Medicaid Reimbursement Policy

CRS Report for Congress
Medicaid Reimbursement Policy
Updated November 2, 2004
Mark Merlis
Contractor to the Congressional Research Service
Domestic Social Policy Division

Congressional Research Service ˜ The Library of Congress

Medicaid Reimbursement Policy
Under Medicaid law, states have considerable freedom to develop their own
methods and standards for reimbursement of Medicaid services. Congress has
periodically intervened to modify the broad guidelines within which states operate,
and the Centers for Medicare and Medicaid Services (CMS) has used its regulatory
authority to restrict certain state practices. Actual payment methodologies, however,
are still left largely to the discretion of the states.
Medicaid reimbursement policies play a central role in determining whether
beneficiaries have access to services of adequate quality, as well as the nature of the
services they receive. Because some providers, such as “safety-net” hospitals and
clinics and nursing facilities are heavily dependent on Medicaid funding, payment
levels can have broad effects on the delivery system and community access to care.
In addition, because Medicaid is a major component of state and federal spending,
decisions about reimbursement policies can have significant budgetary effects.
For both hospitals and nursing homes, Medicaid payment rates in many states
are below the actual costs facilities incur in providing care to Medicaid patients.
Payment rates for other kinds of providers, such as physicians or dentists, cannot be
directly compared to costs; however, Medicaid is often paying less for comparable
services than Medicare or private insurers. Medicaid payment shortfalls have a
variety of possible consequences. Providers may engage in “cost-shifting,” raising
charges to private payers to make up their losses. In addition, the need to subsidize
Medicaid patients may reduce their ability to fund care for people with no coverage
at all. Some providers may adopt cost-cutting measures that potentially affect
quality. Others may refuse to accept Medicaid patients or limit the number they will
treat, since Medicaid law has no requirement prohibiting providers from doing so.
This report provides a snapshot captured primarily through state plan
amendments approved through November 2002, of the methods states use to
establish payment rates for most major types of providers serving Medicaid clients.
It also explores some of the critical issues affecting Medicaid payments rate setting.
Where available, Medicaid rates are compared to other payers such as Medicare and
private insurance. This report will not be updated.
To assist Congress to review policy alternatives and understand the current
status of Medicaid programs, the Congressional Research Service (CRS) is producing
a series of reports on various aspects of Medicaid. This report is one in that series.
This series will address Medicaid programs and policies comprehensively by
covering background subjects including eligibility policy, benefits, and delivery
systems and demonstration projects as well as analytic reports such as Medicaid’s
role for low-income individuals, long-term care, and dual eligibles. Each of the
reports includes a discussion of current issues, background information, data and

In troduction ......................................................1
Organization of This Report.....................................1
Use of State Plan Documents.....................................2
Overview ........................................................3
Basic Federal Rules............................................3
Major Policy Developments, 1980-2003............................4
The Boren Amendment.........................................4
OBRA 81 Waivers.............................................5
Prescription Drug Rebates.......................................6
Disproportionate Share Hospital (DSH) Payments,
Provider Donations, and Provider Taxes........................7
Growth in Managed Care........................................8
Upper Payment Limits (UPLs)....................................9
State Fiscal Problems and Medicaid Cost Containment...............14
Acute Care......................................................15
Hospital Services.............................................15
Inpatient Payment Methods.....................................16
Hospital-Specific Rates....................................22
Peer Group or Statewide Rates..............................22
Use of Case Mix.........................................22
Other Methods...........................................23
Administrative Days/Swing Beds............................24
Outpatient Payment Methods....................................25
DSH Payments...............................................28
Current DSH Requirements.................................28
Amount of DSH Payments..................................32
Adequacy of Hospital Reimbursement............................43
Physician and Dental Care......................................47
Physician Payment........................................48
Dental Payment..............................................57
Federally Qualified Health Centers and Rural Health Clinics...........59
Long-Term Care..................................................61
Nursing Facilities.............................................61
Payment Methods.........................................61
Payment Levels and Adequacy..............................70
Intermediate Care Facilities for the Mentally Retarded................75
Home and Community-Based Services............................81
Non-Waiver Home and Personal Care.........................82
Home and Community-Based Services Waivers.................85
Personal Care............................................85
Case Management........................................87
Budgeted or Bundled Payments..............................88
Managed Care...................................................89
Rate-Setting Methods..........................................90

Reinsurance, Risk Sharing, and Incentive Payments..............92
Risk Adjustment.........................................94
Payment Levels..............................................96
Prescription Drugs................................................99
Pharmacy Reimbursement Methods..................................99
Upper Payment Limits.........................................99
Multiple Source Drugs.....................................99
Other Drugs............................................100
Dispensing Fees and Ingredient Costs............................100
Dispensing Fees.........................................100
Ingredient Cost..........................................100
Drug Rebate Requirements....................................103
Single Source and Innovator Multiple Source Drugs............104
Non-innovator Multiple Source Drugs.......................104
Recent State Initiatives........................................107
Supplemental Rebates....................................107
Pharmacy Plus..........................................107
Purchasing Pools........................................108
Cost Containment.......................................108
Other Payment Requirements......................................109
Federal Rules for Specified Services.............................109
Home and Community-Based Care Option....................109
Hospice Services........................................109
Indian Health Service.....................................110
Laboratory Services......................................110
Programs of All-Inclusive Care for the Elderly (PACE)..........110
Volume Purchasing......................................111
Coordination with Medicare...................................111
List of Tables
Table 1. Selective Contracting Waivers, 2003...........................6
Table 2. Effect of a Typical Provider Donation or Tax Program.............7
Table 3. Medicaid Beneficiaries and Medicaid Managed Care Arrangements,
June 2002....................................................9
Table 4. Typical Enhanced Payment Program...........................10
Table 5. Transition Periods for Compliance with Upper Payment Limits.....12
Table 6. State Enhanced Payment Programs by Provider Type
and Preliminary Transition Period in Years ........................13
Table 7. Number of States Undertaking Medicaid Cost Containment
Strategies, FY2002-FY2004....................................14
Table 8. Number of States Planning Rate Changes
for Selected Services, FY2004...................................15
Table 9. Basic Payment Methodology, Inpatient Hospital Services, 2002.....18
Table 10. Principal Outpatient Hospital Reimbursement Approach..........26
Table 11. Federal DSH Allotments for 1998-2003.......................30
Table 12. Disproportionate Share Hospital Payments, as a Share
of Total Hospital Payments and Total Net Medicaid Spending, 2001.....33

of Hospital and Hospital Ownership, Most Recent Reporting Year......37
Table 14. Hospitals Receiving Disproportionate Share Hospital Payments
by Type of Hospital and Hospital Ownership,
Most Recent Reporting Year....................................40
Table 15. Hospital Payment-to-Cost Ratios, by Source of Revenue,
1991-2001 ..................................................44
Table 16. Estimated Costs and Revenues, Medicaid
and Self-Pay/Other Patients, NAPH Member Hospitals, 2000..........46
Table 17. Medicaid Payment Rates for Selected Physician Procedures,
2001 .......................................................49
Table 18. Medicaid Payment Rate as a Percentage
of Medicare Physician Fee Schedule, 2001.........................53
Table 19. Survey of Pediatricians on Medicaid Participation, 2000..........57
Table 20. Medicaid Fees, 2003, and Median Private Fees, 2002,
for Selected Dental Procedures..................................58
Table 21. Payment Methodologies for Nursing Facility Direct Care
Component, 2002.............................................63
Table 22. Summary of State Wage Pass-Through Programs...............69
Table 23. Average Medicaid Shortfall Per Day, Medicaid Nursing Facility
Payments in Responding States, 1999 and 2000.....................71
Table 24. Change in Daily Medicaid Nursing Facility Payment Rates
and Daily Costs, 1999-2000.....................................73
Table 25. Medicaid Daily Nursing Facility Charges and Payment Rates,
1999 .......................................................74
Table 26. ICF-MR Residents at End of Year by Facility Size
and Ownership, 1977 and 2002..................................76
Table 27. Basic Medicaid Payment Method, Direct Care Component,
Non-State Intermediate Care Facilities for the Mentally Retarded,
2002 .......................................................77
Table 28. Medicaid Spending for Home and Community Care, FY2002.....81
Table 29. Payment Methods for Non-Waiver Home Health Care
and Personal Care Services, January 2003..........................82
Table 30. Principal Approach to MCO Rate-Setting, 2001................90
Table 31. Factors in Capitation Payment, 2002.........................92
Table 32. Reinsurance and Risk-Sharing Arrangements, 2002.............93
Table 33. Risk Adjustment Systems..................................95
Table 34. Change in Medicaid Managed Care Payment Rates,
Section 1931 and Poverty-Related Groups, Selected States,
1998-2001 ..................................................96
Table 35. Commercial and Medicaid-Only MCO Plans
and Enrollment, 1998-2002.....................................98
Table 36. Pharmacy Dispensing Fees
and Ingredient Reimbursement Basis, 2002........................101
Table 37. Effect of Rebates on Medicaid Drug Spending, FY2001.........105
Table 38. Number of States Making Medicaid Prescription Drug Policy
Changes, FY2003 and FY2004.................................109
Table 39. Medicaid Benefits for Low-Income Medicare Beneficiaries......112
Table 40. Medicaid Payment Policies for Medicare Cost-Sharing..........113

Medicaid Reimbursement Policy
Under Medicaid law, states have considerable freedom to develop their own
methods and standards for reimbursement of Medicaid services. Congress has
periodically intervened to modify the broad guidelines within which states operate,
and the Centers for Medicare and Medicaid Services (CMS) has used its regulatory
authority to restrict certain state practices. Actual payment methodologies, however,
are still left largely to the discretion of the states.
Medicaid reimbursement policies play a central role in determining whether
beneficiaries have access to services of adequate quality, as well as the nature of the
services they receive. In addition, because some providers — such as “safety-net”
hospitals and clinics and nursing facilities — are heavily dependent on Medicaid
funding, payment levels can have broader effects on the delivery system and
community access to care. Finally, because Medicaid is a major component of state
and federal spending, decisions about reimbursement policies can have significant
budgetary effects.
Organization of This Report
This report1 begins with a summary of basic federal requirements applicable to
payments for all services and an overview of major developments in federal Medicaid
reimbursement policy over the last 20 years. This overview provides a historical
context for current policies and highlights some issues that have been perennial
concerns for federal and state policymakers.
The next four sections of the report provide a detailed discussion of Medicaid
reimbursement for four basic categories of services or providers:
!Acute care, including hospital inpatient and outpatient services,
services of physicians and dentists, and services of certain federally
defined categories of health centers and clinics;
!Long-term care, including care in nursing facilities, intermediate
care facilities for the mentally retarded (ICFs-MR), and home and
community-based care;

1 The CRS project liaison for this report is Jean Hearne, Specialist in Social Legislation in
the Domestic Social Policy Division. She can be reached at extension 7-7362.

!Managed care organizations (MCOs), which accept financial
responsibility for a range of covered services in return for a fixed
monthly payment per Medicaid enrollee; and
!Prescription drugs.
For each service type, these sections summarize states’ payment methodologies,
review current or recent policy issues, and, to the extent data are available, compare
Medicaid payments to providers’ costs or to payments by other third parties.
The final section of the report describes special federal payment rules for some
specific classes of providers and explains how Medicaid payments coordinate with
Medicare for individuals eligible for benefits under both programs.
Use of State Plan Documents
Most of the state-by-state comparisons of payment methodologies in this report
are based on Medicaid state plans and state plan amendments (SPAs). The state plan
for medical assistance is the basic document each state initially submitted in order
to obtain approval of its Medicaid program. Major policy changes are reflected in
SPAs that must also be approved by CMS. SPAs can be approved retroactively,
meaning that a state can implement a policy before CMS has acted on its submission
(at the risk of a denial of federal funding if the SPA is ultimately disapproved).
CMS maintains a database of state plans and SPA documents on its Web site.2
Full state plans were captured in late 2000, with subsequent plan amendments added
to the database as approved. SPAs reviewed for this document include all those
approved through November 7, 2002. What this report provides, then, is a snapshot
of payment methodologies under each state plan as approved on that date. These will
not necessarily be the methodologies actually in use in November 2002, because
approval of amendments can be retroactive.
Some state plan documents relating to reimbursement methods are lengthy and
complex, and some states have filed numerous SPAs that repeatedly modify the same
sections of the plan. While every effort has been made to track the changes and
identify the most current approved policy, there are undoubtedly errors or omissions.
In a very few cases, it was impossible to ascertain a state’s policy for a particular
service from the state plan, and state regulations or other documents were consulted.
In these cases, which are identified in notes to the tables, the policy described may
be the one in effect at the time the state documents were obtained, rather than in
November 2002.

2 The database can be accessed at [].

Basic Federal Rules
Three basic federal statutory requirements apply to payment for all types of
!Methods and procedures for making payments must be such as to
assure that payments are “consistent with efficiency, economy, and
quality of care.”
CMS relies on this provision as a general authority to regulate state
reimbursement methodologies. In particular, this provision is the basis for the upper
payment limit (UPL) regulations, which require that Medicaid payments for a class
of institutional providers not exceed, in the aggregate, the amount that would have
been paid for comparable services under Medicare principles. Recent revisions in the
UPL rules have had a major effect on state finances; this issue is discussed further
!Payments must be “sufficient to enlist enough providers so that care
and services are available under the plan at least to the extent that
such care and services are available to the general population in the
geographic area.”
This provision explicitly connects the level of Medicaid payment rates with the
willingness of providers to serve Medicaid beneficiaries. While payment levels are
not the only factor affecting provider participation, there has been a tension between
cost containment and access to care throughout the history of Medicaid.
!Providers must accept Medicaid reimbursement as payment in full,
except for any beneficiary cost-sharing amounts provided for by the
state plan or any amount due from a medically needy beneficiary
with a spend-down liability.3
This means that a provider cannot bill a beneficiary when Medicaid’s allowed
payment is less than the provider’s charge for a service. In contrast, Medicare allows
limited balance billing by physicians and some other providers. Private insurance
rules vary; plans with networks commonly restrict balance billing by network
providers and permit it for out-of-network services.
There is an additional set of basic rules for payment of institutional services,
including hospitals, nursing facilities, and intermediate care facilities for the mentally
retarded (ICFs-MR). Rates must be determined through a public process. States
must publish proposed and final rates, including justifications and underlying

3 A Medicaid applicant who is in a state providing optional coverage of the medically needy
population and whose income or resources exceed the limits established by the state may
“spend down” to eligibility by using the excess funds to pay medical bills.

methodologies; and providers, beneficiaries, and the public must be given an
opportunity to comment.
Beyond these general rules, actual payment requirements or methodologies are
prescribed by law for only a few types of providers, such as disproportionate share
hospitals (DSHs, those serving a high proportion of low-income patients), federally
qualified health centers (FQHCs, which are Public Health Service grantees and
similar entities), and hospices. There are also specific rules relating to payment for
prescription drugs. All of these rules are described in later sections of this report.
Major Policy Developments, 1980-2003
Over time, federal Medicaid reimbursement policy has focused on different, and
sometimes conflicting, policy goals, such as cost containment, state flexibility, and
access to care. Congress has set specific minimum or maximum levels of
reimbursement for some types of services, while providing only general guidelines
for others. It has sought to foreclose some payment schemes that have the effect of
shifting financial burdens from states to the federal government. It has acted to
protect some specific classes of providers, while enhancing states’ ability to bargain
with others.
This section provides a brief overview of major developments in Medicaid
reimbursement policy over the last two decades. It is not meant to be a legislative
history, but merely to highlight key issues and some of the shifts in congressional
priorities and concerns.
The Boren Amendment
Until 1980, state Medicaid programs were required to follow Medicare
reimbursement principles in paying institutional providers — hospitals and nursing
facilities. Under the Medicare rules in effect at that time, this meant that states were
required to use a retrospective reasonable cost system. States continued to have to
assure that rates provided access to care. Payment amounts were determined after
services were rendered and were based on the actual costs incurred by the provider
in furnishing those services. In what is known as the “Boren amendment,” the
Omnibus Reconciliation Act of 1980 (P.L. 96-499) repealed this requirement for
nursing facility services, freeing states to establish new methodologies of their own.
The Omnibus Budget Reconciliation Act of 1981 (OBRA 81, P.L. 97-35) applied the
amendment to inpatient hospital services.
The new rules provided simply that payment rates for hospitals and nursing
facilities had to be “reasonable and adequate” to meet the costs of “efficiently and
economically operated” facilities. For hospitals, the law also required payment
adjustments for disproportionate share hospitals (DSHs). Nearly all states responded
to the new flexibility by shifting from retrospective to prospective payment systems
for both hospital and nursing facility services. Under prospective payment systems,
rates may be set in advance and may not be related to the actual costs providers incur
in furnishing services; or the state may set ceilings and pay the lesser of actual costs
or the ceiling amount. States’ interest in these systems stemmed from concerns that

providers paid on a full cost basis had no incentive to perform efficiently and might
furnish unnecessary services.
While the Boren amendment gave states the flexibility to develop new payment
systems, it also established a benchmark against which those systems were to be
measured: the state was required to find, and to provide assurances satisfactory to
the Secretary, that its Medicaid rates were reasonable and adequate. In 1990, the
Supreme Court affirmed that facilities had a right to seek judicial review of the
reasonableness and adequacy of Medicaid rates (Wilder v. Virginia Hospital
Association, 496 U.S. 498, 1990). The Wilder decision merely settled the question
of whether the Boren amendment conferred rights on providers that could be
enforced in court. Even before this decision, hospitals in some states had obtained
court judgments that Medicaid payments were inadequate. Following Wilder,
numerous states faced suits by hospitals and nursing homes. Congress ultimately
responded by repealing the “reasonable and adequate” test in the Balanced Budget
Act of 1997 (BBA, P.L. 105-34). Some hospitals have continued to file suits, relying
on the requirement, still in the statute, that payments for all types of providers be
sufficient to assure access to care.
OBRA 81 Waivers
OBRA 81 authorized the Secretary to waive specified requirements of Medicaid
law so that states could operate innovative service programs. Two types of waivers
were originally permitted: Section 1915(b) freedom-of-choice waivers, under which
states could require beneficiaries to obtain services through a primary care case
manager or a managed care plan, or from a limited set of contracting providers; and
Section 1915(c) home and community-based services waivers, under which states
could provide special services (generally non-medical personal care and supportive
services) to limited populations of beneficiaries who would otherwise need
institutional care.4 Both types of waivers require periodic CMS approval and are
subject to cost-effectiveness tests. Congress has since authorized several other
waiver options.
A number of states have used Section 1915(b) freedom-of-choice waivers to
operate selective contracting systems, under which beneficiaries needing a specified
service are restricted to a limited set of providers whose payment rates are established
by bidding or negotiation. Table 1 lists the selective contracting programs in effect
as of September 2003.

4 The waivers are commonly referred to by the sections of the Social Security Act that set
rules for them.

Table 1. Selective Contracting Waivers, 2003
State Service
ArkansasNon-emergency transportation
CaliforniaInpatient hospital
FloridaNon-emergency transportation
GeorgiaNon-emergency transportation
KentuckyNon-emergency transportation
LouisianaMail order pharmacya
New YorkNon-emergency transportation
OregonNon-emergency transportation
TexasInpatient hospital, psychiatric hospital
UtahNon-emergency transportation
WashingtonInpatient hospital
Source: CMS descriptions of waiver programs, available
at []
a. Asthma and diabetes pharmaceuticals and supplies.
Every state except Arizona has one or more home and community-based
programs, serving the aged, persons with disabilities, and/or persons with mental
retardation or developmental disabilities. Payment for waiver services is discussed
later in this report.
Prescription Drug Rebates
Medicaid programs are major purchasers of prescription drugs, chiefly because
of their role in providing drug coverage to low-income aged and disabled people.
Other large-volume purchasers, such as private insurers, pharmaceutical benefit
managers (PBMs), and hospital buying groups, often get substantial discounts or
rebates from drug manufacturers. To assure that Medicaid programs received similar
benefits, the Omnibus Budget Reconciliation Act of 1990 (OBRA 90, P.L. 101-508)
required manufacturers to give rebates to states for drugs paid for by Medicaid. The
rebate formulas are designed to assure that states pay the lowest price offered by the
manufacturer to any other high-volume purchaser. In return, the state must generally
cover all the drugs marketed by the manufacturer.
There is ongoing debate over how the rebates are calculated and whether
Medicaid programs really are getting the “best price.” Rebates reduced Medicaid
drug spending by 20% in 2001.5 Still, spending for drugs is one of the fastest
growing components of Medicaid budgets. Restricting drug spending has been a
major focus of recent state cost containment efforts (see the last part of this section).

5 See Table 37.

Disproportionate Share Hospital (DSH) Payments, Provider
Donations, and Provider Taxes
In response to the 1981 requirement that hospital payment systems take account
of the situation of DSHs, some states developed plans to make supplemental
payments to these hospitals. These plans potentially conflicted with the Secretary’s
regulation capping aggregate Medicaid reimbursement at Medicare levels. The
Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA, P.L. 99-272)
prohibited the Secretary from limiting states’ payment adjustments to DSHs. Until
1987, states were free to establish their own criteria for classifying facilities as DSHs
and to develop their own reimbursement methods for these hospitals. The Omnibus
Budget Reconciliation Act of 1987 (OBRA 87, P.L. 100-203) defined certain
facilities that states had to designate as DSHs and set minimum payment
requirements for them. These requirements, which have since been amended several
times, are described in the discussion of inpatient hospital reimbursement, below.
The COBRA provision that prohibited the Secretary from limiting DSH
payments meant that these were the only Medicaid payments not subject to any form
of upper limit. Beginning in the late 1980s states began to exploit this loophole by
developing various financing schemes intended to draw extra federal matching funds.
A state might make an extra payment to a hospital, claim federal matching, and then
recapture part or all of the payment by taxing the hospital. Alternatively, the hospital
might agree to donate part of the extra payment to the state. Or, in the case of a
hospital operated by state or local government, the money could be recovered through
an intergovernmental transfer — a transfer of funds from another state agency to the
Medicaid program or from local government to the state.
Table 2 illustrates how a typical provider tax or donation program might work
in a state whose federal matching percentage was 60%. The state pays the hospital
$100. The state reports the payment to CMS and receives $60 in federal matching
funds. The hospital gives back $80 — either as a donation or because the state
imposes a “provider-specific” tax on its payments. The hospital is still ahead by $20,
and the state has gained $40 to spend on other Medicaid services or simply absorb
into its general fund. A state could potentially operate its entire Medicaid program
with no actual state expenditures.
Table 2. Effect of a Typical Provider Donation or Tax Program
(state with 60% federal matching rate)
St a t e Federal
government Hospital government
State pays hospital $100$ (100)$100
State reports payment to CMS, receiving matching$60$(60)
Hospital donation or tax paid to state$80$(80)
Net gain/loss (sum of transactions one to three)$40$20$(60)
Source: Congressional Research Service.

While these schemes could be used with any kind of provider payment, the use
of the DSH loophole was attractive because the state could pay (and then recover)
any amount at all. DSH payments rose from an estimated $569 million in 1989 to
a projected $8 billion, or 12% of total Medicaid spending, by 1992.
The Medicaid Voluntary Contribution and Provider-Specific Tax Amendments
of 1991 (P.L. 102-234) prohibited the use of most provider donations and phased out
the use of provider taxes that were not “broad-based” — that is, taxes that were
levied against a provider’s Medicaid receipts and not receipts from other sources.
The Act did not restrict the use of intergovernmental transfers, on the grounds that
the federal government had no authority to regulate these arrangements, but instead
sought to limit potential federal exposure by capping the total amount of DSH
Beginning in 1992, national aggregate DSH payment adjustments during each
fiscal year were limited to 12% of total Medicaid spending for that year. “High
DSH” states, those whose payments were already above the 12% limit, were allowed
to increase their payments by no more than the projected growth in their overall
Medicaid spending. Other states were allowed larger increases, with each state
receiving an allocation calculated to assure that aggregate national payments did not
exceed the national cap. The BBA replaced this system of calculating DSH limits
with fixed annual limits for each state. These limits, and subsequent amendments,
are described in the discussion of inpatient hospital reimbursement, below.
The Omnibus Budget Reconciliation Act of 1993 (OBRA 93, P.L. 103-66)
further limited DSH payments by capping payments to any single facility. The sum
of regular and DSH payments to a hospital could not exceed the sum of the hospital’s
costs for treating Medicaid beneficiaries and uninsured patients.
Growth in Managed Care
States have been contracting with health maintenance organizations (HMOs) or
similar prepaid capitated plans to enroll Medicaid beneficiaries since the late 1960s.6
OBRA 81 made it easier for states to enter into these contracts and also authorized
a different form of managed care, primary care case management (PCCM). Under
these programs, beneficiaries’ services were still paid on a fee-for-service basis, but
were coordinated by a primary care physician. Using a freedom-of-choice waiver,
states could require beneficiaries to participate in PCCM or to choose between
PCCM and a prepaid plan. Enrollment in managed care arrangements grew steadily
through the 1980s and early 1990s.
By 1996, 40% of beneficiaries received at least some services through some
form of managed care.7 The BBA gave states greater flexibility to contract with

6 Capitated plans receive a fixed per capita payment (usually monthly), in exchange for
which they accept financial risk for providing a defined scope of services to each enrolled
7 CMS, Managed Care Trends, 1991-1996, at [

health maintenance organizations (HMOs) or similar managed care organizations
(MCOs) and to require beneficiaries to enroll in these plans or PCCM programs
without a waiver. By mid-2002, the proportion of beneficiaries in some form of
managed care had reached 58%.
Table 3 gives Medicaid enrollment figures as of June 2002. Forty percent of
enrollees were in some form of full-risk arrangement; that is, a capitated plan
provided their basic Medicaid services. Another 14% were in PCCM programs,
receiving care on a fee-for-service basis, while 25% were in prepaid health plans,
almost all of which provide only one type of service, such as behavioral health care,
dental care, or non-emergency transportation. (Note that enrollees in these special
plans can also be in an MCO or PCCM program.) In sum, then, 60% of beneficiaries
were still receiving most or all of their services on a fee-for-service basis.
Table 3. Medicaid Beneficiaries and Medicaid Managed Care
Arrangements, June 2002
Beneficiaries (thousands)Percentage
Full-risk arrangements16,16840.2
Commercial MCO9,73424.2
Medicaid-only MCO5,72314.2
Health insuring organization5111.3
PACE and other1990.5
Primary care case management5,61514.0
Prepaid health plan10,16625.3
No managed care17,03042.4
To tal 40,175
Source: CMS, Managed Care Enrollment by Program Type, June 30, 2002, at
[], as of Sept. 2003.
Notes: This table provides duplicated figures by plan type. The total number of enrollees includes
8,830,530 individuals who were enrolled in more than one managed care plan.
PACE stands for programs of all-inclusive care for the elderly. Under the PACE programs, Medicare
and state Medicaid programs make integrated capitation payments for preventive, acute and long-term
care services to MCO-like organizations that furnish services to frail elderly people.
Upper Payment Limits (UPLs)
Since the 1970s, federal regulations have required that total Medicaid payments
for a service type, such as hospital or nursing facility services, could not exceed the
amount that would have been spent for the same services under Medicare
reimbursement principles. The UPLs originally applied in the aggregate; a state
could, for example, pay one hospital more than Medicare would have paid and
another hospital less, so long as total payments did not exceed the limit.

