AHCA Memorandum
To:
AHCA Members
From:
Jill Mendlen, Co-Chair, AHCA Finance Committee
Peter Gruhn, Director, Research
William Hartung, VP Research
Elise Smith, Senior VP Finance Policy and Legal Affairs
Re:
Announcement of Assistance to Members for the Centers for Medicare &
Medicaid Innovation Bundled Payment Initiative
Date:
October 10, 2011
_____________________________________________________________________________
Summary
AHCA is announcing our program to assist members who wish to participate in the Centers for
Medicare and Medicaid Services (CMS) Center for Medicare and Medicaid Innovation (CMMI)
recently announced bundled payment initiative. Participation in the initiative will require
considerable information to understand the opportunities in your market, identify conditions and
design the bundles, determine the upside and downside risk for various bundling options and
opportunities. This will be the case whether you seek out others with whom to partner or
whether you are approached by, for example, a hospital.
In Section E, we discuss the data that you will need: data available from your organization,
initiative partners, and CMS, and data made available to you by AHCA -- aggregate patient stay
summary information at the national, state, and county level for selected MS-DRGs to assist
organizations in better understanding opportunities and the cost of acute care, skilled nursing
facility, and other post-acute care services in different markets.
In addition, AHCA will be releasing a series of memos to assist members in completing the
research request package and data use agreement needed to obtain access to the CMS
Medicare Part A claims data. These memorandums will also be posted on the AHCA website the
week of October 17, 2011.
A. Background
The Affordable Care Act (ACA) required the Department of Health and Human Services (HHS) to
establish a national, voluntary pilot program on payment bundling by January 1, 2013. In
preparation to participate in the CMS bundling program mandated by the ACA , AHCA
assembled a member/staff Bundling Workgroup and proactively selected The Moran Company
(Moran) to provide to AHCA with data analytic services related to bundled payment systems. The
project was intended to generate information on post-acute care services that are provided to
Medicare beneficiaries. This information was to be used to educate staff and members on postacute care delivery patterns, develop AHCA advocacy positions on bundling pilot projects, and
provide CMS with a bundling model(s).
The data was also to be used to explore various bundled payment models to help our members
identify opportunities and challenges, and to help us identify what changes providers will need to
make to be responsive to the new models.
1
Instead of establishing a national, voluntary pilot program on payment bundling by January 1,
2013, CMS has instead launched a bundled payment initiative that relies on the
participants to design the bundled payments systems/arrangements.
CMS has released the Innovation Center’s Request for Applications (RFA) which outlines four
broad approaches to bundled payments. Providers can determine which episodes of care and
which services will be bundled together.
Below we provide an overview of the CMMI bundled payment initiative, critical dates, key CMMI
requirements, followed by an outline of the assistance that AHCA will be providing to members.
B. The CMMI Bundled Payment Initiative
As indicated above, CMS has released the Innovation Center’s Request for Applications (RFA)
which outlines four broad approaches to bundled payments. Providers can determine which
episodes of care and which services they would like bundled together.
Briefly the models are as follows:
Retrospective Payment Bundling – Models 1, 2 and 3
In these models, CMS and providers would set a target payment amount for a defined episode of
care. Applicants would propose the target price, which would be set by applying a discount to
total costs for a similar episode of care as determined from historical data. Participants in these
models would be paid for their services under the Original Medicare fee-for-service (FFS) system,
but at a negotiated discount. At the end of the episode, the total payments would be compared
with the target price. Participating providers may then be able to share in those savings. In these
models, CMS and providers would set a target payment amount for a defined episode of care.
Applicants would propose the target price, which would be set by applying a discount to total
costs for a similar episode of care as determined from historical data. Participants in these
models would be paid for their services under the Original Medicare fee-for-service (FFS) system,
but at the discounted rate. At the end of the episode, the total payments would be compared with
the target price. If payments are less than the targeted price, participating providers may then be
able to share in savings. Alternatively, if payments are above the targeted price, providers would
have to reimburse CMS for the overpayments.
Model 1:
Retrospective Acute Care Hospital Stay Only -- In Model 1, the episode of care includes all
hospital services provided to a beneficiary during an acute inpatient hospital stay, where
