
Comment: A commenter stated that analysis available to determine that an 77.082 percent. The labor-related share
the outlier policy is fiscally punitive to increase in the FDL ratio is necessary in includes wages and salaries and
the HH industry and that it appears to order to continue to meet the statutory employee benefits. The proposed non
be a back door mechanism to reduce requirement of having an outlier labor-related share would be 22.918
payments to the industry. The payment outlay that does not exceed 5 percent. The increase in the labor-
commenter suggested eliminating the percent of total HH PPS payments, related share using the 2003-based home
outlier policy and revising the while still providing for an adequate health market basket is primarily due to
standardized rates to include the 5 number of episodes to qualify for outlier the increase in the benefit cost weight.
percent outlier allocation. payments. Comment: Several commenters
Response: Section 1895(b)(5) of the Therefore, in this final rule we are objected to our proposal to change the
Act allows the Secretary to provide an implementing a FDL ratio of 0.89 for FY labor-related share to 77.082 percent
adjustment to the case-mix and wage 2008. To account for an outlier policy and requested that CMS maintain a
adjusted national 60-day episode that estimates outlier payments to be no labor-related share of 76.775 percent.
payment amount when episodes incur more than 5 percent of total HH PPS One commenter noted that the higher
unusually large costs due to patient payments, and to maintain budget labor-related share would have an
home care needs. Section 1895(b)(5) of neutrality, we reduce the national adverse impact on reimbursement
the Act further stipulates that the total standardized 60-day episode payment particularly for rural home health care
outlier payments in a given year may rate. We are revising 42 CFR 484.240(b) providers who have wage indices of less
not exceed 5 percent of total projected (‘‘Methodology used for the calculation than 1.0. The commenter proposed that
estimated HH PPS payments. Again, as of the outlier payment’’) to remove CMS should withdraw its proposal to
stated above, we continue to believe that references to the SCIC adjustment. We increase the labor-related share of the
the benefit to the home health will continue to monitor trends in the HH PPS rate.
community of maintaining an outlier data, along with the effects of the Response: Since the inception of HH
policy is consistent with the statute and refinements, on outlier payments, and PPS, the home health labor-related share
outweighs not having an outlier policy. will update the FDL as needed. We will has been based on the sum of the
Comment: One commenter asked that also continue to review the outlier weights for wages and salaries and
standards for the outlier provision be payments using the administrative data fringe benefits of the home health
changed to allow agencies to recover we monitor yearly. Future reviews will market basket index. We also note the
their costs for those most expensive, wage index is estimated independently
consider the appropriateness of outlier
high needs patients. This would from the labor-related share. The labor-
payments in the entire context of the
encourage agencies to accept these cases related share is calculated based on data
refinements being finalized in this
and provide appropriate care. submitted on the home health Medicare
regulation.
Response: We appreciate the cost reports for both rural and urban
comment. Again, we believe we have set E. The Update of the HH PPS Rates freestanding home health care facilities.
the loss sharing ratio and the fixed The proposed change in the labor-
1. The Home Health Market Basket
dollar loss amount in such a way as to related share is primarily attributable to
Update
preserve a reasonable degree of cost the rebasing of the market basket from
sharing while allowing an appropriate Section 1895(b)(3)(B) of the Act, as base year 2000 to 2003. The 2003 data,
number of episodes to qualify for outlier amended by section 5201 of the DRA, the most recent and comprehensive data
payments. We also believe the FDL ratio requires for CY 2008 that the standard available at the time of this rebasing,
will allow us to better meet the statutory prospective payment amounts be reflect that labor-related costs are
percentage imposed on outlier increased by a factor equal to the increasing faster than aggregate non
payments. applicable home health market basket labor-related costs. Based on the
Comment: A commenter wrote that it percentage increase. The proposed rule submitted cost report data from 2001 to
was unwise to dismiss the need to contained a home health market basket 2003, the weight for wages and salaries
adjust the outlier threshold at the same percentage increase of 2.9 percent. has been declining while the weight for
time that an increase in HH PPS Using revised updated data, we now fringe benefits has been increasing, thus
predictive power was being estimate a home health market basket driving the labor-related share higher
implemented via the refinements. percentage increase of 3.0 percent for overall. We believe the proposed 77.082
Response: Our proposal to keep the CY 2008. percent to be the most technically
FDL at 0.67 for CY 2007 was based upon accurate measure of labor-related costs.
2. The Rebasing and Revising of the
the most recent data analysis at that We will continue to analyze HH cost
Home Health Market Basket
time, and the unknown effects of the HH report data on a regular basis to ensure
PPS refinements on outlier payments. In the proposed rule, we proposed to it accurately reflects the cost structures
As noted above, further analysis and use rebase and revise the home health facing HH providers serving Medicare
of more recent and updated data has led market basket to ensure it continues to beneficiaries.
us to revise the outlier FDL ratio. adequately reflect the price changes of Comment: Several commenters
In summary, since the publication of efficiently providing home health disagreed with the proposed market
the CY 2008 HH PPS proposed rule, we services. Specifically, we proposed to basket update for home health providers
have updated our analysis file, on update the home health market basket of 2.9 percent for CY 2008, which is
which the Abt model is based, to base year from 2000 to 2003. We also lower than the proposed FY inpatient
include 2005 data. Using the best proposed to revise the home health hospital and skilled nursing facility
(SNF) market basket updates. One is derived from the data reported on the appropriate wage index value to the
commenter noted that the lower market 2003 HHA Medicare cost reports. In proposed labor portion (77.082 percent;
basket update relative to other providers determining the market basket see Table 22 of the proposed rule) of the
will have an adverse impact on the percentage increase, these costs are HH PPS rates based on the geographic
industry’s ability to attract health care proxied using the CPI for private area where the beneficiary received the
workers. transportation. Forecasts of this price home health services. As implemented
Response: The final HH market basket proxy reflect the price changes of fuel, under the HH PPS in the July 3, 2000
update for CY 2008 is 3.0 percent, as well as other transportation costs HH PPS final rule, each HHA’s labor
which is based on Global Insight Inc.’s such as vehicle purchase/lease, market area is based on definitions of
(GII) 2007 2nd quarter forecast, the most maintenance, repair, and insurance. We Metropolitan Statistical Areas (MSAs)
current forecast available at the time of believe this is the most appropriate issued by the OMB. We have
publication of the final rule. The update price proxy to use for transportation as consistently used and proposed again in
in the proposed rule was based on GII’s home health providers face all aspects the CY 2008 HH PPS proposed rule to
2006 3rd quarter forecast. GII is a of vehicle expenses and as such, these use the pre-floor and pre-reclassified
nationally recognized economic and costs are appropriately captured in the hospital wage index data to adjust the
financial forecasting firm that contracts rebased and revised home health market labor portion of the HH PPS rates based
with CMS to forecast the components of basket. on the geographic area where the
the market baskets. CMS calculates each Comment: Several commenters stated beneficiary receives home health
market basket (both weight composition that the present wage structure does not services (72 FR 25448). We believe the
and price proxy selection) specific to provide adequate reimbursement for use of the pre-floor and pre-reclassified
the respective industry and independent increased nursing and therapist wages. hospital wage index data results in the
of the other market baskets. Additionally, one commenter suggested appropriate adjustment to the labor
The HH PPS market basket measures CMS should use data from the Bureau portion of the costs as required by
the change in prices for an exhaustive of Labor Statistics (BLS) for clinician statute.
list of categories that represent the costs. In the August 11, 2004 IPPS final rule
inputs required to provide services to Response: The current price proxy [69 FR 49206], revised labor market area
Medicare beneficiaries. The HH index used for the compensation portion of definitions were adopted at § 412.64(b),
weights are based on data reported on the home health market basket was which were effective October 1, 2004 for
the Medicare cost report forms which designed based on the occupational skill acute care hospitals. The new standards,
provide actual cost share data specific to mix specific to the home health Core Based Statistical Areas (CBSAs),
home health agencies. Likewise, the industry. The proxy accounts for all were announced by OMB in late 2000
hospital and SNF market baskets are related compensation expenditures for and were also discussed in greater detail
based on actual cost shares reported on an exhaustive list of occupations within in the July 14, 2005 HH PPS proposed
their respective cost reports. Each cost the home health industry, including but rule. For the purposes of the HH PPS,
category in all market baskets is not limited to, nurses, therapists, and the term ‘‘MSA-based’’ refers to wage
matched to a price proxy that is clinicians. These three occupations fall index values and designations based on
determined to be the most technically into the cost category for skilled the previous MSA designations.
appropriate price proxy for that nursing, therapists, and other Conversely, the term ‘‘CBSA-based’’
category. For example, the HH wage professional/technical workers, a cost refers to wage index values and
price proxy measures price pressures category accounting for 50.506 percent designations based on the new OMB
specific to the occupational skill mix of the total home health wage proxy (72 revised MSA designations which now
within the HH industry while the SNF FR 25440). These wages are proxied by include CBSAs. In the November 9,
wage price proxy measures price a 50/50 blend of the employment cost 2005 HH PPS final rule (70 FR 68132),
pressures specific to the skilled nursing index (ECI) for professional & technical we implemented a 1-year transition
facility industry. (P&T) workers and the ECI for hospital policy using a 50/50 blend of the CBSA-
We believe that HH compensation workers. Accordingly, we believe that based wage index values and the
costs are accurately captured within the the home health occupational wage and Metropolitan Statistical Area (MSA)-
HH market basket. The associated salary index is the most representative based wage index values for CY 2006.
weight is derived directly from the measure of home health wage pressures. The 1-year transition policy ended in
Medicare cost report data, which We are implementing the revised and CY 2006. Currently, wage index values
indicates that compensation in the HH rebased HH market basket as proposed. for CY 2007 are based on CBSA
industry is higher relative to that of designations. For CY 2008, we will
3. Wage Index
other market industries. We believe this continue to use a wage index based on
reflects the labor-intensive nature of the The statute at sections the CBSA designations.
home health industry. Moreover, the 1895(b)(4)(A)(ii) and 1895(b)(4) of the As implemented under the HH PPS in
indices used to proxy changes in the Act requires the Secretary to establish the July 3, 2000 HH PPS final rule, each
price of labor reflect the occupational wage adjustment factors that reflect the HHA’s labor market is determined based
mix of the laborers in the HH industry relevant level of wages and wage-related on definitions of MSAs issued by OMB.
and are thus also technically costs applicable to the furnishing of In general, an urban area is defined as
appropriate. home health services and to provide an MSA or New England County
Comment: Several commenters stated appropriate adjustment to the episode Metropolitan Area (NECMA) as defined
that HH providers face higher payment amount under the HH PPS to by OMB. Under § 412.64(b)(1)(ii)(C), a
transportation costs than other types of account for area wage differences. rural area is defined as any area outside
providers which should be reflected in Section 1895(b)(4)(C) of the Act further of the urban area. The urban and rural
a higher market basket update. provides that the wage adjustment area geographic classifications are
Response: We believe HH factors may be the factors used by the defined in § 412.64(b)(1)(ii)(A) and
transportation costs are accurately Secretary for purposes of section § 412.64(b)(1)(II)(C) respectively, and
captured within the HH market basket. 1886(d)(3)(E) of the Act for hospital have been used under the HH PPS since
The transportation base year cost weight wage adjustment factors. We apply the implementation.
Under the HH PPS, the wage index rule (71 FR 65905), the following wage index is derived from the pre-
value used is based upon the location of methodology for imputing a rural wage floor, pre-reclassified hospital wage
the beneficiary’s home. As has been our index for areas where no hospital wage index which is calculated based on cost
longstanding practice, any area not data are available as an acceptable report data from hospitals paid under
included in an MSA (urban area) is proxy. The methodology that we the hospital inpatient prospective
considered to be non-urban implemented for CY 2007 imputed an payment system (IPPS). All IPPS
§ 412.64(b)(1)(ii)(C) and receives the average wage index value by averaging hospitals must complete the wage index
statewide rural wage index value (see, the wage index values from contiguous survey (Worksheet S–3, Parts II and III)
for example, 65 FR 41173). CBSAs as a reasonable proxy for rural as part of their Medicare cost reports.
As discussed previously and set forth areas with no hospital wage data from Cost reports will be rejected if
in the July 3, 2000 final rule, the statute which to calculate a wage index. We Worksheet S–3 is not completed. In
provides that the wage adjustment believe this methodology best met our addition, our intermediaries perform
factors may be the factors used by the criteria for imputing a rural wage index desk reviews on all hospitals’
Secretary for purposes of section as well as representing an appropriate Worksheet S–3 wage data, and we run
1886(d)(3)(E) of the Act for hospital wage index proxy for rural areas edits on the wage data to further ensure
wage adjustment factors. As discussed without hospital wage data. the accuracy and validity of the wage
in the July 3, 2000 final rule, we Specifically, such a methodology uses data. Furthermore, HHAs have the
proposed to again use the pre-floor and pre-floor, pre-reclassified hospital wage opportunity to submit comments on the
pre-reclassified hospital wage index data, is easy to evaluate, is updateable hospital wage index data during the
data to adjust the labor portion of the from year to year, and uses the most annual IPPS rulemaking period.
HH PPS rates based on the geographic local data available. In determining an Therefore, we believe our review
area where the beneficiary receives imputed rural wage index, we define processes result in an accurate reflection
home health services. We believe the ‘‘contiguous’’ as sharing a border. For of the applicable wages for the areas
use of the pre-floor and pre-reclassified Massachusetts, rural Massachusetts given.
hospital wage index data results in the currently consists of Dukes and Comment: Several commenters
appropriate adjustment to the labor Nantucket Counties. We determined expressed concerns about using the pre-
portion of the costs as required by that the borders of Dukes and Nantucket floor, pre-reclassified hospital wage
statute. For the CY 2008 update to home counties are ‘‘contiguous’’ with index for the home health wage index.
health payment rates, we would Barnstable and Bristol counties. We These commenters believe that CMS has
continue to use the most recent pre-floor again proposed to apply this the regulatory authority to replace the
and pre-reclassified hospital wage index methodology for imputing a rural wage current wage index with one that
available at the time of publication. index for those rural areas without rural achieves parity with hospitals in order
In adopting the CBSA designations, hospital wage data. to compete in the same geographic labor
we identified some geographic areas However, as we noted in the HH PPS markets. Further, these commenters
where there are no hospitals, and thus final rule for CY 2007, we did not support stabilizing the wage index
no hospital wage data on which to base believe that this policy was appropriate through limits on year-to-year changes.
the calculation of the home health wage for Puerto Rico. As noted in the August Specific recommendations include
index. Beginning in CY 2006, we 3, 2006 proposed rule, there are applying a rural floor in addition to
adopted a policy that, for urban labor sufficient economic differences between allowing HHAs to apply for the type of
markets without an urban hospital from the hospitals in the United States and geographic reclassification that IPPS
which a hospital wage index can be those in Puerto Rico, including the fact hospitals are provided.
derived, all of the urban CBSA-wage that hospitals in Puerto Rico are paid on Response: The commenters are
index values within the State would be blended Federal/Commonwealth- referring to rural floor and geographic
used to calculate a statewide urban specific rates, that a separate, distinct reclassification provisions in the IPPS
average wage index to use as a policy for Puerto Rico is necessary. which are only applicable to hospital
reasonable proxy for these areas. Consequently, any alternative payments. The rural floor provision is
Currently, the only CBSA that would be methodology for imputing a wage index provided at section 4410 of Public Law
affected by this policy is CBSA 25980, for rural Puerto Rico would need to take 105–33 and is specific to hospitals. The
Hinesville, Georgia. We proposed to into account those differences. Our reclassification provision provided at
continue this policy for CY 2008. policy of imputing a rural wage index section 1886(d)(10) of the Act is also
Currently, the only rural areas where by using an averaged wage index of specific to hospitals. Because these
there are no hospitals from which to CBSAs contiguous to that rural area floors and reclassifications apply only to
calculate a hospital wage index are does not recognize the unique hospitals, and not to HHAs, we believe
Massachusetts and Puerto Rico. For CY circumstances of Puerto Rico. For CY the use of the most recent available pre-
2006, we adopted a policy in the HH 2008, we again proposed to continue to floor and pre-reclassified hospital wage
PPS November 9, 2005 final rule (70 FR use the most recent wage index index data results in the most
68138) of using the CY 2005 pre-floor, previously available for Puerto Rico appropriate adjustment to the labor
pre-reclassified hospital wage index which is 0.4047. portion of home health costs as required
value. In the August 3, 2006 proposed Comment: A commenter supported at 1895(b)(4)(C). We also note that the
rule, we again proposed to apply the CY ensuring that the hospital cost reports HH PPS wage adjustment is based on
2005 pre-floor/pre-reclassified hospital that are used to calculate the wage index the geographic area where the
wage index to rural areas where no are accurate. The commenter stated that beneficiary is located, not where the
hospital wage data is available. In CMS should not accept or utilize faulty HHA is located.
response to commenters’ concerns and cost report data. Comment: One commenter
in recognition that, in the future, there Response: We appreciate the recommended that CMS adopt a ‘‘rural
may be additional rural areas impacted comment and note CMS utilizes floor’’ policy for the home health wage
by a lack of hospital wage data from efficient means to ensure and review the index, comparable to the policy that
which to derive a wage index, we accuracy of the cost report data and exists for hospitals. The commenter
adopted, in the November 9, 2006 final resulting wage index. The home health believed that CMS has the authority to
make the change in the regulation. The health specific wage index. Similarly, requirements under Pub. L. 109–432.
commenter expressed that its proposal other commenters recommended that We are reviewing MedPAC’s Report to
would be the simplest, fairest, and most CMS develop a home health specific Congress and the wage index
cost effective solution to the ‘‘wage wage index to reflect the true costs of methodology recommended therein. We
index problems’’ and would serve as an HHAs. will carefully consider MedPAC’s
important bridge to any legislative Response: While we appreciate the recommendations as they apply to the
revision to the wage index provisions, commenters’ desire to use a home HH PPS. Finally, we note that MedPAC
which is likely to take years to enact. health specific wage index, we note that released its June 2007 report to Congress
Response: Sections 1895(b)(4)(A)(ii) our previous attempts at either on June 15, 2007. As the statute
and (b)(4)(C) of the Act require the proposing or developing a home health requires, the report includes MedPAC’s
Secretary to establish area wage specific wage index were not well analysis and recommendations on
adjustment factors that reflect the received by commenters or the industry. alternatives to the method to compute
relative level of wages and wage-related Generally, the volatility of the home the wage index. The full report can be
costs applicable to the furnishing of health wage data and the resources downloaded from MedPAC’s Web Site
home health services and to provide needed to audit and verify that data, at http://www.medpac.gov/documents/
appropriate adjustments to the episode make it difficult to ensure that such a Jun07_EntireReport.pdf.
payment amounts under the HH PPS to wage index accurately reflects the wages Comment: A commenter expressed
account for area wage differences. The and wage-related costs applicable to the concern because the wage index for
wage adjustment factors may be the furnishing of services. Thus, we are not CBSA 25180, Berkeley County, WV is
factors used by the Secretary for adopting a home health specific wage lower than other nearby CBSAs in the
purposes of section 1886(d)(3)(E) of the index at this time. We believe it is Washington, DC area. In addition, the
Act. We believe the use of the hospital important that a home health specific commenter stated that CBSA 25180 is
wage data, without application of a rural wage index be more reflective of the one of the fastest growing areas in the
floor, results in appropriate adjustment wages and salaries paid in a specific nation, thereby increasing property
to the labor portion of costs based on an area, be based upon a stable data source, values and hence labor costs.
appropriate wage index as required and significantly improve our ability to Response: CBSA 25180 ‘‘Hagerstown–
under section 1895(b)(3)(A)(i), determine home health payments Martinsburg, MD–WV’’ includes not
(b)(4)(A)(ii), and (b)(4)(C) of the Act. without being overly burdensome. We only Berkeley County, WV but also
Additionally, as stated above, the rural continue to believe that using the most Morgan County, WV and Washington
floor provision provided at section 4410 recent available pre-floor, pre- County, MD. Prior to our adoption of
of Pub. L. 105–33 is specific to hospital reclassified hospital wage index results OMB’s revised geographic area
payments. in the appropriate adjustment to the designations in CY 2006, Morgan
Comment: Several commenters labor portion of the costs as required by County was classified as rural. Prior to
expressed concern that in FY 2004, we the statute. CY 2006, Berkeley County was grouped
dropped Critical Access Hospitals Comment: Several commenters with 24 other geographic areas (23
(CAHs) from our calculation of the proposed that CMS adopt MedPAC’s counties and the District of Columbia)
hospital wage index. Commenters stated proposed method for calculating the in order to calculate a wage index for
that wage cost data from over 1,000 hospital wage index and apply it to the this area, which was classified as MSA
CAHs are no longer included in the HH PPS. Chapter 6 of MedPAC’s June 8840 ‘‘Washington, DC–MD–VA–WV.’’
calculation of the hospital wage index. 2007 Report to Congress, entitled After adopting OMB’s revised
These hospitals are located in rural ‘‘Promoting Greater Efficiency in geographic area designations, Morgan,
areas and therefore impact the Medicare’’ discusses MedPAC’s Berkeley, and Washington counties’
calculation of the rural wage indexes. proposed methodology. Under hospital wage data are now added
The commenters believed not including MedPAC’s system, HHAs and hospitals together to calculate the wage index for
CAH cost report data in the wage index in the same market would have the CBSA 25180. We were aware that
calculation has had a significant impact same wage index. The new methodology changes to wage index values might
on HHAs that serve beneficiaries in would be available for all labor areas, result from adopting the revised OMB
rural areas. eliminating the need for imputing an designations. Therefore, we provided a
Response: As noted previously, we index for agencies in areas with no one-year transition period in CY 2006 as
adopted the pre-floor, pre-classified hospital wage data. One commenter a means to phase in the changes and to
hospital wage index data as we believe urged CMS to begin implementing mitigate the resulting adverse impact of
they most appropriately reflect the MedPAC’s proposed wage index a CBSA-based wage index on certain
relative level of wages and wage-related methodology for home health in CY HHAs. As to the appropriateness of
costs applicable to the furnishing of 2009. what CBSA a particular area has been
home health services and provide Response: Section 106(b)(1) of the designated into, CBSA designations are
appropriate adjustments to the episode MIEA–TRHCA (Pub. L. 109–432) determined by the Office of
payment amounts under the HH PPS to requires MedPAC to submit to Congress, Management and Budget (OMB). This
account for area wage differences. not later than June 30, 2007, a report on information is available at the following
Therefore, for this final rule, we are the Medicare wage index classification Web site address: http://
adopting the pre-floor, pre-reclassified system applied under the Medicare www.whitehouse.gov/omb/bulletins/
hospital wage index. Comments as to Prospective Payment System. Section b03–04.html. We continue to believe
how the IPPS should construct that 106(b) of MIEA–TRHCA requires the that OMB’s CBSA designations reflect
wage index are beyond the scope of this report to include any alternatives that the most recent available geographic
rule. MedPAC recommends to the method classifications and are a reasonable and
Comment: One commenter stated that used to compute the wage index under appropriate way to define geographic
we should use the HHA wage data that section 1886(d)(3)(E) of the Act. areas for purposes of determining wage