7 (...continued)
managedcare/trends1.asp] as of Sept. 2003.

After the use of provider taxes and donations was limited in 1991, states could
still recover Medicaid payments made to governmental providers through
intergovernmental transfers. States’ ability to use these mechanisms was limited by
the cap on total DSH payments and on DSH payments to any one facility. However,
states found that they could draw extra Federal matching funds by exploiting the fact
that UPL limits were aggregate rather than facility-specific.
Table 4 illustrates how what came to be known as “enhanced payment”
programs worked. Private hospitals have actual costs of $80 million, while
county-owned hospitals have costs of $20 million. The state pays private hospitals
80% of their costs, or $64 million, meaning that it can pay the county hospitals $36
million and still be within the aggregate UPL of $100 million. (The shortfall in
payments to the private hospitals might be made up through DSH payments, which
do not count toward the UPL.) The state claims $60 million in federal
reimbursement, and the county hospitals return the excess payment to the state.
While the state has nominally spent $40 million on hospital services, it has actually
spent only $24 million, while the county hospitals have been paid their full costs.
The federal government has spent $60 million to the state’s $24 million; in effect, the
federal share of hospital spending is 71% instead of 60%.
Table 4. Typical Enhanced Payment Program
(state with 60% federal matching rate; millions of dollars)
Cost underNominal
MedicareMedicaidFederalstateIntergovernmentalNet state
pr i nc i pl e s pay me nt s ma t c hi ng s pe ndi ng transfer s pe ndi ng
hospitals $80 $64 $38.4 $25.6 $25.6
hospitals $20 $36 $21.6 $14.4 $16.0 $ (1 .6)
Total $100 $100 60.0 40.0 $24.0
Source: Congressional Research Service.
By FY2000, 28 states had adopted enhanced payment programs, making an
estimated $10.3 billion in extra payments to hospitals, nursing facilities, and, in one
state, community mental health centers. The DHHS Office of the Inspector General
estimated that states were drawing $5.8 billion in excess federal matching payments
through these programs.8
The Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act
of 2000 (BIPA, P.L. 106-554) required the Secretary to adopt a new regulation that
would establish three separate UPLs for each category of institutional care (hospital,
nursing facility, ICF-MR, and clinic): one for state facilities, one for private

8 U.S. Department of Health and Human Services (HHS), Office of the Inspector General,
Review of Medicaid Enhanced Payments to Local Public Providers and the Use of
Intergovernmental Transfers, A-03-00-002 16, Sept. 2001.

facilities, and one for non-state governmental facilities.9 The Act specified that there
was to be a five-year transition period for programs operating under a state plan or
state plan amendment approved or in effect before October 1, 1992.
In January 2001, the Clinton Administration published a final rule that limited
payments to 100% of the UPLs for state and private facilities and 150% for non-state
governmental facilities. To lessen the fiscal shock to states that had become
dependent on the extra Federal matching, states making payments above the 150%
limit were allowed a transition period to phase down to the limit. The length of the
phase-down depended on when the state had adopted its plan. (Table 5 reflects
modifications in the transition periods adopted in a final rule published in September


9 A separate UPL for state facilities had been established by regulation in 1987, but private
and non-state public facilities had been left under a single UPL, allowing for programs like
the one illustrated.

Table 5. Transition Periods for Compliance
with Upper Payment Limits
Group 1AGroup 1BGroup 2Group 3
Group definitionPlan effectivePlan effectiveApproved planApproved plan
on or afteron or aftereffective aftereffective on or
October 1, 1999October 1,October 1, 1992before October
and approved1999, submittedand before1, 1992
before Januarybefore MarchOctober 1, 1999
22, 200113, 2001, and
approved on or
after January 22,
Base period for
d e ter mining
amount ofState fiscal year 2000
excess payments
that must be
phased out
When phase-outMarch 13, 2001SFY2003First state fiscal
beginyear that begins
after September
30, 2002, i.e.,
SFY2003 or
PercentageNot specified; states must be inExcessExcess
reduction incompliance by end of phase-outpayments mustpayments must
excess paymentsperiodbe reduced inbe reduced in
each year of the25% increments15% increments
phase-outover each ofover each of five
four yearsyears SFY2004-a
SFY2003-SFY2008, plus
SFY200615% reduction
for the portion
of SFY2009
occurring before
October 1,2008
with the final
10% reduction
achieved as of
October 1, 2008
When phase-outSeptember 30,November 5,End of SFYSeptember 30,
ends date by20022001 or one20062008
which fullyear from
compliance witheffective date of
UPLs isplan, whichever
requiredis later
Source: CRS Report RL31021, Medicaid Upper Payment Limits and Intergovernmental Transfers:
Current Issues and Recent Regulatory and Legislative Action, by Elicia Herz.
a. This schedule applies to states that begin the phase-out in SFY2004. For states that begin the
phase-out in SFY2003, the schedule is modified accordingly (i.e., the process begins in SFY2003).
See row labeled “when phase-out begins.”

In January 2002, the Bush Administration issued a new final rule that reduced
the UPL for non-state governmental facilities to the same 100% applicable to state
and private facilities. The transition periods for states paying above 150% were
modified only slightly, except that they now had to reach 100% instead of 150% by
the end of the period — meaning larger cuts in payments and federal matching at
each step. (States that were paying more than 100% but less than 150% of the UPL
were allowed no transition to bring their payments within the 100% limits.) The
change was projected to save $9 billion in federal funds for FY2002-FY2006.
Table 6 shows CMS’s preliminary analysis of state enhanced payment programs
and their phase-out periods, as of January 22, 2004.
Table 6. State Enhanced Payment Programs by Provider Type
and Preliminary Transition Period in Years
(as of January 22, 2004)
Inpa t i ent Outpatient Nursing
St a t e ho spit a l ho spit a l f a cilit y
Alabama 5 * 5 * 5 *
Al a s k a 2
Ar ka nsa s 2
Califo r ni a 8
Geo r gia 5 a
I llino is 8 8
Iowa 2
Kansas 2
Lo ui s i a n a 2
Michigan 55
M i sso ur i 1 2
Neb r aska 8
New Hampshire5
New Jersey2
New York5
North Dakota5
Orego n 5
P e nnsyl va ni a 8
South Dakota2
T e nne sse e 2
Vir ginia 1
W a shi ngt o n 1 5
W i sc o nsin 8
Programs 7 5 19
Source: CMS communication to the Congressional Research Service (CRS), Feb. 17,
Note: One and two-year transition periods have expired.
* May not qualify for a transition period.

State Fiscal Problems and Medicaid Cost Containment
Because of revenue shortfalls resulting from the economic downturn and rising
spending pressures, most states have faced serious budget imbalances beginning in
FY2002. They have responded by cutting expenditure growth, raising revenues, and
drawing on reserve funds. For FY2004, two-thirds of states plan expenditure
increases of less than 5%, and 19 plan to spend less in FY2004 than in FY2003.10
A recent survey has found that every state and the District of Columbia took
some measures to control Medicaid spending growth in FY2003 and that each plans
further measures for FY2004. Table 7 shows the types of cost containment
measures implemented in FY2002 and FY2003 and planned for FY2004. Nearly
every state has reduced or frozen payment rates for some types of providers. Most
have also acted to control prescription drug spending; as will be discussed in the
section on drug payment, below, these measures have not always involved changes
in payment methods. States have so far been slightly less likely to drop coverage
of beneficiaries, reduce benefits, or increase copayments paid by beneficiaries for
Table 7. Number of States Undertaking Medicaid Cost
Containment Strategies, FY2002-FY2004
P la nned
FY2002 FY2003 FY2004
Controlling drug costs324644
Reducing/ freezing provider payment225049
Reducing/restricting eligibility82518
Reducing benefits91820
Increasing copayments41721
Source: V. Smith, et al., States Respond to Fiscal Pressure: State Medicaid Spending Growth and
Cost Containment in Fiscal Years 2003 and 2004, Kaiser Commission on Medicaid and the
Uninsured, 2003.
Table 8 shows the major service types for which states plan rate changes for
FY2004. States are more likely to freeze or decrease payments to hospitals and
physicians, and more likely to increase payments for nursing homes and MCOs. In
the case of nursing home rates, the survey authors note that some states have
statutory requirements for annual inflation increases. Another possible factor is that
nursing homes and MCOs with Medicaid contracts may rely much more heavily on
Medicaid than other providers and may be unable to cost-shift to other purchasers.

10 National Governors Association and National Association of State Budget Officers, The
Fiscal Survey of States 2003, June 2003.

Table 8. Number of States Planning Rate Changes for Selected
Services, FY2004
Increase Freeze Decrease
Hospitals 19 22 10
P hysicians 1 1 3 5 3
Nursing homes29136
Managed care organizations (MCOs)20145
Any of these374722
Source: V. Smith, et al., States Respond to Fiscal Pressure: State Medicaid Spending Growth and
Cost Containment in Fiscal Years 2003 and 2004, Kaiser Commission on Medicaid and the
Uninsured, (Washington, 2003).
It should be emphasized that the information in this report generally reflects
payment methods and payment levels in effect before most states faced budget
shortfalls. Many states that improved their provider payments during periods of
economic growth may now have cut back. General payment methodologies — how
states decide the relative amounts paid to different providers for different services
— may or may not have been affected. However, states that have not changed
methodologies may achieve savings by imposing uniform cuts, or simply by
granting rate increases below the rate of inflation in the cost of goods and services
providers must buy. Therefore, the discussions in this report of the adequacy of
Medicaid payment to assure access or quality may not reflect current conditions.
Acute Care
Hospital Services
Medicaid payments for hospital services take three forms:
!Payments for services to individual inpatients and outpatients,
!Lump-sum disproportionate share hospital (DSH) payment
adjustments, and
!In states with UPL plans, additional lump sum distributions.
In FY2001, DSH payments made up close to one-third of all direct payments
to general or community hospitals (including inpatient and outpatient payments),
and half of payments to inpatient psychiatric facilities. (See Table 13) Comparable
figures for UPL plans are not available. However, the HHS Office of the Inspector
General estimated that total UPL payments to hospitals for inpatient and outpatient
services were about $4.5 billion in FY2000. This would have been 14% of total
inpatient non-DSH and outpatient spending in that year.
This section describes states’ basic methodologies for establishing payments
for individual inpatients and outpatients. It then summarizes current rules relating
to DSH and UPL payments and provides data on the size and distribution of these

payments. Finally, it reviews available evidence on the extent to which Medicaid
payments meet hospitals’ costs for Medicaid beneficiaries, as well as whether DSH
or UPL supplements help hospitals that treat uninsured patients.
Inpatient Payment Methods
All states now use some form of prospective system as their basic method for
setting inpatient hospital payments. That is, payment amounts per day or per case
are fixed at the start of a year and are generally not subject to retrospective
adjustment on the basis of actual costs incurred. States may establish a different rate
for each participating hospital, may use one rate for all hospitals in a defined peer
group, or may have one statewide rate. Two-thirds of the states have adopted some
form of case mix adjustment, under which reimbursement varies according to the
intensity of services required or the expected resources used by each individual
patient. These adjustments, discussed further below, can be applied regardless of
the state’s method for setting basic rates.
Some states’ systems allow additional reimbursement for “outliers,” patients
whose costs or length of inpatient stay are significantly higher than the average for
comparable patients. Medicaid law requires states with prospective systems —
effectively all states now — to make outlier adjustments for high-cost or
long-staying infants under one year old in any hospital, and for children under six
in a DSH hospital.11
Table 9, based on an analysis of state Medicaid plans, shows the method in use
and approved by CMS as of November 2002. The table classes states according to
whether hospitals receive hospital-specific rates, receive rates set for a whole group
of hospitals or for all hospitals in the state, or are paid under some other method.
For states using some form of case mix adjustment, the table indicates the method.
Finally, where applicable, it identifies the facility characteristics states use in
establishing peer groups of hospitals.
Several general points about the table and the accompanying discussion should
be noted:
!The systems described in the table and in the following discussion
are those used for most acute general hospitals in the state. States
may use different modes of payment for particular classes of
facilities. For example, states may use prospective payment for
acute general hospitals and a reasonable cost system for psychiatric,
rehabilitation, or other specialized hospitals. Some general
hospitals — for example, those that are state-owned, or small
hospitals in rural areas — may receive special treatment. In
addition, states that negotiate rates with preferred providers under
a selective contracting system may have a separate payment

11 The provision, in Section 1902(s) of the Social Security Act, also prohibits imposition for
such cases of any day limit or (for infants) any dollar limit in the state plan.

methodology for emergency or other services obtained outside that
!Many states that use peer-group or statewide payment systems
provide hospital-specific add-ons for certain categories of costs,
such as capital costs (interest, depreciation, and other costs related
to owning a physical facility) and graduate medical education costs
(costs directly or indirectly related to training residents).
!In two states, Arizona and Tennessee, nearly all beneficiaries are
enrolled in MCOs, and some other states have very high rates of
MCO enrollment. The methods shown are for cases in which the
state pays a hospital directly (for example, because a beneficiary is
in an aid category exempt from MCO enrollment), not the methods
used by MCOs in paying their contracting hospitals.

Table 9. Basic Payment Methodology, Inpatient Hospital Services, 2002
Basic payment methodology
Hospital specific ratePeer group or statewide rate
SubjectBlend of
to ratehospital-
ofSubjectSubjectspecificMethod of caseFacility
increaseto peertoandPeerOthermixcharacteristics
limitsgroupstatewidestatewidegroupStatewidepaymentadjustment (ifused to define
Stateonlyceilingceilingraterateratemethodany)peer groupsNotes
SelectiveRegional hospital coalitions
abamacontractingreceive per eligible amount
a s k a x
Only for services outside
izonaxAdmission typeAHCCCS
ka nsa s x
iki/CRS-RL32644SelectiveRates negotiated with each
g/wo r ni a contracting contractor
s.or Dia gno sis-
leakrelated groupLocation,
lo r a d o x (DRG) sp ecialty
://wikinne c t i c ut x
httpare x
lumb ia x D RG
r id a x Co unt y
rgiaxSpecialtyHospital loss limited to 10%
T eaching,
numb e r
waiixAdmission typedischarges
Hospitals <41 beds
guaranteed cost, larger
oxguaranteed 85% of cost
o i s x DRG
diana x DRG

Basic payment methodology
Hospital specific ratePeer group or statewide rate
SubjectBlend of
to ratehospital-
ofSubjectSubjectspecificMethod of caseFacility
increaseto peertoandPeerOthermixcharacteristics
limitsgroupstatewidestatewidegroupStatewidepaymentadjustment (ifused to define
Stateonlyceilingceilingraterateratemethodany)peer groupsNotes
a x DRG
Urb an/rural,
sas x DRG siz e
Size, Medicaid
vo l ume ,
nt uc ky x sp ecialty
Size, teaching,
ui s i a n a x sp ecialty
iki/CRS-RL32644e x
g/wState rate-setting commission
s.orlandAll-payersets hospital-specific rates
leakBonus for hospital with lower
assachusettsxDRGcosts or lower rate of increase
://wikiBonus for hospital with lower
httpigan x DRG costs
Co llap sed
i nne so t a x DRGs
i ssissip p i x Size
i sso ur i x
tana x DRG
Urb an/rural,
r aska x D RG siz e
Ad mi ssio n
type, length of
adaxstay range
p shire x DRG

Basic payment methodology
Hospital specific ratePeer group or statewide rate
SubjectBlend of
to ratehospital-
ofSubjectSubjectspecificMethod of caseFacility
increaseto peertoandPeerOthermixcharacteristics
limitsgroupstatewidestatewidegroupStatewidepaymentadjustment (ifused to define
Stateonlyceilingceilingraterateratemethodany)peer groupsNotes
Teaching, referral, regional,
w MexicoxDRG4 typescommunity
YorkxDRGteaching, size
rth CarolinaxDRG
rth DakotaxDRG
iki/CRS-RL32644ioxDRGspecialty, size
g/wlahomax8 care levels
s.orLower rate of increase for
leakhospital with higher profit
://wikio n x DRG ma r g i nnnsyl va ni a x DRG
Multi-Maxicap: state and Blue Cross
payer (seenegotiate rates with hospital
ode Islandnote)association
Hospital-specific per diem for
infrequent or highly variable
uth CarolinaxDRGDRGs
uth DakotaxDRG
Only for services outside
nne sse e x T e nnCa r e
SelectiveRates negotiated with each
xas contracting contractor
Statewide rate used for low
xDRGvariability or low-cost DRGs

Basic payment methodology
Hospital specific ratePeer group or statewide rate
SubjectBlend of
to ratehospital-
ofSubjectSubjectspecificMethod of caseFacility
increaseto peertoandPeerOthermixcharacteristics
limitsgroupstatewidestatewidegroupStatewidepaymentadjustment (ifused to define
Stateonlyceilingceilingraterateratemethodany)peer groupsNotes
Shared savings when costs
ontxAdmission typeTeaching, sizebelow rates
ginia x D RG
DRG weight times fixed
Selectiveamount negotiated with each
a shi ngt o n contracting DRG contractor
Urb an/rural,
est VirginiaxDRGsize
iki/CRS-RL32644i sc o nsin x DRG Lo catio n,sp ecialty
s.orIncentive for cost below
leakyomingx10 care levelsceiling
mber of
://wiki using
httpthod105631011633 (27 DRG)
: Medicaid state plans and amendments approved as of Nov. 7, 2002, except as follows: Alabama Medicaid Administrative Code, at [
NUALS/AdminCode/ad_ch_37.htm], as of Aug. 20, 2003. Maryland Health Services Cost Review Commission, Report to the Governor Fiscal Year 2001, at
://], as of July 31, 2003. Nevada Medicaid Rates and Cost Containment Unit Rate Matrix, at], as of Aug. 1, 2003. Rhode Island Medicaid Program, Annual Report, Fiscal Year 2002, at
://], as of Aug. 20, 2003.
AHCCCS = the Arizona Health Care Cost Containment System, the managed care program that serves most Medicaid beneficiaries in Arizona.
= diagnosis-related groups. DRGs represent a system of classifying any inpatient stay into groups for purposes of payment. DRG systems relate the type of patients a hospital
the costs incurred by the hospital. According to this classification system, patients who have similar diagnoses and undergo similar procedures are placed together in the same
nosis-related group. DRG definitions may also take into account other patient characteristics, such as common sex, age, and discharge status. []

Hospital-Specific Rates. In 24 states, fixed per diem or per case payment
rates are established for each hospital, using historic data on that hospital’s Medicaid
costs and some form of fixed update factor for inflation. A hospital whose costs rise
faster than the update will therefore lose money. Some states use an objective
inflation index, such as CMS-released estimates of price changes for a “market
basket” of goods and services commonly purchased by hospitals. Often, however,
annual updates are set by legislation and regulation and may be higher or lower than
actual inflation. Oregon uses update factors that vary inversely with each hospital’s
operating margin (or profit); the effect is to grant lower increases to hospitals earning
a profit on their Medicaid patients.
Of the states using hospital-specific rates, five use peer group ceilings; the
hospital’s rate is based on the lesser of its own costs or some percentile of costs for
similar hospitals. Hospital characteristics used to establish peer groups include size,
location, presence of a teaching program, specialized services (for example, pediatric
hospitals), and volume of Medicaid services. Six states use a statewide ceiling for
all general hospitals, based on a percentile of all hospitals’ costs or, in the case of
Arkansas, a legislatively fixed per diem limit ($675 for 2002). Finally, three states
use a blend of hospital-specific and peer group or statewide experience to set
payment ceilings. For example, the operating cost component of Iowa rates is based
on 50% of the hospital’s cost and 50% of the statewide average.
One effect of systems using ceilings is that, while a hospital with costs above
the ceiling will lose money, a hospital with costs below the ceiling will receive a rate
derived from its base-year costs. It can earn a profit only if it can reduce its costs still
further; it is not rewarded for being more efficient than its competitors. Wyoming
provides incentive payments to hospitals with costs below the statewide ceiling. Two
states, Georgia and Idaho, limit the losses that can be incurred by hospitals.
Peer Group or Statewide Rates. In 21 states, a fixed rate is set for an
entire class of hospitals or for all hospitals in the state. In most of these states, part
or all of the fixed rate is adjusted (as in Medicare’s inpatient prospective payment
system, or PPS) for higher or lower labor costs in the hospital’s market area. In fixed
rate systems, unlike ceiling systems, a hospital with costs below the rate can realize
a profit. Three states, Massachusetts, Michigan, and Vermont, provide additional
bonuses to lower-cost hospitals.
Use of Case Mix. Nearly two-thirds of the states have adopted some form of
case mix adjustment, under which reimbursement varies according to some measure
of the intensity of services required or the resources used by each individual patient.
Most of these use the diagnosis-related groups (DRGs) developed for Medicare
hospital reimbursement. Patients are assigned to one of 540 DRGs on the basis of
admitting diagnosis, procedures performed, presence of complications, or other12
characteristics. Each DRG has an assigned weight — for example, 0.8889 for an
uncomplicated appendectomy or 9.7823 for a liver transplant-which is then
multiplied by the fixed rate established for the hospital. So, if a hospital’s standard

12 Not all of the 540 codes are actually in use.

rate were $5,000, it would be paid $4,445 for the appendectomy and $48,912 for the
liver transplant.
Because Medicaid patients may be different from Medicare patients, the
weighting factors established for DRGs under Medicare may not be appropriate for
Medicaid reimbursement. Most states using DRGs have developed their own
weights on the basis of Medicaid-specific data. Some states use alternative DRG
classification systems, such as the DRGs developed for the Civilian Health and
Medical Program of the Uniformed Services (CHAMPUS) or New York’s All
Patient DRGs. These groupings add additional categories for types of patients, such
as maternity cases or newborns, rarely treated under Medicare. Minnesota has
collapsed the DRGs into a smaller number of diagnostic categories.
Some states that have not adopted DRG classifications nevertheless modify
reimbursement according to the type of patient served. Some of these use admission
types — for example medical/surgical, maternity, psychiatric — while others assign
cases to a limited number of level-of-care groupings. Nevada additionally adjusts its
per case rates using length-of-stay ranges.
Other Methods. Four states have used 1915(b) freedom-of-choice waivers
to develop hospital contracting systems, while two states have systems under which
Medicaid and other payers use common reimbursement methods.
Selective Contracting. Under Section 1915(b), a state may receive a waiver
of Medicaid requirements, including the requirement that beneficiaries be allowed
a free choice of medical providers, in order to allow the development of innovative
delivery or reimbursement systems. One of the available options for states is to limit
program participation (except for emergency services) to providers who meet
reimbursement, quality, and utilization standards approved by the state. Certain
payment rules cannot be waived under this option, including requirements for
additional payment to disproportionate share hospitals (see below) and requirements
for prompt payment to providers.
Alabama, California, Texas, and Washington have used this authority to restrict
the inpatient hospitals from which beneficiaries may obtain services. (Illinois
operated a similar system until 1991.) Alabama’s program is statewide. In the other
states the waiver applies only in selected counties or areas; however, a large share of
beneficiaries live in the covered area. Except in emergencies or other exceptional
cases, these beneficiaries may use only hospitals selected for participation through
a system of competitive negotiation. In California and Texas, reimbursement rates
for the participating hospitals are established in the course of the negotiation. In
Washington, what is negotiated is the hospital’s “conversion factor,” a fixed dollar
amount that is multiplied by the weighting factor for a DRG to produce a final
payment amount for each case.
Under Alabama’s Partnership Hospital Program, groups of hospitals in a
geographic area form a prepaid inpatient health plan that is reimbursed on a capitated
(fixed per beneficiary per month) basis; the plan in turn makes payments to its
participating hospitals. All Medicaid beneficiaries are automatically enrolled, except

those who are also Medicare beneficiaries and certain pregnant women participating
in a separate Maternity Care program.13
Multi-payer Systems. Beginning in the 1970s, several states established
“all-payer” hospital rate-setting systems. In these systems, all insurers or other
payers in the state, including Medicare and Medicaid, agreed to pay uniform rates or
use a standard reimbursement methodology for inpatient services. Only one state,
Maryland, still has an all-payer system in which Medicare participates.14 A state
rate-setting commission sets each hospital’s allowable prices for specific service
units, such as a day of routine care or a particular laboratory test. The prices are set
at levels expected to result in a target average charge per case for each facility. A key
feature of the system is that every payer contributes to hospitals’ costs for treating
uninsured patients.
In Rhode Island, the state and Blue Cross jointly negotiate with the state hospital
association an annual statewide ceiling (the “Maxicap”) on reimbursable expenses
for the 12 voluntary hospitals in the state. Within this ceiling, an operating budget
is developed for each hospital, and rates paid by Medicaid and Blue Cross are set to
meet these budgets.
Administrative Days/Swing Beds. Under Medicare, small rural hospitals
may enter into “swing bed” agreements with CMS, under which beds may be used
either for inpatient hospital care or for care equivalent in intensity to that furnished
by a nursing facility. Costs are allocated and reimbursement adjusted to reflect the
level of care furnished to each patient. A Medicaid program may also allow for
swing beds, but only in hospitals that have entered into a Medicare swing bed
arrangement. The state may develop a specific payment methodology for swing bed
days of care at the nursing facility level or may pay at a rate based on average
payments for comparable services in freestanding nursing facilities. The swing bed
program assists hospitals that are underused and also helps to meet local shortages
of nursing facility beds.
Sometimes a hospital which is not a swing bed facility will provide care to a
patient at the nursing facility level of intensity because a place cannot be found for
the patient in an appropriate facility and the patient cannot be discharged. The days
of inpatient care received by patients in this situation are known as “administrative
days.” Prior to 1997, Medicaid payment for an administrative day was limited to the
statewide average Medicaid payment rate for a day of care in a skilled nursing
facility. Most states have continued this practice despite the repeal of the provision
in the BBA.