physicians are partners in improving care. In Model 1, the episode of care would be defined as
the inpatient stay in the general acute care hospital. Medicare will pay the hospital a discounted
amount based on the payment rates established under the Inpatient Prospective Payment
System (IPPS). Medicare will pay physicians separately for their services under the Medicare
Physician Fee Schedule. Hospitals and physicians will be permitted to share gains arising from
better coordination of care.
Model 2:
Retrospective Acute Care Hospital Stay plus Post-Acute Care -- In Model 2, the episode of
care includes hospital, physician, post-acute provider (i.e. skilled nursing facility, inpatient
rehabilitation facility, long-term care hospital, and home health agency), and other Medicarecovered services provided during the inpatient hospital stay as well as during post-acute hospital
discharge to the home or another care setting. In Model 2, the episode of care would include the
2
inpatient stay and post-acute care and would end, at the applicant’s option, either a minimum of
30 or 90 days after discharge, while in Model 3, the episode of care would begin at discharge
from the inpatient stay and would end no sooner than 30 days after acute care hospital discharge.
Organizations interested in receiving historical Medicare claims data for use in developing
episode definitions for Model 2 should complete a research request packet, data use agreement,
and data use agreement signature addendum (if needed). Data will be provided for approved
requests before the application is due.
Model 3:
Retrospective Post-Acute Care Only -- In Model 3, the episode of care includes hospital,
physician, post-acute provider, and other Medicare-covered services beginning with the initiation
of post-acute care services after discharge from an acute inpatient hospital stay. Organizations
interested in receiving historical Medicare claims data for use in developing episode definitions for
Model 3 should complete a research request packet, data use agreement, and data use
agreement signature addendum (if needed). Data will be provided for approved requests before
the application is due.
In both Models 2 and 3, the bundle would include physicians’ services, care by a post-acute
provider, related readmissions, and other services proposed in the episode definition such as
clinical laboratory services; durable medical equipment, prosthetics, orthotics and supplies
(DMEPOS); and Part B drugs. The target price will be an amount based on the applicant’s
historical fee-for-service payments for the episode. Payments will be made at the negotiated
discounted fee-for-service payment rates, after which the aggregate Medicare payment for the
episode will be reconciled against the target price. Any reduction in expenditures beyond the
discount reflected in the target price will be paid to the participants to share among the
participating providers. Alternatively, providers will be responsible for reimbursing the Medicare
program for expenditures in excess of the target price.
Prospective Bundling – Model 4
Model 4:
Acute Care Hospital Stay Only -- In Model 4, the episode of care includes all services furnished
during the inpatient hospital stay. CMS would make a single, prospectively determined bundled
payment to the hospital that would encompass all of these services. Physicians and other
practitioners would submit “no-pay” claims to Medicare and would be paid by the hospital out of
the bundled payment.
Organizations interested in receiving historical Medicare claims data for use in developing
episode definitions for Model 4 should complete a research request packet, data use agreement,
and data use agreement signature addendum (if needed).
Data will be provided for approved requests before the application is due. Under Model 4, CMS
would make a single, prospectively determined bundled payment to the hospital that would
encompass all services furnished during the inpatient stay by the hospital, physicians and other
practitioners. Physicians and other practitioners would submit “no-pay” claims to Medicare and
would be paid by the hospital out of the bundled payment.
Gainsharing Arrangements: In addition to streamlining care through the use of bundles, the
proposals for this initiative may include gainsharing arrangements. Gainsharing refers to
payments that may be made by hospitals and other providers to physicians and other
practitioners as a result of collaborative efforts to improve quality and efficiency.
3
C. Critical Dates
Organizations interested in applying to the Bundled Payments for Care Improvement initiative
must first submit a non-binding letter of intent and may submit a research request packet to
obtain CMS data. Organizations subsequently submit their application to participate in the
initiative. Deadlines for the submission of the letter of intent, research request packet, and the
application are below.
Model 1
Model 2
Model 3
Model 4
Letter of Intent (non-binding)
9/22/2011
11/4/2011
11/4/2011
11/4/2011
Data Request Packet and
Data Use Agreement
9/22/2011
11/4/2011
11/4/2011
11/4/2011
Application Submitted
10/21/2011
3/15/2012
3/15/2012
3/15/2012
D. Additional Information about Applying for the Bundle Payments
for Care Improvement initiative