we collected and analyzed to rebase the We thank the commenters for their index values.
labor share of the home health market ideas and suggestions on the wage index Comment: A commenter pointed out
basket in order to develop a home in response to the statutory that the CY 2007 wage index for rural
Massachusetts is listed as 1.0661 in the early investments in data infrastructure public comment on a version of the
proposed rule but that it should be and supporting software that CMS and OASIS that we plan to begin testing in
1.1661. HHAs have made over the past several early 2008 (72 FR 41328).
Response: This was an inadvertent years in order to create this quality Comment: A number of commenters
typographical error in the proposed reporting structure have been successful requested that we eliminate OASIS item
rule. The HH PPS Pricer for CY 2007 in making quality reporting and M0175. Commenters also requested
contains the correct value of 1.1661. measurement an integral component of numerous item-specific revisions to the
Accordingly, payments made to HHAs the HHA industry. For CY 2007, we OASIS.
who serve patients residing in rural specified 10 OASIS quality measures as Response: We are presently unable to
areas of Massachusetts are being paid appropriate for measurements of health accommodate the request to delete
based upon the correct wage index care quality. These measures were to be OASIS item M0175. OASIS item M0175
value of 1.1661. submitted by HHAs to meet their has a critical role in risk adjusting many
For the CY 2008 update to home statutory requirement to submit quality quality measures as it is used to
health payment rates, we are finalizing data for a full increase in their market determine the type of facility the patient
the wage index and associated policies basket percentage increase amount. The was discharged from in the previous 14
in that we will continue to use the most 10 measures are: days before HH admission. However, we
recent pre-floor and pre-reclassified will continue to look for ways to reduce
(1) Improvement in ambulation/
hospital wage index. In addition, we the overall burden to providers and
locomotion
note that we plan to evaluate any determine if this information can be
(2) Improvement in bathing
policies adopted in the FY 2008 IPPS obtained in a more simplified or
(3) Improvement in transferring
final rule that affect the wage index, automated manner as we re-examine the
(4) Improvement in management of oral
including how we treat certain New OASIS instrument.
England hospitals under § 601(g) of the medications The remainder of the item-specific
Social Security Amendments of 1983 (5) Improvement in pain interfering comments received relate to data items
(Pub. L. 98–21). We continue to believe with activity that will be addressed in an upcoming
that the use of the pre-floor and pre- (6) Acute care hospitalization notice concerning revisions of the
reclassified hospital wage index data for (7) Emergent care OASIS mentioned above. These
HH PPS results in the appropriate (8) Improvement in dyspnea revisions are currently planned for an
adjustment to the labor portion of the (9) Improvement in urinary OASIS update in calendar year 2009.
costs as required by statute. incontinence These changes are responsive to the
(10) Discharge to community comments we have received, and reflect
4. Home Health Care Quality For CY 2007, we specified 10 OASIS months of development and analysis, as
Improvement quality measures as appropriate for well as industry input and concerns.
Section 5201(c)(2) of the DRA added measurements of health care quality. On July 27, 2007, a notice was
section 1895(b)(3)(B)(v)(II) to the Act, These measures were to be submitted by published in the Federal Register
requiring that ‘‘each home health agency HHAs to meet their statutory (CMS–10238) which seeks public
shall submit to the Secretary such data requirement to submit quality data for a comment on a version of the OASIS that
that the Secretary determines are full increase in their market basket we plan to begin testing in early 2008.
appropriate for the measurement of percentage increase amount. For CY Based on the finding from the testing,
health care quality. Such data shall be 2008, we proposed to expand the we may pursue adopting the
submitted in a form and manner, and at existing set of 10 quality measures by commenter’s suggested changes in
a time, specified by the Secretary for adding up to 2 NQF-endorsed measures. future payment rule notices.
purposes of this clause.’’ In addition, The proposed additional measures for Comment: Some commenters were
section 1895(b)(3)(B)(v)(I) of the Act, as 2008 were: concerned about the proposed quality
also added by section 5201(c)(2) of the • Emergent Care for Wound measure regarding emergent care for
DRA, dictates that ‘‘for 2007 and each Infections, Deteriorating Wound Status wound infections.
subsequent year, in the case of a home • Improvement in the Status of Response: We note that the title and
health agency that does not submit data Surgical Wounds (For a complete list description of the quality measure do
to the Secretary in accordance with and description of the quality measure not fully reflect the breadth of the issue
subclause (II) with respect to such a requirements see the proposed rule (72 being measured. Specifically, the
year, the home health market basket FR 25449–25452)). quality measure entitled ‘‘Emergent Care
percentage increase applicable under Comment: Several commenters for Wound Infections, Deteriorating
such clause for such year shall be suggested that CMS continue to refine Wound Status’’ is calculated using a
reduced by 2 percentage points.’’ and enhance the OASIS assessment data item that includes new pressure
The OASIS data currently provide instrument and associated Quality ulcers and lesions, and therefore the
consumers and HHAs with 10 publicly- Measures, and suggested item-specific title of the measure may cause some
reported home health quality measures or quality measure-specific items in use confusion. Nonetheless, we feel that the
which have been endorsed by the in the home health quality reporting quality measure is an important
National Quality Forum (NQF). requirement. indicator and we intend to conform the
Reporting these quality data has also Response: CMS is constantly working title of the measure to more accurately
required the development of several to improve the OASIS instrument and reflect the concepts being measured.
supporting mechanisms such as the the quality measures that are built upon Comment: Several commenters
HAVEN software used to encode and it. We will continue to pursue suggested that we delete two quality
transmit data using a CMS standard improving the assessment instrument’s items to compensate for the two new
electronic record layout, edit accuracy in reflecting both the health quality items added. Some also
specifications, and data dictionary. The status and improvements in condition of suggested that we reduce the total
HAVEN software includes the required our beneficiaries. On July 27, 2007, a number of OASIS items. Another
OASIS data set that has become a notice was published in the Federal suggested we develop quality measures
standard part of HHA operations. These Register (CMS–10238) which seeks for fall prevention.
Response: CMS is not adding new intent of M0830, Emergent Care for Medicare, Medicare Advantage, and
OASIS quality items to be reported in Wound Infections, before publicly Medicaid. This commenter suggested
this rule. CMS is adding two quality reporting data. If the focus is only on we use the information to monitor
measures to expand the number of infections or deteriorating status, then outcomes from Medicare Advantage
measures currently being reported for the commenter suggested we revise the plans compared to traditional Medicare,
quality reporting purposes by using wording of the data element. or require Medicare Advantage plans to
existing OASIS data. The data elements Response: This measure addresses pay agencies according to the HH PPS
used to calculate these measures are high-risk, high-volume, high-cost rule, thereby putting the physician and
already captured by the OASIS conditions. These conditions are agency back in control of managing the
instrument and do not require identifiable, preventable and serious in patient. This commenter also suggested
additional reporting or burden to HHAs. their consequences and they can cause removing ‘‘private duty’’ Medicaid
We believe that through this expansion serious harm to beneficiaries. Public patients, such as ventilator dependent
of measures for the HH PPS quality reporting of the measure will continue patients, from the CMS Compare data.
reporting segment, we are providing the to enable providers to investigate and Response: We appreciate the
public with a wider array of comparable take corrective actions to improve safety comment and we will consider this with
and consensus-based (endorsed by the and quality of care delivered. In regard to future changes to the Home
National Quality Forum in 2005) addition, it is responsive to the NQF Health Compare site. However, it is
information on health care quality. proposed priority for measures beyond the scope of this rule to address
CMS will continue to review the associated with the frail elderly specific issues concerning Home Health
OASIS items collected for the purposes population. CMS continues to believe Compare.
of quality to determine if any changes, that the additional measures selected for Comment: Numerous commenters
additions, or deletions are appropriate, the reporting of quality are appropriate. wrote that many of the Medicaid waiver
and the public will have the On July 27, 2007, a notice was programs authorize ‘‘skilled nursing
opportunity to comment on proposed published in the Federal Register (72 services’’ based on their payment
changes to the OASIS items. FR 41328) which seeks public comment terminology, when in reality, the
CMS agrees with the commenter that on a version of the OASIS that we plan services are not ‘‘skilled’’ by Medicare’s
the domain of falls prevention is a to begin testing in early 2008. This new definition. Clients on waiver programs
critical aspect of health care quality. On version of the OASIS addresses many of tend to be chronically ill and show no
July 27, 2007, a notice was published in the item-specific and quality measure improvement in outcomes, but rather
the Federal Register (CMS–10238) specific comments that we have show stabilization in their condition.
which seeks public comment on a received, including those of the Under current regulations, these waiver
version of the OASIS that we plan to commenters. A critical element of this clients are required to have OASIS
begin testing in early 2008. This version testing will be the gathering of data collection performed. With the
of OASIS incorporates several process necessary to make a more accurate inclusion of these waiver clients, the
measures, one of which is geared estimate of the provider burden that the data skews provider outcomes as well as
specifically toward fall prevention OASIS and the anticipated revisions aggregate state outcomes. The
outcome measures in future updates of would require. commenters suggested eliminating the
the OASIS instrument for the purpose of Comment: Numerous commenters requirement to complete OASIS
pay for reporting. noted that data submitted for Home assessments on non-Medicare clients.
Comment: A commenter was in favor Health Compare reporting include both OASIS should be for traditional
of adding Improvement of Status of Medicare and Medicaid patients. They Medicare only.
Surgical Wound to the home health noted that inclusion of Medicaid data Response: The request to change the
compare quality measures, but he felt can skew the data as Medicaid and regulation in § 484.55 concerning
adding an adverse event (Emergent Care Medicare admission criteria are not the OASIS collection requirements is
for Wound Status) was not appropriate. same. One commenter stated that many beyond the scope of this rule and will
Outcome Based Quality Management Medicaid patients are seen in lieu of not be addressed here.
(OBQM) instructs the agency to audit more costly nursing home placement; Comment: One commenter wrote that
the record to determine if an adverse therefore at discharge, their outcomes in New York, there is a 1915 waiver
event occurred. With the definition of (especially those related to activities of program called the Long Term Home
emergent care being an unplanned daily living) have deteriorated. Health Care Program (LTHHCP), which
physician visit within 24 hours, this Several commenters felt that HHAs provides an intensive array of Medicaid
reporting could be detrimental. In the with high Medicaid caseloads will most home and community-based services to
commenter’s area there is physician likely be damaged in the public nursing home eligible patients. The
office availability that encourages reporting process because these patients majority of patients in LTHHCP are
appointments to be made within 24 are less likely to show marked dually eligible, but Medicaid is the
hours. It is seen as good practice rather improvement due to their chronic appropriate payer of services
than an adverse event. The commenter conditions. The public reporting does approximately 90 percent of the time.
recommended removing ‘‘Emergent Care not give an accurate picture of the Patients must also meet the
for Wound Infections, Deteriorating agency’s performance or outcomes. requirements of a mandatory state
Wound Status’’ from the home health When pay for performance begins, this assessment every 120 days, which is
quality measures. Another commenter negative impact could create issues of separate from the federal OASIS
suggested we revise the instructions so access to care for Medicaid patients. requirements. The commenter is
only visits to an emergency room or These commenters suggested only concerned that CMS does not
outpatient emergency clinic constitute including Medicare patients in the differentiate between LTHHP and
emergent care. Two commenters noted publicly reported data and Home Health traditional Medicare providers regarding
that it is not appropriate to present Compare. submitted OASIS data. The commenter
outcomes that are not risk adjusted or Another commenter suggested that we urges CMS to exclude LTHHCPs and
Adverse Event Outcomes. One stratify CMS Compare information into any Special Needs Certified Home
commenter asked that we clarify the at least three categories: traditional Health Agencies from the OASIS
Quality Reporting and Pay for Reporting OASIS and the anticipated revisions We received no comments on this
Initiative. would require. proposal. Accordingly, we are adopting,
Response: For the purposes of the We are adopting, as final, the two as final, that those agencies do not need
Home Health quality reporting quality measures and note that a total of to submit quality measures for reporting
requirements, HHAs are required to 12 quality measures are necessary to purposes for those patients who are
submit quality measures to CMS meet the statutory submission of quality excluded from OASIS submission as a
through the OASIS instrument. CMS has data to maintain the full home health CoP.
also specified the circumstances under market basket percentage increase. We also proposed that agencies newly
which home health agencies would be Additionally, section certified (on or after May 31, 2007 for
excluded from the HH PPS quality 1895(b)(3)(B)(v)(II) of the Act provides payments to be made in CY 2008) be
reporting requirement (72 FR 25449). the Secretary with the discretion to excluded from the quality reporting
The existing LTHHCP does not fall submit the required data in a form, requirement as data submission and
under any of those exclusions. manner, and time specified by him/her. analysis will not be possible for an
Comment: A commenter is concerned We proposed, for CY 2008, to consider agency certified this late in the reporting
that the OASIS was designed to measure OASIS data submitted by HHAs to CMS time period. In future years, agencies
outcomes by asking nurses to assess the for episodes beginning on or after July that certify on or after May 31 of the
ability of the patient to perform a task, 1, 2006 and before July 1, 2007 as preceding year involved would be
rather than by using performance based meeting the reporting requirement for excluded from any payment penalty for
measures. The commenter gave the calendar year 2008. This reporting time quality reporting purposes for the
example of activities of daily living period will allow 12 full months of data following CY. We note, these exclusions
(ADL) measures. and will provide CMS the time only affect quality reporting
Response: The instrument was requirements and do not affect the
necessary to analyze and make any
designed to collect the information agency’s OASIS reporting
necessary payment adjustments to the
needed to measure changes in health responsibilities under the CoP (72 FR
CY 2008 payment rates. HHAs that meet
status over several designated time 25449). We received no comments on
the reporting requirement shall be
points. The OASIS data set was this proposal, and are adopting it as
eligible for the full home health market
designed for the purpose of enabling final.
basket percentage increase. We received
rigorous and systematic measurement of We note that all HHAs, unless
no comments and are adopting this
patient home health outcomes. We covered by these specific exclusions,
proposal as final.
believe that the quality measures must meet the reporting requirement, or
selected from the OASIS accurately As noted in the proposed rule (72 FR
be subject to a 2 percent reduction in
reflect measures of quality, and that 25449), the home health CoPs (part 484)
the home health market basket
those measures meet the statutory that require OASIS submission also percentage increase in accordance with
requirement to report quality data. provide for exclusions from this section 1895(b)(3)(B)(v)(I) of the Act.
Comment: A commenter wrote that requirement. Generally, agencies Section 1895(b)(3)(B)(v)(III) of the Act
pay for performance would have a excluded from the OASIS submission further requires that the ‘‘Secretary shall
negative effect on whether high acuity requirement do not receive Medicare establish procedures for making data
patients would be able to find agencies payments as they either do not provide submitted under subclause (II) available
willing to help them. services to Medicare beneficiaries or the to the public.’’ Additionally, the statute
Response: Currently, CMS only patients are not receiving Medicare- requires that ‘‘such procedures shall
requires reporting of the specified covered home health services. Under ensure that a home health agency has
quality measures for the HH PPS quality the CoP, agencies are excluded from the the opportunity to review the data that
report for reporting. At this time, there OASIS reporting requirement on is to be made public with respect to the
is no ‘‘Pay for Performance’’ individual patients if: agency prior to such data being made
requirement in HH PPS. However, we • Those patients are receiving only public.’’ To meet the requirement for
believe the current reporting non-skilled services, making such data public, we proposed,
requirements and any future work on • Neither Medicare nor Medicaid is to continue for CY 2008 to use the Home
‘‘Pay for Performance’’ initiatives will paying for home health care (patients Health Compare Web site whereby
help ensure that Medicare beneficiaries receiving care under a Medicare or HHAs are listed geographically.
continue to have access to the highest Medicaid Managed Care Plan are not Currently the 10 quality measures are
quality care possible. excluded from the OASIS reporting posted on the Home Health Compare
Comment: A few commenters were requirement), Web site, and this site would be
concerned that the estimates of burden • Those patients are receiving pre-or updated to reflect the performance level
on reporting the reporting burden have post-partum services, and of the proposed 2 additional quality
been underestimated. • Those patients are under the age of measures. Consumers can search for all
Response: We believe our 18 years. Medicare-approved home health
determination of the collection burden We believe that the rationale behind providers that serve their city or zip
is based upon our best estimates given the exclusion of these agencies from code and then find the agencies offering
the information and data available to us submission of OASIS on patients which the types of services they need as well
at this time. CMS published a notice in are excluded from OASIS submission as as the proposed quality measures. See
the Federal Register that begins the a CoP is equally applicable to HHAs for http://www.medicare.gov/HHCompare/
process of testing a new version of the quality purposes. Therefore, we again Home.asp. HHAs currently have access
OASIS instrument which addresses proposed for CY 2008 that if an agency (through the Home Health Compare
many of the item-specific and quality is not submitting OASIS for patients contractor) to their own agency’s quality
measure specific comments that we excluded from OASIS submission for data (updated periodically), thus
have received. A critical element of this purposes of a CoP, that the submission enabling each agency to know how it is
testing will be the gathering of data of OASIS for quality measures for performing before public posting of data
necessary to make a more accurate Medicare purposes is likewise not on Home Health Compare (72 FR
estimate of the provider burden that the necessary. 25452). We received no comments on
the proposed process and are adopting site of service for the beneficiary as set update for those HHAs that submit
it in the final rule with comment period forth in § 484.230. We adjust the labor quality data as required by the
for CY 2008. portion of the updated national per-visit Secretary. The applicable home health
amounts by discipline used to calculate market basket update will be reduced by
5. CY 2008 Payment Updates
the LUPA by the most recent pre-floor 2 percentage points for those HHAs that
The Medicare HH PPS has been and pre-reclassified hospital wage fail to submit the required quality data.
effective since October 1, 2000. As set index, as discussed in section III.E.3. of
forth in the final rule published July 3, • CY 2008 Adjustments.
this rule.
2000 in the Federal Register (65 FR Medicare pays the 60-day case-mix In calculating the annual update for
41128), the unit of payment under the and wage-adjusted episode payment on the CY 2008 national standardized 60-
Medicare HH PPS is a national a split percentage payment approach. day episode payment rates, we first look
standardized 60-day episode payment The split percentage payment approach at the CY 2007 rates as a starting point.
rate. As set forth in § 484.220, we adjust includes an initial percentage payment The CY 2007 national standardized 60-
the national standardized 60-day and a final percentage payment as set day episode payment rate is $2,339.00.
episode payment rate by a case-mix forth in § 484.205(b)(1) and (b)(2). We In order to calculate the CY 2008
grouping and a wage index value based may base the initial percentage payment national standardized 60-day episode
on the site of service for the beneficiary. on the submission of a request for payment rate, we first increase the CY
The CY 2008 HH PPS rates use the case- anticipated payment and the final 2007 national standardized 60-day
mix methodology discussed in the percentage payment on the submission episode payment rate ($2,339.00) by the
proposed rule (72 FR 25395), of the claim for the episode, as rebased and revised home health market
incorporating the changes discussed in discussed in § 409.43. The claim for the basket update of 3.0 percent for CY
III.B of this rule and application of the episode that the HHA submits for the 2008.
wage index adjustment to the labor final percentage payment determines
portion of the HH PPS rates as set forth Given this updated rate, we would
the total payment amount for the then take a reduction of 2.75 percent to
in the July 3, 2000 final rule. As stated episode and whether we make an
in section III.E.2. of this rule, we are account for change in case-mix not
applicable adjustment to the 60-day related to actual change in case-mix. We
rebasing and revising the home health case-mix and wage-adjusted episode
market basket, resulting in a revised and would multiply the resulting value by
payment. The end date of the 60-day 1.05 and 0.95 to account for the
rebased labor related share of 77.082 episode as reported on the claim
percent and a non-labor portion of estimated percentage of outlier
determines which CY rates Medicare payments for CY 2008 (that is, $2,339.00
22.918 percent. We multiply the will use to pay the claim.
national standardized 60-day episode * 1.030 * 0.9725 * 1.05 * 0.95), to yield
We may also adjust the 60-day case-
payment rate by the patient’s applicable a CY 2008 national standardized 60-day
mix and wage-adjusted episode
case-mix weight. We divide the case- episode payment rate of $2,337.06 for
payment based on the information
mix adjusted amount into a labor and episodes that begin in CY 2007 and end
submitted on the claim to reflect the
non-labor portion. We multiply the in CY 2008 (see Table 11A below). For
following:
labor portion by the applicable wage • A LUPA provided on a per-visit episodes that begin in CY 2007 and end
index based on the site of service of the basis as set forth in § 484.205(c) and in CY 2008, the new 153 HHRG case-
beneficiary. For CY 2008, we are basing § 484.230. mix model (and associated Grouper)
the wage index adjustment to the labor • A PEP adjustment as set forth in would not yet be in effect. For that
portion of the HH PPS rates on the most § 484.205(d) and § 484.235. reason, episodes that begin in CY 2007
recent pre-floor and pre-reclassified • An outlier payment as set forth in and end in CY 2008 will be paid at the
hospital wage index as discussed in § 484.205(f) and § 484.240. rate of $2,337.06, and be further
section III.E.3. of this rule (not including As discussed in section III.C.3 of this adjusted for wage differences and for
any reclassifications under section final rule with comment period, we are case-mix, based on the current 80 HHRG
1886(d)(8)(B) of the Act). implementing the removal of the SCIC case-mix model. We recognize that the
As discussed in the July 3, 2000 HH adjustment from the HH PPS. annual update for CY 2008 is for all
PPS final rule, for episodes with four or This rule reflects the updated CY episodes that end on or after January 1,
fewer visits, Medicare pays the national 2008 rates that will become effective 2008 and before January 1, 2009. By
per-visit amount by discipline, referred January 1, 2008. paying this rate ($2,337.06) for episodes
to as a LUPA. We update the national Section 1895(b)(3)(B) of the Act, as that begin in CY 2007 and end in CY
per-visit amounts by discipline annually amended by section 5201 of the DRA, 2008, we will have appropriately
by the applicable home health market requires for CY 2008 that the standard recognized that these episodes are
basket percentage. We adjust the prospective payment amounts be entitled to receive the CY 2008 home
national per-visit amount by the increased by a factor equal to the health market, even though the new
appropriate wage index based on the applicable home health market basket case-mix model will not yet be in effect.
TABLE 11A.—NATIONAL 60-DAY EPISODE AMOUNTS UPDATED BY THE HOME HEALTH MARKET BASKET UPDATE FOR CY
2008, BEFORE CASE-MIX ADJUSTMENT, WAGE INDEX ADJUSTMENT BASED ON THE SITE OF SERVICE FOR THE BENE-
FICIARY OR APPLICABLE PAYMENT ADJUSTMENT FOR EPISODES BEGINNING IN CY 2007 AND ENDING IN CY 2008
National standard-
ized 60-day epi-
Multiply by the Reduce by 2.75 Adjusted to ac- sode payment rate
Total CY 2007 national standardized 60-day episode home health mar- percent for nominal count for the 5 per- for episodes begin-
payment rate ket basket update change in case-mix cent outlier policy ning in CY 2007
(3.0 percent) 1 and ending in CY
2008
$2,339.00 ........................................................................... × 1.030 .................. × 0.9725 ................ × 1.05 × 0.95 $2,337.06