13 Note that, because the plans provide inpatient services only, they are not subject to the
Section 1903(m)(2) requirements for Medicaid managed care organizations.
14 To retain the Medicare waiver, a system must hold cumulative growth in cost per
Medicare admission from 1981 to the present at or below national average growth.

Outpatient Payment Methods
Because hospitals furnish a wide variety of services on an outpatient basis —
from emergency room visits to surgery to diagnostic tests — many states use several
different payment methodologies. For example, a state might pay a flat per-visit fee
for a clinic visit, use a fee schedule for surgery, and pay on a cost basis for some
specialized services. Because states vary in their service definitions, there is no ready
way of comparing methods for particular services across states. Table 10 attempts
to identify the “principal” payment approach in each state, with notes on variants in
some states, without depicting the full complexity of state systems.
About half of the states still base outpatient reimbursement largely on
hospital-specific costs. Of these, 15 pay actual costs or prospective rates based on
historic costs with a limit on annual increases. One state, Florida, uses a peer group
ceiling comparable to those common in inpatient hospital and nursing facility
payment. Another 11 states pay a fixed percentage of actual costs; that is, their
systems explicitly pay each facility less than its costs. One goal of such systems may
be to discourage use of hospitals for services that could be rendered in a
noninstitutional setting.
Sixteen states use fee schedules, varying payment by the surgical or other
procedures performed. For at least some services, several states pay the same rates
regardless of whether the service is performed in a hospital or in a physician’s office.
Again, the aim is to avoid incentives for use of the more costly setting.
Only four states have adopted systems comparable to Medicare’s new
prospective system for outpatient hospital services. Under this system, services are
classified into one of 383 ambulatory patient classifications (APCs), groups of
services expected to require comparable resources. As in the inpatient DRG system,
payment for each APC is at a fixed rate times a weight that reflects resource use for
the APC relative to that of other APCs. One state has adopted Medicare’s system
directly; others use their own classification system or prices.
Of the remaining states, Maryland uses the same all-payer system, and Rhode
Island the same multi-payer negotiation, as for inpatient care. (Hawaii also
negotiates some rates.) Utah pays a percentage of charges, rather than costs, while
Arkansas uses Blue Cross customary charge screens. Finally, Delaware pays blended
rates based on a mix of hospital-specific and statewide experience.

Table 10. Principal Outpatient Hospital
Reimbursement Approach
specific rate
based on
Fee Ca sePercent
StateschedulepaymentOtherNotesCostof cost
AlaskaxRate-of-increase limit
Alabama x
ArkansasxPercent of Blue Cross
customary charges
ArizonaxOnly for
non-AHCCCS patients
Califo r ni a x
Co lo rado 72%
Co nne c t i c ut x
District ofxRate-of-increase limit
Co lumb ia
DelawarexVisit rates blend of
hospital-specific, state
average; other services
cost-b ased
FloridaxCost up to ceiling set
at 80th percentile for
c o unt y
Georgia 90%
Hawaii75%Some rates negotiated
Iowa x B lend ed
hospital/statewide rate
Idaho x Radiology/surgery
based on schedule for
comp arable
non-hospital service
I llino is x
Indiana x
KansasxBased on schedule for
comp arable
non-hospital service
K e nt uc ky x
Lo uisiana 83%
MassachusettsxState-developed prices
MarylandxState rate-setting
commission sets
hospital-specific rates
Maine x
Michigan x
MinnesotaxUses Medicare prices
M i sso ur i 9 0 %

specific rate
based on
Fee Ca sePercent
StateschedulepaymentOtherNotesCostof cost
M i ssissip p i x
Montanax93% of cost for
services not on
sc he d ule
North Carolina80%
North Dakotax
Nebraska 85%
New x
Hamp shire
New Jersey94.2%
New Mexico97%
Neva d a x
New Yorkx
Ohio x
Oklaho ma x
Oregon 59%
P e nnsyl va ni a x
Rhode IslandxState and Blue Cross
negotiate rates with
hospital association
South Carolinax
South Dakotax
TennesseexOnly for services
outside TennCare
Texas80.3%84.48% of cost for
high-volume providers
UtahxPercent of charges:
77% urban, 93% rural
VirginiaxEmergency room paid
at all-inclusive rate
VermontxServices available in
physicians’ offices
paid at physician rate
WashingtonxState-developed prices
WisconsinxPer visit rates based
on past hospital-
specific costs, rate of
increase limit
West Virginiax
Wyoming x
Number of
states using
method 15 11 16 4 5

Source: Medicaid state plans and amendments approved as of Nov. 7, 2002, except as follows:
Maine: MaineCare Benefits Manual, Chapter III, Section 45, 01-015 CMR (Code of Maine Rules)
Chapter 101, at [], as of July 31, 2003.
Maryland: Maryland Health Services Cost Review Commission, Report to the Governor Fiscal Year
2001, at [], as of July
31, 2003. Nevada: Nevada Medicaid Rates and Cost Containment Unit Rate Matrix,
[], as of July 31, 2003. Rhode
Island: Rhode Island Medicaid Program, Annual Report, Fiscal Year 2002, at
[], as of Aug. 20, 2003.
Washington, at [], as of Aug. 6,
DSH Payments15
Current DSH Requirements. Federal Medicaid law requires that states
make additional payments to hospitals that serve a disproportionate share of
Medicaid and other low-income patients. The statute defines which hospitals must
receive DSH payments and which hospitals may never receive DSH payments.
States can decide on their own whether to make payments to hospitals that are in
neither category. Similarly, the law sets minimum payment amounts that must be
made for certain hospitals and maximum payment amounts for individual hospitals
and for all hospitals in the state. Again, states are free to set their payments at any
level between the minimum required and the maximum permitted.
Individual state plan specifications for DSH payments are often extremely
complex, defining numerous classes of facilities and varying payment amounts; some
states amend this section of their plan every year. Accordingly, this section will not
offer a comparison of the way different states have designed their DSH programs, but
will merely summarize the current requirements.
Hospitals That Must Receive DSH Payments. A hospital must be
deemed a DSH hospital if either of the following is true:
! Its Medicaid utilization rate is more than one standard deviation
above the average Medicaid utilization rate for all
Medicaid-participating hospitals in the state.16 The Medicaid
utilization rate is defined as the number of days of care furnished to
Medicaid beneficiaries during a given period divided by the total
number of days of care provided during the period.
!Its low-income utilization rate is at least 25%. The low-income
utilization rate is the sum of two fractions: Medicaid payments plus
state and local subsidies divided by total patient care revenues, and
inpatient charges attributable to charity care (other than charity care
subsidized by state or local government) divided by total inpatient

15 For additional information on DSH payments, see CRS Report 97-483, Medicaid
Disproportionate Share Payments, by Jean Hearne.
16 The “standard deviation” used in the first criterion is a statistical measure of the
dispersion of hospitals’ utilization rates around the average; the use of this measure
identifies hospitals whose Medicaid utilization is unusually high.

In computing either of these measures, states are now required to include
Medicaid patients whose stays were paid for by an MCO, rather than directly by the
Hospitals That May Not Receive DSH Payments. A state may not make
DSH payments to a hospital whose Medicaid utilization rate is less than 1%. In
addition, a hospital may not be deemed a DSH hospital unless it has on staff at least
two obstetricians who are prepared to accept Medicaid patients. This requirement
does not apply to children’s hospitals or to those that do not furnish non-emergency
obstetrical care; rural hospitals may use other attending physicians for obstetrical
Minimum DSH Payment. In computing the amount of the supplementary
payment, the state must use one of three methods. It may (a) use the formula for
comparable payments under Medicare, with special adjustments for children’s
hospitals; (b) provide for a fixed payment increase or percentage increase for DSHs
plus an additional increase for hospitals whose Medicaid utilization is more than one
standard deviation above the statewide mean; or (c) develop its own methodology
which may vary payments to different types of hospitals, so long as all hospitals of
each type are treated equally and payments are reasonably related to hospitals’
Medicaid or low-income volume. The payments are required even if they result in
Medicaid payments to a hospital in excess of the hospital’s usual charges to the
public for similar services.
Maximum Payment to an Individual Hospital. The DSH payment cannot
exceed the sum of (a) the hospital’s costs for Medicaid patients that are not already
met through non-DSH Medicaid hospital payments and (b) the hospital’s costs for
patients without health insurance or other third-party coverage.17 (Third-party
payment does not include state and local subsidies for indigent care.) California has
a permanent waiver to pay certain “high disproportionate share” public hospitals up
to 175% of this limit. BIPA granted a similar exemption to all states, but only for the
two state fiscal years beginning on or after September 30, 2002.
Maximum DSH Payments to Mental Hospitals. The BBA limited total
DSH payments to mental hospitals during a year to the lesser of the dollar amount
of such payments in FY1995 or a percentage of the state’s DSH allotment (see
below) for the year. This percentage was initially based on the percentage of the
state’s FY1995 payments that went to mental hospitals, then was phased down to

50% for FY2001, 40% for FY2002, and 33% for FY2003 and later years.

DSH Allotments. The Medicaid Voluntary Contribution and Provider-Specific
Tax Amendments of 1991 (P.L. 102-234) limited national aggregate spending for
DSH payments to 12% of total Medicaid program spending, roughly the level
projected for FY1992. “High” DSH states — those with DSH payments already
exceeding 12% of their Medicaid spending — could not increase the percentage of

17 Note that non-DSH Medicaid payments include enhanced payments under UPL
arrangements. Centers for Medicare and Medicaid Services, State Medicaid Director Letter,
no. 02-013, Aug. 16, 2003.

spending devoted to DSH payments. That is, those payments could not increase
faster than the rate of growth in the state’s overall Medicaid spending. Other states
were allowed to raise their DSH payments, subject to an allocation system that would
keep aggregate national payments within the cap.
The BBA of 1997 replaced this formulaic allocation with a table of specified
allotments for each of the years FY1998 through FY2002. These allotments
effectively froze states with very low DSH payments at their 1995 payment levels
and required higher-spending states to gradually reduce their payments. After 2002,
each state’s annual allotment would increase at the rate of the medical care
component of the CPI-U. BIPA froze the allotments for FY2001 and FY2002 at the
FY2000 levels, meaning high-DSH states would not have to reduce their spending
so rapidly. For FY2003, however, the DSH allotment returned to the level prescribed
by the BBA — that is the original published FY2002 allotment plus inflation. Table
11 shows the allotments for FY1998 through FY2003. The reversion to the BBA
rules for FY2003 meant that total allotments dropped about 11% in a single year, and
some states’ allotments dropped by as much as 25%.
Table 11. Federal DSH Allotments for 1998-2003
(millions of dollars)
1998 1999 2000 2001 2002 2003
Alabama 293 269 248 257 263 250
Alaska 10 10 10 10 11 9
Arizona 818181 84 8682
Arkansas 2 2 2 19 19 19
California 1 ,085 1,068 986 1,021 1,047 890
Co lo rado 93 85 79 82 84 75
Co nnecticut 200 194 164 170 174 162
Delawre 444 4 44
District of Columbia232332333432
Florid a 207 203 197 204 209 162
Georgia 253 248 241 249 256 218
Hawa ii a 000 0 00
Idaho 111 7 77
I llino is 2 0 3 1 9 9 1 9 3 2 0 0 2 0 5 1 7 5
Indiana 201 197 191 198 203 174
Iowa 8 8 8 17 17 18
Kansas 51 49 42 43 45 33
Kentucky 137 134 130 135 138 118
Lo uisiana 880 795 713 713 713 631
Maine 103 99 84 87 89 85
Maryland 72 70 68 70 72 62
Massachusetts 288 282 273 283 290 248
Michigan 249 244 237 245 252 215

1998 1999 2000 2001 2002 2003
Minneso ta 33 33 33 34 35 33
Mississippi 143 141 136 141 144 124
Misso uri 436 423 379 392 402 385
Monta .2.2.2 5 5 5
Nebraska 5 5 5 12 13 13
Nevada 37 37 37 38 39 38
New Hampshire140136130130132132
New Jersey600582515533547523
New Mexico5999109
New York1,5121,4821,4361,4861,5251,304
North Carolina278272264273280240
North Dakota111444
Ohio 382 374 363 376 385 330
Oklahoma 16 16 16 17 17 16
Oregon 20 20 20 21 21 20
Pennsylvania 529 518 502 520 533 456
Rhode Island626058606253
South Carolina313303262271278266
South Dakota111555
T e nne sse e a 000 0 00
T exas 979 950 806 834 856 776
Utah 333 8 99
Vermont 181818 19 1918
Virginia 70 68 66 68 70 71
Washington 174 171 166 172 176 150
West Virginia646361636555
Wisconsin 7 7 7 41 42 42
Wyoming bbb b bb
To tal 10,272 9,958 9,278 9,662 9,893 8,748
Source: U.S. Department of Health and Human Services, Centers for Medicare and Medicaid
Services, “Medicaid Program; Disproportionate Share Hospital Payments,” 69 Federal Register
15850-15884, Mar. 26, 2004.

a. Does not make DSH payments
b. Allotments round to less than $1 million.

The Medicare Prescription Drug, Improvement, and Modernization Act of 2003
(P.L. 108-173) provides a one-time 16% increase in the FY2004 allotment for each
state. A state’s allotment will be frozen at this 2004 level until the year for which the
Secretary estimates that the allotment that would have been available under the
previous rules (that is, the FY2003 allotment plus cumulative inflation) equals or
exceeds the 2004 amount. For that year, the state would receive the allotment
determined under the previous rules; for subsequent years, allotments would again
increase at the rate of the medical care component of the CPI-U. There is an
exception for a “low DSH” state, one whose FY2000 DSH spending was greater than
zero but less than 3% of the state’s total Medicaid spending. A low DSH state will
receive the 16% increase in its allotment for FY2004 and an additional 16% increase
for each fiscal year through FY2008. For FY2009 and later years, the allotment
would increase with the medical care component of the CPI-U.
Amount of DSH Payments. Table 12 shows DSH payments as a percentage
of total spending for general and mental hospital services, and as a percentage of all
Medicaid spending, in FY2001. Overall, DSH payments account for about a third
of payments for general hospitals and nearly half of payments for mental hospitals.18

About 21% of total DSH payments went to mental hospitals.
18 Some states show payments to inpatient mental hospitals that exceed the 50% of total
DSH ceiling as imposed under the BBA. This may be because CMS-64 reports for a given
year can reflect spending related to a prior year.

Table 12. Disproportionate Share Hospital Payments,
as a Share of Total Hospital Payments and Total Net Medicaid Spending, 2001
General hospitalInpatient mental hospitalIMH asDSH aspercent of
percent oftotal MedicaidaPercent
Statetotal DSHspendingRegularDSHDSHRegularDSHPercent DSH
aska 161 5 3 % 1 4 9 39% 65% 2%
abama31636353%36 39%1%13%
kansas 364 22 6% 68 1 1 % 4 % 1 %
izona1417435% 02899%28%4%
lifornia4,3561,92631%1,021 0%0%8%
iki/CRS-RL32644lo rado 303 186 38% 2 0 1% 0% 9%
s.ornnecticut 224 203 48% 9 8 8 90% 30% 9%
leakstrict of Columbia2467924%29412%5%8%
://wikilaware26 0%11428%100%1%
httporid a 1 ,661 189 10% 9 150 94% 44% 4%
orgia1,87442518%28 0%0%8%
a242146%21 0%0%1%
o126107%7 0%0%1%
o is 2 ,7 5 7 2 6 4 9 % 4 0 1 1 5 7 4 % 3 0 % 5 %
diana 737 514 41% 182 142 44% 22% 16%
nsas 171 11 6% 31 36 54% 77% 3%
ntucky 613 155 20% 44 36 46% 19% 6%
uisiana 687 795 54% 7 7 7 92% 9% 21%

General hospitalInpatient mental hospitalIMH asDSH aspercent of
percent oftotal MedicaidaPercent
Statetotal DSHspendingRegularDSHDSHRegularDSHPercent DSH
sachusetts 1,098 382 26% 44 103 70% 21% 7%
land 447 31 7% 146 31 18% 50% 2%
ne200 0%194973%100%4%
chigan 1,079 217 17% 37 215 85% 50% 6%
nneso ta 272 62 18% 31 3 8 % 4 % 2 %
sso uri 897 279 24% 11 176 94% 39% 10%
ssissippi65917921%40 0%0%7%
iki/CRS-RL32644tana9300% 0%0%0%
g/wrth Carolina1,48424014%2617587%42%7%
leakrth Dakota5501%3118%57%0%
://wikibraska15500%3 0%0%0%
httpw Hampshire7213164% 32890%17%18%
w Jersey89170544%102 41380%37%16%
w Mexico229156%2013%2%1%
vada1407635%15 0%0%11%
w York6,4021,88123%50057453%23%8%
io 1,486 544 27% 279 93 25% 15% 8%
lahoma 138 21 13% 25 1 5% 6% 1%
egon 188 13 7% 40 17 30% 57% 1%
nsylvania 570 361 39% 152 400 72% 53% 7%
ode Island18179 30%1928%2%7%

General hospitalInpatient mental hospitalIMH asDSH aspercent of
percent oftotal MedicaidaPercent
Statetotal DSHspendingRegularDSHDSHRegularDSHPercent DSH
uth Carolina60432135%345160%14%12%
uth Dakota8500%3119%70%0%
xas 1,111NA5223582%17%12%
ah 124 1 0 % 1 1 0 2% 25% 0%
rginia 520 235 31% 163 2 1 % 1 % 8 %
rmont752726%0 0%0%4%
iki/CRS-RL32644hington 564 213 27% 60 115 66% 35% 8%
g/wt Virginia2367925%222352%23%7%
leakoming4401%15 0%0%0%
://wikiS. total (excluding territories)34,84812,44826%3,4563,40650%21%7%
: Medicaid Financial Management Report (CMS-64), FY 2001. General hospital includes inpatient and outpatient spending.
otal does not include administrative spending.

Note that the non-DSH figures include any enhanced payments under UPL
programs. In addition, the figures on the DSH share of general hospital spending
should be viewed with caution, because states with large numbers of enrollees in
MCOs may make DSH payments, but not regular payments, on behalf of those
Tables 13 and 14 provide some further perspective on how states are allocating
their DSH funds. These tables are based on the most recent annual DSH report filed19
by each state and posted on the CMS website. While submission of annual reports
is required by the statute, one state with DSH spending, Georgia, has never filed a
report, while others have not done so for some years. (Some states have filed reports
that do not fully categorize all hospitals by type or ownership. These omissions have
been corrected when there were only a few instances in a state and the information
was readily available from other sources.)
States vary widely in the degree to which they have targeted payments at public
hospitals and mental hospitals. In some states, nearly all the payments went to
private general hospitals; in others, nearly all payments went to public mental
hospitals. (These are states for which only older reports are available, so that the
figures do not reflect the BBA-required phase-down of the share of payments going
to mental hospitals.) In addition, some states are distributing the funds among a large
number of hospitals, while other make DSH payments only to a handful of facilities.

19 [].

Table 13. Disproportionate Share Hospital Payments by Type of Hospital and Hospital Ownership,
Most Recent Reporting Year
DSHGeneral hospitalInpatient mental health facility
StateYear(millions)PublicPrivateUnknown ownershipPublicPrivateUnknown ownership
abama1998$ 140.1%0.0%0.0%99.9%0.0%0.0%
aska 2000 395 10.8% 0.0% 87.8% 1.4% 0.0% 0.0%
izona 1998 2 9 .7% 68.8% 9.6% 0.0% 11.9% 0.0%
kansas 1999 122 86.8% 13.2% 0.0% 0.0% 0.0% 0.0%
lifornia 2000 1,908 75.4% 24.4% 0.2% 0.0% 0.0% 0.0%
iki/CRS-RL32644lo rado 1999 175 90.3% 9.7% 0.0% 0.0% 0.0% 0.0%
s.ornnecticut 2001 291 0.0% 69.9% 0.0% 30.1% 0.0% 0.0%
leaklaware 1999 33 14.9% 76.7% 0.0% 8.4% 0.0% 0.0%
://wikistrict of Columbia1999350.0%0.0%0.0%100.0%0.0%0.0%
httporid a 1998 371 11.6% 88.4% 0.0% 0.0% 0.0% 0.0%
r gia a
a ii b
o 2000 14 92.8% 7.2% 0.0% 0.0% 0.0% 0.0%
ois 2000 1 41.3% 58.7% 0.0% 0.0% 0.0% 0.0%
diana 1998 433 0.0% 0.0% 62.8% 0.0% 0.0% 37.2%
a 2001 116 29.5% 6.0% 0.0% 64.5% 0.0% 0.0%
nsas 1999 44 5.9% 6.0% 0.0% 88.0% 0.0% 0.0%
ntucky 2000 184 0.0% 0.0% 80.9% 17.7% 1.4% 0.0%
uisiana 1998 734 98.9% 0.4% 0.7% 0.0% 0.0% 0.1%

DSHGeneral hospitalInpatient mental health facility
StateYear(millions)PublicPrivateUnknown ownershipPublicPrivateUnknown ownership
ne 2001 553 26.7% 55.3% 0.0% 17.8% 0.3% 0.0%
land 2001 81 0.2% 49.8% 0.0% 50.0% 0.0% 0.0%
sachusetts 2000 50 0.0% 0.0% 0.0% 76.6% 23.4% 0.0%
chigan 1998 215 0.8% 1.1% 0.0% 98.1% 0.0% 0.0%
nneso ta 1997 56
ssissippi 1998 455 0.0% 0.0% 60.8% 0.0% 0.0% 39.2%
sso uri 2001 183 98.7% 1.1% 0.0% 0.0% 0.2% 0.0%
iki/CRS-RL32644tana 2000 0 c 0.0% 0.0% 97.8% 0.0% 0.0% 2.2%
g/wbraska 1999 339 44.3% 6.6% 0.0% 49.1% 0.0% 0.0%
leakvada 1998 1 0 .0% 43.3% 0.0% 56.7% 0.0% 0.0%
://wikiw Hampshire2001445.6%1.2%0.0%0.0%53.2%0.0%
httpw Jersey199917574.9%10.3%0.0%14.8%0.0%0.0%
w Mexico20009830.0%0.0%51.9%0.0%0.0%48.1%
w York20011280.0%19.5%0.0%0.0%0.4%0.0%
rth Carolina19987490.6%9.4%0.0%0.0%0.0%0.0%
rth Dakota20011,19144.3%31.2%0.0%24.0%0.5%0.0%
io 2001 636 0.0% 0.0% 85.3% 0.0% 0.0% 14.7%
laho ma 1999 23 1.8% 83.7% 0.0% 12.4% 2.1% 0.0%
egon 2000 25 19.6% 1.6% 0.0% 78.8% 0.0% 0.0%
nsylvania 1999 52 0.0% 0.0% 100.0% 0.0% 0.0% 0.0%
ode Island20018110.3%87.8%0.0%0.0%1.9%0.0%

DSHGeneral hospitalInpatient mental health facility
StateYear(millions)PublicPrivateUnknown ownershipPublicPrivateUnknown ownership
uth Carolina199943456.5%35.0%0.0%8.3%0.2%0.0%
uth Dakota200114.1%26.1%0.0%69.9%0.0%0.0%
nne sse e b
xas 2001 1,183 56.8% 23.3% 0.0% 19.8% 0.1% 0.0%
ah 2001 4 43.7% 36.5% 0.0% 19.8% 0.0% 0.0%
rmont 2001 164 81.1% 13.6% 0.0% 0.0% 5.3% 0.0%
rginia 1999 26 0.0% 100.0% 0.0% 0.0% 0.0% 0.0%
iki/CRS-RL32644hington 2000 277 54.7% 5.4% 0.0% 39.6% 0.4% 0.0%
g/wt Virginia2000110.2%75.0%0.0%24.0%0.8%0.0%
leaksconsin 1998 80 14.2% 71.5% 0.0% 14.4% 0.0% 0.0%
b 0.9% 99.1% 0.0% 0.0% 0.0% 0.0%
://wikioming1999 $ 0
http: State DSH reports, latest available year.
DSH report filed.
es not make DSH payments.