Applicants will be required to identify the clinical condition(s) through MS-DRGs, define
the time period for the episode of care, and identify the services included in the bundled
payment, among other criteria. In addition, applicants may also explicitly exclude
conditions from the MS-DRG as well.

Applicants will also be required to plan and implement quality assurance and
improvement activities as a condition of participation in this initiative and participate in
CMS quality monitoring by reporting appropriate quality measures.

During the demonstration, CMS will carefully monitor the program to ensure improved
clinical quality, patient experience, and outcomes of care throughout participation in the
initiative. Applicants will be required to propose strong patient protections that preserve
beneficiary choice in seeking care from the provider of their choice.

Applicants are asked to submit their own episode definitions and bundled payment
proposals. By submitting a research request packet and receiving the necessary CMS
approval, CMS will provide historical Medicare claims data to potential applicants
planning to apply for Models 2-4. The data are intended to enable potential applicants to
develop well-defined episodes and discount proposals based on the experience of
providers in the applicant’s area.

In order to be considered for receipt of data, applicants must submit a Research Study
Protocol along with their letter of intent (LOI) and will later be expected to submit and
comply with a Data Use Agreement (DUA). Both of these forms are available on the
Bundled Payments for Care Improvement website.
E. Helping Members Access CMS Data
Participation in the initiative will require considerable information to understand the opportunities
in your market, identify conditions and design the bundles, determine the upside and downside
risk for various bundling options and opportunities, and so forth. Data from within your
organization as well as key data from the providers that will be partnering with your organization
on the bundling initiative will be critical.
4
To help organizations identify opportunities, develop the bundling proposal, estimate target
prices, determine the discount factor, etc., providers may also request access to Medicare Part A
claims level data from CMS. These data are available for this initiative by completing a research
request packet and entering into a data use agreement with CMS. The data that CMS will make
available are patient-level data, and provide information about a patient’s acute care
hospitalization, post-acute care services, physician services, and durable medical equipment
related costs. Though the data sets are very large and difficult to work with, these data could help
organizations better understand their market, develop bundled payment proposals, estimate
target prices, and determine payment discount factors.
In addition to the data available from your organization, initiative partners, and CMS, AHCA will
be making available aggregate patient stay summary information at the national, state, and
county level for selected MS-DRGs to assist organizations in better understanding opportunities
and the cost of acute care, skilled nursing facility, and other post-acute care services in different
markets.
As part of our bundled payment initiative, AHCA identified seven high-volume, high-cost
conditions that could be candidates for a bundled payment system. These seven conditions
include:
Acute Myocardial Infarction (AMI)
 Heart Failure and Shock
 Lower Extremity Procedures
o Revision of Hip/Knee
o Major Joint Replacement
o Hip/Femur Procedures Not Joint
o Lower Extremity Procedures Not Foot, Etc.
o Local Excision/Removal Interior Fixture Device
 Stroke and Related
 Respiratory
o Respiratory Infections/Inflammations
o Chronic Obstructive Pulmonary Disease (COPD)
o Pneumonia/Pleurisy
 Other Lower Extremity
 Septicemia and Other Post-Operative Infections
o Post-Operative Trauma Infections
o Septicemia or Severe Sepsis With or Without Mechanical Ventilation 96+ Hours
Summary data for these conditions at the county, state and national level will be posted on the
AHCA website the week of October 17, 2011 (See the bundled payment section of the Medicare
web-page under the Facility Operations section of the AHCA website). The summary data are not
intended to be a substitute for data that will be available under data use agreements from CMS
through the CMMI’s Bundled Payments for Care Improvement initiative. The summary data will
show the distribution and average acute care and post-acute care service costs. The summary
data do not, however, include cost information on physician services, outpatient services, and
durable medical equipment that will be necessary to develop a proposal and establish target
prices and discount factors. Use and interpret the summary data carefully.
In addition, AHCA will be releasing a series of memos to assist members in completing the
research request package and data use agreement needed to obtain access to the CMS
Medicare Part A claims data. These memorandums will also be posted on the AHCA website the
week of October 17, 2011.
5
F. CMMI Bundled Payment Initiative Materials
Background materials and forms needed to participate in the CMMI bundled payment initiative