1 The estimated home health market basket update of 3.0 percent for CY 2008 is based on Global Insight, Inc, 2nd Qtr, 2007 forecast with his-
torical data through 1st Qtr, 2007.
Next, in order to establish new rates in $2,529.63. Next, we need to reduce rate, for episodes beginning and ending
based on a new case-mix system, we this amount to pay for each of our final in CY 2008, of $2,270.32 (see Table
again start with the CY 2007 national policies. As noted previously, based 11B). These episodes would be further
standardized 60-day episode payment upon our change to the LUPA payment, adjusted for case-mix based on the 153
rate and increase that rate by the the NRS redistribution, and the HHRG case-mix model for episodes
rebased and revised home health market elimination of the SCIC policy, the beginning and ending in CY 2008. As
basket update (3.0 percent) ($2,339.00 * amounts needed to account for outlier we noted in section II.A.2.d. of the
1.030 = $2,409.17). We next have to put payments, and the reduction to account proposed rule, we increased the case-
dollars associated with the outlier for the 2.75 percent case-mix change mix weights by a budget neutrality
targeted estimates back into the base adjustment, we reduce the national
factor of 1.194227193. In this final rule,
rate. In the 2000 HH PPS final rule (65 standardized 60-day episode payment
the case-mix weights were increased by
FR 41184), we divided the base rate by rate by $5.70, $45.87, $10.96, $127.22,
a budget neutrality factor of
1.05 to account for the outlier target and $69.56, respectively. This results in
policy. Therefore, we proposed to a CY 2008 updated national 1.238848031.
multiply the $2,409.17 by 1.05, resulting standardized 60-day episode payment
TABLE 11B.—NATIONAL 60-DAY EPISODE AMOUNTS UPDATED BY THE HOME HEALTH MARKET BASKET UPDATE
FOR CY 2008, BEFORE CASE-MIX ADJUSTMENT, WAGE INDEX ADJUSTMENT BASED ON THE SITE OF SERVICE
FOR THE BENEFICIARY OR APPLICABLE PAYMENT ADJUSTMENT FOR EPISODES BEGINNING AND ENDING IN CY
2008
Changes to account
for LUPA adjustment
($5.70), NRS pay-
ment ($45.87), elimi- CY 2008 national
Total CY 2007 Adjusted to return the nation of SCIC policy
Multiply by the home Updated and outlier standardized 60-day
national stand- outlier funds to the ($10.96), outlier pol-
health market basket adjusted national episode payment rate
ardized 60-day national standardized icy ($127.22), and
update (3.00 per- standardized 60-day for episodes begin-
episode pay- 60-day episode pay- 2.75 percent reduc-
cent) 1 episode payment ning and ending in
ment rate ment rate tion for nominal CY 2008
change in case-mix
(69.56) for episodes
beginning and ending
in CY 2008
$2,339.00 ......... X 1.030 ..................... X 1.05 ....................... $2,529.63 .................. ¥$259.31 ................. $2,270.32
1 The estimated home health market basket update of 3.0 percent for CY 2008 is based on Global Insight, Inc, 2nd Qtr, 2007 forecast
with historical data through 1st Qtr, 2007.
Under the HH PPS, NRS payment, non-adjusted national standardized 60- adjusted labor portion to get the case-
which was $49.62 at the onset of the HH day episode payment rate and multiply mix and wage adjusted national 60-day
PPS, has been updated yearly as part of it by the appropriate case-mix weight episode payment without NRS.
the national standardized 60-day from Table 5 of this rule. Next, multiply To calculate the NRS amount,
episode payment rate. As discussed the case-mix adjusted national multiply the episode’s NRS weight
previously in section III.C.4., we are standardized 60-day episode payment (taken from Table 9 of this rule) by the
removing the current NRS payment by the labor portion (77.082 percent); NRS conversion factor ($52.35). This
amount portion from the national multiply this result by the appropriate adjusted NRS payment is added to the
standardized 60-day episode payment wage index factor listed in Addendum case-mix and wage-adjusted national
rate and adding a severity-adjusted NRS A or B to wage-adjust the 60-day standardized 60-day episode payment.
payment amount subject to case-mix episode payment. Next multiply the The resulting amount is the case-mix
and wage adjustment to the national case-mix adjusted national standardized and wage-adjusted national
standardized 60-day episode payment 60-day episode payment by 22.918 standardized 60-day episode payment
rate. To calculate an episode’s percent to compute the non-labor rate including NRS for that particular
prospective payment amount, take the portion. Add this result to the wage- episode.
The following example illustrates the $2,270.32 (see Table 11B). The HHA under NRS severity level #4. The NRS
computation described above: determines that the beneficiary is in his or Severity Level #4 weight = 3.9686 and the
Example 1. An HHA is providing services her 3rd episode and thus falls under the NRS Conversion Factor = $52.35 (see Table
to a Medicare beneficiary in Grand Forks, C1F3S3 HHRG for 3rd+ episodes with 0 to 13 9).
ND; the episode begins and ends in 2008. therapy visits (Case-Mix Weight = 1.4674). It BILLING CODE 4120–01–P
The national standardized payment rate is is also determined that the beneficiary falls
BILLING CODE 4120–01–C