Table 14. Hospitals Receiving Disproportionate Share Hospital Payments by Type of Hospital
and Hospital Ownership, Most Recent Reporting Year
HospitalsGeneral hospitalInpatient mental health facility
receiving DSHUnknownUnknown
St a t e Y ea r payments Public Priv a t e o w nership P ublic P r iv a t e ow nership
abama 1998 12 25% 0% 67% 8% 0% 0%
aska 2000 2 50% 0% 0% 50% 0% 0%
kansas 1999 11 18% 45% 27% 0% 9% 0%
izona 1998 30 7% 93% 0% 0% 0% 0%
lifornia 2000 131 29% 62% 8% 1% 0% 0%
iki/CRS-RL32644lo rado 1999 65 35% 58% 0% 3% 3% 0%
g/wnnecticut 2001 33 0% 94% 0% 6% 0% 0%
leakstrict of Columbia1999911%78%0%11%0%0%
://wikilaware 1999 1 0 % 0 % 0 % 100% 0% 0%
httporid a 1998 78 18% 82% 0% 0% 0% 0%
r gia a
a ii b
a 2001 27 59% 41% 0% 0% 0% 0%
o 2000 36 67% 33% 0% 0% 0% 0%
ois 2000 88 0% 0% 82% 0% 0% 18%
diana 1998 10 30% 20% 0% 50% 0% 0%
nsas 1999 31 61% 29% 0% 10% 0% 0%
ntucky 2000 117 0% 0% 90% 3% 8% 0%
uisiana 1998 85 61% 11% 16% 0% 0% 12%

HospitalsGeneral hospitalInpatient mental health facility
receiving DSHUnknownUnknown
St a t e Y ea r payments Public Priv a t e o w nership P ublic P r iv a t e ow nership
sachusetts 2000 83 8% 81% 0% 10% 1% 0%
land 2001 20 10% 55% 0% 35% 0% 0%
ne 2001 4 0 % 0 % 0 % 50% 50% 0%
chigan 1998 22 9% 64% 0% 27% 0% 0%
nneso ta 1997 1 0 % 0 % 100% 0% 0% 0%
sso uri 2001 141 0% 0% 89% 0% 0% 11%
ssissippi 1998 55 45% 45% 0% 0% 9% 0%
iki/CRS-RL32644tana 2000 8 0 % 0 % 75% 0% 0% 25%
g/wrth Carolina199813353%32%6%6%4%0%
s.orrth Dakota200170%86%0%14%0%0%
braska 1999 12 25% 25% 0% 0% 50% 0%
httpw Hampshire20012983%14%0%3%0%0%
w Jersey1999770%0%82%0%0%18%
w Mexico20002524%68%0%0%8%0%
vada 1998 11 73% 27% 0% 0% 0% 0%
w York200126510%80%0%9%0%0%
io 2001 173 0% 0% 97% 0% 0% 3%
lahoma 1999 14 7% 21% 0% 36% 36% 0%
egon 2000 11 18% 64% 0% 18% 0% 0%
nsylvania 1999 1 0% 0% 100% 0% 0% 0%
ode Island2001147%79%0%0%14%0%
uth Carolina19995250%42%0%6%2%0%

HospitalsGeneral hospitalInpatient mental health facility
receiving DSHUnknownUnknown
St a t e Y ea r payments Public Priv a t e o w nership P ublic P r iv a t e ow nership
uth Dakota2001128%83%0%8%0%0%
nne sse e b
xas 2001 171 53% 39% 0% 6% 2% 0%
ah 2001 29 3% 93% 0% 3% 0% 0%
rginia 1999 42 5% 76% 0% 0% 19% 0%
rmont 2001 14 0% 100% 0% 0% 0% 0%
hington 2000 61 70% 25% 0% 2% 3% 0%
iki/CRS-RL32644sconsin 1998 25 4% 68% 0% 16% 12% 0%
g/wt Virginia2000593%95%0%2%0%0%
s.oroming 1999 3 33% 67% 0% 0% 0% 0%
://wiki: State DSH reports, latest available year.
DSH report filed.
es not make DSH payments.

These data alone cannot indicate which states are using DSH payments for the
intended purpose of helping hospitals with low-income and uninsured patients, and
which are probably recovering the funds through transfers or using them to pay for
non-Medicaid residents of psychiatric facilities. States vary, for example, in the
share of general hospitals operated by government units, and those with few public
general hospitals (such as Maryland) are likely to make more payments to private
providers than states where public hospitals are more common.
Adequacy of Hospital Reimbursement
Since the shift away from cost-based reimbursement that began in the 1980s,
aggregate Medicaid payments to hospitals (including regular inpatient and outpatient
payments and DSH payments) have consistently been less than the total costs
hospitals incur in treating Medicaid beneficiaries. However, the gap narrowed
dramatically during the 1990s.
The only comprehensive source of data on Medicaid hospital costs and
payments is an annual survey of community hospitals conducted by the American
Hospital Association (AHA). The survey includes questions about gross Medicaid
charges and actual Medicaid payments received by each hospital. Hospitals’ charges
are generally in excess of their actual costs. AHA estimates actual costs for Medicaid
patients at each hospital by using that hospital’s overall cost-to-charge ratio; the
estimate may be inaccurate if the ratio is actually different for Medicaid and
non-Medicaid patients. It should also be noted that 35% of hospitals — especially
public and for-profit hospitals — did not participate in the most recent survey; values
for these hospitals have been imputed.
Table 15 shows payments by Medicaid and other major payers as a percentage
of costs in 1991 through 2001. Nationally, aggregate Medicaid payments were
81.6% of estimated costs for Medicaid beneficiaries in 1991. The ratio rose steadily
through the decade; by 2001, aggregate Medicaid payments — including regular and20
DSH payments — equaled 98% of costs. While this is a significant improvement,
hospitals overall are still losing money on Medicaid patients. AHA estimates that21

73% of hospitals had negative inpatient margins in 2000.

One likely factor in Medicaid losses is that states have been granting annual
increases lower than the rate of inflation. A study for Oregon’s hospital association
by Lewin Associates contends that, over a 10-year period, Medicaid payment rates
increased 13%, while inflation was 33%. In addition, the study notes that rates for
each hospital continued to be based on data from 1987. This means that rates do not
account for changes in case mix or other factors that might cause one hospital’s costs
to rise faster than another’s. Examination of state plan documents indicates that a
number of other states allow long intervals to pass without “rebasing” their hospital
cost data.

20 The Medicaid payment figures are net revenues; that is, they do not include any amounts
that might have been returned to the state through intergovernmental transfers or other
21 AHA, Cracks in the Foundation: Averting a Crisis in America’s Hospitals, Aug. 2000.

Table 15. Hospital Payment-to-Cost Ratios,
by Source of Revenue, 1991-2001
(in percentages)
Uncompensa ted Private
Year M e dicare M e dicaid care payers
1991 88.4 81.6 19.6 129.7
1992 88.8 90.9 18.9 131.3
1993 89.4 93.1 19.5 129.3
1994 96.9 93.7 19.3 124.4
1995 99.3 93.8 18.0 123.9
1996 102.4 94.8 17.3 121.5
1997 103.6 95.9 14.1 117.6
1998 102.6 97.9 13.2 113.6
1999 101.1 96.7 13.2 112.3
2000 100.2 96.1 12.1 112.5
2001 99.4 98.0 12.2 113.2
Source: Medicare Payment Advisory Commission (MedPAC), Report to the Congress, Medicare
Payment Policy, Mar. 2003, based on data from the American Hospital Association annual survey of
ho sp itals.
Note: Payment-to-cost ratios indicate the relative degree to which payments from each payer cover
the costs of treating its patients. Operating subsidies from state and local governments are considered
payments for uncompensated care, up to the level of each hospital’s uncompensated care costs. Data
are for community hospitals and reflect all types of patient care services. Imputed values are used for
missing data (about 35% of observations),which corrects for underrepresentation of proprietary and
public hospitals relative to voluntary institutions. Most Medicare and Medicaid managed care patients
are included in the private payers category. The costs allocated to Medicare and Medicaid include
CMS’s allowed and nonallowed costs. [This note by MedPAC means that its method for estimating
hospitals costs for Medicare and Medicaid patients does not take account of federal rules for
determining whether specific costs are reimbursable.]
Hospital losses are not due solely to reimbursement methods. Coverage
limitations also play a role. For example, under Maryland’s all-payer system,
Medicaid pays the same daily rates as Medicare and private insurers. However, the
state has imposed a limit on the number of covered inpatient days. The result is a
loss for uncovered days that is passed on to all payers in the form of higher rates.22
Hospitals and private insurers have always contended that Medicaid losses must
be made up through higher charges to other payers, a phenomenon known as
cost-shifting. Private insurers pay more than the costs of treatment for their
enrollees, while both Medicaid and Medicare pay less than cost. What is striking is
how much cost-shifting has diminished. As Table 15 shows, charges to private
payers were nearly 30% above costs in 1991, largely to compensate for Medicare and
Medicaid losses. By 2001, private payers paid 13% above costs. Some of the change

22 “Maryland Health Cuts to Mean Higher Insurance Rates,” Baltimore Sun, Aug. 3, 2003,
p. 1D.

is probably attributable to pressure for lower prices from major managed care plans
and other insurers. But hospitals might have been more willing to accept lower
prices from private payers because their losses from the public programs had dropped
so much.
One final point to be made about the figures in Table 15 is that the proportion
of uncompensated care costs made up through state or local subsidies has gone from
19.6% in 1991 to 12.2% in 2001, a drop of more than one-third. Yet losses from
uncompensated care have not risen proportionately; as a percent of total hospital
costs, the losses were 4.8% in 1991 and 5.3% in 2000. One possible explanation is
that non-Medicaid subsidies were partially replaced by DSH payments to safety net
The possible role of Medicaid in offsetting some hospitals’ losses from bad
debts or charity care appears to vary by state. Table 16 is based on the 2000 annual
member survey of the National Association of Public Hospitals and Health Systems
(NAPH). This organization chiefly represents large state and local hospitals; a few
members are operated by private, non-profit corporations but function as “safety net”
providers, treating substantial numbers of Medicaid and uninsured patients. The
table thus illustrates the experience of a few major safety net providers in each state
listed, and may not be representative of all hospitals, or even comparable safety net
hospitals, in each state.23
The table first compares estimated costs for Medicaid patients and Medicaid
revenues, including DSH payments. Of the 20 states listed, 10 were paying the
reporting hospitals less than their Medicaid costs. In the other 10 states, Medicaid
revenues exceeded costs, sometimes substantially. The table then compares costs and
revenues for patients classed as “self pay/other.” This group is made up of all
patients without private insurance, Medicaid, or Medicare, including the uninsured
and people with coverage through CHAMPUS, workers compensation, and other
sources. Offsetting revenues for this group include various forms of non-Medicaid
public funding, such as local subsidies or state indigent care pools), and other
funding sources for the self-pay/other population. In nearly all the states, the
reporting hospitals incurred sizeable losses for this population. (Exceptions may be
artifacts of the method of estimating costs from gross charges.)

23 The survey, like the AHA survey, ascertains Medicaid and self-pay/other gross charges
but not costs. As in the AHA data, the estimates in the table assume that the Medicaid and
self-pay/other cost/charge ratios are the same as the overall cost/charge ratio reported by the

Table 16. Estimated Costs and Revenues, Medicaid and Self-Pay/Other Patients,
NAPH Member Hospitals, 2000
(millions of dollars)
MedicaidSelf-pay and otherCombinedMedicaid/
self-pay/otherCount ofPercent gain/Percent gain/
gain/lossStateentitiesEstimated costRevenuelossEstimated costRevenueloss
lifornia15$1,812$1,797-1%$1,323$947 -28%-12%
lo rado 2 $112 $165 48% $198 $45 -77% -32%
orid a 9 $387 $301 -22% $374 $416 11% -6 %
orgia 1 $153 $194 27% $214 $101 -53% -20%
waii 3 $30 $24 -21% $14 $15 10% -11%
iki/CRS-RL32644a 1 $13 $6 -52% $45 $38 -15% -23%
g/wo is 1 $ 4 0 $ 4 1 3 % $ 3 7 $ 2 3 -3 9 % -1 7 %
s.ordiana 1 $81 $105 29% $100 $66 -34% -6 %
leakuisiana 9 $184 $624 238% $456 $4 -99% -2 %
://wikisachusetts 2 $185 $120 -35% $210 $208 -1 % -17%
httpnneso ta 1 $140 $128 -9 % $24 $41 71% 3%
sso uri 1 $84 $77 -9 % $65 $64 -2 % -6%
w Mexico1$64$8737%$91$55-39%-8%
vada 1 $65 $66 1% $78 $67 -14% -7%
w York12$1,772$1,7881%$731$496-32%-9%
io 2 $136 $154 13% $102 $56 -45% -12%
nnessee 3 $184 $112 -39% $86 $65 -25% -34%
xas5$513$495-4%$1,022 $ 958-6%-5%
rginia 1 $64 $79 24% $128 $109 -15% -2 %
hington 1 $116 $98 -16% $46 $33 -28% -19%
: Authors calculations from National Association of Public Hospitals and Health Systems, America’s Safety Net Hospitals and Health Systems, 2000.
: Assumes Medicaid and self-pay/other cost/charge ratio equal to overall cost/charge ratio. The table omits five hospitals, one in Alabama, two in Illinois, and two in New York,
eported total costs greater than total charges.

In some states it appears clear that Medicaid reimbursement is reducing the
burden of uncompensated care. Louisiana is paying the responding hospitals, all
operated by Louisiana State University, more than twice their Medicaid costs. (The
nine hospitals received about one-third of Louisiana’s DSH funding in 1998.) The
excess nearly exactly offsets the hospitals’ bad debt and charity care costs. Hospitals
in several other states, such as Colorado, Indiana, and New Mexico, received
considerably more in Medicaid payments than their Medicaid costs. In these states,
however, the Medicaid payments and other public subsidies were insufficient to
offset costs for uncompensated care.
The two states with the most responding hospitals, California and New York,
paid these hospitals amounts roughly equal to their estimated Medicaid costs. The
15 California hospitals received $1.2 billion in DSH funds in 2000, or 64% of the
state’s total DSH spending of $1.9 billion. In New York, the 12 hospitals received
$407 million in 2001, or 34% of the state’s $1.2 billion in DSH spending. But these
payments were just sufficient to bring Medicaid payments close to Medicaid costs,
with no excess available to subsidize other patients.
Physician and Dental Care
Medicaid payment levels for physician and dental care, and their effects on
provider participation and beneficiary access, have been issues since the earliest years
of the program. States have commonly paid independent practitioners using fixed fee
schedules, often at rates well below those paid by Medicare or private insurers.
Many physicians refused to accept Medicaid patients or limited their Medicaid
caseloads, leaving beneficiaries to rely on more costly hospital outpatient
departments and emergency rooms as a primary source of care.
Medicaid payments to physicians and other providers are subject to the general
requirement that payments be sufficient to attract enough providers to ensure that
covered services will be as available to Medicaid beneficiaries as they are to the
general population. OBRA 89 codified this requirement (previously established only
by regulation) and established specific reporting requirements with respect to
payment rates for obstetric and pediatric services, to allow the Secretary to determine
the adequacy of state payments for these services. These special reporting
requirements were repealed by the BBA, but the requirement that payments be
sufficient to assure access remains in the statute. It has been the basis for numerous
lawsuits by groups of physicians, dentists and other providers.24
This section provides data on payment levels for physician and dental services
and summarizes some recent literature on how these payment levels affect access to

24 For reviews of recent litigation, see National Health Law Program, Docket of Medicaid
Cases to Improve Provider Participation, Feb. 23, 2003, at [
conf2003/provider-docket.htm] and summaries by the American Dental Association, at
[], as of Sept. 2003.

Physician Payment. Every state except Hawaii now pays physicians the
lesser of actual charges or a fixed fee schedule amount for each visit or procedure,25
whether performed in offices, hospitals, or other settings. States set these fee
schedules in various ways. Some were originally based on physicians’ actual charges
for services, while others are set arbitrarily by the state or negotiated with provider
groups. Others use systems comparable to Medicare’s, under which each procedure
is assigned a weight on a resource-based relative value scale (RBRVS); the
weightings reflect relative physician work, practice expenses, and malpractice costs
associated with different procedures. A brief physician office visit might have a
value of three, an appendectomy a value of 150. The state then multiplies the
different values by a single standard dollar amount. If a unit is valued at $5, the state
will pay $15 for the brief office visit and $450 for the appendectomy. Some states
have adopted Medicare’s scales, while others use different weighting systems. The
effect is the same as under a fee schedule, except that the Medicaid agency has an
external reference for its pricing decisions.
However the schedule is established, basic rates and/or inflation increases are
fixed by the state and may bear no relation to what physicians ordinarily charge or
what they are paid by Medicare or private insurers. Table 17 compares each state’s
Medicaid rates in 2001 for selected procedures. The rates are those reported by state
Medicaid directors in an annual survey conducted by the American Academy of
Pediatrics (AAP). The AAP collects data on a large number of different procedures;
the five shown here were selected arbitrarily as representative of broad classes of
services: primary care, mental health, and so on. (Unfortunately, because AAP
focuses on pediatric care, its procedure list does not include obstetric services,
payment for which has been a long-standing issue in Medicaid programs.)
As the table shows, states’ payment rates vary enormously. Leaving aside
Alaska, an outlier because of its high cost of living, rates for an initial pediatric
preventive office visit range from $20 in Pennsylvania to $114.87 in New Mexico,
almost six times as much. Payment for a complex procedure like a cardiac
catheterization ranges from $80 in New York to $1,688 in Arizona, a twenty-fold
difference. 26

25 Hawaii continues to use the “reasonable charge” method used by Medicare before
Medicare adopted its own fee schedule: the reasonable charge for a specific service is the
lowest of (a) the provider’s actual charge for that service; (b) the provider’s customaryth
charge for comparable services; or (c) the “prevailing” charge in the area, fixed at the 75
percentile of charges for comparable services.
26 This rate applies to individuals not enrolled in the Arizona Health Care Cost Containment
System (AHCCCS), the managed care program that serves most Medicaid beneficiaries in

Table 17. Medicaid Payment Rates for Selected Physician Procedures, 2001
P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GILeft heart
patient, age 1-4aminutes (90806,amoderate complexityaendoscopy, biopsyaAppendectomyacatheterizationa
(99382)nonfacility)(99222)(43239, nonfacility)(44950)(93510)
abama$70.00$63.00$76.00$169.00$405.00 $894.00
aska $160.07 $142.29 $174.65 $378.70 $823.44 $2,761.37
izona $101.18 $94.11 $114.86 $240.85 $534.28 $1,687.99
kansas $51.28 $88.13 $84.00 $373.00 $488.00 $356.00
lifornia $47.13 $46.44 $73.20 $234.18 $400.59 $1,038.99
lo rado $55.05 NA $88.37 $20.06 $334.30 $441.55
iki/CRS-RL32644nnecticut $50.00 $50.00 $51.40 $161.36 $374.70 $172.11
s.orlaware $97.52 NA $107.73 $227.94 $496.04 $1,596.11
leakstrict of Columbia*$45.00*NP$36.00*$123.00*$267.00*$108.00*
://wikiorid a NA $50.34 $61.42 $129.63 $527.99 $145.57
httporgia $55.38* NP $104.28 $219.02 $463.03 $1,526.14
waii $31.50 $84.71 $73.90 $233.73 $429.07 $267.77
o $59.20 $57.54 $117.35 $327.29 $557.96 $1,644.41
ois $44.30 $50.25 $54.43 $264.35 $396.45 $770.05
diana $34.52 $63.67 $80.67 $181.60 $314.84 $1,167.49
a $44.36 $60.28 $60.31 $394.19 $688.77 $526.22
nsas $35.00 $60.00 $69.54 $220.00 $268.00 $1,431.45
ntucky $79.91 $65.73 $84.07 $179.22 $333.52 $1,223.29
uisiana $36.90 $76.70 $41.40 $177.66 $343.81 $232.65
ne $50.20* $73.60 $63.05 $166.95 $297.19 $423.91

P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GILeft heart
patient, age 1-4aminutes (90806,amoderate complexityaendoscopy, biopsyaAppendectomyacatheterizationa
(99382)nonfacility)(99222)(43239, nonfacility)(44950)(93510)
land $37.00 $40.50 $24.50 $234.00 $206.00 $80.00
sachusetts $90.86 $66.22 $81.05 $192.99 $397.28 $220.84
chigan $62.13 $59.89 $66.39 $85.01 $322.09 $1,013.16
nneso ta $34.82 $67.77 $100.42 $325.99 $610.27 $463.50
ssissippi $37.63 $78.80 $60.59* $175.64* $304.20* $1,310.01
sso uri $23.00 NC $25.00 $110.00 $251.00 $165.00
tana $58.47 $82.03 $99.65 $205.69 $455.63 $923.99
iki/CRS-RL32644braska $72.80 $79.49 $71.28 $201.20 $467.90 $194.40
g/wvada $59.07 $81.62 $99.66 $17.72 $799.87 $1,541.19
leakw Hampshire$40.00$65.00$86.00$126.00$284.00$900.00
://wikiw Jersey$22.00$37.00$22.00$163.00$211.00$1,045.00
httpw Mexico$114.87$85.72$104.17$228.71$476.03$1,473.62
w York$30.00NP$10.00$100.00$160.00$80.00
rth Carolina$77.75$89.97$109.26OM$495.82$1,550.72
rth Dakota$84.59$78.38$95.04$196.29$433.40$1,332.53
io $57.61 $57.10 $55.71 $172.53 $353.21 $1,175.83
laho ma $67.97 $63.03 $76.40 $157.11 $345.25 $1,054.26
egon $71.88 NC $76.81 $118.59 $372.64 $1,172.16
nsylvania $20.00 NA $29.50 $211.50 $301.50 $187.50
ode Island$37.00*NP$44.00$184.80$248.30*$235.20
uth Carolina$38.00$55.94$38.00$152.44$321.72$1,027.53

P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GILeft heart
patient, age 1-4aminutes (90806,amoderate complexityaendoscopy, biopsyaAppendectomyacatheterizationa
(99382)nonfacility)(99222)(43239, nonfacility)(44950)(93510)
uth Dakota$28.30NP$86.30$317.50$529.20$1,293.81*
nnesseeNo fee-for-service program
xas $49.01 $64.10 $82.65 $207.84 $343.68 $1,307.07
ah $61.94 $55.23 $71.26 $148.80 $316.88 $867.91
rmont $62.46 NP $77.05 $116.10* $270.60* $147.60*
rginia $74.21 $67.44 $82.29 $173.68 $380.70 $1,218.66
hington $67.58 $57.27 $70.02 $156.14 $321.01 $1,040.21
iki/CRS-RL32644t Virginia$76.64$74.01$81.91$186.75$398.26$1,252.20
g/wsconsin $31.39 NA $64.72 $471.66 $522.39 $402.25
leakoming $45.00 $60.00 $92.34 $270.90 $630.00 $260.00
://wiki: American Academy of Pediatrics, Medicaid Reimbursement Survey, 2001.
: NA=Not applicable. NC=Not covered. OM=Other method. NP=Information not provided by state.
a provided by state in 1998/1999 survey.
he codes are from Current Procedural Terminology (CPT), Fourth Edition, developed by the American Medical Association and used by CMS in determining physician payment
a mo unt s.

Table 18 shows the reported Medicaid rates as a percentage of the 2001
Medicare rate for the same state. Rates under the Medicare fee schedule are partially
adjusted using Geographic Practice Cost Indices (GPCIs), which reflect differences
in the costs of practicing medicine in different areas. Sometimes there is one GPCI
for a whole state, in which case the Medicare rate used for comparison is the
statewide rate. Sometimes there are different GPCIs for different parts of a state, in27
which case the Medicare comparator is for the specific area noted in the table.
Although states generally pay less than Medicare for the listed services, the gap
varies considerably by procedure. In states that have systems similar to Medicare’s
RBRVS, but use a different dollar multiplier to establish Medicaid rates, the ratio of
Medicaid to Medicare rates will be roughly constant. In other states that have
assigned their own values to different procedures, rates may be far below Medicare’s
for some services and higher for others. The payment gap tends to be larger for
preventive office visits and for cardiac catheterization than for the other listed
The gap between Medicaid physician payment rates and rates paid by private
insurers is likely to be even greater. Studies done for the Medicare Payment
Advisory Commission (MedPAC) estimate that Medicare physician rates were about28

83% of average private rates in 2001.

The relationship between Medicaid fee-for-service payment rates and
physicians’ willingness to accept Medicaid patients was extensively studied in the
1970s and 1980s, but has received less attention recently — perhaps because policy
focus has shifted to access by enrollees in managed care arrangements. While these
earlier studies did show a positive relationship between payment levels and
participation, changes in the health care marketplace and other factors might mean
that the results would be different now.

27 AAP’s own published comparison uses the national rate before application of the GPCIs.
As local Medicare rates for the five listed procedures varied by as much as 44% in 2001, use
of local rates seemed preferable.
28 The difference is much smaller than in 1994, when Medicare paid 66% of average private
rates. MedPAC attributes the change to shifts from indemnity plans to lower-paying HMOs
and PPS. MedPAC, Report to the Congress: Medicare Payment Policy, 2003.