can be found on the CMMI website (http://innovations.cms.gov). These include the Request
for Application, the Letter of Intent, the Research Request Packet, Data Use Agreement (DUA),
DUA Signature Addendum, the Model 2 Application, the Model 2 Application tables, the Model 3
Application, the Model 3 Application tables, and a FAQ. For more information, please send your
questions to BundledPayments@cms.hhs.gov.
APPENDIX
Appendix 1 -- Key features of bundled payment models compared
Appendix 2 -- Details of Model 2
Appendix 3 – Details of Model 3
6
APPENDIX 1
Key Features Of Bundled Payment Models Compared
MODEL 2 –
MODEL 3 -
MODEL 4 –
Post-discharge
Services Only
Inpatient Stay
Only
Inpatient Stay
Only
Physician group
practices
Inpatient Stay
plus Postdischarge
Services
Physician group
practices
Physician group
practices
Physician group
practices
Acute care
hospitals paid
under the IPPS
Acute care
hospitals paid
under the IPPS
Acute care
hospitals paid
under the IPPS
Acute care
hospitals paid
under the IPPS
MODEL
MODEL 1 –
FEATURE
Eligible
Awardees
Health systems
Physicianhospital
organizations
Conveners of
participating
health care
providers
Health systems
Physicianhospital
organizations
Post-acute
providers
Conveners of
participating
health care
providers
Health systems
Long-term care
hospitals
Inpatient
rehabilitation
facilities
Skilled nursing
facilities
Health systems
Physicianhospital
organizations
Conveners of
participating
health care
providers
Home health
agency
Physicianhospital
organizations
Payment of
Bundle and
Target Price
Discounted IPPS
payment; no
separate target
price
Clinical
Conditions
All MS-DRGs
Retrospective
comparison of
target price and
actual FFS
payments
Applicants to
propose based on
7
Conveners of
participating
health care
providers
Retrospective
comparison of
target price and
actual FFS
payments
Applicants to
propose based on
Prospectively set
payment
Applicants to
propose based on
Targeted
Types of
Services
Included in
Bundle
Inpatient hospital
services
MS-DRG for
inpatient hospital
stay
Inpatient hospital
and physician
services
Related postacute care
services
Related
readmissions
Expected
Discount
Provided to
Medicare
To be proposed
by applicant;
CMS requires
minimum
discounts
increasing from
0% in first 6 mos.
to 2% in Year 3
Payment
from CMS to
Providers
Acute care
hospital: IPPS
payment less
pre-determined
discount
Quality
Measures
Physician:
Traditional fee
schedule
payment (not
included in
episode or
subject to
discount)
All Hospital IQR
measures and
additional
measures to be
proposed by
applicants
Other services
defined in the
bundle
To be proposed
by applicant; CMS
requires minimum
discount of 3% for
30-89 days postdischarge
episode; 2% for
90 days or longer
episode
Traditional fee-forservice payment
to all providers
and suppliers,
subject to
reconciliation with
predetermined
target price
MS-DRG for
inpatient hospital
stay
Post-acute care
services
Related
readmissions
MS-DRG for
inpatient hospital
stay
Inpatient hospital
and physician
services
Related
readmissions
Other services
defined in the
bundle
To be proposed
by applicant
To be proposed
by applicant;
subject to
minimum discount
of 3%; larger
discount for MSDRGs in ACE
Demonstration
Traditional fee-forservice payment
to all providers
and suppliers,
subject to
reconciliation with
predetermined
target price
Prospectively
established
bundled payment
to admitting
hospital; hospitals
distribute
payments from
bundled payment
To be proposed by applicants, but CMS will ultimately
establish a standardized set of measures that will be aligned
to the greatest extent possible with measures in other CMS
programs
8
Appendix 2
Details of Model 2
Model 2: Retrospective Bundled Payment Models For Hospitals, Physicians, And PostAcute Providers For An Episode Of Care Consisting Of An Inpatient Hospital Stay
Followed By Post-Acute Care , August 29, 20111
This model extends the episode of care beyond the acute care inpatient hospitalization to include
post-acute care following and associated with the acute care episode. This model uses typical
FFS payment with retrospective payment reconciliation against the predetermined target price for
the episode.
Applicants will have two options under this model based on the length of the episode of the care.
In both options, the episode anchor is an acute care hospital admission for an agreed-upon MSDRG. All beneficiaries admitted to an awardee or Bundled Payment participating provider (acute
care hospital) for agreed-upon MS-DRGs will be included in the episode.
The Episode
The episode will begin with the inpatient hospital admission to a participating provider and will
continue through a minimum of 30 days following discharge from the hospital.
The episode will include all hospital diagnostic testing and all related therapeutic services
furnished by an entity wholly owned or wholly operated by the admitting hospital in the three days
prior to hospital admission, Part A and Part B services that are furnished during the hospital stay,
and Part A and Part B services in the post-discharge period related to the episode anchor.
All Part A services for related readmissions and all related Part B services furnished during the
post-discharge period including during related and unrelated readmissions must be included in
the episode in both options that are described below.
Options