ER29AU07.051</GPH>
National Per-Visit Amounts Used To $87.93 for initial and only episode 2008 by the rebased and revised home
Pay LUPAs and Compute Imputed Costs LUPAs during CY 2008. In calculating health market basket update (3.0
Used in Outlier Calculations the CY 2008 national per-visit amounts percent), then multiply by 1.05 and 0.95
As discussed previously in the CY used to calculate payments for LUPA to account for the estimated percentage
2008 HH PPS proposed rule, the episodes and to compute the imputed of outlier payments (see Table 12
policies governing LUPAs and the costs in outlier calculations, we start below). LUPA rates are not being
outlier calculations set forth in the July with the CY 2007 per-visit amounts. We reduced due to the increase in case-mix
3, 2000 HH PPS final rule will continue increase the CY 2007 per-visit amounts since they are per-visit rates and hence
(65 FR 41128) with an increase of for each home health discipline for CY are not subject to changes in case-mix.
TABLE 12.—NATIONAL PER-VISIT AMOUNTS FOR LUPAS (NOT INCLUDING THE INCREASE IN PAYMENT FOR A BENE-
FICIARY’S ONLY EPISODE OR THE INITIAL EPISODE IN A SEQUENCE OF ADJACENT EPISODES) AND OUTLIER CALCULA-
TIONS UPDATED BY THE HOME HEALTH MARKET BASKET UPDATE FOR CY 2008, BEFORE WAGE INDEX ADJUSTMENT
BASED ON THE SITE OF SERVICE FOR THE BENEFICIARY
Final CY 2007 Multiply by
per-visit the home Adjusted to CY 2008 per-
amounts per health mar- account for visit payment
Home health discipline type 60-day epi- ket basket the 5 percent amount per
sode for (3.0 per- outlier policy discipline
LUPAs cent) 1
Home Health Aide ................................................................................................ $46.24 × 1.030 ........ × 1.05 .......... $47.51
× 0.95
Medical Social Services ........................................................................................ 163.68 × 1.030 ........ × 1.05 .......... 168.17
× 0.95
Occupational Therapy ........................................................................................... 112.40 × 1.030 ........ × 1.05 .......... 115.48
× 0.95
Physical Therapy .................................................................................................. 111.65 × 1.030 ........ × 1.05 .......... 114.71
× 0.95
Skilled Nursing ...................................................................................................... 102.11 × 1.030 ........ × 1.05 .......... 104.91
× 0.95
Speech-Language Pathology ............................................................................... 121.22 × 1.030 ........ × 1.05 .......... 124.54
× 0.95
torical data through 2nd Qtr, 2007.
Payment for LUPA episodes is 2008 HH PPS proposed rule and final that occur as the only episode or initial
changed in that for LUPAs that occur as amount in section III.C.2. of this rule) is episode in a sequence of adjacent
initial episodes in a sequence of to be added to the LUPA payment. Table episodes are adjusted by adding $87.93
adjacent episodes or as the only 12 rates below are before that to the LUPA payment before adjusting
episode, a revised payment amount (see adjustment and are the rates paid to all for wage index.
our proposal in section II.A.5. of the CY other LUPA episodes. LUPA episodes
ER29au07.052</GPH>
BILLING CODE 4120–01–C added to the sum of the wage and case-mix from Table 9, the NRS payment amount =
Outlier payments are determined and adjusted 60-day episode amount. The steps to $551.00
calculated using the same methodology calculate the total episode payment, from Table 5, the case-mix weight = 1.9413
that has been used since the including an outlier payment, are given from Addendum B, the wage index = 0.9860
implementation of the HH PPS. below.
For this example, assume that a beneficiary 1. Calculate case-mix and wage-adjusted
Example 3 details the calculation of an 60-day episode payment, including NRS.
outlier payment. lives in Greenville, SC and that the episode
in question began and ended in CY 2008. The National standardized 60-day episode
Example 3. Calculation of an Outlier episode has a case-mix severity = C3F3S5, payment amount for episodes beginning and
Payment and is a second episode with 63 visits (30 ending in CY 2008:
The outlier payment amount is the product skilled nursing, 20 home health aide visits, = $2,270.32
of the imputed amount in excess of the and 13 physical therapy visits). The
Calculate the case-mix adjusted episode
outlier threshold absorbed by the HHA and beneficiary had 105 NRS points, for an NRS
payment:
ER29au07.053</GPH>
the loss sharing ratio. The outlier payment is severity level = 6. Therefore,
Multiply the national standardized 60-day Calculate the total wage-adjusted fixed portion for the total wage-adjusted imputed
episode payment by the applicable case-mix dollar loss amount including NRS by adding costs for home health aide visits:
weight: the wage-adjusted fixed dollar loss amount of $1,133.38 + $341.76 = $1,475.14
$2,270.32 × 1.9413 = $4,407.37 NRS to the wage-adjusted fixed dollar loss
Total wage adjusted imputed per-visit costs
Divide the case-mix adjusted episode amount without NRS:
for skilled nursing, home health aide, and
payment into the labor and non-labor $485.10 + $1,998.78 = $2,483.88 physical therapy visits during the 60-day
portions: Add the case-mix and wage-adjusted 60- episode:
Labor portion: 0.77082 × $4,407.37 = day episode amount including NRS and the $3,113.34 + $939.95 + $1,475.14 = $5,528.43
$3,397.29 wage-adjusted fixed dollar loss amount
including NRS to get the wage-adjusted 4. Calculate the amount absorbed by the
Non-labor portion: 0.22918 × $4,407.37 = HHA in excess of the outlier threshold.
$1,010.08 outlier threshold:
Subtract the outlier threshold from (2) from
Wage-adjust the labor portion by $4,910.81 + $2,483.88 = $7,394.69
the total wage-adjusted imputed per-visit
multiplying it by the wage index factor for 3. Calculate the wage-adjusted imputed costs for the episode from (3).
Greenville, SC: cost of the episode.
$5,528.43 ¥ $4,910.81 = $617.62
0.9860 × $3,397.29 = $3,349.73 Multiply the total number of visits by the
national average per-visit amounts listed in 5. Calculate the outlier payment and total
Add wage-adjusted labor portion to the episode payment.
non-labor portion to calculate the total case- Table 12:
30 skilled nursing visits × $104.91 = Multiply the imputed amount in excess of
mix and wage-adjusted 60-day episode the outlier threshold absorbed by the HHA
payment before NRS added: $3,147.30
20 home health aide visits × $47.51 = $950.20 from (4) by the loss sharing ratio of 0.80:
$3,349.73 + $1,010.08 = $4,359.81 $617.62 × 0.80 = $494.10 = outlier payment
13 physical therapy visits × $114.71 =
Add NRS amount to get the total case-mix $1,491.23 Add the outlier payment to the case-mix
and wage-adjusted 60-day episode payment, and wage-adjusted 60-day episode payment,
Calculate the wage-adjusted labor and non-
including NRS: including NRS, calculated in (1):
labor portions for the imputed skilled
$551.00 + $4,359.81 = $4,910.81 nursing visit costs: $494.10 + $4,910.81 = $5,404.91
2. Calculate wage-adjusted outlier Labor portion: 0.77082 × $3,147.30 = $5,404.91 equals the total payment for the
threshold. $2,426.00 episode, including the outlier payment.
Fixed dollar loss amount = national Non-labor portion: 0.22918 × $3,147.30 =
standardized 60-day episode payment $721.30 For episodes that begin in CY 2007
multiplied by 0.89 FDL: Adjust the labor portion of the skilled and end in CY 2008, the new 153 HHRG
$2,270.32 × 0.89 = $2,020.58 nursing visits by the wage index: case-mix model (and associated
Divide fixed dollar loss amount into labor 0.9860 × $2,426.00 = $2,392.04 Grouper) would not yet be in effect. For
and non-labor portions: Add the wage-adjusted labor portion of the that reason, for HHAs that do not submit
Labor portion: 0.77082 × $2,020.58 = skilled nursing visits to the non-labor portion required quality data (for episodes that
$1,557.50 for the total wage-adjusted imputed costs for begin in CY 2007 and end in CY 2008),
Non-labor portion: 0.22918 × $2,020.58 = skilled nursing visits: HH PPS rates are calculated as follows
$463.08 $2,392.04 + $721.30 = $3,113.34 (see section III.E.4., of this rule, for an
Wage-adjust the labor portion by Calculate the wage-adjusted labor and non- explanation of the DRA requirement for
multiplying the labor portion of the fixed labor portions for the imputed home health
dollar loss amount by the wage index:
submission of quality data and the
aide visits: minus 2 percentage points for failure to
$1,557.50 × 0.9860 = $1,535.70 Labor portion: 0.77082 × $950.20 = $732.43 submit that quality data): First, we
Calculate the wage-adjusted fixed dollar Non-labor portion: 0.22918 × $950.20 = update the CY 2007 rate of $2,339.00 by
loss amount without NRS by adding the $217.77
wage-adjusted portion of the fixed dollar loss
the home health market basket
Adjust the labor portion of the home health percentage update (3.0 percent) minus 2
amount to the non-labor portion of the fixed aide visits by the wage index:
dollar loss amount: percent, reduced by 2.75 percent to
0.9860 × $732.43 = $722.18 account for the case-mix change
$1,535.70 + $463.08 = $1,998.78
Add the wage-adjusted labor portion of the adjustment, and multiplied by 1.05 and
Calculate the fixed dollar loss amount of home health aide visits to the non-labor
NRS by multiplying the NRS payment 0.95 to account for the estimated
portion for the total wage-adjusted imputed
amount by the FDL ratio: costs for home health aide visits:
percentage of outlier payments
$551.00 × 0.89 = $490.39 ($2,339.00 * 1.010 * 0.9725 * 1.05 *
$722.18 + $217.77 = $939.95
Divide NRS fixed dollar loss amount into
0.95), to yield an updated CY 2008
Calculate the wage-adjusted labor and non- national standardized 60-day episode
labor and non-labor portions: labor portions for the imputed physical
Labor portion: 0.77082 × $490.39 = $378.00 therapy visits:
payment rate of $2,291.68 for episodes
Non-labor portion: 0.22918 × $490.39 = that begin in CY 2007 and end in CY
Labor portion: 0.77082 × $1,491.23 =
$112.39 $1,149.47
2008 for HHAs that do not submit
Wage-adjust the labor portion by Non-labor portion: 0.22918 × $1,491.23 = required quality data (see Table 13A).
multiplying the labor portion of the NRS $341.76 As stated in the CY 2008 HH PPS
fixed dollar loss amount by the wage index: Adjust the labor portion of the home health proposed rule, these episodes would be
$378.00 × 0.9860 = $372.71 aide visits by the wage index: further adjusted for case-mix based on
Add the wage-adjusted labor portion to the 0.9860 × $1,149.47 = $1,133.38 the 80 HHRG case-mix model for
non-labor portion for the total NRS amount: Add the wage-adjusted labor portion of the episodes beginning in CY 2007 and
$372.71 + $112.39 = $485.10 home health aide visits to the non-labor ending in CY 2008 (72 FR 25450).
TABLE 13A.—FOR HHAS THAT DO NOT SUBMIT THE REQUIRED QUALITY DATA—NATIONAL 60-DAY EPISODE AMOUNTS
UPDATED BY THE HOME HEALTH MARKET BASKET UPDATE FOR CY 2008, MINUS 2 PERCENTAGE POINTS, FOR EPI-
SODES THAT BEGIN IN CY 2007 AND END IN CY 2008 BEFORE CASE-MIX ADJUSTMENT, WAGE INDEX ADJUSTMENT
BASED ON THE SITE OF SERVICE FOR THE BENEFICIARY OR APPLICABLE PAYMENT ADJUSTMENT
National
standardized
60-day epi-
Multiply by sode payment
the home Reduce by rate for epi-
health mar- 2.75 percent Adjusted to account sodes begin-
ket basket
Total CY 2007 national standardized 60-day episode payment rate for nominal for the 5 percent ning in CY
update (3.0 change in outlier policy 2007 and end-
percent) 1 case-mix ing in CY 2008
minus 2 per- for HHAs that
cent do not submit
required qual-
ity data
$2,339.00 .................................................................................................... × 1.010 ........ × 0.9725 ...... × 1.05 × 0.95 $2,291.68