Table 18. Medicaid Payment Rate as a Percentage of Medicare Physician Fee Schedule, 2001
P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GI endoscopy,Left heart
patient, age 1-4aminutes (90806,amoderate complexityabiopsy (43239,aAppendectomyacatheterizationa
(99382) no nf a c ilit y ) (99222) no nf a c ilit y ) (44950) (93510)
abama 56% 57% 59% 52% 65% 44%
aska 156% 149% 157% 146% 158% 182%
izona 102% 102% 108% 96% 108% 117%
kansas 47% 88% 72% 132% 88% 21%
lifornia (Los Angeles)38%43%58%73%66%53%
lo rado 50% NA 77% 7% 61% 26%
iki/CRS-RL32644nnecticut 41% 46% 41% 50% 62% 9%
s.orlaware 79% NA 85% 71% 82% 81%
leakstrict of Columbia40%*NA30%*42%*47%*6%*
://wikiorida (Miami)NA48%49%42%85%8%
httporgia (Atlanta)48%*NA87%74%82%85%
waii 26% 81% 61% 75% 73% 14%
o 58% 61% 108% 127% 110% 109%
ois (Chicago)43%54%50%102%78%51%
diana 29% 60% 65% 59% 52% 61%
a 42% 63% 54% 148% 133% 34%
nsas 34% 63% 63% 84% 51% 92%
ntucky 78% 69% 76% 69% 64% 81%
uisiana (New Orleans)34%77%36%64%62%14%
ne (southern)40%*67%50%51%49%21%

P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GI endoscopy,Left heart
patient, age 1-4aminutes (90806,amoderate complexityabiopsy (43239,aAppendectomyacatheterizationa
(99382) no nf a c ilit y ) (99222) no nf a c ilit y ) (44950) (93510)
yland (Baltimore)32%39%20%79%36%4%
sachusetts (Boston)82%67%70%68%73%13%
chigan (Detroit)53%56%52%28%50%53%
nneso ta 32% 69% 88% 118% 115% 28%
ssissippi 35% 80% 53%* 64%* 56%* 80%
ssouri (St. Louis)23%NA23%44%50%11%
tana 58% 88% 92% 80% 89% 61%
iki/CRS-RL32644braska 69% 83% 64% 75% 90% 12%
g/wvada 58% 88% 92% 7% 158% 102%
leakw Hampshire40%70%80%49%57%61%
://wikiw Jersey (northern)20%37%19%56%38%59%
httpw Mexico92%78%81%70%78%74%
w York (Manhattan)29%NA9%38%30%5%
rth Carolina68%88%92%NA87%87%
rth Dakota62%66%69%54%64%58%
io 54% 58% 49% 63% 65% 72%
lahoma 67% 67% 70% 61% 68% 71%
egon (Portland)64%NA66%41%68%68%
nnsyl va ni a
iladelp hia) 17% NA 24% 69% 51% 10%
ode Island32%*NA36%62%43%*13%
uth Carolina37%59%35%58%63%68%

P sy c ho t hera py ,
Preventive visit, newoffice, 45-50Initial hospital care,Upper GI endoscopy,Left heart
patient, age 1-4aminutes (90806,amoderate complexityabiopsy (43239,aAppendectomyacatheterizationa
(99382) no nf a c ilit y ) (99222) no nf a c ilit y ) (44950) (93510)
uth Dakota28%NA81%125%107%87%
xas (Houston)43%62%69%71%59%74%
ah 59% 57% 64% 55% 60% 55%
rmont 59% NA 68% 43% 51% 9%
rginia 68% 69% 73% 62%* 72%* 74%*
ashington (Seattle)58%56%58%52%56%57%
iki/CRS-RL32644t Virginia72%76%72%69%74%78%
g/wsconsin 31% NA 59% 183% 99% 27%
leakoming 44% 63% 83% 103% 120% 17%
://wiki: American Academy of Pediatrics, Medicaid Reimbursement Survey, 2001, and 2001 Medicare Physician Fee Schedule, available at
http: / / www. c m s . h h s . g o v / p h y s i c i a n s / mp f s a p p / s t e p 0 . a s p ] .
: NA = Not available.
a provided by state in 1998/1999 survey.
he codes are from Current Procedural Terminology (CPT), Fourth Edition, developed by the American Medical Association and used by CMS in determining physician payment
a mo unt s.

A recent study by the Center for Studying Health System Change found that the
proportion of physicians accepting Medicaid patients dropped from 87.1% in 1997
to 85.4% in 2001. The share accepting no new Medicaid patients increased slightly,
from 19.4% to 20.9%. However, there was no consistent relationship between these
measures and relative physician payment levels. The author suggests that capacity
constraints, the prevalence of Medicaid managed care, administrative rules, and other
market factors might play a role.29 Similarly, a multi-variate analysis of factors
affecting access and use by adult beneficiaries in 13 states in 1996 found that those
in states with above-average Medicaid physician payments were no more likely than
others to have a usual source of care or to have had a doctor’s visit in the last year.30
One recent survey of pediatricians, summarized in Table 19, found that, while
nearly 90% had some Medicaid patients, only 61% accepted all Medicaid patients
seeking care. Part of the difference may be related to capacity; nearly a quarter of the
respondents were not accepting all privately insured patients, either. While
physicians clearly felt that payments were inadequate — over half reported that
Medicaid payments were insufficient even to cover their overhead, leaving aside any
compensation for the physician’s time — fewer than one-third said that they would
see more Medicaid patients if payments were raised. On average, pediatricians said
that Medicaid rates would have to exceed 82% of their customary charge before they
would accept additional Medicaid patients. (As Table 18 showed, the Medicaid rate
for a well child visit was at or above 82% of the Medicare rate in only four states in


Although low reimbursement was the most commonly cited reason for limited
Medicaid participation, there were other issues, including paperwork, unpredictable
and delayed payments, and a perception that Medicaid patients miss appointments.

29 P. Cunningham, Mounting Pressures: Physicians Serving Medicaid Patients and the
Uninsured, 1997-2001, Center for Studying Health System Change, Tracking Report no. 6,
Dec. 2002.
30 T. Coughlin, and S. Long, “Adult Health Care Access and Use Under Medicaid: Does it
Vary by State?” Journal of Health Care for the Poor and Underserved, vol. 14, no. 2
(2003), pp. 208-228.

Table 19. Survey of Pediatricians
on Medicaid Participation, 2000
Currently accept any patients covered by
Currently accept all patients covered by-
Private insurance74.3%
Medicaid payments cover overhead
Don’t know32.4%
Would see more Medicaid patients with increased reimbursement31.0%
Percent of customary fee for well-child visit needed to —
Accept more, or any, Medicaid patients82.0%
Accept all Medicaid patients86.5%
Source: American Academy of Pediatrics, Division of Health Policy Research, Pediatrician
Participation in Medicaid/SCHIP: Survey of Fellows of the American Academy of Pediatrics, 2000.
Dental Payment
Adequacy of payment for dental care, as for physician care, has been a constant
issue in Medicaid programs. States have always used fixed fee schedules for dental
services, and payments are commonly below dentists’ usual fees. While no national
estimates are available, Table 20 compares Medicaid fees to median private fees for
selected services in 12 states. The private fees are drawn from the 2002 annual
survey of dentists by Dental Economics; the Medicaid fees are from the states’ most
recent fee schedule. The states are those (a) for which Dental Economics’ sample
was sufficient to allow reporting of state-level medians and (b) whose fee schedules
were readily accessible through the state’s web site.31

31 The General Accounting Office (GAO) has done its own comparison of state fee
schedules and provider charges, using 1999 fees and American Dental Association survey
data. The results are somewhat different, perhaps because GAO uses different procedures
and compares Medicaid fees to mean (rather than median) private fees for entire regions
(rather than individual states). Generally, GAO found higher Medicaid/private fee ratios,
but the relative ranking of states was fairly similar. U.S. General Accounting Office, Oral
Health: Factors Contributing to Low Use of Dental Services by Low-Income Populations
(GAO/HEHS-00-149), Sept. 2000.

Table 20. Medicaid Fees, 2003, and Median Private Fees, 2002, for Selected Dental Procedures
Comprehensive oralaComplete x-ray seriesaaFilling (amalgam), twoaRoot canal, molara
evaluation (D0150)(D0210)Cleaning, child (D1120)surfaces (D2150)(D3330)
MedicaidPercent ofMedicaidPercent ofMedicaidPercent ofMedicaidPercent ofMedicaidPercent of
feemedian feefeemedian feefeemedian feefeemedian feefeemedian fee
nne c t i c ut b $13.00 20% $24.75 25% NA NA $22.00 19% $192.50 23%
orid a $16.00 32% $32.00 40% $14.00 29% $41.00 43% $235.00 34%
ois $21.05 41% $30.10 35% $25.40 56% $48.15 48% $202.30 29%
diana $35.50 93% $72.25 99% $34.50 93% $72.25 84% $524.00 84%
a ssa c husse ttsc $36.00 56% $63.00 66% $33.00 66% $80.00 80% $613.00 77%
iki/CRS-RL32644chigan $14.89 31% $40.95 48% $19.53 47% $31.21 39% $378.00 62%
g/wsso uri $38.50 101% $33.50 45% $18.50 51% $37.00 49% $241.00 44%
s.or JerseydNANA$26.0031%$13.0026%$38.0036%$247.0034%
w YorkNANA$58.0075%$43.0086%$84.0084%$ 406.0058%
://wikirth Carolinae$45.00102%$75.1994%$21.6251%$79.4184%NANA
nsylvania $20.00 40% $45.00 58% $22.00 49% $50.00 53% $270.00 41%
xas $18.02 46% $36.04 47% $18.75 48% $43.73 52% $312.13 50%
: R. Willeford, “2002 Practice, Salary, and Fee Surveys,Dental Economics, vol. 92, no.2 (Dec. 2002), pp.28-44, and state Medicaid dental fee schedules as of Aug. 2003.
: NA = Medicaid fee schedule used does not include procedure.
he codes are from Current Procedural Terminology (CPT), Fourth Edition, developed by the American Medical Association and used by CMS in determining physician payment
a mo unt s.
chedule for adults; childrens services covered under HuskyCare plans.
for early and periodic screening, diagnosis and treatment-related services (EPSDT), when higher.
for specialist services, when higher.
reported in proposed settlement of McCree v. Odom, a beneficiary lawsuit on dental access (U.S. District Court — Eastern District of North Carolina Case No.:
4:00-CV-173-H(4)); at [].

As with physician payment, there is wide variation among states. However,
there appears to be somewhat more consistency across different procedures within
a single state — that is, some of the states are low payers and others high payers
across the board. It should be emphasized that the comparison here is not between
what Medicaid pays and what some insurer or third-party payer pays for the same
service, but between Medicaid rates and providers’ charges. Private dental insurance
plans also commonly use fixed fee schedules, and these, too, may often be well
below providers’ charges. The difference is that, while patients with private
insurance may have to pay the balance, dentists who treat Medicaid patients must
accept the Medicaid rate as payment in full.
Low fees probably play a role in limited use of dental services by Medicaid
beneficiaries, especially children. (Some state plans offer little or no dental coverage
for adults.) One recent study found that fewer than one in five children with
Medicaid received a dental visit over the course of a year.32 Whether increasing fees
would improve access is not certain. Responding to a GAO survey of 40 states that
increased rates between 1997 and 2000, 14 reported increases in participation or
utilization, 15 reported no increase, and 11 indicated that not enough time had
elapsed or the state did not have reliable data. In states reporting changes, the
improvement was often marginal. GAO concluded that the size of rate increases was
less important in explaining access improvements than the absolute amount of the
fees after the increase.
A 2002 study by the National Conference of State Legislatures identified a
number of factors in dentist participation unrelated to Medicaid reimbursement,
including stigmatization of Medicaid beneficiaries and a perception that they fail to
keep appointments. In addition, the study noted that many practices are already at
capacity with private-pay patients and have no need to accept Medicaid
Federally Qualified Health Centers and Rural Health Clinics
Under OBRA 89, states were required to cover services in federally qualified
health centers (FQHCs) and to pay full reasonable cost for these services. FQHCs
include community health centers, migrant health centers, and health care for the
homeless programs receiving funding from the Health Resources and Services
Administration, as well as centers that meet the standards for a grant but are not
actually receiving federal funding. States were also required, under a 1977
amendment, to pay reasonable costs for services of rural health clinics (RHC), which
provide services of nurse practitioners and physician assistants in medically
underserved rural areas.
The BBA provided for a gradual phase-out of mandatory cost reimbursement
for FQHCs and RHCs, with required reimbursement dropping to 70% of cost by

32 U.S. Surgeon General, Surgeon General’s Report on Oral Health, May 2000.
33 S. Gehshan, and T. Straw, Access to Oral Health Services for Low-Income People: Policy
Barriers and Opportunities for Intervention, National Conference of State Legislatures, Oct.


FY2003 and with no minimum beginning in FY2004. During the transition, states
were required to make supplemental payments to FQHCs and RHCs that had
contracts with Medicaid MCOs under which the MCO paid less than the required
percentage of costs. In 1999, the Medicare, Medicaid, and SCHIP Balanced Budget
Refinement Act of 1999 (BBRA, included by reference in the Consolidated
Appropriations Act for FY2000, P.L. 106-113) slowed the phase-out and delayed
repeal of minimum payment rules until FY2005.
BIPA established a new prospective payment system for FQHCs and RHCs.
Each center or clinic received a per visit rate for 2001 based on its own average
reasonable costs for 1999 and 2000. For later years, this rate is updated by the
“Medicare economic index” (MEI) used to update Medicare physician payments for
primary care; the rate may be modified to reflect a change in the scope of services
provided by the facility. The state and the center may agree to an alternative payment
methodology, but only if aggregate payments under the alternative method are at least
equal to those that would have been made under the standard method. BIPA
continued the requirement that states make supplemental payments to FQHCs and
RHCs who are paid less than the minimum by a Medicaid MCO. Finally, GAO was
required to report on whether rates should be periodically rebased or refined and on
how to do so.
The GAO report is not due until the end of 2004, but GAO has already
concluded that the new system is likely to pay many facilities less than their costs,
for several reasons. First, the initial 2001 rates were based on costs for the two
preceding years, with no allowance for inflation.34 Second, the MEI index used to
update the rates rises less rapidly than other measures of inflation, and centers may
have difficulty holding increases to these limits, especially if they began with a low
per-visit rate. GAO suggests that it will be difficult to develop a system that assures
the continued viability of FQHCs and RHCs while maintaining incentives for

34 In addition, some states were already paying less than full actual costs, because some costs
were disallowed as not “reasonable under various tests.”
35 U.S. General Accounting Office, Health Centers and Rural Clinics: Payments Likely to
be Constrained Under Medicaid’s New System, (GAO-01-577), June 2001.

Long-Term Care
Nursing Facilities
Payments to nursing facilities (NFs) are the single largest component of
Medicaid expenditures, accounting for 19% of spending in 2002. At the same time
Medicaid is the key funding source for NFs; in 2002, 67% of NF residents relied on
Medicaid as their principal payer.36 Thus ability to control growth in NF spending
has an important effect on state budgets, while the adequacy of Medicaid
reimbursement can determine whether a facility can offer high-quality care. The
tension between the competing goals of cost containment and quality assurance has
been present almost since the beginning of the Medicaid program.
Payment Methods. Since the 1980 Boren amendment allowed states to move
away from Medicare’s retrospective cost-based reimbursement rules, states have
evolved very complex NF payment systems. These systems commonly distinguish
among direct patient care costs; costs for various operating, support, and
administrative functions; and capital costs, such as interest, rent, and depreciation.
A state may treat each component differently: for example, payment to a particular
facility might be the sum of a case-mix adjusted fixed amount for direct care, a
facility-specific cost-based payment subject to a peer group ceiling for other
operating costs, and a “fair rental value” payment for capital costs.
The following discussion of payment methods cannot capture the full
complexity of states’ systems. Instead, it focuses chiefly on how states pay for the
direct care component, the actual delivery of services to individual residents by
nursing staff. Obviously the totality of the state payment, relative to costs, can affect
quality or access. Still, a payment system that has incentives to hold down
administrative costs or that limits the rate of return on capital investment has
different implications from a system that squeezes direct care spending.
Payment for Direct or Nursing Care. Table 21 shows each state’s basic
payment method for the direct care component or, in states classifying costs37
differently, its nearest equivalent. The table classes states according to whether NFs
receive facility-specific rates subject to a peer group ceiling, receive rates set for a
whole group of facilities, or are paid under some other method. The table also notes
which states offer incentive payments to facilities with costs below the ceiling or rate.
For states using some form of case mix adjustment, the table indicates the method.
Finally, where applicable, it identifies the facility characteristics states use in
establishing peer groups of nursing facilities.
Over two-thirds of the states pay the lesser of the facility’s actual costs for
Medicaid residents or a fixed ceiling based on the cost experience of comparable

36 American Health Care Association analysis of CMS OSCAR data for 2002 at
[] as of Sept. 2003.
37 States have different definitions of this component and of the types of personnel and other
costs it may include.

NFs. States may set statewide ceilings or define peer groups on a number of
dimensions, including size, location, ownership, and whether a facility is
hospital-based or freestanding. A handful of states continue, for payment purposes,
the distinction between intermediate care facilities (ICFs) and skilled nursing
facilities (SNFs) that was eliminated by the nursing home reform provisions of
OBRA 87. Two states treat facilities granted a waiver of OBRA 87 minimum
staffing requirements as a separate group.
Of the states paying the lesser of cost or fixed ceilings for direct care, 11 have
incentive arrangements, under which facilities whose costs are below the ceiling
share in the savings. Many more states use these arrangements for other cost
components, such as administration, where rewards for cost-cutting may arguably be
less likely to affect patient care.

Table 21. Payment Methodologies for Nursing Facility Direct Care Component, 2002
Basic method
for facilitiesMethod ofFacility
NF-specific rateFixed ratewith costscase mixcharacteristics
subject to peerfor peerbelowadjustmentused to define
Stategroup ceilinggroupOtherceiling/rate(if any)peer groupsNotes
abama110% of medianxBed size
askaNF-specific, rate
of increase limits
izonaCapitation3 levelsALTCS; levels are NF, HCBS,
ve ntilato r -d e p e nd ent
iki/CRS-RL32644kansas105% of medianStatewideStatewide fixed rate for other
g/wcomponents; direct has 90% floor
leakorniaMedian7 levelsLocation, size
://wikilorado125% of averageStatewide
httpnnecticut135% of medianLocation, size
areMedian4 levelsPublic, private by
c o unt y
ict of ColumbiaMedianHospital-based/
fr eestand ing
oridaMedian plus 1.75Location, size
rgia90th percentilexSize, skilled/
intermediate mix
waii115% of average3 levelsFull cost for highest acuity level
o M edian RUGs Lo cation,
Ho sp ital-b a sed /
fr eestand ing

Basic method
for facilitiesMethod ofFacility
NF-specific rateFixed ratewith costscase mixcharacteristics
subject to peerfor peerbelowadjustmentused to define
Stategroup ceilinggroupOtherceiling/rate(if any)peer groupsNotes
oisUnspecifiedSee noteLocationFixed hourly rate times estimated
basisnursing hours, based on resident
a sse ssme nt
dianaMedianRUGsStatewideProfit-sharing if cost below rate
a60th to 70thHospital-based/SNFs get add-on for high case
percentile,freestanding,mix, high Medicaid share
depending on classSNF/ICF
iki/CRS-RL32644nsas115% of medianRUGsStatewide
g/wnt uc ky U nsp e c i fi e d RUGs Urb an/rural
s.or basis
uisiana62nd6 levelsStatewide
://wiki p e r centile
httpne110%-150% of45 state-Hospital-based/
median, dependingdevelopedfreestanding, size
on classgroups
arylandSee note at rightx6 levelsLocation, sizeFixed nursing cost ceiling for
care level based on survey data
usettsBlend6 levelsBlend of facility-specific,
igan80th percentilexHospital-based/
fr e e s t a nd i ng,
o wne r s hi p ,
specialty, size

Basic method
for facilitiesMethod ofFacility
NF-specific rateFixed ratewith costscase mixcharacteristics
subject to peerfor peerbelowadjustmentused to define
Stategroup ceilinggroupOtherceiling/rate(if any)peer groupsNotes
nesotaNF-specific, rate11 classesHospital-based/Facility with costs above median
of increase limitsfreestanding,gets lower increase
lo catio n
ssissippi120% of medianRUGsSize
ssouri120% of medianStatewide
tana Unspecified RUGs Statewide
g/wbraska125% of median19 levelsUrban/rural,waiver
leakvada60th6 levelsStatewideMethod for SNF; ICF method not
p e r centile available
://wiki HampshireMedianRUGsStatewideStatewide fixed rate for other
http components
w Jersey115%-120% ofOwnership,
average, sp ecialty
depending on class
w Mexico110% of medianxState, non-state
w York105% of meanRUGsSize, location, highDirect care floor at 95% of mean.
or low case mixMedicare maximization incentive
rth Carolina80th percentileSNF/ICFStatewideStatewide fixed rate for other
rth DakotaUnspecified basisxRUGsStatewideFixed statewide ceiling; basis not
sp ecified
io85th percentileRUGsSize, location

Basic method
for facilitiesMethod ofFacility
NF-specific rateFixed ratewith costscase mixcharacteristics
subject to peerfor peerbelowadjustmentused to define
Stategroup ceilinggroupOtherceiling/rate(if any)peer groupsNotes
laho ma Aver age Statewid e
egonAverageStatewideFixed 40% add-on for residents
with complex needs
nsylvania117% of medianRUGsLocation, size,Add-on for high-Medicaid NF
Ho sp ital-b a sed /
fr e e s t a nd i ng,
sp ecialty
iki/CRS-RL32644ode Island80th percentilexStatewide
g/wuth Carolina105% of meanStatewide
leakuth Dakota115% of medianRUGsWaiverRisk corridor: pays full direct
care cost to 115% of median,
://wiki80% of excess up to 125% of
http me d i a n
nnessee65th percentilexSNF/ICFIncentive reduced for occupancy
(equivalents)below 80%
xas107% of mean11 levelsStatewide
ah120% of medianxStatewide
rmont115% of medianRUGsStatewide
rginia112% of median3 levelsLocation, size
hington105% of medianRUGsLocation95% floor; scheduled shift to
fixed price at median for peer
gr o up

Basic method
for facilitiesMethod ofFacility
NF-specific rateFixed ratewith costscase mixcharacteristics
subject to peerfor peerbelowadjustmentused to define
Stategroup ceilinggroupOtherceiling/rate(if any)peer groupsNotes
est VirginiaAveragexSee noteSizeUnspecified case mix scoring
based on minimum data set
isconsinUnspecified5 levelsLocation, sizeCMI adjusted upward for small
basis NF
oming125% of medianxStatewide
mber of states using361141131 (14 RUG)
iki/CRS-RL32644t h o d
s.or: Medicaid state plans and amendments approved as of Nov. 7, 2002, except as follows: Nevada Medicaid Rates and Cost Containment Unit Rate Matrix,], as of July 2003. Ohio Administrative Code 5101-3-3.
://wiki: ALTCS = Arizona Long Term Care System; CMI = case mix index; HCBS = home and community-based services; ICF = intermediate care facility; SNF = skilled nursing facility;
httpesource utilization groups.

Eleven states pay fixed per diem amounts for direct care. In some states the per
diem rates are based, as in states using cost ceilings, on the experience of comparable
facilities. In other states, the per diems are fixed by law or regulation. (While the
rates may be derived through some form of cost analysis, a specific formula is not
described in the state plan.)
Of the remaining states, two use facility-specific ceilings based on the NF’s
historical costs and fixed updates, while one, Massachusetts, pays a rate based on a
blend of facility-specific and statewide experience. Finally, Arizona provides NF and
other long-term care services through the Arizona Long Term Care System (ALTCS),
under which contracting plans receive fixed capitation payments for care of each
enrollee. Payment amounts vary for two classes of nursing home residents and for
enrollees receiving home care.
Case Mix. More than half the states now use some form of case mix
adjustment in paying NFs. Of these, 14 use the resource utilization groups (RUGs)
developed by CMS for Medicare SNF payment. The Medicare system assigns each
resident to one of 44 groups based on a resident assessment that measures physical
function, rehabilitation needs, cognitive impairment, and other factors. Medicaid
programs commonly use a set of 34 RUGs; these have fewer distinct categories of
rehabilitative care, because fewer Medicaid residents are receiving such care. The
remaining states have developed their own classification systems, usually grouping
residents into a much smaller number of care categories.
Case mix is often used differently in Medicaid NF payment than in Medicare
SNF payment. Under Medicare’s PPS for SNF services, there is a fixed daily rate for
each resident; part of this rate, the nursing and therapy case-mix components, varies
by the resident’s RUG class.38 In Medicaid programs case mix is often used, not to
establish payment for a particular resident, but to adjust the per diem cost ceiling for
the entire facility. A facility that has served residents with more intensive needs
during some base period will be allowed a higher ceiling than other facilities in its
peer group.
Case mix adjustment is intended both to treat NFs fairly and to reduce
incentives to refuse heavy care patients. At least some observers contend that the
adjustments may create perverse incentives of its own. For example, an NF might
be penalized for promoting resident independence, because payment is greater for
residents requiring more assistance.39
Labor Cost Adjustments. There are concerns that the supply of nurses’
aides and other direct care workers in nursing homes, as well as in-home and
personal care programs, is not keeping pace with demand. Providers have difficulty
retaining these workers because of low wages and benefits and physically demanding

38 There are separate urban and rural rates. Payment is further adjusted to reflect local wage
39 C. Harrington, et al., 1998 State Data Book on Long Term Care Program and Market
Characteristics, at [].

work; some studies have found turnover rates approaching 100%.40 State constraints
on growth in the direct care component of NF payment have limited providers’
ability to improve compensation or benefits; an individual facility that does so on its
own may risk exceeding a class-based cost ceiling. Some states have sought to
address this problem through “wage pass-through programs,” which directly
compensate providers that increase wages for direct-care workers. (Some other states
have periodically made general rate adjustments to reflect increases in federal or state
minimum wage requirements.)
Table 22 summarizes the pass-through programs in 28 states. Most programs
target NFs, though some also reach home care and/or personal care programs. Some
are voluntary — the provider receives enhanced reimbursement if it shows that it has
raised wages — while others require participation by all providers in the targeted
group. Information on the effectiveness of the programs is limited. Some states have
reported modest improvements in turnover rates, while others have found no change
or have not measured the effects.41
Table 22. Summary of State Wage Pass-Through Programs
Target provider typeParticipation
Nursing Home Personal No t
St a t e f a cilit y care ca re Vo lunt a r y M a nda t o ry available
Ar i z o n a x x x x
Califo r ni a x x
Co lo r a d o x x
I llino is x x
Kansas x x x
Lo ui s i a n a x x
M a ssa c huse t t s x x x
Maine x x x x
Michigan x x
M i nne so t a x x
M i sso ur i x x
Montana x x x
North Dakotaxx
Oklaho ma x x
Rhode Islandxxxx

40 U.S. General Accounting Office, Nursing Workforce: Recruitment and Retention of
Nurses and Nurse Aides is a Growing Concern, statement of Williams J. Scanlon in U.S.
Congress, Senate Committee on Health, Education, Labor and Pensions, (GAO-01-750T),
May 17, 2001.
41 Paraprofessional Healthcare Institute, State Wage Pass-Through Legislation: An Analysis,
Workforce Strategies, no. 1, Apr. 2003.