Under the first option, applicants may propose an episode that extends 30 to 89 days
following the hospital discharge.
Under the second option, applicants must propose an episode that extends 90 days or
longer following the hospital discharge.
The applicant will propose the length of the episode and need not propose a strict prorated episode definition (for example, services starting within the episode time period and
continuing past the episode conclusion, based on days following hospital discharge,
could still be included in the episode).
Further Definitions of the Episode
Applicants under both options will be expected to propose further definitions of the episode,
including:


Beneficiary identification (through MS-DRGs)
Excluded unrelated Part A services such as certain readmissions (identified by MS-DRGs
designated as unrelated), and
1
I have taken this information from Bundled Payments for Care Improvement Initiative Request for
Application; Center for Medicare and Medicaid Innovation, pages 13-16 and put it in outline form.
9

Excluded unrelated Part B services (identified by principal ICD-9 diagnosis codes
designated as unrelated).
Physicians’ services furnished throughout the episode period and post-acute services related to
the episode anchor and furnished during the episode period must be included in the episode
under either option.
Discounts

Applicants under the first option must offer a minimum 3% discount off of all included MSDRGs and other Part A and Part B services within the episode.

Applicants under the second option must offer a minimum 2% discount off of all included
MS- DRGs and other Part A and Part B services within the episode. The lower minimum
level of discount under the second option is meant to balance the increased financial risk
(due to the longer episode) under that option.

CMS encourages applicants to propose an episode with a longer period post-hospital
discharge, because CMS is interested in understanding how redesign care extends to a
beneficiary’s transition back into the community.

Applicants should factor expected Medicare outlier payments into financial models when
proposing a target price that reflects a discount. CMS will consider applicant proposals
using risk adjustment which include a description of the methodology.
Risk Adjustment
A key issue that CMS will critically assess when considering risk adjustment models that
incorporates diagnosis data is that risk scores can be affected not just by changes in the health
status of the population or patient but also by changes in coding intensity and by the mix of
specialists and other providers furnishing services.

The experience in Medicare Advantage shows that health plans can significantly increase
the complexity score of their populations by focusing on more complete coding.

Similarly, CMS experience with the Physician Group Practice demonstration shows that
participating awardees have an incentive to code more fully or intensely because of the
potential impact on performance payments and to provide more accurate measurement
and reporting of quality measures, as well as to provide more complete and accurate
information that can be used for population management.
Target Price
Applicants will be expected to propose a target price for the episode that includes a single rate of
discount on the expected Medicare payments for all included Part A and Part B services.

With respect to the inpatient hospital payment considered in the target price, this
proposed target price should consider the base MS-DRG payment accounting for all
payment adjustors and applicable outlier payments, except disproportionate share
hospital (DSH) payments, hospital capital, and indirect medical education (IME)
payments, for either option.

Over the model years, the negotiated discount reflected in the target price will remain
constant, while we index the target price each year to FFS payment changes as the
systems are updated (positively or negatively) annually according to the applicable
10
standard IPPS, PFS, and post-acute provider prospective payment system updates and
other adjustments that apply.

CMS will give preference to applications that offer a greater discount to Medicare, in the
context of a robust programmatic design that ensures high quality care for beneficiaries.
All beneficiaries eligible for the episode (based on MS-DRG) admitted to the awardee or its
Bundled Payment participating providers (acute care hospitals must be included in the bundled
payment model under either option.
Financial Liability of Awardees
While hospitals, physicians, and post-acute providers are encouraged to engage with each other
as participating partners in the episode of care, CMS indicates that applicants should recognize
that awardees will be financially liable for Medicare payment in aggregate beyond the
predetermined target price, including care for included beneficiaries that is furnished by providers
who are not participating in testing the bundled payment model.

No outlier payments beyond the usual FFS outlier payments that would have been paid
for qualifying individual cases will be made above the agreed-upon target price for
catastrophic cases at reconciliation.