Next, in order to establish new rates multiply the $2,362.39 by 1.05, resulting subtract the payment adjustment
based on a new case-mix system, we in $2,480.51. Next, we need to reduce amount of $254.27 from $2,480.51, for a
again start with the CY 2007 national this amount to pay for each of our final final rate of $2,226.24 for HHAs that do
standardized 60-day episode payment policy changes. To do this, we take the not submit quality data, for episodes
rate and increase that rate by the payment adjustment amount to pay for that begin and end in CY 2008 (see
rebased and revised home health market our policy changes of this rule, Table 13B).
basket update (3.0 percent) minus 2 determined in Table 11A of $259.31,
percent ($2,339.00 * 1.010 = $2,362.39). multiply it by (1/1.030) to take away the These episodes would be further
We next have to put dollars associated 3.0 percent increase, and multiply that adjusted for case-mix based on the 153
with the outlier target estimate back into number by 1.010 to impose the 1.0 HHRG case-mix model for episodes
the base rate. In the 2000 HH PPS final percent update for episodes where beginning and ending in CY 2008. We
rule (65 FR 41184), we divided the base HHAs have not submitted the required increase the case-mix weights by a
rate by 1.05 to account for outlier quality data. This results in a payment budget neutrality factor of 1.238848031.
payments. Therefore, we proposed to adjustment amount of $254.27. Finally,
TABLE 13B.—FOR HHAS THAT DO NOT SUBMIT THE REQUIRED QUALITY DATA—NATIONAL 60-DAY EPISODE AMOUNTS
UPDATED BY THE HOME HEALTH MARKET BASKET UPDATE FOR CY 2008, MINUS 2 PERCENTAGE POINTS, FOR EPI-
SODES THAT BEGIN AND END IN CY 2008, BEFORE CASE-MIX ADJUSTMENT, WAGE INDEX ADJUSTMENT BASED ON
THE SITE OF SERVICE FOR THE BENEFICIARY OR APPLICABLE PAYMENT ADJUSTMENT
Changes to account for LUPA ad- CY 2008 national