Target provider typeParticipation
Nursing Home Personal No t
St a t e f a cilit y care ca re Vo lunt a r y M a nda t o ry available
South Carolinaxxxx
Texas x x
Vir ginia x x x
W a shi ngt o n x x
W i sc o nsin x x
Wyoming x x
Total 17 10 6 6 10 5
Source: Paraprofessional Healthcare Institute, State Wage Pass-Through Legislation: An Analysis
(Workforce Strategies n. 1), Apr. 2003.
Payment Levels and Adequacy. Analyses performed for the nursing home
industry indicate that state Medicaid programs are, in the aggregate, paying less than
the full costs of caring for Medicaid residents. However, it is difficult to know
whether payments are insufficient across the board or just for the most costly
facilities. Table 23 presents the results of studies by BDO Seidman, LLP for the
American Health Care Association, the organization of proprietary nursing homes.
The studies, based on data collected from state affiliates, show the difference
between average daily rates and daily costs (excluding capital costs) in 36 states in

1999 and 37 states in 2000. Averages are weighted by Medicaid days in each facility.

In 2000, Medicaid payments as a percent of cost ranged from a low of 83% in South
Dakota to 100% in Alabama. The average for the 37 states was 92%.

Table 23. Average Medicaid Shortfall Per Day,
Medicaid Nursing Facility Payments in Responding States, 1999 and 2000
1999 2000
StateRateCostDifferenceRate as percentage of costRateCostDifferenceRate as percentage of cost
abama 102.78 100.30 2.48 102% 107.13 106.99 0.14 100%
kansas64.5270.48-5.96 92%69.4075.34 -5.94 92%
lifornia 88.47 95.58 -7 .11 93% 97.54 104.74 -7 .20 93%
lo rado 111.39 119.12 -7 .73 94% 113.57 120.87 -7 .30 94%
nnecticut156.06165.00-8.9495%No response
lawareNo response118.89138.56-19.6786%
orid a 106.99 119.15 -12.16 90% 112.82 123.99 -11.17 91%
iki/CRS-RL32644orgiaNo response90.1192.80-2.6997%
g/woisNo response87.4495.56-8.1292%
s.ordiana 92.80 103.64 -10.84 90% 105.14 112.56 -7 .42 93%
a 84.83 89.99 -5 .16 94% 83.21 89.08 -5 .87 93%
://wikinsas 85.28 93.91 -8 .63 91% 91.34 97.01 -5 .67 94%
httpne 113.04 122.87 -9 .83 92% 119.12 130.68 -11.56 91%
land 123.46 133.16 -9 .70 93% 127.96 138.22 -10.26 93%
sachusetts 120.76 135.47 -14.71 89% 128.59 145.02 -16.43 89%
chigan 103.94 111.81 -7 .87 93% 109.24 119.64 -10.40 91%
sso uri 93.06 101.03 -7 .97 92% 97.26 109.91 -12.65 88%
tana92.26103.04-10.7890%No response
braska 99.13 106.03 -6 .90 93% 105.01 111.07 -6 .06 95%
vada102.15116.02-13.8788%No response
w Hampshire117.43127.50-10.0792%119.25139.87-20.6285%
w Jersey124.95146.06-21.1186%131.78154.11-22.3386%
w Mexico99.72105.32-5.6095%101.23105.89-4.6696%

1999 2000
StateRateCostDifferenceRate as percentage of costRateCostDifferenceRate as percentage of cost
rth Carolina94.3197.00-2.6997%97.72101.36-3.6496%
rth Dakota95.9199.70-3.7996%102.74105.33-2.5998%
io 115.81 125.31 -9 .50 92% 122.64 131.33 -8 .69 93%
lahomaNo response66.5774.17-7.6090%
egon 91.10 104.24 -13.14 87% 94.97 110.03 -15.06 86%
nsylvania 125.14 135.03 -9 .89 93% 131.13 143.85 -12.72 91%
ode Island111.79121.83-10.0492%117.46129.59-12.1391%
uth Dakota79.9993.80-13.8185%83.2199.75-16.5483%
nnessee 81.48 86.63 -5 .15 94% 88.39 94.13 -5 .74 94%
xas 78.47 82.07 -3 .60 96% 83.06 88.25 -5 .19 94%
iki/CRS-RL32644ah 88.55 101.01 -12.46 88% 90.24 106.74 -16.50 85%
s.orrmont 103.02 122.97 -19.95 84% 108.24 127.03 -18.79 85%
leakrginia 82.12 92.68 -10.56 89% 87.51 99.93 -12.42 88%
a shi ngt o nb 106.96 118.92 -11.96 90% 109.68 126.23 -16.55 87%
://wikit Virginia109.10117.26-8.1693%110.97117.74-6.7794%
httpsconsin 99.57 109.85 -10.28 91% 104.05 118.48 -14.43 88%
ghted average-9.0592%-9.7892%
: BDO Seidman LLP, A Briefing Chartbook on Shortfalls in Medicaid Funding for Nursing Home Care, 1999 and 2000 eds., American Health Care Association, 2001 and 2002.
at data are based on reports from state affiliate associations; not all states reported in one or both years.
he data represent single level nursing facilities only. Multilevel facilities providing non-nursing home services such as housing, adult day care and home health were excluded since
the reported costs did not reflect allocations between nursing home and non-nursing home services.
ates and costs are exclusive of property costs and property rates which were not included in the available database.

As in the case of hospital payments, Medicaid payments to nursing facilities are
usually based on historical cost data with periodic updates. Annual increases in rates
or ceilings may or may not keep pace with inflation, and some states may go for long
intervals without “rebasing” — updating cost data to reflect changes in facility case
mix, occupancy levels, or other factors that may affect costs. As a result, even
facilities whose costs were at one time fully covered by Medicaid reimbursement
may gradually see shortfalls. Table 24 shows the change between 1999 and 2000 in
average payment rates and average costs in the 33 states for which BDO Seidman has
survey responses in both years. Costs rose more rapidly than rates in 21 of the 33
Table 24. Change in Daily Medicaid Nursing Facility Payment
Rates and Daily Costs, 1999-2000
Percent change in
StateAverage daily rateAverage daily cost
Alabama 4 .2% 6 .7%
Arkansas 7.6% 6.9%
California 10.3% 9.6%
Co lo rado 2.0% 1.5%
Florida 5 .4% 4 .1%
Indiana 13.3% 8.6%
Iowa -1 .9% -1.0%
Kansas 7.1% 3.3%
Maine 5 .4% 6 .4%
Maryland 3.6% 3.8%
Massachusetts 6.5% 7.0%
Michigan 5.1% 7.0%
Misso ur i 4 .5% 8 .8%
Nebraska 5.9% 4.8%
New Hampshire1.5%9.7%
New Jersey5.5%5.5%
New Mexico1.5%0.5%
New York4.6%5.6%
North Carolina3.6%4.5%
North Dakota7.1%5.6%
Ohio 5.9% 4.8%
Oregon 4.2% 5.6%
P e nnsyl va ni a 4 . 8 % 6 . 5 %
Rhode Island5.1%6.4%
South Dakota4.0%6.3%

Percent change in
StateAverage daily rateAverage daily cost
T exas 5 .8% 7 .5%
Utah 1.9% 5.7%
Vermont 5 .1% 3 .3%
Virginia 6.6% 7.8%
Washington 2.5% 6.1%
West Virginia1.7%0.4%
W i sc o nsin 4 . 5 % 7 . 9 %
Source: BDO Seidman LLP, A Briefing Chartbook on Shortfalls in Medicaid Funding for Nursing
Home Care, 1999 and 2000 eds., Washington, 2001 and 2002.
Still, even in states whose payments keep pace with inflation and other changes
affecting costs, some facilities may still lose money. Nearly all state systems pay the
lesser of actual costs or a fixed ceiling, in order to create pressure for greater
efficiency in the most costly providers. If a state pays nearly all facilities their full
costs and underpays a small number of facilities, the average payment to cost ratio
will inevitably be less than 100%.
Table 25, based on the 1999 National Nursing Home Survey, suggests that cost
ceilings may have affected overall Medicaid payment/cost ratios. The table shows
daily charges reported for residents whose current principal source of payment was
Medicaid. (Charges would ordinarily be at or above actual costs, although some
facilities might have reported what they actually expected Medicaid to pay.) It then
shows the actual Medicaid payment rates reported by facilities.42 Median rates are
about 95% of median charges, but the gap widens at the upper end of the distribution.
Table 25. Medicaid Daily Nursing Facility Charges
and Payment Rates, 1999
Average Medicaid dailyMedicaid daily payment
charge per current residentrate for the average facility
Mean $112 $105
Median $102 $97
75th percentile$128$116
90th percentile$167$139
Source: Authors calculations from the 1999 National Nursing Home Survey.

42 The average charge is weighted according the number of residents for which each charge
level was reported, while average Medicaid payment rates are weighted by the number of
facilities reporting a given rate. While it would have been preferable to use the same
method for both numbers, the design of the NNHS precludes this. In practice, some states
that calculate medians to set ceilings use the median facility, while others use the median

In some states, peer group or statewide ceilings, or flat rates based on group
experience, may be at least as important as limited annual increases in explaining
current payment shortfalls. Whether these ceilings are set at appropriate levels is a
difficult policy question, the answer to which depends in part on how sensitive the
states’ system is to differences in facility and resident characteristics. If a state uses
a single statewide ceiling with no case mix adjustment, then a very efficient facility
might be penalized because it is in a high-cost urban area or has residents with
complex needs.
Some observers have suggested that state use of complex ceiling systems for NF
cost containment, in preference to simple rate cuts or freezes, was driven in part by
concerns about litigation during the period when the Boren amendment was still in
effect. It was easier for states to show that their payments were adequate to meet the
costs of “efficiently and economically operated” facilities if only a minority of NFs
were affected by payment constraints.43 With the repeal of the Boren requirements,
across-the-board limitations may have greater appeal for states. As Table 24 showed,
many states limited general 1999-2000 rate increases to levels below average cost
growth; more may do so in response to current budgetary problems.
Intermediate Care Facilities for the Mentally Retarded
Until the 1980s, most Medicaid services for people with mental retardation or
developmental disabilities were provided in large state-operated Intermediate Care
Facilities for the Mentally Retarded (ICFs-MR). As a result of the availability of the
home and community-based services waivers, court decisions requiring treatment in
less restrictive settings, and other factors, over three-fourths of people receiving
Medicaid-funded MR/DD services were in the community in 2002.44 However, the
absolute number of ICF-MR residents actually grew slightly between 1977 and 2002,
and ICF-MR spending still accounts for nearly 5% of Medicaid spending.
As Table 26 shows, the major change in the use of facilities to provide care for
this population is that many residents are now in smaller facilities or in facilities
operated by local government or private organizations. The share of ICF-MR
residents in non-state facilities went from 12.5% in 1977 to 59.7% in 2002. Only
1.6% of residents were in facilities with 15 or fewer beds in 1977, compared to

37.8% in 2002.

43 CMS Report to Congress: Appropriateness of Minimum Nurse Staffing Ratios in Nursing
Homes (known as the “Phase I” report) ch. 2, at [

2.pdf], as of Sept. 2003.

44 Of 489,138 service recipients, 378,566 were served under HCBS waivers as of June 2002.
Research and Training Center on Community Living, Residential Services for Persons with
Developmental Disabilities, Status and Trends Through 2002, University of Minnesota,
Minneapolis, 2003.

Table 26. ICF-MR Residents at End of Year by Facility Size
and Ownership, 1977 and 2002
Facility type/size19772002
State Residents Percent Residents Percent
1-15 residents3560.3%1,0130.9%
16 or more residents92,49887.1%43,53039.4%
All state92,85487.5%44,54340.3%
No n-sta t e
1-15 residents 1,3541.3%40,74836.9%
16 or more residents11,95811.3%25,28122.9%
All non-state13,31212.5%66,02959.7%
All facilities
1-15 residents1,7101.6%41,76137.8%
16 or more residents104,45698.4%68,81162.3%
To t a l 106,166 100.00% 110,572 100.00%
Source: Research and Training Center on Community Living, Institute on Community
Integration/UCEED, Residential Services for Persons with Developmental Disabilities: Status and
Trends Through 2002, University of Minnesota, 2003.
Most states that use both state-operated and non-state ICF-MRs have different
payment rules for the two classes of facilities. States generally pay state-operated
ICF-MRs their full operating costs, for the obvious reason that paying less would
mean forgoing federal matching funds without reducing state expenditures. While
some states also reimburse the full costs of non-state facilities, many have developed
alternate systems.
Table 27 shows the payment methods used by the 44 states that make payments
to non-state facilities, whether public or private. (Note that two states made no
ICF-MR payments in FY 2001, while five others used only state facilities.) The table
classes states according to whether non-state facilities receive full costs or costs
subject to a peer group ceiling, or are paid under some other method. For states using
some form of case mix adjustment, the table indicates the method. Finally, where
applicable, it identifies the facility characteristics states use in establishing peer
groups of ICF-MRs.

Table 27. Basic Medicaid Payment Method, Direct Care Component,
Non-State Intermediate Care Facilities for the Mentally Retarded, 2002
Basic methodFacilityMethod of case
characteristicsmixFullCost with direct care
used to define peeradjustment (ifreasonablecomponent subject to
groupsany)Notescostpeer group ceilingOther method
abama90th percentileBed sizes 4-15, 16+
askaNo ICF-MR payments in FY2001
izonaNo ICF-MR payments in FY2001
kansasUnder 16 beds: fixed
iki/CRS-RL32644statewide rate 16+ beds: costwith ceiling for nonpatient
g/w care
leakornia65th percentileBed sizes 1-59, 60+Separate groups for small
rehabilitative or nursing
://wikilorado125% of averageAllThree care
http leve ls
nnecticut135% of medianAll
are75th percentileAll private
ict of ColumbiaMedianAllAdjustment for
high acuity
ridaBase cost plus fixed inflationAllFour levels
rgia90th percentileAll
waii115% of averageAll
oSee noteFixed cap to assure aggregate
payments below UPL

Basic methodFacilityMethod of case
characteristicsmixFullCost with direct care
used to define peeradjustment (ifreasonablecomponent subject to
groupsany)Notescostpeer group ceilingOther method
oisFixed hourly rate timesBed sizes 4-16, 17+,
estimated nursing hours,location
based on resident assessment
dianaSee noteICF-MR, communitySeven carePercent of median for community
residential (CRF)levelsresidential facility ceiling varies
by level of care; ceiling for
ICF-MR at 125% of median
wa80th percentileAll privatePenalty for occupancy < 80%
iki/CRS-RL32644sasFixed dollar limitsBed sizes 4-8, 9-16,Five levels
g/w 17+
s.ornt uc ky x
uisianaFixed rates based on averageBed sizes 1-8, 9-32,
://wikiper diem33
httpe x
landNo non-state ICF-MR in FY2001
assachusettsNo non-state ICF-MR in FY2001
iganNo non-state ICF-MR in FY2001
innesotaBase cost plus fixed inflation
ssissippi110% of medianAll
ssouri135% of meanAll private
tana x
kaSee noteLevels notPersonnel ceilings based on
described inmodel” staffing hours for facility
state plansize/level of care

Basic methodFacilityMethod of case
characteristicsmixFullCost with direct care
used to define peeradjustment (ifreasonablecomponent subject to
groupsany)Notescostpeer group ceilingOther method
vada1-6 beds: ceiling at 60th
percentile; 6+ beds, full cost
HampshireUnder 16 beds: full cost; 16+
beds: ceiling based on median
w MexicoBase cost plus fixed inflationThree levelsNon-care components subject to
iki/CRS-RL32644 YorkFixed dollar or staffingscreensBed sizes 1-30/30+,locationStaffing screens include disabilitymeasures
s.orrth CarolinaMedianBed sizes 1-32, 33+Five levels
leakrth DakotaxUp to approved budget
://wikiio~82nd percentileBed sizes 1-8/9+Four levels
httplahomaFixed rate based on meanStandard/ specialized
costs <17
onNo non-state ICF-MR in FY 2001
nnsylvaniaxUp to approved budget
ode Islandx
uth Carolinax
uth Dakotax
nnessee65th percentileAll
xasFixed rate for level of careBed sizes <9, 9-13,Five levels
Negotiated rate

Basic methodFacilityMethod of case
characteristicsmixFullCost with direct care
used to define peeradjustment (ifreasonablecomponent subject to
groupsany)Notescostpeer group ceilingOther method
ontxUp to approved budget
giniaxRate cannot exceed highest rate
for state facility
ashingtonCost with ceiling for noncare
t VirginiaMeanBed sizes 1-8, 8, 9+Four levels
isconsinFixed dollar ceilingLocationFour levels
iki/CRS-RL32644omingNo non-state ICF-MR in FY 2001
g/wmber of states using11191412
s.ort h o d
://wiki: Medicaid state plans and amendments approved as of Nov. 7, 2002, except as follows: 114.1 Code of Massachusetts Regulations 29; MaineCare Benefits Manual 50.07;
httpada Medicaid Rates and Cost Containment Unit Rate Matrix, at [], as of July 2003; Ohio Administrative Code3-3; Rules of the Tennessee Dept. of Health (1200-13-6)

Eleven of the states pay non-state ICFs-MR their full costs; a few of these
require that operating budgets be approved in advance. Another 19 use cost ceilings,
comparable to those used in NF payment, for the direct care component of costs or
for the entire per diem rate. Some of these use peer groups, based on bed size or
other characteristics. Others group all facilities together (partly because some states
have few or no residents in larger facilities). Of the remaining states, three use
different methods for smaller and larger facilities. Most of the rest use some form
of fixed rate, often based on peer group means.
Twelve states use some form of case mix adjustment, classifying residents in a
small number of level of care groups. Illinois estimates needed hours of nursing
home care as part of the resident assessment and uses this estimate to set the direct
care component for each patient.
Home and Community-Based Services
In FY2002, Medicaid spent nearly $25 billion for community long-term care
services for the aged, the disabled, the mentally retarded/developmentally disabled,
and other defined groups. As Table 28 shows, two-thirds of the spending was made
through section 1915(c) home and community-based services (HCBS) waivers. The
rest went for non-waiver personal care and home health services; these are optional
services defined in the state plan and must be furnished to all Medicaid beneficiaries
who require them. Waiver services, on the other hand, are available to a limited
number of individuals approved for waiver participation.
Table 28. Medicaid Spending for Home
and Community Care, FY2002
Spending (millions)Percentage
Home and community-based services waivers$16,40866%
Non-waiver personal care$5,54722%
Non-waiver home health care$2,76511%
T o tal $24,720 100%
Source: Brian Burwell, Kate Sredl, and Steve Eiken, Medicaid Long Term Care Expenditures in
FY2002, The Medstat Group, May 2003, available at [], as of Nov.
In addition, each approved waiver is subject to a “budget neutrality” test; total
spending on waiver services cannot exceed what would have been spent to provide
institutional services to the same population. Some waiver programs apply this test
on an aggregate basis; that is, they allow some participants to incur costs greater than
the cost of institutional care, so long as overall costs meet the cap. Other programs
establish individual budget limits. In either case, states can keep spending within the
established budgets by limiting the scope of services provided, the number of units
of service any individual can receive, and/or the payment for each service.

Non-Waiver Home and Personal Care. Payment methodologies for
non-waiver services are specified in state plans. Because these services account for
a comparatively small share of overall Medicaid spending, information from state
plan documents was not collected for this report. Table 29 provides information
compiled by Health Management Associates from state plans and other sources,
reflecting methods in effect as of January 2003.
Table 29. Payment Methods for Non-Waiver Home Health Care
and Personal Care Services, January 2003
StateHome health carePersonal care
AlabamaCost based payment for governmentNot covered
providers, fee for service using time
units for private providers, medical
equipment and supplies paid fee for
se r vic e
AlaskaPercentage of chargeFee for service using hourly
ArizonaFee for serviceNot covered
ArkansasFee for serviceFee for service using hourly
rates, transportation paid rate
per mile
CaliforniaFee for serviceFee for service using hourly
rates, or negotiated rates
ColoradoFee for service, using maximum dailyNot covered
ConnecticutFee for serviceNot covered
DelawareFee for serviceNot covered
District of ColumbiaFee for service using Medicare costFee for service using hourly
ceilingsrates, adjusted for multiple
beneficiaries same address
FloridaFee for serviceNot covered
GeorgiaProspective cost based rate per visitNot covered
HawaiiFee for service using Medicare costNot covered
IdahoFee for service using Medicare costHourly rates based on nursing
ceilings, medical equipment rentalfacility wages, rates vary for
paid at one-twelfth purchase price forindependent providers and
12 monthsagencies
IllinoisFee for serviceNot covered
IndianaProspective cost based ratesNot covered
IowaCost based payment for most servicesNot covered
with some paid on fee for service
KansasFee for serviceFee for service using hourly
KentuckyFee for serviceNot covered

StateHome health carePersonal care
LouisianaProspective rates based on historicalNot covered
MaineFee for service using Medicare costFee for service using hourly
ceilingsrates with annual payment
MarylandFee for service with rates setPer diem varied by level of care
MassachusettsFee for service using peer groups toFee for service
set maximum payments
MichiganFee for serviceFee for service using hourly
rates adjusted for level of care
MinnesotaFee for serviceFee for service
MississippiFee for service with nursing facilityNot covered
rate as upper limit, medical
equipment rented if cost over $150
MissouriFee for serviceFee for service with monthly
payment ceiling at 60% to
100% of average nursing
facility rate depending on
services provided
MontanaPercentage of charge using aNegotiated hourly rates
percentage of Medicare allowable
cost as ceiling
NebraskaFee for serviceFederal minimum hourly wage,
increased following training or
NevadaFee for serviceNegotiated hourly rates
New HampshireFee for serviceFee for service
New JerseyCost based payment per time unit,Fee for service using hourly
medical supplies paid fee for servicerates
New MexicoCost based payment using MedicareFee for service
upper limits
New YorkProspective cost based rates, servicesFee for service
provided on long term basis paid
75% nursing facility rate
North CarolinaProspective cost based rates forNegotiated hourly rates up to
nursing, home health aide andreasonable cost
therapies, other services paid on
reasonable charge basis using
Medicare limits
North DakotaProspective cost based rate per visitNot covered
OhioFee for service for nursing, homeFee for service
health aide and therapies, medical
supplies paid 75% list price if no
payment limit available
OklahomaProspective cost based hourly ratePer diem

adjusted for cost of travel and
medical supplies

StateHome health carePersonal care
OregonFee for serviceEstablished hourly rate for
individual providers and
negotiated rate for agencies
PennsylvaniaFee for serviceNot covered
Rhode IslandFee for serviceFee for service using hourly
South CarolinaCost based payment using MedicareNot covered
upper limits for visits, medical
equipment paid at 50th percentile of
Medicare allowable charge
South DakotaFee for service, medical equipmentCost based payment
paid at 75% of charge
TennesseeVisits paid fee for service usingNot covered
Medicare cost ceilings, medical
equipment and supplies paid
percentage of charge
TexasCost based payment for visits,Not covered
medical equipment and supplies paid
fee for service
UtahFee for service, payment for medicalFee for service
equipment and supplies may be
ne go tiated
VermontFee for serviceNot covered
VirginiaFee for service using geographicNot covered
adj ustments
WashingtonFee for service using prevailingHourly rate up to 54.5% of
charge as limitnursing facility rate
West VirginiaVisits paid at Medicare rates, medicalMonthly rate based on hours of
equipment and supplies paid 90% ofcare
Medicare rates
WisconsinFee for service using Medicare costFee for service using hourly
ceilingsrate for care and visit rate for
sup e r visio n
WyomingVisits paid fee for service, medicalNot covered
supplies paid reasonable charge
Source: Kaiser Commission on Medicaid and the Uninsured and National Conference of State
Legislatures, Medicaid Benefits: Services Covered, Limits, Copayments and Reimbursement
Methodologies, at [], as of Nov. 2003.
Note: Payment methods indicated in this table are defined by the authors as follows: “Fee for service”
means the state has established a maximum payment amount for a particular service, or uses the
maximum applicable to the Medicare program for the service, and pays the lesser of the provider’s
charge or this amount. Often the payment is capped by an estimate of cost. “Cost based payment”
means there is a year-end settlement process or some documentation of actual cost is required to justify
payment, while “prospective payment means there is not such a process although the payment rates
are generally based on historical cost. Some states make payment using a “percentage of charge” to
reflect cost, typically using some documentation of a provider’s historical cost to charge ratio. Some
states negotiate payment rates.