Awardees may not restrict beneficiary choice of provider and must notify beneficiaries of
their participation in this initiative.
Applicants will be expected to provide evidence about how they would ensure beneficiaries have
complete freedom of choice of providers, including post-acute providers. Payments to all
physicians and other practitioners who provide care to included beneficiaries during the episode
will be considered in the retrospective reconciliation for the episode.
Gainsharing
In this model, awardees may gainshare with all providers treating patients during the episode,
although participation in the gainsharing element of the payment model (if the proposal includes
gainsharing) must be voluntary. Given this, applicants will be expected to provide evidence of
active participation by physicians and post-acute providers in the initiative, including evidence of
disclosure to physicians that the hospital and post-acute providers are participating.
Payment
Under this payment model, claims for all services will continue to be processed under the relevant
IPPS, PFS, and post-acute payment system rules under either option.
There will be a regular retrospective reconciliation against the predetermined target price. If
aggregate FFS payments for included services exceed the predetermined target price, the
awardee must repay Medicare.

If aggregate FFS payments for included services are less than the predetermined price, the
awardee will be paid the difference, which may be shared among the participants.

All payments for the episode of care, including outlier payments, will be included in the
episode reconciliation of actual payments made against the target price.

The awardee bears full risk for any expenditures beyond the target price of the episode.
11
Care and Expenditure Measurement
The CMS will measure and monitor care throughout to ensure project objectives are met in
redesigning care, meeting quality and patient experience of care standards, and demonstrating
improved care coordination.
The CMS or its contractor will measure care provided during a 30 day post-episode monitoring
period to ensure the aggregate Medicare Part A and Part B spending for included beneficiaries
does not increase as a result of this initiative. This will include measuring expenditures for
included beneficiaries at non-participating providers.

Aggregate Medicare Part A and Part B expenditures for included beneficiaries during the
post-episode monitoring period will be compared to a historical baseline payment that
has been trended over time and which will include a risk threshold. If spending exceeds
the risk threshold, the awardee must pay Medicare for the excess.

Beneficiaries who receive care from Model 2 participants may benefit from increased
communication and coordination between their treating providers, improved hospital
discharge and facility transfer planning, fewer re-operations, fewer avoidable
readmissions, more appropriate post-acute care, higher quality of care throughout the
episode, and shorter average lengths of stay in the acute care hospital and in post-acute
care facilities.
12
Model 2: Retrospective Acute Care Hospital Stay plus Post-Acute Care *
Entities eligible to be awardees:
• Acute care hospital.
• Health systems.
• Post-acute providers.
• Physician hospital organizations.
• Physician group practices.
• Conveners of participating health care providers.
Episode definition
Criteria for beneficiary inclusion in episode:
• Organized around reason for hospitalization
(MS-DRG).
• Exact identification criteria to be proposed.
Episode anchor:
• Acute care hospital admission at awardee or
Bundled Payment participating organization for
included clinical conditions (identified via MSDRG).
End of episode:
• Option 1: Minimum 30 days post-hospital
discharge; maximum of 89 days post-hospital
discharge.
• Option 2: Minimum 90 days post-hospital
discharge.
Types of services included in bundle:
• Physicians’ services.
• Inpatient hospital services (episode anchor).
• Inpatient hospital readmission services.
• Long term care hospital services (LTCH).
• Inpatient rehabilitation facility services (IRF).
• Skilled nursing facility services (SNF).
• Home health agency services (HHA).
• Hospital outpatient services.
• Independent outpatient therapy services.
• Clinical laboratory services.
• Durable medical equipment.
• Part B drugs.
Payment from CMS to providers:
• Traditional FFS (ultimate reconciliation with
predetermined target price).
Expected discount provided to Medicare:
• Option 1: Minimum 3% discount on included
Part A and Part B allowed charges for episodes
that include a post-hospital discharge period of 30
days to 89 days.
• Option 2: Minimum 2% discount on included
13
Part A and Part B allowed charges for episodes
that include a post-hospital discharge period of 90
days or longer.
• Exact discount rate to be proposed under either
option.
Reconciliation, spending calculation,
disbursement, and post-episode monitoring
period:
• Episode reconciliation:If aggregate FFS
payments for included services during the
episode are less than the predetermined target
price, Medicare will pay the difference to
awardee. If aggregate FFS payments for included
services during the episode exceed the
predetermined target price, awardee must repay
Medicare.
• Post-episode monitoring:Medicare Part A and
Part B payment for included beneficiaries during
the post-episode monitoring period that exceeds
trended historical aggregate Part A and Part B
payment beyond a risk threshold will be paid by
the awardee to Medicare.
Post-episode monitoring period:
• 30 days following the end of the episode.
Gainsharing;
Other payment arrangements between Bundled
Payment participating organizations:
• To be proposed.
• To be proposed.
Quality measures:
• To be proposed, but a standardized set will
ultimately be required and agreed upon by CMS
and the awardee. These measures will be aligned
with other CMS programs to the greatest extent
possible.
14
Apendix 3
Details of Model 3
Model 3: Retrospective Bundled Payment Models For Post-Acute Care Where The Bundle
Does Not Include The Acute Inpatient Hospital Stay
Elise Smith (esmith@ahca.org), August 29, 2011 2
This model will test the potential for reducing Medicare expenditures and improving quality via
bundled payment for an episode of care consisting of post-acute care following an acute inpatient
hospital stay, but where the initial inpatient hospital stay is not included in the episode.
The Episode