Multiply by Adjusted to justment ($5.70), NRS payment standardized 60-
the home return the ($45.87), elimination of SCIC policy
Updated and day for episode
health mar- outlier funds ($10.96), outlier policy ($127.22),
Total CY 2007 national stand- outlier adjusted payment rate for
ket basket to the na- and 2.75 percent reduction for nomi-
ardized 60-day episode pay- national standard- episodes begin-
update (3.0 tional stand- nal change in case-mix ($69.56) =
ment rate ized 60-day epi- ning and ending in
percent) 1 ardized 60- $259.31; minus 2 percentage points
sode payment CY 2008 that do
minus 2.0 day episode off of the home health market basket not submit re-
percent payment rate update (3.0 percent) 1 for episodes quired quality data
beginning and ending in CY 2008
$2,339.00 .................................. × 1.010 ........ × 1.05 .......... $2,480.51 ¥$254.27 $2,226.24

In calculating the CY 2008 national we start with the CY 2007 per-visit outlier payments, to yield the updated
per-visit amounts used to calculate rates. We multiply those amounts by the per-visit amounts for each home health
payments for LUPA episodes for HHAs home health market basket update (3.0 discipline for CY 2008 for HHAs that do
that do not submit required quality data percent) minus 2 percentage points, not submit required quality data (see
and to compute the imputed costs in then multiply by 1.05 and 0.95 to Table 14).
outlier calculations for those episodes, account for the estimated percentage of
TABLE 14.—FOR HHAS THAT DO NOT SUBMIT THE REQUIRED QUALITY DATA—NATIONAL PER-VISIT AMOUNTS FOR
LUPAS (NOT INCLUDING THE INCREASE IN PAYMENT FOR A BENEFICIARY’S ONLY EPISODE OR THE INITIAL EPISODE IN
A SEQUENCE OF ADJACENT EPISODES) AND OUTLIER CALCULATIONS UPDATED BY THE HOME HEALTH MARKET BAS-
KET UPDATE FOR CY 2008, MINUS 2 PERCENTAGE POINTS, BEFORE WAGE INDEX ADJUSTMENT BASED ON THE SITE
OF SERVICE FOR THE BENEFICIARY
CY 2008 per-
visit payment
Multiply by amount per
Final CY 2007 the home discipline for a
per-visit Adjusted to
health mar- beneficiary
amounts per account for
Home health discipline type ket basket who resides in
60-day epi- the 5 percent
(3.0 per- a non-MSA for
sode for outlier policy
cent) 1 minus HHAs that do
LUPAs 2.0 percent not submit re-
quired quality
data
Home Health Aide ................................................................................................ $46.24 × 1.010 ........ × 1.05 .......... $46.59
× 0.95 ..........
Medical Social Services ........................................................................................ 163.68 × 1.010 ........ × 1.05 .......... 164.90
× 0.95 ..........
Occupational Therapy ........................................................................................... 112.40 × 1.010 ........ × 1.05 .......... 113.24
× 0.95 ..........
Physical Therapy .................................................................................................. 111.65 × 1.010 ........ × 1.05 .......... 112.48
× 0.95 ..........
Skilled Nursing ...................................................................................................... 102.11 × 1.010 ........ × 1.05 .......... 102.87
× 0.95 ..........
Speech-Language Pathology ............................................................................... 121.22 × 1.010 ........ × 1.05 .......... 122.13
× 0.95 ..........
IV. Provisions of the Final Rule With • The need for the information proposed changes for the information
Comment Period collection and its usefulness in carrying collection requirements (ICRs) as
out the proper functions of our agency. summarized and discussed below. For
In this final rule with comment • The accuracy of our estimate of the the purposes of soliciting public review
period, we are adopting the provisions information collection burden. and comment, we also placed a draft of
as set forth in the CY 2008 HH PPS • The quality, utility, and clarity of the proposed changes to the OASIS on
proposed rule, except as noted in the the information to be collected. the CMS Web site at:
specific response to comments in the • Recommendations to minimize the
applicable sections of this rule (for http://www.cms.hhs.gov/
information collection burden on the PaperworkReductionActof1995/PRAL/
example, case-mix refinements; affected public, including automated
payment adjustments to include the list.asp#TopOfPage.
collection techniques.
LUPA, SCIC, and NRS; outlier policy; We solicited public comments on As discussed in section II.A.(2)(a) of
and the update of the HH PPS rates to each of aforementioned issues for the the proposed rule, and further clarified
include the home health market basket information collection requirements in section III.B.2 of this rule, in order for
and the wage index). We are specifically discussed below. In this final rule with the OASIS to have the information
soliciting comments on the 2.71 percent comment period, we are restating the necessary to allow the grouper to price-
reduction to the HH PPS payment rates discussion of the information collection out the claim, we proposed to make the
schedule in 2011, to account for changes requirements as it appeared in the HH following changes to the OASIS to
in coding that were not related to an PPS proposed rule that published on capture whether an episode is an early
underlying change in patient health May 4, 2007 (72 FR 25356). or later episode.
status (see Section III.B.6.) To implement the OASIS changes The creation of a new OASIS item to
V. Collection of Information discussed in sections II.A.(2)(a), capture whether a particular assessment
Requirements II.A.(2)(b), and II.A.(2)(c) of the is for an episode considered to be an
proposed rule, and further discussed early episode or a later episode in the
Under the Paperwork Reduction Act and clarified in sections III.B.2, III.B.3, patient’s current sequence of adjacent
(PRA) of 1995, we are required to and III.B.4 of this rule in the analysis of Medicare home health payment
provide 30-day notice in the Federal and public response to public comments episodes. As defined in section II.A.1. of
Register and solicit public comment on the proposed rule, which are the proposed rule, and further clarified
before a collection of information currently approved in § 484.55, in section III.B.2 of this rule, we define
requirement is submitted to the Office of § 484.205, and § 484.250, a few items in a sequence of adjacent episodes for a
Management and Budget (OMB) for the OASIS will need to be modified, beneficiary as a series of claims with no
review and approval. In order to fairly deleted, or added. The requirements and more than 60 days without home care
evaluate whether an information burden associated with the OASIS are between the end of one episode, which
collection should be approved by OMB, currently approved under OMB control is the 60th day (except for episode that
section 3506(c)(2)(A) of the PRA of 1995 number 0938–0760 with an expiration have been PEP-adjusted), and the
requires that we solicit comment on the date of August 31, 2007. We solicited beginning of the next episode. This
following issues: public comment on each of the definition holds true regardless of
whether or not the same HHA provided The burden associated with the Mandates Reform Act of 1995 (Pub. L.
care for the entire sequence of adjacent proposed changes discussed in sections 104–4), and Executive Order 13132.
episodes. The HHA will chose from the II.A.(2)(a), II.A.(2)(b), and II.A.(2)(c) of Executive Order 12866 (as amended
options: ‘‘Early’’ for single episodes or the proposed rule, and further discussed by Executive Order 13258, which
the first or second episode in a sequence and clarified in section III.B.2, III.B.3, merely reassigns responsibility of
of adjacent episodes, ‘‘Later’’ for third or and III.B.4 of this rule, includes possible duties) directs agencies to assess all
later episodes, ‘‘UK’’ for unknown if the training of staff, the time and effort costs and benefits of available regulatory
HHA is uncertain as to whether the associated with downloading a new alternatives and, if regulation is
episode is an early or later episode (the form and replacing previously pre- necessary, to select regulatory
payment grouper software will default printed versions of the OASIS, and approaches that maximize net benefits
to the definition of an ‘‘early’’ episode), utilizing updated vendor software. (including potential economic,
and ‘‘NA’’ for not applicable (no However, as stated above, CMS is environmental, public health and safety
Medicare case-mix group to be defined removing or modifying existing effects, distributive impacts, and
by this assessment). questions in the OASIS data set to equity). A regulatory impact analysis
As discussed in section II.A.(2)(b) of accommodate the proposed (RIA) must be prepared for major rules
the proposed rule, we proposed to make requirements referenced above. In with economically significant effects
changes to the OASIS in order to enable addition, as a result of the proposed ($100 million or more in any 1 year).
agencies to report secondary case-mix changes, we expect that the claims This final rule will be a major rule, as
diagnosis codes. The proposed changes processing system will automatically defined in Title 5, United States Code,
clarify how to appropriately fill out adjust the therapy visits both upward section 804(2), because we estimate the
OASIS items M0230 and M0240, using and downward on the final claim, impact to the Medicare program, and
ICD–9–CM sequencing requirements if according to the information on the final the annual effects to the overall
multiple coding is indicated for any claim. Consequently, the HHA would no economy, will be more than $100
diagnosis. Additionally, if a V-code is longer have to withdraw and resubmit a million. The update set forth in this
reported in place of a case-mix revised claim when the number of proposed rule would apply to Medicare
diagnosis for OASIS item M0230 or therapy visits delivered to the patient is payments under the HH PPS in CY
M0240, then the new optional OASIS higher than the level report on the RAP. 2008.
Therefore, CMS believes the burden Accordingly, the following analysis
item (which is replacing existing OASIS
increase associated with these changes describes the impact in CY 2008 only.
item M0245) may then be completed. A
is negated by the removal or We estimate that the net impact in this
case-mix diagnosis is a diagnosis that
modification of several current data rule, including a 2.75 percent reduction
determines the HH PPS case-mix group. to the payment rate to account for the
Further discussion or clarification of items.
We have submitted a copy of this final case-mix change adjustment in case-
these proposed changes can be found in mix, is estimated to be approximately
section III.B.3 of this rule. rule to OMB for its review of the
information collection requirements $20 million in CY 2008 expenditures.
As discussed in section II.A.(2)(c) of That estimate incorporates the 3.0
the proposed rule, we proposed to make described above. These requirements are
not effective until OMB has approved percent home health market basket
changes to the OASIS to capture the increase (an estimated additional $430
projected total number of therapy visits them.
If you comment on any of these million in CY 2008 expenditures
for a given episode. With the projected attributable only to the CY 2008 home
total number of therapy visits, the information collection and record
keeping requirements, please mail health market basket update), and the
payment grouper would be able to group 2.75 percent decrease (¥$410 million
that episode into the appropriate case- copies directly to the following:
for the first year of a 4-year phase-in) to
mix group for payment. The existing Centers for Medicare & Medicaid
the HH PPS national standardized 60-
OASIS item M0825 asks an HHA if the Services, Office of Strategic
day episode rate to account for the case-
projected number of therapy visits Operations and Regulatory Affairs, mix change adjustment under the HH
would meet the therapy threshold or Regulations Development Group, PPS. The $20 million is reflected in
not. As noted previously, we proposed Attn.: Melissa Musotto, CMS–1541– column 7 of Table 15 as a 0.2 percent
to delete OASIS item M0825 and FC, Room C4–26–05, 7500 Security increase in expenditures when
replace it with a new OASIS item. The Boulevard, Baltimore, MD 21244– comparing the current CY 2007 system
OASIS item would ask the following: 1850; and to the revised CY 2008 system. In the
‘‘In the plan of care for the Medicare Office of Information and Regulatory
proposed rule, the difference between
payment episode for which this Affairs, Office of Management and
the proposed 2.9 percent update ($410
assessment will define a case-mix Budget, Room 10235, New Executive
million) and the 2.75 percent decrease
group, what is the indicated need for Office Building, Washington, DC ($400 million) was $10 million. The
therapy visits (total of reasonable and 20503, Attn: Carolyn Lovett, CMS additional $130 million difference, in
necessary physical, occupational, and Desk Officer, (CMS–1541–FC), the proposed rule, between estimated
speech-pathology visits combined)?’’ carolyn_lovett@omb.eop.gov. Fax CY 2007 and CY 2008 total payments
The HHA would provide the total (202) 395–6974. resulted from the differential treatment
number of projected therapy visits for VI. Regulatory Impact Analysis of the outlier offsets to the payment
that Medicare payment episode, unless rates and the percent of outlier
not applicable (that is, no case-mix A. Overall Impact payments between the two simulations.
group defined by this assessment). The We have examined the impacts of this Specifically, the $130 million difference
HHA would enter ‘‘000’’ if no therapy rule as required by Executive Order reflected the lower payments estimated
visits were projected for that particular 12866 (September 1993, Regulatory for CY 2007 resulting from the estimated
episode. Further discussion and Planning and Review), the Regulatory outlier payments of only 4.14 percent
clarification of these proposed changes Flexibility Act (RFA) (September 19, rather than 5 percent. Our analysis of
can be found in section III.B.4 of this 1980, Pub. L. 96–354), section 1102(b) of more recent data than the CY 2005 data
rule. the Social Security Act, the Unfunded available for both the CY 2007 and CY
2008 impact analysis simulations proposed rule (and subsequent final of a budget neutrality factor for the case-
strongly suggests that outlier payments rule) that imposes substantial direct mix weights. Column one of this table
in CY 2007 and CY 2008 are or will be requirement costs on State and local classifies HHAs according to a number
greater than 5 percent of total payments. governments, preempts State law, or of characteristics including provider
Since the CY 2005 data show outlier otherwise has Federalism implications. type, geographic region, and urban
payments of only about 4.1 percent, the We have determined that this final rule versus rural location. Column two
CY 2005 data are not informative about will not have substantial direct effects displays the average case-mix weight for
actual outlier experience in CY 2007 on the rights, roles, and responsibilities each type of agency under the current
and CY 2008. For the final rule impact of States. payment system. Column three displays
analysis, we have set the FDLs in the CY the average case-mix weight for each
B. Anticipated Effects
2007 and CY 2008 simulations to be type of agency incorporating all of the
consistent with outlier payments of 5 This final rule with comment period changes/refinements discussed above.
percent so that outlier payments have updates the HH PPS rates contained in The average case-mix weight for
similar effects in all of the impact the CY 2007 final rule (71 FR 65884, proprietary (for profit) agencies is
simulations. We believe that this November 9, 2006). The impact analysis estimated to decrease from 1.2821 to
approach comes as close as possible to of this final rule presents the refinement 1.2620. Comparatively, the average case-
estimating the desired impacts in a related policy changes in this rule. We mix weight for voluntary non-profit
comparable manner, given the recent use the latest data and best analysis agencies is estimated to increase from
changes in outlier payments. available, but we do not attempt to 1.1875 to 1.2334. Rural agencies are
The RFA requires agencies to analyze predict behavioral responses to these estimated to experience a decrease in
options for regulatory relief of small changes, and we do not make their average case-mix from 1.2047 to
businesses. For purposes of the RFA, adjustments for future changes in such 1.1798. It is estimated that urban
small entities include small businesses, variables as days or case-mix. agencies would see a slight increase in
nonprofit organizations, and small This analysis incorporates the latest
their average case-mix weight from
governmental jurisdictions. Most estimates of growth in service use and
1.2520 to 1.2616. In particular, the New
hospitals and most other providers and payments under the Medicare home
England, Mid-Atlantic, South Atlantic,
suppliers are small entities, either by health benefit, based on the latest
East North Central, West North Central,
nonprofit status or by having revenues available Medicare claims from 2005.
and Mountain areas of the country are
of $6 million to $29 million in any 1 We note that certain events may
estimated to see their average case-mix
year. For purposes of the RFA, combine to limit the scope or accuracy
increase under the proposed
approximately 75 percent of HHAs are of our impact analysis, because such an
refinements of this rule. Conversely, the
considered small businesses according analysis is future-oriented and, thus,
susceptible to forecasting errors due to East South Central, West South Central,
to the Small Business Administration’s and Pacific areas of the country are
size standards with total revenues of other changes in the forecasted impact
time period. Some examples of such estimated to see their average case-mix
$11.5 million or less in any 1 year. decrease as a result of refinements of
Individuals and States are not included possible events are newly-legislated
general Medicare program funding this rule. Both small and large agencies
in the definition of a small entity. As are estimated to see decreases in their
stated above, this final rule will have an changes made by the Congress, or
changes specifically related to HHAs. In average case-mix under the new
estimated positive effect upon small proposed case-mix system, the only
entities that are HHAs. addition, changes to the Medicare
program may continue to be made as a exception being much larger agencies
In addition, section 1102(b) of the Act
result of the BBA, the BBRA, the (200+ first episodes), which are
requires us to prepare a regulatory
impact analysis if a rule may have a Medicare, Medicaid, and SCHIP estimated to see an increase of their
significant impact on the operations of Benefits Improvement and Protection average case-mix from 1.2376 to 1.2398.
a substantial number of small rural Act of 2000, the MMA, the DRA, or new For the purposes of analyzing impacts
hospitals. This analysis must conform to statutory provisions. Although these on payments, we performed five
the provisions of section 603 of the changes may not be specific to the HH simulations and compared them to each
RFA. For purposes of section 1102(b) of PPS, the nature of the Medicare program other.
the Act, we define a small rural hospital is such that the changes may interact, Based on our estimate that outliers, as
as a hospital that is located outside of and the complexity of the interaction of a percentage of total HH PPS payments,
a Metropolitan Statistical Area and has these changes could make it difficult to will be at least 5 percent in CY 2007, the
fewer than 100 beds. We have predict accurately the full scope of the 2007 baseline, for the purposes of these
determined that this final rule will not impact upon HHAs. simulations, we assumed that the full 5
have a significant economic impact on Table 15 represents how home health percent outlay for outliers will be paid.
the operations of a substantial number agencies are likely to be affected by the The first simulation estimates 2008
of small rural hospitals. policy changes described in this rule. payments under the current system (to
Section 202 of the Unfunded For each agency type listed below, Table include the 2007 wage index and labor
Mandates Reform Act of 1995 also 15 displays the average case-mix index, share). The second simulation estimates
requires that agencies assess anticipated both under the current HH PPS case-mix 2008 payments under the current
costs and benefits before issuing any system and the CY 2008 HH PPS case- system, but with the 2008 wage index
rule that may result in expenditure in mix system. For this analysis, we used and the new 2008 labor share. The
any 1 year by State, local, or tribal the most recent data available that second simulation produces an estimate
governments, in the aggregate, or by the linked home health claims and OASIS of what total payments using the sample
private sector, of $110 million. We assessments, a 20-percent sample of data will be in 2008 without making any
believe this final rule will not mandate episodes occurring in CY 2005. In Table of the refinement-related changes
expenditures in that amount. 15, the average case-mix is the same, in described in this final rule. The third
Executive Order 13132 establishes the aggregate, between the current HH simulation estimates 2008 payment with
certain requirements that an agency PPS system and the proposed revised the old, 2007 labor share and a 2008
must meet when it promulgates a HH PPS system, due to our application wage index. The fourth simulation
estimates 2008 payments with a new expenditures. However, the CY 2007 the decrease 2.37 percent for proprietary
2008 labor share and a 2007 wage index. payment simulation in our proposed HHAs is the free-standing proprietary
These first four simulations allow us rule predicted outlier payments of only HHAs, which are estimated to see a
to demonstrate the effects of a new 2008 4.14 percent with the CY 2007 FDL of decrease of 2.49 percent in the
wage index and a new 2008 labor share 0.67. Since in the CY 2007 simulation percentage change in estimated total
as a percentage change in estimated we made no upward adjustment to the payment from CY 2007 to the revised
expenditures. Specifically, the fourth rates similar to the offsetting adjustment CY 2008 system.
column of Table 15 shows the percent we made in the CY 2008 simulation, We note that some of these impacts
change due to the combined effects of estimated CY 2007 total payments with are partly explained by practice patterns
the new 2008 labor share and the 2008 the .67 FDL were lower than they would associated with certain types of
wage index. Column five shows the have been had outlier payments been 5 agencies. For example, LUPA episodes
percent change due to the effects of the percent of total payments. This are relatively common among nonprofit
new labor share. And finally, column 6 asymmetrical approach to the agencies and freestanding government-
shows us the percent change due to the comparative simulations for CY 2007 owned agencies. Our implementing an
effects of updated wage data (2008 wage and CY 2008 yielded an estimated $130 additional payment for certain LUPA
index). million in additional payments from episodes would tend to increase
The fifth, and final, simulation moving to the new system. payments for such classes of agencies
estimates what total payments would be We have revised the final rule’s with higher-than-average LUPA rates,
in 2008, using the final case-mix model, impact analysis by simulating CY 2007 while tending to decrease payments for
the additional payment for initial and and CY 2008 payments in a consistent agencies with comparatively low LUPA
only episode LUPA episodes, the manner with respect to outlier policy. rates. Similarly, the elimination of the
removal of SCIC adjustments, and the We made no adjustment to the rates in SCIC policy would tend to favorably
revised approach to making NRS either simulation of the kind we made affect total payments for agencies with
payments. The fifth simulation also to the proposed regulation’s CY 2008 relatively high rates of SCIC episodes,
assumes payments will incorporate the simulation. In other words, both sets of such as facility-based proprietary
rebased and revised home health market rates and the FDL ratios assume outlier agencies and facility-based government
basket increase of 3.0 percent, the new payments reach the 5 percent target. The agencies.
outlier threshold determined by an basis for taking this approach is that our The percentage change in estimated
updated FDL ratio of 0.89, and the 2.75 supplementary analysis of more recent total payments from CY 2007 to a CY
percent reduction in the national data than the CY 2005 data available for 2008 system that incorporates all of the
standardized 60-day episode payment both the CY 2007 and CY 2008 refinements to the HH PPS for rural
rate to account for the case-mix change simulations strongly suggests that HHAs is a decrease of 1.77 percent,
adjustment. All five simulations use a outlier payments in CY 2007 and CY while for urban HHAs an increase of
CBSA-based wage index (we used a 2008 are or will be greater than 5 0.80 percent is expected. Urban agencies
crosswalk from the MSA reported on the percent of total payments. Since the CY have somewhat higher LUPA rates than
2005 claims to the CBSA to determine 2005 data show outlier payments of rural agencies, so urban agencies would
the appropriate wage index). only about 4.1 percent, the CY 2005 data be expected to benefit, relative to rural
Column seven shows the percentage are not informative about actual outlier agencies, from the proposal to make an
change in estimated total payments in experience in CY 2007 and CY 2008. For additional payment for certain LUPA
moving from the current CY 2007 to the the final rule impact analysis, we have episodes. Urban agencies are also more
revised CY 2008 system outlined in this set the FDLs in the CY 2007 and CY likely to benefit from elimination of the
final rule. As a result of changes in our 2008 simulations to be consistent with SCIC policy. Urban agencies are less
approach to the impact analysis outlier payments of 5 percent so that likely to bill a SCIC episode than rural
simulations between the proposed rule outlier payments have similar effects in agencies. However, when urban
and this rule, our estimate of the change all of the impact simulations. We agencies do bill a SCIC episode the
in total payments between CY 2007 and believe that this approach comes as payment is reduced more, on average,
CY 2008 is substantially less than what close as possible to estimating the than when rural agencies bill a SCIC.
we presented in the proposed rule. The desired impacts in a comparable The net effect of these two components
percentage change in estimated total manner, given the recent changes in (relative frequency and payment impact
payments from CY 2007 to the revised outlier payments. As a result of these per SCIC episode) is a larger expected
CY 2008 system is now the difference changes in approach, our estimate of the reduction for urban agencies under the
between the 3.0 percent update and the change in total payments between CY SCIC adjustment policy. Therefore,
2.75 percent reduction in the rates for 2007 and CY 2008 is an increase of $20 while both urban and rural agencies
an increase of $20 million, or million or approximately 0.1 to 0.2 benefit from eliminating the SCIC
approximately 0.2 percent). percent. policy, urban agencies benefit more.
In the proposed rule, we stated that In general, voluntary non-profit HHAs HHAs in the North are expected to
the estimated additional $130 million (3.60 percent), facility-based HHAs (3.66 experience a percentage change increase
yielding the $140 million in estimated percent), and government owned HHAs of 4.57 percent in estimated total
spending for CY 2008 is due to the fixed (3.04 percent) are estimated to see an payments from CY 2007 to the revised
dollar loss ratio at 0.67 (72 FR 25454). increase in the percentage change in CY 2008 system. One region, the South,
What that means is that the CY 2008 estimated total payments from CY 2007 is estimated to experience a decrease in
simulation compensated for fixing the to the revised CY 2008 system. the percentage change in estimated total
FDL at 0.67 by raising all the payment Proprietary and freestanding HHAs, on payments from CY 2007 to the revised
rates to meet the target expenditure the other hand, are estimated to see CY 2008 system. That percentage
total. In the CY 2008 simulation, this decreases of 2.37 percent and 0.64 change is an estimated decrease of 2.91
compensatory adjustment raised total percent, respectively, in estimated total percent.
payments by an amount that would payments from CY 2007 to the proposed It is estimated that New England and
have been equivalent to spending the revised CY 2008 system. As it was in the Mid Atlantic area HHAs will experience
entire outlier target of 5% of total proposed rule, the major contributor to percentage change increases
approaching 4 or 5 percent, respectively estimated total payments from CY 2007 CY 2007 to the revised CY 2008 system.
(New England, 3.83 percent and the to the revised CY 2008 system. In Conversely, the largest HHAs (those
Mid-Atlantic, 4.96 percent) in estimated general, HHAs with less than 200 with 200 or more Medicare home health
total payments from CY 2007 to the Medicare home health initial episodes initial episodes per year) are estimated
revised CY 2008 system. Conversely, per year are expected to experience a to experience a slight increase of 0.36
West South Central HHAs are expected decrease (ranging from ¥0.78 percent to percent change in estimated total
to experience a decrease (¥6.32 1.93 percent) for their percentage payments from CY 2007 to the CY 2008
percent) in the percentage change in change in estimated total payments from system.
TABLE 15.—IMPACT BY AGENCY TYPE