Home and Community-Based Services Waivers. The HCBS waiver
programs offer a wide variety of services. One study found that services offered in
at least some states included: adult day care, adult day habilitation, adult day health
services, assistive technology, adaptive equipment, case management, personal care
attendants, habilitation services, homemaker services, home health aide, nursing care
service, personal care services, respite care, training for the family, day treatment or45
other partial hospitalization, and vocational services.
There has been little study of how states pay for particular home and community
services. While reimbursement methodologies for regular Medicaid services are
described in state plans, methodologies for waiver services are not. Instead, they are
described in individual waiver applications and waiver amendments, which are
maintained by CMS regional offices and are not readily accessible. Occasional
surveys of state waiver programs have not addressed reimbursement, in part because
service coverage and service definitions vary so widely that cross-state comparisons
are impossible.46 Even within a state, there may be multiple waiver programs, each
with its own service definitions and payment rules. Finally agencies, providers, and
researchers in this field often tend to think of Medicaid as a “funding stream,”
merged with other sources of funding for clients, and do not focus on individual
Medicaid payment transactions.
Given the complexity of waiver programs and the limits of available
information on reimbursement, a full review of payment rules is beyond the scope of
this report. Instead, the following discussion provides illustrations of how different
states pay for two key services: personal care and case management. These may
serve to highlight some of the policy considerations states must address in paying for
waiver services. The section concludes with a brief review of state systems that use
bundled payments for overall care of an HCBS waiver participant.
Most of the information has been drawn from a set of case studies performed for
CMS by the Lewin Group, the Urban Institute, and the University of Minnesota
Research and Training Center on Community Living. The studies, completed in

2001, reviewed programs for the aged and disabled (AD) in seven states (Alabama,

Indiana, Kentucky, Maryland, Michigan, Washington, and Wisconsin) and programs
for the mentally retarded and developmentally disabled (MRDD) in six states
(Indiana, Kansas, Louisiana, New Jersey, Vermont, and Wyoming).47
Personal Care. Personal care services (both waiver and non-waiver) are
furnished by home health agencies, personal care agencies, and local government
agencies (such as area agencies on aging, AAAs). In addition, many states directly
pay individual personal care attendants. A 1998-1999 survey found that states paid
an average of $13 an hour for agency services, compared to $8-$9 an hour for

45 S. Lutzky, et al., Review of the Medicaid 1915(c) Home and Community Based Services
Waiver Program Literature and Program Data, report prepared for the Health Care
Financing Administration, Reston, VA, June 2000.
46 Personal communication with Brian Burwell.
47 The complete set of studies is available at [].

independent workers. While agency rates were 27% to 39% higher than those for
independent providers, the attendants employed by the agencies earned only a little
more than the minimum wage.48
Some states vary payment, not according to whether services are performed by
an agency or an individual provider, but according to some measure of intensity. For
example, Louisiana pays a slightly higher rate per hour of care for “high need”
participants, $11.36 versus $10.05 per hour of regular care. Maryland’s non-waiver
personal care program uses three daily rates for personal care based on the
complexity of the beneficiary’s needs and informal support system: $10 per day for
minimal assistance furnished in one visit; $20 for extensive help provided in one or
two daily visits; and $50 for constant supervision and assistance throughout the day.
There are concerns that daily, as opposed to hourly, rates give providers incentives
to minimize the hours of care they provide.
However rates are set, case studies found that states often failed to increase them
over time, even to account for inflation. One conclusion was that, given the choice
between increasing enrollment and raising rates, states preferred to cover more
beneficiaries. As is the case for direct care workers in nursing homes, low payment
or wage rates for personal care workers have led to problems in recruiting and high
turnover. The problems are made worse if waiver limits on covered hours per
individual mean that the workers cannot even work full-time.49 In the 1998-1999
survey, 61% of states reported recruiting problems. The authors suggest that supply
problems may in turn lead to poor quality, lack of access, and ultimately premature
institutional placement for participants. High turnover may also mean limited
training and lack of skills needed to provide necessary support services, as well as
difficulty in establishing worker-client relationships.
A few states have made additional payments to agencies that agree to pay higher
wages. However, as Table 23 showed, of 28 states with wage pass-through
programs, only six had extended these programs to personal care services. Some
states, such as Washington, have also acted to improve direct payment to independent
workers (but only by $1 per hour, leaving rates well below those paid to agencies).
At least one state, New Jersey, considered more sweeping action. A 2003
legislative proposal would have shifted all personal care attendants, now employed
by 100 different agencies, to state-administered regional home care councils;
minimum wages would be $10 an hour.50

48 A. LeBlanc, M. Tonner, and C. Harrington, “State Medicaid Programs Offering Personal
Care Services,” Health Care Financing Review, vol. 22, no. 4, summer 2001, pp. 155-173.
49 Paraprofessional Health Care Institute, The Personal Assistance Services and Direct-
Support Workforce: A Literature Review, report prepared for CMS, June 2003.
50 Henry J. Kaiser Family Foundation, “New Jersey Assembly Bill to Establish State-
Administered Home Health Care System Leads to ‘Fierce Debate’,” Daily Health Policy
Report, June 25, 2003, at [

Case Management. Case management includes “assessing the beneficiary’s
needs, developing the plan of care, arranging for the delivery of services, monitoring
the beneficiary, and conducting periodic reassessments of the beneficiary’s needs and
modifying the plan of care as needed.”51 While most waiver programs provide case
management, states differ in how it is structured and paid for. Some states use their
own employees, whether in the Medicaid unit or in another unit such as the agency
responsible for MRDD services. Some contract with agencies or individuals
specifically to provide case management, while others pay public or nonprofit
agencies a bundled rate that includes case management and service provision.
However case management is paid for, a recurring issue is whether the funding
is adequate to hire sufficient qualified personnel. GAO has reported that, in 20 of 51
waiver programs reviewed by CMS regional offices or state audits, case managers
were not providing ongoing assessment and monitoring of waiver beneficiaries or
follow-up of changes in beneficiaries’ care needs was inadequate. This may be due
in part to the fact that managers in some states carry a larger caseload than in others;
the variation may in turn be a function of reimbursement levels.
The state of Washington pays area agencies on aging (AAAs) to manage
participants in both the AD and MRDD waivers. The state treats case management
as part of the AAA’s administrative budget, rather than as a service, and thus allows
a fixed amount set at the start of the year; the amount does not change even if
caseloads rise. In addition the annual budgets assume an average caseload per
manager. This has gone from 100 in the late 1990s to a proposed 75 in
FY2001-FY2003, but some advocates say this is still too high for adequate contact
and quality monitoring.
In New Jersey’s MRDD program, budgets for county-level community services
offices assume three different levels of managers with different caseloads and visit
frequency: primary (direct supervision of an individual living alone), with a caseload
of 35-45 and monthly visits; program (management of people in residential settings
where other workers attend the person), with a caseload of 90-100 and quarterly
visits, and resource (for persons living with the family), providing only annual
contact with up to 250 cases. Indiana’s MRDD program, instead of setting levels of
care, has a fixed fee schedule for different management services, from $8.00 per
quarter hour of ongoing case management to $355 for an initial diagnostic and
evaluation service.
Wyoming allows participants in its MRDD program to choose their own service
coordinator from a listing of qualified individuals. The state has reportedly made a
substantial investment in case management and has one of the lowest caseloads in the
U.S., 25 to 30 participants per manager. The state pays managers a monthly flat fee
of $200 per child and $150 per adult — meaning that managers can make an
adequate living even with the limited caseload. Indiana’s AD program also allows

51 U.S. General Accounting Office, Long-Term Care: Federal Oversight of Growing
Medicaid Home and Community-Based Waivers Should be Strengthened, (GAO-03-576),
Washington, June 2003.

participants to choose between a private case manager and one supplied by the local
AAA; most choose the AAA.
Budgeted or Bundled Payments. Some states, instead of paying for
individual services under their waiver programs, make fixed payments to agencies
for complete services to individual clients or to entire populations.
The Kansas MRDD program originally had a system that paid fixed daily rates
to providers for in-home support, residential services, or day services; rates varied
depending on the client’s assignment to one of five tiers reflecting assessed need. In
2000 Kansas shifted to a system that pays bundled rates (again by acuity tier) for
family/individual supports to an agency selected by the family. While some choose
community service providers, others enroll with a community organization that will
receive the payments and pool them for participating families; each family hires its
own helpers and is reimbursed by the organization. Some groups of families have
formed “participant alliances,” which receive the bundled payments and directly hire
and oversee helpers.
In Vermont’s MRDD program, a single nonprofit agency is designated to serve
most waiver participants in each of 10 geographic regions. The agencies receive
capped annual budgets, based on the previous year’s allocation with inflation or other
adjustments, and are expected to provide all services for their current caseload within
this budget. The allocation is increased when an agency takes on a new participant;
the budgeted amount for new entrants is generally at one of 10 flat rates (from $7,191
to $71,376 per year in 2000) based on the client’s assessed service needs. Within the
overall budget, the agency develops and manages individual budgets in consultation
with families or guardians.
Michigan’s AD program makes fixed daily payments to one of 23 regional
waiver agents, chiefly AAAs or nonprofit organizations. The payment in 2000
included $32 per day for services and $9 or $10 for administration. The agency is at
risk if average service costs for all clients exceed the allowance; there are concerns
that some agencies might treat the $32 figure as an individual ceiling rather than an
aggregate one. To limit potential losses for clients with complex needs, there is an
exceptions process under which an agency can obtain additional funding for clients
whose services cost more than $96 per day.

Managed Care
Over 15 million beneficiaries, or 38% of the Medicaid population, were enrolled
in managed care organizations (MCOs) or other full-risk capitated arrangements as
of June 2002.52 Contracting organizations agree to provide or accept financial
liability for a broad range of Medicaid-covered acute care services in return for a
fixed monthly payment for each enrollee. (Capitation for long-term care is used in
Arizona’s ALTCS program and in a small number of programs of all-inclusive care
for the elderly or PACE programs — programs under which Medicare and Medicaid
make integrated capitation payments for preventive, acute and long-term care
services to MCO-like organizations). Many states require some classes of Medicaid
beneficiaries — such as families eligible on the basis of meeting old welfare program
rules, poverty-related groups of pregnant women and children, or higher-income
families enrolled through Medicaid expansion waivers — to enroll in MCOs, either
statewide or in selected areas. In other states, enrollment is voluntary; beneficiaries
may choose between an MCO and fee-for-service (with or without a primary care
case management feature). Some states have mandatory enrollment in some
geographic areas and voluntary enrollment in others.
Federal Medicaid law requires simply that “prepaid payments to the entity [be]
made on an actuarially sound basis.” Until recently, federal regulations provided that
state payments for enrollees in MCOs could not exceed the “fee-for-service
equivalent” — the estimated amount the state would have spent for a comparable
population not enrolled in the MCO and continuing to receive services on a
fee-for-service basis. (This limit was similar to the adjusted average per capita cost
(AAPCC) formerly used in setting Medicare HMO payments.) The use of the
fee-for-service equivalent as an upper limit was dropped in 2001, partly because
some states had enrolled so many beneficiaries in MCOs that they no longer had
reliable data on fee-for-service experience. Instead the regulations now provide
detailed specifications of what would constitute “actuarially sound” payment rates.
Under the new rules, a qualified actuary must certify that the state’s capitation
rates have been developed in accordance with generally accepted actuarial principles
and practices. Rates must be based only on services covered under the state’s
Medicaid plan; that is, the state may not pay extra for services available under the
MCO contract but not provided to other beneficiaries. Finally, the state must provide
CMS with documentation of the basis for the rates and with an explanation of any
incentive arrangements, or stop-loss, reinsurance, or other risk-sharing
methodologies. (These contractual options are explained below.)

52 Prepaid plans providing only specific Medicaid services, such as behavioral health, dental
care, or non-emergency transportation, had 9.5 million enrollees; some of these were also
enrolled in MCOs.

Rate-Setting Methods
This section provides information on how states establish payment rates for
MCOs. It relies on two national surveys: a 2001 survey by the Urban Institute that
collected information from 36 of the 39 states with full-risk MCO contracts,53 and a
2003 survey of 41 states by the National Academy for State Health Policy. This
survey included the 39 states with full-risk contracts and two additional states,
Alabama and South Dakota, that have prepaid contracts for limited services.
Basic Approach. Capitation rates can be established in three basic ways: the
state can simply set them administratively, it can negotiate with plans, or it can use
a process of competitive bidding. Table 30 shows the basic method in use in states
in 2001. As the Urban Institute Report points out, states may actually use a mix of
methods; a state that sets rates may also engage in negotiation, while one that relies
on competitive bids might have some limits on what bids will be deemed acceptable.
The Urban Institute Report suggests that some states have moved away from
competitive bidding, because it is administratively burdensome and the results are
often subject to disputes. States with larger numbers of enrollees are somewhat more
likely to use administered rate-setting, as are states with a higher proportion of
enrollees in commercial MCOs.
Table 30. Principal Approach to MCO Rate-Setting, 2001
of MCO
MCOenrollees in
Co mpet it iv e enrollment commercial
Administ ered Negotiated bidding ( t ho usa nds) MCOs
Arizona x 528 7%
California x 2,838 84%
Co lo rado x 123 49%
Co nnecticut x 240 84%
Delaware x 8 3 100%
District ofx70%
Co lumb ia
Floridax526 76%
Hawaii x 118 77%
I llino is x 1 3 6 3 8 %
Indiana x 109 0%
Iowa x 5 9 100%
Kansas x 4 6 100%

53 Under a full-risk contract, the MCO assumes liability for most or all covered Medicaid
services. Contracts may have “carve-outs” or exclusions for specific services, such as
mental health care or prescription drugs, which the state buys in some other way; in addition
contracts commonly exclude long-term care services.

of MCO
MCOenrollees in
Co mpet it iv e enrollment commercial
Administ ered Negotiated bidding ( t ho usa nds) MCOs
K e nt uc ky x 1 2 0 0 %
Maryland x 421 0%
Massachusetts x 192 59%
Michigan x 704 63%
Minneso ta x 326 99%
Missourix379 37%
Nevada x 48 100%
New x 6 100%
Hamp shire
New Jerseyx45928%
New Mexicox212100%
New Yorkx69657%
No rth x 44 100%
Car o lina
North Dakotax0100%
Ohio x 278 61%
Oklahoma x 162 87%
Pennsylvaniax860 24%
Rhode Islandx112100%
South Carolinax280%
T exas x 428 68%
Utah x 8 3 84%
Virginia x 157 83%
Washington x 415 78%
West Virginiax43100%
Wisconsin x 266 100%
To tal 1 9 7 10 11,324 62%
Source: John Holahan and Shinobu Suzuki, Medicaid Managed Care Payment Methods and
Capitation Rates in 2001: Results of a New National Survey, Urban Institute, 2002, and CMS
enrollment data as of June 2001.
Note: Enrollment is total in commercial MCOs, Medicaid MCOs, and (in California only), health
insuring organizations (HIOs).

However rates are established, states commonly pay different amounts for
different enrollees, depending on enrollee characteristics. Most states use age, sex,
and eligibility category to set rates. Somewhat fewer use geography, possibly
because some states have contracts that only cover one geographic area. Only a few
states consider whether an enrollee is in an institution or is dually eligible for
Medicare and Medicaid. This is presumably because many states do not enroll
beneficiaries who are aged or disabled in MCOs. Finally 18 states consider health
status in plan payment. Of these, 11 states adjust rates for a few specific conditions
(e.g., HIV) and nine have a comprehensive system of risk adjustment (New Jersey
and Utah use both types of adjustment). Risk adjustment systems are considered
further below.
Table 31. Factors in Capitation Payment, 2002
Number of states usingPercent of states with
Factorfactorrisk contracts
Age 37 90%
Eligibility category3585%
Sex 33 80%
Geography 2 8 68%
Health status1844%
Institutional/Medicare status1229%
States with risk programs41100%
Source: Unpublished data collected by the National Academy for State Health Policy for the
Congressional Research Service in 2003.
In addition to using more or fewer factors, states vary widely in the number of
different “rate cells” they establish — that is, how many categories within each
grouping they use. The Urban Institute Report found that states had as few as two
rate cells (for example, under age 19 and 19 or over) or as many as 126. In a state
with only a few rate cells, plans may have an incentive to seek out lower-cost
populations within each cell. In a state with very many cells, data used to set rates
for each cell might be statistically unreliable, and values could fluctuate widely from54
year to year. Federal regulations currently require that states make adjustments for
eligibility category, age, sex, and geography, but allow a state to explain why an
element is not applicable; inadequate sample size would be one acceptable
explanation. States are encouraged but not required to use adjustments based on
health status or diagnosis.
Reinsurance, Risk Sharing, and Incentive Payments. While MCOs
agree to accept financial liability for the services required by their enrollees, there are
a number of reasons why it may be necessary to limit their potential losses on
individual cases or in the aggregate. First, some plans have very small numbers of
enrollees and could be overwhelmed by a few high-cost cases. In 2002, 52 full-risk

54 This was the case under Medicare’s former HMO payment system, which recalculated
rates annually for every county in the U.S.

contractors had fewer than 5,000 Medicaid enrollees. Second, Medicaid-only MCOs,
especially recently established ones, may have little capital and a limited ability to
sustain losses. Third, a large proportion of spending for MCO enrollees — who are
disproportionately women and children — is related to pregnancy and to problems
of newborns. These unpredictable events cannot be corrected for in prospective
rate-setting, and a plan that gets fewer or more than its share of pregnancies in a
particular year might be considerably over- or underpaid. Table 32 summarizes the
ways in which states limit losses for risk contractors.
Table 32. Reinsurance and Risk-Sharing Arrangements, 2002
Number of states usingPercent of states with
Methodmethodrisk contracts
Reinsurance: commercial1946%
Reinsurance: statea1332%
Condition specific limits717%
Risk corridors512%
Risk pools410%
Recalculate upper payment limit410%
States reporting any arrangement2971%
Source: Unpublished data collected by the National Academy for State Health Policy for the
Congressional Research Service in 2003.
a. New York reported both voluntary and mandatory state reinsurance programs.
One way in which MCOs — whether or not contracting with Medicaid —
commonly limit losses is by purchasing reinsurance (sometimes known as stop-loss
coverage) from a commercial reinsurer. The reinsurer assumes some or all of the
liability when costs for an individual enrollee exceed a deductible amount, such as
$100,000 or $250,000. Less commonly, the entity might obtain aggregate
reinsurance, which limits its overall losses for the year; for example, the reinsurer
might begin paying claims when the entity’s costs exceed 110% of revenues.
In 2002, 19 states reported that their risk contractors obtained commercial
reinsurance; six of these states required contractors to do so. Other states act as
reinsurers themselves. For example, an MCO contract might provide that the
contractor will be liable for the first $25,000 in costs for a Medicaid enrollee, after
which the state would assume liability. Of 13 states with these arrangements,
reinsurance was voluntary in six and mandatory in six; one state reported both
voluntary and mandatory coverage. Seven states provide what amounts to
reinsurance by limiting contractors’ liability for enrollees with specific conditions,
such as HIV.

Five states had contracts with “risk corridors.” These work somewhat like
aggregate reinsurance: the contractor is liable for all costs up to some fixed
percentage of its capitation revenue. Costs above the limit are shared by the
contractor and the state or may be assumed entirely by the state.55 Four states had
risk pools, in which multiple contractors make payments (usually a per capita
assessment) into a combined fund that is used to help pay for high-cost cases.
Finally, there are four states that base their prospective capitation rates for a year on
projected fee-for-service expenditures for a comparable population and recalculate
those rates retrospectively if fee-for-service costs are greater than expected. (NASHP
treats this as a risk-sharing arrangement; however, it is different from the others in
that it does not relate to the contractors’ actual spending.)
Finally, a state may provide supplemental incentive payments to an MCO for
meeting specified performance targets — for example, providing more child health
screenings under the early and periodic screening, diagnostic, and treatment program
(EPSDT). Or the state may share savings with the MCO; that is, it may split the
difference between the MCO’s capitation rates and the estimated amount that would
have been spent for the enrollees in the fee-for-service program. Federal rules
require that these arrangements cannot provide payments greater than 105% of
approved capitation rates, must be offered equally to public and private contractors,
and may not be linked to intergovernmental transfers.
Risk Adjustment. Rate cells based only on age, sex, eligibility category, or
location can capture differences in average spending for entire population groups but
cannot predict the level of risk represented by individual enrollees. Within each
population defined by a rate cell, there can be significant variation in health status
and likely costs. This has two consequences. First, in states in which MCO
enrollment is voluntary, the group choosing to enroll may have better or worse health
status than the group choosing to remain in the fee-for-service system. If capitation
rates are based on fee-for-service experience, overall payments to MCOs might not
reflect the level of risk they are assuming. Second, when multiple MCOs are
competing, any one entity might be over- or underpaid, depending on the health
status of the beneficiaries it enrolled. This creates incentives to market to healthier
beneficiaries and/or promote disenrollment by sicker ones.56
As mentioned earlier, NASHP reports that nine states have attempted to deal57
with this problem by adding risk adjustment to their rate-setting methods. Risk
adjustment systems assign enrollees to some form of health status or diagnostic
category; capitation rates are increased or decreased to reflect higher or lower

55 Federal regulations provide that total payments to the MCO under a risk corridor
arrangement may not exceed what would have been paid under Medicaid fee-for-service for
the services received by enrollees, plus an administrative allowance. Note that this rule
relates to services actually furnished to enrollees, while the repealed fee-for-service
equivalent rule compared capitation rates to average expenditures for a comparable non-
enrolled population.
56 While there are rules prohibiting these practices, they may be difficult to detect, and there
is a long history of such abuses under both Medicaid and Medicare.
57 See page 93.

expected costs for enrollees in each category. NASHP does not provide further
details on what systems states are using or what groups of enrollees the system
applies to. Table 33 has been pieced together from the 2001 Urban Institute survey
and from information reported by Kronick et al., in 2000.
Table 33. Risk Adjustment Systems
StatePopulation coveredClassification system
ColoradoSSI + 1931DPS
DelawareSSI + 1931CDPS
MarylandSSI + 1931ACGs
Michigan SSI CD P S
Minneso ta 1931 ACGs
New JerseySSIDPS
Orego n SSI DP S
Washington 1931 CDPS
Source: Holahan and Suzuki, 2002, and Richard Kronick et al., “Improving Health-Based Payment
for Medicaid Beneficiaries: CDPS,” Health Care Financing Review, vol. 21, no. 3, spring 2000, p.
Note: SSI = Supplemental Security Income. 1931= Families eligible under Section 1931 of Title XIX
of SSA. These are families eligible on the basis of meeting welfare program rules in place in 1996.
ACGs = Adjusted Clinical Groups. DPS = Disability Payment System. CDPS = Chronic Illness and
Disability Payment System.
Four of the states adjust payments for aged and disabled SSI recipients; two
adjust payments for recipients who are members of families with dependent children
qualifying under old welfare program rules; and three adjust for both groups.
Maryland and Minnesota use adjusted clinical groups (ACGs). This system,
originally developed for use in the Medicare+Choice program, uses diagnostic
information from past inpatient and ambulatory claims (or encounter reports), along
with age and sex, to assign each enrollee to a risk category. (There are several ACG
models, using different numbers of groupings.) The remaining states use the
Disability Payment System (DPS) or the more recent refinement, the Chronic Illness
and Disability Payment System (CDPS). These systems, developed by researchers
at the University of California, San Diego, also use data from past claims to classify
enrollees by diagnoses and complications.
Kronick et al., report that the CDPS method performs better than ACGs in
predicting costs for the disabled. Neither system was very good at predicting costs
for families eligible on the basis of meeting former welfare program rules (Section
1931 eligibles). One key problem in risk adjustment for Medicaid is that many
beneficiaries are enrolled only for a limited period. This means that it may not be
possible to collect sufficient data to classify short-term enrollees under a given

system, and the data are likely to be less reliable than data collected for longer-term
enrol l ees. 58
Most states that have not adopted broader risk adjustment have made special
provisions for costs related to pregnancies and newborns. Of the 36 states
responding to the Urban Institute survey, 28 had some provision. Of these, 21 made
lump-sum payments to MCOs for each pregnancy. Twelve had established separate
rates for women qualifying for Medicaid on the basis of pregnancy; one of these also
paid a higher rate for a Section 1931-eligible pregnant woman. Finally, eight states
paid higher rates for newborns or infants under one year old.
Payment Levels
Table 34 shows 1998 and 2001 Medicaid managed care rates for Section 1931
and poverty-related populations in selected states; the rates are adjusted for some
differences in population and in services covered under contracts. In both years, the
highest-paying state (North Dakota) paid nearly twice as much as the lowest-paying
state (Oklahoma). Rates increased an average of 18% over three years, but some
states were cutting rates while others were increasing substantially.
Differences were not found to correlate with differences in Medicare AAPCCs
for the areas served, suggesting that general geographic variation in health spending
does not explain the wide variation in rates. The authors note that there is
considerable variation in states’ Medicaid fee-for-service spending as well,
depending in part on state cost containment efforts. This variation carries over into
MCO rates in states that base rates on fee-for-service experience or consider that
experience in negotiating or evaluating bids. Rates may also be affected by biased
selection (differences in health risk of MCO and non-MCO beneficiaries not
corrected for through risk adjustment) and by whether a state enrolls beneficiaries
statewide or only in high-cost urban areas.
Table 34. Change in Medicaid Managed Care Payment Rates,
Section 1931 and Poverty-Related Groups, Selected States,
State19982001Percentage change
Arizona $126.29 $131.54 4.0
California 94.68 137.79 45.5
Co lo rado 137.57 134.36 -2 .3
Co nnecticut 161.57 169.37 4.8
District of Columbia143.39186.430.0
Florid a 117.29 135.82 15.8
Georgia 94.29 NA NA

58 E. Adams, J. Bronstein, and C. Raskiind-Hood, “Adjusted Clincal Groups: Predictive
Accuracy for Medicaid Enrollees in Three States,” Health Care Financing Review, vol. 24,
no. 1, fall 2002, pp. 43-61.