The episode anchor is the initiation of post-acute care services at a skilled nursing facility,
inpatient rehabilitation facility, long-term care hospital, or with a home health agency
within 30 days of beneficiary discharge from an acute care hospital for an agreed-upon
MS-DRG.

All beneficiaries who initiate post-acute care services with an awardee or Bundled
Payment participating organization of the type specified previously will be included in the
Model 3 episode.

The episode will begin on the date post-acute services are initiated with an awardee or
Bundled Payment participating organization and will continue through a minimum of 30
days following initiation of the episode.

The episode must include all related Part A and Part B services furnished during the
episode period, including related readmissions.

All Part A services for related readmissions and all related Part B services (e.g.,
physicians’ services and post-acute services related to the episode anchor) furnished
during the episode period, including during related and unrelated readmissions, must be
included in the episode.
Further Definitions of the Episode
Applicants will be expected to propose further definitions of the episode, including:



Beneficiary identification(through MS-DRGs),
Length of the episode, excluded unrelated Part A services such as certain readmissions
(identified by MS-DRGs designated as unrelated), and
Excluded unrelated Part B services (identified by principal ICD-9 diagnosis codes
designated as unrelated).
CMS will give preference to applications that propose an episode definition longer than 30 days
because CMS is interested in understanding how care redesign extends to the beneficiary’s
transition back into the community.
2
I have taken this information from Bundled Payments for Care Improvement Initiative Request for
Application; Center for Medicare and Medicaid Innovation, pages 17-19and put it in outline form.
15
Target Price and Discounts
Applicants will be expected to propose a target price for the episode that includes a single rate of
discount off of the expected Medicare payments for all included services.
Applicants should factor expected readmissions during the episode period into their financial
proposals. In addition, applicants should factor expected outlier payments into their financial
models when proposing a target price that reflects a discount.
Over the model years, the negotiated discount reflected in the target price will remain constant,
while we index the target price each year to FFS payment changes as the systems are updated
(positively or negatively) annually according to the applicable standard PFS and post-acute
provider prospective payment system updates and other adjustments that apply.
Risk Adjustment
CMS will consider applicant proposals around risk adjustment which must include a description of
the methodology and plans for updating risk adjustment on a yearly basis based on new
information.
A key issue that CMS will critically assess when considering risk adjustment models that
incorporates diagnosis data is that risk scores can be affected not just by changes in the health
status of the population or patient but also by changes in coding intensity and by the mix of
specialists and other providers furnishing services.

The experience in Medicare Advantage shows that health plans can significantly increase
the complexity score of their populations by focusing on more complete coding.

Similarly, CMS experience with the Physician Group Practice demonstration shows that
participating awardees have an incentive to code more fully or intensely because of the
potential impact on performance payments and to provide more accurate measurement
and reporting of quality measures, as well as to provide more complete and accurate
information that can be used for population management.
Discounts
The CMS will give preference to applications that offer highly competitive discounts to Medicare,
in the context of a robust programmatic design that ensures high quality care for beneficiaries.
Financial Liability of Awardees

All beneficiaries eligible for the episode initiating post-acute care services with the
awardee or its Bundled Payment participating providers (SNF, IRF, LTCH, or HHA) must
be included in the episode payment model.

While post-acute providers and physicians are encouraged to engage with each other as
partners in the episode, applicants should recognize that awardees will be financially
liable for Medicare payment in aggregate beyond the predetermined target price,
including care for included beneficiaries that is provided by providers who are not directly
participating in testing the episode payment model.

No outlier payments beyond the usual FFS outlier payments that would have been paid
for qualifying individual cases will be made above the agreed-upon target price for
catastrophic cases at reconciliation. Awardees may not restrict beneficiary choice of
16

provider and must notify beneficiaries of their participation in this initiative. Applicants will
be expected to provide evidence about how they would ensure beneficiaries have
complete freedom of choice of providers, including post-acute providers.
Payments
Payments to all physicians and other practitioners who provide care to included beneficiaries
during the episode will be considered in the retrospective reconciliation for the episode.