Case-Mix Comparisons
Percent
Change Due
to the Com- Percent Percent
Percent
bined Effects Change Due Change from
Change Due
Group Case-Mix Case-Mix of the New to the Effects the Current
to the Effects
Index Current Index, Revised Labor Share of the Updated CY 2007 Sys-
of the New
80 HHRGs 153 HHRGs (0.77082) and Wage Data tem to the Re-
Labor Share
the Updated (2008 Wage vised CY 2008
(0.77082)
Wage Data Index) System
(2008 Wage
Index)
Type of Facility
Unknown .................................................. 1.5011 1.4848 0.10 0.02 0.07 ¥1.64

Free-Standing/Other Vol/NP .................... 1.1982 1.2467 0.09 0.00 0.08 3.47
Free-Standing/Other Proprietary .............. 1.2841 1.2625 ¥0.06 ¥0.02 ¥0.04 ¥2.49
Free-Standing/Other Government ........... 1.2038 1.2576 0.04 ¥0.05 0.09 2.84
Facility-Based Vol/NP .............................. 1.1736 1.2162 0.04 ¥0.02 0.05 3.78
Facility-Based Proprietary ........................ 1.2145 1.2439 ¥0.03 ¥0.05 0.01 2.79
Facility-Based Government ...................... 1.1513 1.1857 ¥0.10 ¥0.05 ¥0.05 3.28
Subtotal: Freestanding ...................... 1.2551 1.2576 ¥0.02 ¥0.02 0.00 ¥0.64
Subtotal: Facility-based .................... 1.1737 1.2146 0.02 ¥0.02 0.04 3.66
Subtotal: Vol/PNP ............................. 1.1875 1.2334 0.07 ¥0.01 0.07 3.60
Subtotal: Proprietary ......................... 1.2821 1.2620 ¥0.06 ¥0.02 ¥0.04 ¥2.37
Subtotal: Government ....................... 1.1796 1.2244 ¥0.02 ¥0.05 0.03 3.04
TOTAL ....................................... 1.2388 1.2388 ¥0.01 ¥0.02 0.00 0.20
Type of Facility (Rural* Only)
Unknown .................................................. 0.8205 0.8221 0.05 0.05 0.00 ¥0.15

Free-Standing/Other Vol/NP .................... 1.1746 1.1895 0.09 ¥0.05 0.14 1.14
Free-Standing/Other Government ........... 1.1883 1.2490 0.08 ¥0.07 0.14 2.74
Facility-Based Vol/NP .............................. 1.1588 1.1790 ¥0.04 ¥0.06 0.02 2.12
Facility-Based Proprietary ........................ 1.2073 1.2242 ¥0.09 ¥0.08 ¥0.01 1.98
Type of Facility (Urban* Only)
Unknown .................................................. 1.5025 1.4861 0.10 0.02 0.07 ¥1.64

Free-Standing/Other Vol/NP .................... 1.2037 1.2598 0.09 0.01 0.07 3.92
Free-Standing/Other Government ........... 1.2312 1.2749 ¥0.01 ¥0.02 0.00 2.99
Facility-Based Vol/NP .............................. 1.1803 1.2332 0.07 0.00 0.06 4.41
Facility-Based Proprietary ........................ 1.2225 1.2655 0.02 ¥0.02 0.03 3.54
Type of Facility: Urban* or Rural*
Rural* ....................................................... 1.2047 1.1798 ¥0.06 ¥0.07 0.00 ¥1.77

Urban* ...................................................... 1.2520 1.2616 0.01 0.00 0.01 0.80
TOTAL .............................................. 1.2388 1.2388 ¥0.01 ¥0.02 0.00 0.20
Type of Facility: Region
North ........................................................ 1.1499 1.2090 0.12 0.02 0.10 4.57

South ........................................................ 1.2761 1.2351 ¥0.19 ¥0.04 ¥0.15 ¥2.91

Midwest .................................................... 1.2249 1.2645 0.16 ¥0.02 0.18 3.12
West ......................................................... 1.2423 1.2382 0.18 0.02 0.15 0.03
Other ........................................................ 1.2716 1.2933 ¥0.04 ¥0.06 0.02 2.13
TOTAL .............................................. 1.2388 1.2388 ¥0.01 ¥0.02 0.00 0.20
TABLE 15.—IMPACT BY AGENCY TYPE—Continued

Case-Mix Comparisons
Percent
Change Due
to the Com- Percent Percent
Percent
bined Effects Change Due Change from
Change Due
Group Case-Mix Case-Mix of the New to the Effects the Current
to the Effects
Index Current Index, Revised Labor Share of the Updated CY 2007 Sys-
of the New
80 HHRGs 153 HHRGs (0.77082) and Wage Data tem to the Re-
Labor Share
the Updated (2008 Wage vised CY 2008
(0.77082)
Wage Data Index) System
(2008 Wage
Index)
Type of Facility: Area of the Country
New England ............................................ 1.1106 1.1611 0.10 0.02 0.07 3.83