Under this payment model, all claims for all services will continue to be processed under
the relevant physician, post-acute provider, and other provider and supplier payment
systems and rules.

There will be a regular retrospective reconciliation against the predetermined target price.
If aggregate FFS payments for included services exceed the predetermined target price,
the awardee must repay Medicare.

If aggregate FFS payments for included services are less than the predetermined price,
the awardee will be paid the difference, which may be shared among the participants.

All payments for the episode of care, including outlier payments, will be included in the
episode reconciliation of actual payments made against the target price. The awardee
bears full risk for any expenditures beyond the target price of the episode.
Gainsharing
In this model, awardees may gainshare with all providers treating patients during the episode,
although participation in the gainsharing element of the payment model (if the proposal includes
gainsharing) must be voluntary. Given this, applicants will be expected to provide evidence of
active participation by physicians and other practitioners in this initiative.
Care and Expenditure Measurement
The CMS will measure and monitor care throughout to ensure project objectives are met in
redesigning care, meeting quality and patient experience of care standards, and demonstrating
improved care coordination.

The CMS or its contractor will measure care provided during a 30 day post-episode
monitoring period to ensure the aggregate Medicare Part A and Part B spending for
included beneficiaries does not increase as a result of this initiative.

This will include measuring expenditures for included beneficiaries at non-participating
providers. Aggregate Medicare Part A and Part B expenditures for included beneficiaries
during the post-episode monitoring period will be compared to a historical baseline
payment that has been trended over time and which will include a risk threshold.

If spending exceeds the risk threshold, the awardee must pay Medicare for the excess.

Beneficiaries who receive care from Model 3 participants may benefit from increased
communication and coordination between their treating providers, improved hospital
discharge and facility transfer planning, fewer avoidable readmissions, higher quality
post-acute care, and shorter average lengths of stay in post-acute care facilities.
17
Model 3: Retrospective Post-Acute Care Only *
Entities eligible to be awardees:
• Physician group practices.
• Acute care hospitals.
• Health systems.
• Long term care hospitals (LTCH).
• Inpatient rehabilitation facilities (IRF).
• Skilled nursing facility (SNF).
• Home health agency (HHA).
• Physician hospital organizations.
• Conveners of participating health care
providers.
Episode definition
Criteria for beneficiary inclusion in episode:
• Organized around reason for hospitalization
(MS-DRG).
• Exact criteria to be proposed.
Episode anchor:
• Initiation of SNF, IRF, HHA, or LTCH services
with awardee or Bundled Payment participating
organization within 30 days following discharge
from an acute care inpatient hospital for an
included MS-DRG.
End of episode:
• Minimum 30 days following the episode
anchor.
• Exact duration to be proposed.
• Physicians’ services.
Types of services included in bundle:
• Inpatient hospital readmission services.
• Long term care hospital services (LTCH).
18
• Inpatient rehabilitation facility services (IRF).
• Skilled nursing facility services (SNF).
• Home health agency services (HHA).
• Hospital outpatient services.
• Independent outpatient therapy services.
• Clinical laboratory services.
• Durable medical equipment.
• Part B drugs.
Payment from CMS to providers:
• Traditional FFS (ultimate reconciliation with
predetermined target price).
Expected discount provided to Medicare:
• To be proposed.
Reconciliation, spending calculation,
disbursement, and post-episode monitoring
period:
• Episode reconciliation: If aggregate FFS
payments for included services during the
episode are less than the predetermined target
price, Medicare will pay the difference to
awardee. If aggregate FFS payments for
included services during the episode exceed
the predetermined target price, awardee must
repay Medicare.
• Post-episode monitoring:Medicare Part A and
Part B payment for included beneficiaries
during the post-episode monitoring period that
exceeds trended historical aggregate Part A
and Part B payment beyond a risk threshold will
be paid by the awardee to Medicare.
Post-episode monitoring period:
• 30 days following the end of the episode.
Gainsharing;
• To be proposed.
Other payment arrangements between
Bundled Payment participating
• To be proposed.
19
organizations:
• To be proposed, but a standardized set will
ultimately be required and agreed upon by
CMS and the awardee. These measures will be
aligned with other CMS programs to the
greatest extent possible.
Quality measures:
20