Mid Atlantic .............................................. 1.1706 1.2343 0.14 0.01 0.12 4.96
South Atlantic ........................................... 1.2862 1.2877 ¥0.09 ¥0.03 ¥0.07 0.44
East South Central ................................... 1.2897 1.2667 ¥0.22 ¥0.07 ¥0.16 ¥1.99
West South Central .................................. 1.2618 1.1781 ¥0.27 ¥0.05 ¥0.23 ¥6.32
East North Central ................................... 1.2409 1.2818 0.22 ¥0.01 0.23 3.14
West North Central .................................. 1.1705 1.2055 ¥0.04 ¥0.04 ¥0.01 3.04
Mountain .................................................. 1.2660 1.3161 ¥0.06 ¥0.04 ¥0.03 3.22
Pacific ....................................................... 1.2305 1.1992 0.28 0.05 0.22 ¥1.21
Other ........................................................ 1.2716 1.2933 ¥0.04 ¥0.06 0.02 2.13
TOTAL .............................................. 1.2388 1.2388 ¥0.01 ¥0.02 0.00 0.20
Type of Facility: Size (Number of First Episodes/Year)
Unknown .................................................. 1.0130 0.8895 ¥0.27 ¥0.03 ¥0.24 ¥7.85

1 to 5 ........................................................ 1.2056 1.1866 ¥0.02 ¥0.02 0.00 ¥1.05
6 to 9 ........................................................ 1.2145 1.1806 0.00 ¥0.03 0.02 ¥1.83
10 to 14 .................................................... 1.2297 1.2128 ¥0.07 ¥0.02 ¥0.05 ¥0.78
15 to 19 .................................................... 1.2335 1.2186 ¥0.05 ¥0.02 ¥0.03 ¥1.10
20 to 29 .................................................... 1.2412 1.2065 ¥0.05 ¥0.02 ¥0.03 ¥1.93
30 to 49 .................................................... 1.2463 1.2335 ¥0.05 ¥0.02 ¥0.03 ¥0.86
50 to 99 .................................................... 1.2505 1.2360 ¥0.04 ¥0.02 ¥0.02 ¥0.84
100 to 199 ................................................ 1.2489 1.2334 ¥0.03 ¥0.02 ¥0.01 ¥0.92
200 or More ............................................. 1.2376 1.2398 ¥0.01 ¥0.02 0.01 0.36
TOTAL .............................................. 1.2388 1.2388 ¥0.01 ¥0.02 0.00 0.20
Note: Based on a 20 percent sample of CY 2005 claims linked to OASIS assessment. Due to sample differences, national average case-mix
weight in this table differs slightly from national average for CY 2005 reported in the text (1.2361).
*Urban/rural status, for the purposes of these simulations, is based on the wage index on which episode payment is based. The wage index is
based on the site of service of the beneficiary.
C. Accounting Statement have prepared an accounting statement the HH PPS as a result of the changes
showing the classification of the presented in this final rule with
As Required by OMB Circular A-4 expenditures associated with the comment period based on the data for
(available at http:// provisions of this final rule. This table 8,164 HHAs in our database. All
www.whitehouse.gov/omb/circulars/ provides our best estimate of the expenditures are classified as transfers
a004/a-4.pdf), in Table 16 below, we increase in Medicare payments under to Medicare providers (that is, HHAs).
TABLE 16.—ACCOUNTING STATEMENT: CLASSIFICATION OF ESTIMATED EXPENDITURES, FROM CY 2007 TO CY 2008

[In millions]
Category Transfers
Annualized Monetized Transfers .............................................................. $20.

From Whom to Whom .............................................................................. Federal Government to HHAs.
In accordance with the provisions of ■ For the reasons set forth in the Authority: Secs. 1102 and 1871 of the
Executive Order 12866, this regulation preamble, the Centers for Medicare & Social Security Act (42 U.S.C. 1302 and
was reviewed by the Office of Medicaid Services amends 42 CFR 1395(hh)).
Management and Budget. chapter IV as set forth below:
Subpart E—Prospective Payment

List of Subjects in 42 CFR Part 484 System for Home Health Agencies
PART 484—HOME HEALTH SERVICES
Health facilities, Health professions, § 484.205 [Amended]
Medicare, and Reporting and ■ 1. The authority citation for part 484 ■ 2. Amend § 484.205 by—
recordkeeping requirements. continues to read as follows: ■ A. Removing paragraph (a)(3).
■ B. Redesignating paragraph (a)(4) as § 484.220 Calculation of the adjusted claims with no more than 60 days
paragraph (a)(3). national prospective 60-day episode without home care between the end of
payment rate for case-mix and area wage one episode, which is the 60th day
■ C. Revising paragraph (b) introductory levels.
text. (except for episodes that have been PEP-
CMS adjusts the national prospective adjusted), and the beginning of the next
■ D. Removing paragraph (e). 60-day episode payment rate to account episode. This additional amount will be
■ E. Redesignating paragraph (f) as for the following: updated annually after 2008 by a factor
paragraph (e). (a) HHA case-mix using a case-mix equal to the applicable home health
index to explain the relative resource market basket percentage.
The revisions read as follows: utilization of different patients. To
§ 484.205 Basis of payment.
address changes to the case-mix that are § 484.237 [Removed]
a result of changes in the coding or ■ 5. Remove § 484.237.
* * * * * classification of different units of
(b) Episode payment. The national service that do not reflect real changes ■ 6. Amend § 484.240 by revising
in case-mix, the national prospective 60- paragraph (b) to read as follows:
prospective 60-day episode payment
represents payment in full for all costs day episode payment rate will be § 484.240 Methodology used for the
associated with furnishing home health adjusted downward as follows: calculation of the outlier payment.
services previously paid on a reasonable (1) For CY 2008, the adjustment is * * * * *
cost basis (except the osteoporosis drug 2.75 percent. (b) The outlier threshold for each
listed in section 1861(m) of the Act as (2) For CY 2009 and CY 2010, the
case-mix group is the episode payment
adjustment is 2.75 percent in each year.
defined in section 1861(kk) of the Act) (3) For CY 2011, the adjustment is amount for that group, the PEP
as of August 5, 1997 unless the national 2.71 percent. adjustment amount for the episode plus
60-day episode payment is subject to a (b) Geographic differences in wage a fixed dollar loss amount that is the
low-utilization payment adjustment set levels using an appropriate wage index same for all case-mix groups.
forth in § 484.230, a partial episode based on the site of service of the * * * * *
payment adjustment set forth at beneficiary. (Catalog of Federal Domestic Assistance
§ 484.235, or an additional outlier ■ 4. Amend § 484.230 by adding a third, Program No. 93.773, Medicare—Hospital
payment set forth in § 484.240. All fourth, and fifth sentence after the Insurance; and Program No. 93.774,
payments under this system may be second sentence to read as follows: Medicare—Supplementary Medical
subject to a medical review adjustment Insurance Program)
reflecting beneficiary eligibility, medical § 484.230 Methodology used for the
calculation of the low-utilization payment Dated: August 17, 2007.
necessity determinations, and HHRG adjustment. Herb. B. Kuhn,
assignment. DME provided as a home
* * * For 2008 and subsequent Acting Deputy Administrator, Centers for
health service as defined in section Medicare & Medicaid Services.
calendar years, an amount will be added
1861(m) of the Act continues to be paid to low-utilization payment adjustments Approved: August 20, 2007.
the fee schedule amount. for low-utilization episodes that occur Michael O. Leavitt,
* * * * * as the beneficiary’s only episode or Secretary.
■ 3. Revise § 484.220 to read as follows: initial episode in a sequence of adjacent
episodes. For purposes of the home Note: The following addenda will not be
health PPS, a sequence of adjacent published in the Code of Federal Regulations.
episodes for a beneficiary is a series of BILLING CODE 4120–01–P
ER29AU07.054</GPH>
ER29AU07.055</GPH>
ER29AU07.056</GPH>
ER29AU07.057</GPH>
ER29AU07.058</GPH>
ER29AU07.059</GPH>
ER29AU07.060</GPH>
ER29AU07.061</GPH>
ER29AU07.062</GPH>
ER29AU07.063</GPH>
ER29AU07.064</GPH>
ER29AU07.065</GPH>
ER29AU07.066</GPH>
ER29AU07.067</GPH>
ER29AU07.068</GPH>
ER29AU07.069</GPH>
ER29AU07.070</GPH>
ER29AU07.071</GPH>
ER29AU07.072</GPH>
ER29AU07.073</GPH>
ER29AU07.074</GPH>
ER29AU07.075</GPH>
ER29AU07.076</GPH>
ER29AU07.077</GPH>
ER29AU07.078</GPH>
ER29AU07.079</GPH>
ER29AU07.080</GPH>
ER29AU07.081</GPH>
ER29AU07.082</GPH>
ER29AU07.083</GPH>
ER29AU07.084</GPH>
ER29AU07.085</GPH>
ER29AU07.086</GPH>
ER29AU07.087</GPH>
ER29AU07.088</GPH>
ER29AU07.089</GPH>
ER29AU07.090</GPH>
ER29AU07.091</GPH>
ER29AU07.092</GPH>
ER29AU07.093</GPH>
ER29AU07.094</GPH>
ER29AU07.095</GPH>
ER29AU07.096</GPH>
ER29AU07.097</GPH>
ER29AU07.098</GPH>
ER29AU07.099</GPH>
ER29AU07.100</GPH>
ER29AU07.101</GPH>
ER29AU07.102</GPH>
ER29AU07.103</GPH>
ER29AU07.104</GPH>
ER29AU07.105</GPH>
ER29AU07.106</GPH>
ER29AU07.107</GPH>
ER29AU07.108</GPH>
ER29AU07.109</GPH>
ER29AU07.110</GPH>
ER29AU07.111</GPH>
ER29AU07.112</GPH>
ER29AU07.113</GPH>
ER29AU07.114</GPH>
ER29AU07.115</GPH>
ER29AU07.116</GPH>
ER29AU07.117</GPH>
ER29AU07.118</GPH>
[FR Doc. 07–4184 Filed 8–22–07; 4:00 pm]

BILLING CODE 4120–01–C
ER29AU07.119</GPH>

Rule: Medicare: Home Health Prospective Payment System 2008 CY Refinement and Rate Update

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Rule: Medicare: Home Health Prospective Payment System 2008 CY Refinement and Rate Update

Hochgeladen von

Copyright:

Verfügbare Formate

Wednesday,

August 29, 2007

42 CFR Part 484

four ways (no duplicates, please):

under Medicare. 1999. The requirements include the points.

nominal coding practices. Conversely, HHAs some protection against episodes

1st and 2nd Episodes, 6–13—Therapy Visits:

TABLE 7.—AVERAGE RESOURCE COST AND CMI

1999Q4 ............................................................................................................................ $451.11 $451.39 1.1165 1.0796

2000Q4 ............................................................................................................................ N/A N/A N/A N/A

2002Q1 ............................................................................................................................ 453.76 426.42 1.2086 1.1843

TABLE 7.—AVERAGE RESOURCE COST AND CMI—Continued

2003Q3 ............................................................................................................................ 460.15 427.58 1.2333 1.2044

TABLE 9. RELATIVE WEIGHTS FOR NON-ROUTINE MEDICAL SUPPLIES—SIX-GROUP APPROACH

1 ....................................................................... 63.7 0 .................................................................... 0.2698 $14.12

TABLE 10A.—NRS CASE-MIX ADJUSTMENT VARIABLES AND SCORES

SELECTED SKIN CONDITIONS

1 ....................................... Primary diagnosis = Anal fissure, fistula and abscess ................................................................................. 15

OTHER CLINICAL FACTORS

(NRS) CASE-MIX ADJUSTMENT MODEL

566 .................... ........................... ABSCESS OF ANAL AND RECTAL REGIONS.

941 except ........................... BURN OF FACE, HEAD, AND NECK.

$2,339.00 ........................................................................... × 1.030 .................. × 0.9725 ................ × 1.05 × 0.95 $2,337.06

BILLING CODE 4120–01–C

$2,339.00 .................................................................................................... × 1.010 ........ × 0.9725 ...... × 1.05 × 0.95 $2,291.68

Changes to account for LUPA ad- CY 2008 national

$2,339.00 .................................. × 1.010 ........ × 1.05 .......... $2,480.51 ¥$254.27 $2,226.24

TABLE 15.—IMPACT BY AGENCY TYPE

Unknown .................................................. 1.5011 1.4848 0.10 0.02 0.07 ¥1.64

Type of Facility (Rural* Only)

Unknown .................................................. 0.8205 0.8221 0.05 0.05 0.00 ¥0.15

Type of Facility (Urban* Only)

Unknown .................................................. 1.5025 1.4861 0.10 0.02 0.07 ¥1.64

Type of Facility: Urban* or Rural*

Rural* ....................................................... 1.2047 1.1798 ¥0.06 ¥0.07 0.00 ¥1.77

Type of Facility: Region

North ........................................................ 1.1499 1.2090 0.12 0.02 0.10 4.57

South ........................................................ 1.2761 1.2351 ¥0.19 ¥0.04 ¥0.15 ¥2.91

TABLE 15.—IMPACT BY AGENCY TYPE—Continued

Type of Facility: Area of the Country

New England ............................................ 1.1106 1.1611 0.10 0.02 0.07 3.83

Type of Facility: Size (Number of First Episodes/Year)

Unknown .................................................. 1.0130 0.8895 ¥0.27 ¥0.03 ¥0.24 ¥7.85

TABLE 16.—ACCOUNTING STATEMENT: CLASSIFICATION OF ESTIMATED EXPENDITURES, FROM CY 2007 TO CY 2008

Annualized Monetized Transfers .............................................................. $20.

Subpart E—Prospective Payment

[FR Doc. 07–4184 Filed 8–22–07; 4:00 pm]

Das könnte Ihnen auch gefallen