State19982001Percentage change
Hawaii 150.25 147.64 -1 .7
Illinois 112.22 146.36 30.4
Indiana 94.98 164.84 73.5
Iowa 142.6 181.43 27.2
Kansas 115.07 134.84 17.2
Kentucky 156.42 191.95 22.7
Maine 102.74 NA NA
Maryland 145.75 180.05 23.5
Massachusetts 169.32 170.96 1.0
Minneso ta 151.78 202.36 33.3
Mississippi 126.92 NA NA
Michigan 140.32 151.79 8.2
Nevada 96.74 128.06 32.4
New Hampshire148.9175.9518.2
New Jersey152.16143.04-6.0
New Mexico138.04186.9435.4
New York108.24149.4138.0
North Dakota221.83209.34-5.6
Ohio 147.91 162.3 9 .7
Oklaho ma 92.18 118.32 28.4
Rhode Island133.49159.2919.3
South Carolina132.45141.386.7
T ennessee 101.2 NA NA
T exas 133.2 127.63 -4 .2
Utah 137.22 140.21 2.2
Virginia 146.33 190.35 30.1
Washington 130.47 154.67 18.6
West Virginia132.87143.818.2
Wisconsin 116.13 132.44 14.1
Average 132.02 157.21 18.1
Source: John Holahan and Shinobu Suzuki, “Medicaid Managed Care Payment Methods and
Capitation Rates In 2001,” Health Affairs, vol. 22, no. 1 (Jan.-Feb. 2003), pp. 204-218. Rates are
adjusted for some differences in population and in services covered under contracts.
Whether Medicaid managed care rates are “adequate” is difficult to assess. As
Table 35 shows, the number of MCOs participating in Medicaid has dropped, with
the number of commercial plans dropping especially sharply. (Some of this change
may reflect plan mergers, rather than withdrawal from Medicaid.) At the same time,
Medicaid MCO enrollment has been growing, so that more enrollees are served by
fewer plans. This suggests that small plans, or plans with a small number of

Medicaid enrollees, might have been more likely to drop out. One reason might be
that Medicaid imposes special administrative requirements; plans with few Medicaid
enrollees cannot justify the costs. Or they may fail to develop the care systems
needed to effectively manage Medicaid beneficiaries.59
The commercial MCO share of enrollment peaked in 1999 and has since been
dropping. This trend is not necessarily related to payment rates; this was a period
when many MCOs were losing money and reevaluating their overall business
strategies. At one time, it would have been assumed that declining commercial
participation raised concerns about access and quality. Until passage of the BBA,
most MCO plans participating in Medicaid were required to have commercial
enrollees as well, on the theory that employers and private enrollees were in a better
position to press for high-quality care than Medicaid beneficiaries. If this was ever
true, since the passage of the BBA, states have arguably been at least as active in
monitoring quality as private purchasers.
However, even if increasing reliance on Medicaid-only plans does not
necessarily mean quality problems, it is not clear how these plans will be affected by
the current financial pressures on states. As was shown in Table 8, above, about half
of states with MCO programs plan to freeze or reduce their capitation rates for
FY2004. Medicaid-only plans cannot shift losses to other purchasers and would need
to reduce their provider payments or limit access to care.60
Table 35. Commercial and Medicaid-Only MCO Plans
and Enrollment, 1998-2002
PlansEnrollees (thousands)
Percent of
enrollees in
Medicaid Medicaid commercial
Commercial only Commercial only MCOs
1998 283 136 7,248 4,645 60.9%
1999 237 136 8,488 3,524 70.7%
2000 210 127 8,396 4,016 67.6%
2001202122 8,846 4,61765.7%
20021881209,734 5,72363.0%
Source: CMS annual summaries, Managed Care Enrollment by Program Type,
[], as of Sept. 2003.

59 For a discussion of these issues, see M. McCue, et al., “Financial Performance and
Participation in Medicaid and Medi-Cal Managed Care.” Health Care Financing Review,
vol. 23, no. 2, winter 2001, pp. 69-81.
60 R. Hurley and D. Draper, “Medicaid Confronts a Changing Managed Care Marketplace,”
Health Care Financing Review, vol. 24, no. 1, fall, winter 2002, pp. 11-25.

Prescription Drugs
Medicaid payment for a prescription drug furnished to a beneficiary on an
outpatient basis has two components: an amount to cover the cost of the ingredients
(the acquisition or ingredient cost) and an amount to cover the pharmacy’s costs to
fill the prescription (the dispensing fee). Medicaid regulations establish upper limits
on payment for acquisition costs, but do not limit dispensing fees; these must merely
be “reasonable.” Two separate limits on acquisition costs are used, one for certain
multiple source drugs — those for which therapeutic equivalents or “generic”
versions are available from three or more suppliers — and one for all other drugs.
The limits are designed to encourage the substitution of lower cost generic61
equivalents for more costly brand name drugs.
Since 1991, pharmaceutical manufacturers have been required to give rebates
to states for drugs paid for by Medicaid. The rebate formulas are designed to assure
that states pay the lowest price offered by the manufacturer to any other high-volume
purchaser. In return, the state must generally cover all the drugs marketed by the
This section reviews the regulatory limits on reimbursement and summarizes
state policies on acquisition costs and dispensing fees. It then provides an overview
of recent state initiatives in prescription drug purchasing.
Pharmacy Reimbursement Methods
Upper Payment Limits
Multiple Source Drugs. For purposes of the upper payment limits, a
“multiple source drug” is one that meets the following two requirements: (a) the
drug is made available by at least three different suppliers, and (b) the Food and Drug
Administration (FDA) has determined that at least three approved formulations of the
drug are “therapeutically equivalent” — that is, contain identical doses of the active
ingredient and have the same biological effects. For each multiple source drug,
CMS establishes a price limit (known as the maximum allowable cost, or MAC)
equal to 150% of the estimated wholesale cost of the least expensive therapeutic
equivalent. A state’s payments for such drugs during a given period may not exceed
what would have been spent if the state had paid the price limit plus a reasonable
dispensing fee. (Note that the federal MAC limits apply to aggregate spending for
the listed drugs. Many states have established their own MAC systems, which may
have higher price limits for some drugs and lower limits for others.)
The effect of the MAC limits is that, when a lower-cost “generic” equivalent
exists for a brand-name drug, a pharmacy will be paid the generic price even if the

61 For additional information on Medicaid’s payment for and coverage of prescription drugs
see CRS Report RL30726, Prescription Drug Coverage Under Medicaid, by Jean Hearne.

brand-name drug is actually furnished. The pharmacy therefore has a financial
incentive to substitute the generic equivalent for the brand-name drug. If the
prescribing physician specifies that generic substitution is unacceptable (for example,
by writing “dispense as written” or “no substitution” on the prescription), the CMS
price limits do not apply. The pharmacy must supply the brand-name drug and may
be paid the full brand-name cost.
Other Drugs. For all other drugs (including multiple source drugs for which
the prescribing physician has requested no substitution), statewide payments may not
exceed the lesser of (a) the pharmacies’ usual and customary charge to the general
public and (b) the estimated acquisition cost (EAC) plus a dispensing fee. The EAC
or ingredient cost is the state’s best estimate of what pharmacies are generally paying
for a drug.
Dispensing Fees and Ingredient Costs
Table 36 shows the dispensing fee established by each state as well as the
state’s method of computing the ingredient cost or EAC.
Dispensing Fees. Most states pay fixed dispensing fees ranging from about
$3 to $6 per prescription. Some states pay different fees for brand-name and generic
drugs; for drugs dispensed in nursing facilities (or for drugs provided in the unit dose
systems often used in NFs), or for drugs compounded by the pharmacist from
multiple ingredients. A few states pay different fees to different pharmacies,
depending on area, the pharmacy’s historic costs, or volume of Medicaid or other
state-paid prescriptions. New Jersey pays extra for pharmacies providing 24-hour
emergency service.
Wisconsin pays higher fees to pharmacists providing “pharmaceutical care,”
added services such as patient assessment, counseling, or contact with a physician.
New Jersey has a much smaller add-on for patient counseling, and three other states
pay pharmacists for services related to specific treatments or conditions: Alabama
(counseling on Clozaril, a medication for schizophrenia), Missouri (diabetes
education), and Washington (emergency contraceptive counseling).
It is possible that many states’ dispensing fees are less than the actual overhead
cost to pharmacies. GAO has cited studies done for the California and Texas
Medicaid programs finding median 2002 costs of $6.95 and $5.95 per prescription,
respectively, and a study by the National Association of Chain Drug Stores finding
average 2001 costs of $7.26.62
Ingredient Cost. Most states base the EAC for a particular drug on the
average wholesale price (AWP). This measure is not, as its name would suggest, the
average price charged by wholesalers to retail pharmacies. The AWP is a figure
reported by the drug’s manufacturer to several firms that maintain pricing databases

62 U.S. General Accounting Office (GAO), Federal Employees’ Health Benefits: Effects of
Using Pharmacy Benefit Managers on Health Plans, Enrollees, and Pharmacies, (GAO-03-

196), Jan. 2003.

used by states and other purchasers. It resembles a list price or sticker price, and does
not reflect what pharmacies are actually paying wholesalers after discounts or rebates.
To correct for this, states commonly arrive at an EAC by applying a fixed percentage
reduction to the AWP.
Some states also consider the wholesaler’s acquisition cost (WAC), what the
wholesaler paid for the drug, and then add a fixed percentage amount to reflect the
wholesaler’s markup. A few states ascertain the actual acquisition cost (AAC), what
a specific retailer actually paid for a drug, or the “direct price” charged to the retailer.
Several states use different methods to establish the EAC for different classes of
drugs. Two, Louisiana and Michigan, pay lower prices to chain pharmacies, on the
assumption that high-volume pharmacies pay lower wholesale prices.
Table 36. Pharmacy Dispensing Fees and Ingredient
Reimbursement Basis, 2002
StateDispensing feeIngredient reimbursement basis
Alabama$5.40AWP- 10%; WAC+9.2%
Alaska$3.45 minimumaAWP-5%
ArizonaMost drugs paid through AHCCCS plans
Arkansas $5.51 AW P-10.5%
California $4.05 AW P-10%
Colorado$4.00; $1.89 for institutionsAWP-13.5% or WAC+18%,
whichever is lowest; AWP-35%
(for generics)
Co nnecticut $3.85 AW P-12%
Delaware $3.65 AW P-12.9%
District of Columbia$3.75AWP-10%
Florida$4.23-$4.73 (LTC)AWP-13.25%; WAC+7%
Georgia$4.63 + $0.50 (for generics)AWP-10%
Hawaii $4.67 AW P-10.5%
Idaho$4.94 ($5.54 for unit dose)AWP-12%
IllinoisG: $5.10, B: $4.00B: AWP-11%, G: AWP-20%
Indiana$4.90B: AWP-13.5%, G: AWP-20%
Iowa $5.17 AW P-10%
Kansas$3.40 B: AWP-15%, G: AWP-27% IV
AWP-50%, blood AWP-30%
Kentucky $4.51 AW P-12%
Louisiana$5.77AWP-13.5% (AWP-15% for
Maine$3.35 (+extra fees forAWP-13%
c o mp o und i ng)
Maryland$4.21Lowest of : WAC+10%,
direct+10%, AWP-10%
MassachusettsB: $3.50 G: $5.00WAC+5%
Michigan$3.72 AWP-13.5% (1-4 stores),
AWP-15.1% (5+stores)
Minnesota$3.65 AWP-9%
Mississippi$3.91AWP-12 %

StateDispensing feeIngredient reimbursement basis
Missouri$4.09AWP-10.43%, WAC+10%
Montana$2.00 - $4.70bAWP-15%, direct price for some
lab e ler s
Nebraska$3.27 - $5.00cAWP-11%
Nevada$4.76 AWP-15%
New Hampshire$2.50AWP-12%
New Jersey$3.73 - $4.07dAWP-10%, WAC+30%, AAC for
inj ectables
New Mexico$3.65AWP-12.5%
New YorkB: $3.50 G: $4.50AWP-10%
North CarolinaB: $4.00 G: $5.60AWP-10%
North Dakota$5.10AWP-10%
Ohio$3.70WAC + 9%
Oklahoma $4.15 AW P-12.0%
OregonRetail: $3.50 Inst./NF: $3.80AWP-13%
Pennsylvania$4.00 ($5.00 for compounds)AWP-10%
Rhode IslandOP: $3.40, LTC: $2.85WAC+5%
South Carolina$4.05AWP-10%
South Dakota$4.75 ($5.55 for unit dose)AWP-10.5%
TennesseeMost drugs paid through TennCare plans
TexasFormulaeAWP-15% or WAC+12%,
whichever is lowest
Utah$3.90-$4.40 (based on area)AWP-15%
Vermont$4.25 AWP-11.9%
Virginia $4.25 AW P-10.25%
Washington$4.20-$5.20 (based on annual #AWP-14%
of Rx)
West Virginia$3.90 (+ extra $1.00 forAWP-12%
c o mp o und i ng)
Wisconsin$4.88 (to a maximum $40.11)AWP-11.25%
Wyoming $5.00 AW P-11%
Source: National Pharmaceutical Council, Pharmaceutical Benefits Under State Medical Assistance
Programs 2002, 2003.
Note: B = Brand; G = Generic; AWP = Average Wholesale Price; WAC = Wholesalers Acquisition
Co st.
a. Pharmacy-specific, using formula that considers prescription volume and square footage.
b. Based on pharmacys reported costs, with ceiling and floor.
c. Peer group comparison.
d. Highest rate includes add-ons for emergency service, consultation, and high Medicaid/other state
program volume.
e. Average dispensing expense (ADE) formula for payment: (Estimated acquisition cost + 5.27)
divided by 0.980 = amount paid + $0.15 delivery service.

In recent years the DHHS Office of the Inspector General (OIG), the Department
of Justice, and some states have expressed concerns about the use of the AWP in
prescription drug payment. Because it is an arbitrary number that does not reflect
what pharmacies actually pay wholesalers, all states that use AWP take a fixed
percentage reduction. But there have been allegations that some manufacturers report
highly inflated AWPs. As a result, even though the state pays less than the full AWP,
it may pay the pharmacy much more than the pharmacy actually paid for the drug.
The OIG has recomputed AWPs for some commonly prescribed drugs, using actual
wholesale transaction data; 30 states had used some of these revised AWPs as of
2001.63 The OIG has also suggested that states take larger fixed percentage
discounts. On the basis of sample pharmacy pricing information, it found that actual
1999 acquisition costs were 21.84% less than AWP for brand name drugs and

65.93% less than AWP for generics.64

Drug Rebate Requirements
OBRA 90 required that drug manufacturers, as a condition of Medicaid
coverage of their prescription drug products, enter into agreements with the Secretary
of HHS, under which they pay state Medicaid programs rebates for
Medicaid-reimbursed drugs.65 In return, states are required to cover under Medicaid
all of the drugs marketed by that manufacturer, with certain exceptions. States may
require prior authorization to dispense certain drugs or can establish a formulary, a
listing of preferred drugs, and require authorization for all drugs not on the list.
There are also certain categories of drugs which can be excluded from coverage
ent i rel y. 66
Rebate requirements do not apply to products dispensed as a part of a service
provided in a hospital, physician’s or dentist’s office, or similar setting, or to drugs
provided by MCOs when payment is included in the capitation rate. However, they
do apply to drugs dispensed in nursing facilities if the cost of the drugs is not
included in the NF’s Medicaid per diem payment and is instead reimbursed by
Medicaid to the dispensing pharmacy. (For this reason, as of 1998, only three states
still included prescription drugs in NF payments.) Rebate requirements may also
apply to a nonprescription item such as aspirin, if it is covered in a state’s Medicaid

63 U.S. Department of Health and Human Services (DHHS), Office of the Inspector General,
Medicaid’s Use of Revised Average Wholesale Prices, (OEI-03-01-00010), Sept. 2001.
64 Ibid., Medicaid Pharmacy — Actual Acquisition Cost of Brand Name Prescription Drug
Products, (A-06-00-00023), Aug. 2001, and Medicaid Pharmacy — Actual Acquisition Cost
of Generic Prescription Drug Products, (A-06-01-00053), Mar. 2002.
65 Manufacturers are also required (again as a condition of Medicaid reimbursement) to
provide similar rebates to FQHCs and other PHS-funded entities, public disproportionate
share hospitals, and various other specified purchasers.
66 There are 10 categories of drugs that a state is permitted to exclude. Examples include
medications for weight loss, fertility drugs, drugs for hair loss, and barbiturates.

Rebates are computed and paid to states each quarter on the basis of price
information supplied by manufacturers to CMS and utilization information furnished
to the manufacturers by state Medicaid agencies. (The federal share is recovered
through an adjustment in federal matching payments to states.)
In setting the amount of required rebates, the law distinguishes between two
classes of drugs. The first includes single source drugs (generally, those still under
patent) and “innovator” multiple source drugs (drugs originally marketed under a
patent but for which generic competition now exists). The second class includes all
other, “non-innovator” multiple source drugs (generics).
Single Source and Innovator Multiple Source Drugs. For these drugs,
manufacturers are required to pay state Medicaid programs a basic rebate for each
covered drug, along with an additional rebate if drug product prices increase faster
than inflation, as measured by the Consumer Price Index for all urban consumers
Basic rebate amounts are determined by comparing the average manufacturer
price (AMP) for a drug — the average price paid by wholesalers — to the “best
price,” the lowest price offered by the manufacturer in the same period to any
wholesaler, retailer, provider, HMO, nonprofit entity, or governmental entity in the
U.S. Prices offered to federal agencies, state pharmaceutical assistance programs,
and certain other entities are not considered in determining the best price. Under the
Medicare Prescription Drug, Improvement, and Modernization Act of 2003, prices
negotiated from manufacturers for discount card drugs and prices negotiated by
private plans providing the new Medicare prescription drug benefit will also be
excluded from the calculation.
The basic rebate is the greater of 15.1% of the AMP or the difference between
the AMP and the best price. The additional rebates are required for any drug whose
price increases faster than the CPI-U. In determining the rebate, prices in effect on
October 1, 1990 — or the date of introduction, if later — are used as a base; these are
then compared with prices as of the month before the start of the period for which the
rebate is to be issued.
Non-innovator Multiple Source Drugs. For multiple source drugs, basic
rebates are a fixed 11% of the AMP. Prices offered to other payers are not
considered, nor is there any additional rebate for excess price increases.
Table 37 shows prescription drug spending in FY2001 before and after rebates
(including supplemental rebates in some states; see below). Nationally, rebates
reduced spending by about 20%.

Table 37. Effect of Rebates on
Medicaid Drug Spending, FY2001
Prescription drug spending
($ thousands)
StateBefore rebateAfter rebatePercent reduction
Alabama 387 310 19.8%
Alaska 56 44 20.3%
Ar i z o n a a 3 3 0.0%
Arkansas 242 196 18.9%
Califo r ni a b 2,984 2,198 26.3%
Co lo rado 166 132 20.6%
Co nnecticut 305 243 20.3%
Delaware 81 64 21.0%
District of Columbia645316.5%
Flo r id a b 1,476 1,178 20.1%
Georgia 736 626 15.0%
Hawaii75 6119.2%
Idaho 103 84 18.3%
I llino is 8 8 4 7 1 3 1 9 . 3 %
Indiana 562 458 18.4%
Iowa 235 192 18.2%
Kansas 185 145 21.5%
Kentucky 592 487 17.7%
Lo uisiana 585 470 19.7%
Maine 192 150 21.8%
Maryland b 244 210 14.0%
Massachusetts 798 617 22.6%
Michigan 585 473 19.1%
Minneso ta 266 211 20.5%
Mississippi 493 405 17.9%
Missouri676542 19.8%
Montana 7 3 5 9 18.4%
Nebraska171 14117.7%
Nevada 62 45 26.6%
New Hampshire927815.2%
New Jersey65152719.1%
New Mexico584620.9%
New York2,9862,44218.2%
North Carolina98577721.1%
North Dakota443519.9%
Ohio1,100 88219.8%

Prescription drug spending
($ thousands)
StateBefore rebateAfter rebatePercent reduction
Oklaho ma 171 131 23.5%
Oregon 229 194 15.3%
Pennsylvania 693 563 18.7%
Rhode Island103 8120.9%
South Carolina43934321.7%
South Dakota524218.2%
T ennessee 681 579 15.1%
T exas 1 ,326 1,057 20.3%
Utah 117 95 18.7%
Vermont 104 82 21.1%
Virginia 418 338 19.0%
Washington 458 367 19.9%
West Virginia26020720.2%
Wisconsinb382 30320.8%
Wyoming 312618.5%
U.S. total24,65719,70920.1%
Source: Medicaid Financial Management Reports (CMS 64).
a. Arizona did not report rebates on its negligible drug spending for non-MCO services.
b. Includes rebates under state supplemental agreements.
The rebate formula has been criticized on several grounds. First, there may be
inconsistencies in the way different manufacturers compute the AMP or the best
price. Second, because the rebates are based on AMP while most states use the AWP
to determine pharmacy reimbursement, there is no relationship between the rebate67
amount and the amount the state actually spent on the drug.
The President’s FY2003 budget proposal would have changed the formula for
computing rebates for single source and innovator multiple source drugs. Instead of
comparing the manufacturer’s best price to the AMP, it would compare the best price
to the AWP. The Bush Administration estimated five-year savings of $5.5 billion.
The FY2004 budget proposal did not repeat this specific proposal but included
unspecified reforms of the rebate system intended to produce $13.2 billion in savings68

over10 years.
67 DHHS, Office of the Inspector General, 2002 Red Book (The Cost-Saver Handbook).
68 DHHS, Budget in Brief, FY2003 and Budget in Brief, FY2004.

Recent State Initiatives
Supplemental Rebates. Some states have negotiated supplemental rebates
from manufacturers, in return for which the state might agree to include all the
manufacturer’s products on its formulary or waive prior authorization. As of
FY2001, only California showed significant savings from these agreements, with
supplemental rebates equal to 7% of gross drug spending.
At least three states, Florida, Michigan, and Maine, automatically require prior
authorization unless the manufacturer agrees to pay a supplemental rebate. Florida
requires a 10% rebate, while Michigan requires a rebate equal to the difference
between the drug’s price and the lowest price for any drug in the same “therapeutic
class” (a group of drugs with similar formulas, effects, or clinical use).69 CMS has70
notified states of its willingness to approve similar systems. Federal courts have
rejected suits by manufacturers and patient advocates seeking to block the Florida
and Maine programs.
Maine’s program, MaineRx, is not intended to produce savings for Medicaid but
to help residents without prescription drug coverage. The state plans to negotiate
supplemental rebates with manufacturers in return for an exemption from prior
authorization requirements. Instead of retaining the rebate, the state would pass it on
to pharmacies, which would in turn give discounts to uninsured consumers. The
Pharmaceutical Research and Manufacturers Association (PhRMA) obtained an
injunction blocking implementation of the program, partly on the ground that
Medicaid law prohibited Maine from limiting access to services by Medicaid
beneficiaries to benefit other citizens. The Supreme Court lifted the injunction in
May 2003, but did not actually rule that the program was legal; it could still be71
subject to scrutiny by CMS or lower courts.
Pharmacy Plus. CMS is sponsoring a demonstration program under which
states may extend pharmacy benefits to Medicare beneficiaries and disabled people
with incomes below 200% of poverty who do not already receive Medicaid drug
coverage.72 As of September 2003, programs have been approved in Florida, Illinois,
South Carolina, and Wisconsin. The programs must be budget-neutral; that is, the
state must achieve some savings elsewhere in the Medicaid program that will offset73
the costs of the coverage. Because the benefits are treated as Medicaid coverage,

69 Two drug companies, Pfizer and Bristol-Myers Squibb, were exempted in return for
offering special programs to improve care for Medicaid beneficiaries. D. Gencarelli,
Medicaid Prescription Drug Coverage: State Efforts to Control Costs, National Health
Policy Forum, NHPF Issue Brief 790, May 2003.
70 CMS, State Medicaid Directors Letter, no. 02-014, Sept. 2002.
71 Pharmaceutical Research and Manufacturers of America v. Walsh, U.S. Supreme Court,
no. 01-188, May 19, 2003.
72 The program is operated under Section 1115 of the Social Security Act, which authorizes
the Secretary to waive provisions of Medicaid law to conduct demonstrations.
73 Some of the states with approved plans have met this requirement by arguing that the

the programs will automatically get Medicaid-level rebates without the side
agreements planned for the MaineRx program. (At this writing, it is unknown how
these demonstrations will be affected by the enactment of the Medicare prescription
drug benefit, which provides low-income subsidies to people with incomes below

150% of poverty.)

Purchasing Pools. In February 2003, Michigan and Vermont announced the
formation of a joint purchasing pool that would operate in essentially the same way74
as Michigan’s existing supplemental rebate program. Manufacturers would bid for
inclusion on a preferred drug list to be used by both states. South Carolina has since
joined the initiative, and about 10 other states are reportedly discussing
participation.75 A number of other states have been considering similar coalitions.
Georgia and Texas have been exploring a different approach, under which the state
Medicaid agency and other agencies within the same state, possibly including the76
state employees’ health benefit plan, would engage in joint purchasing.
Cost Containment. Because growth in prescription drug costs has been a
major factor in Medicaid spending increases, most states have taken measures to
control spending growth. Table 38 shows changes implemented in FY2003 or
planned for FY2004. Relatively few states have directly reduced AWP-based
payments, but more are seeking supplemental rebates. As the preceding discussion
would suggest, some of the measures designed to steer beneficiaries toward preferred
drugs may also be intended to give manufacturers incentives to negotiate rebates.
States’ ability to command larger discounts for Medicaid beneficiaries or
leverage their buying power on behalf of other state residents may have been
diminished by passage of the Medicare Prescription Drug, Improvement, and
Modernization Act of 2003. Dual eligibles (beneficiaries eligible for both Medicare
and full Medicaid benefits) will receive covered drugs through the new Medicare
benefit beginning in 2006. This will significantly reduce the volume of drugs
Medicaid programs pay for directly.77 Spending for beneficiaries aged 65 and older
accounted for 32% of Medicaid drug spending in 1999, and this figure does not

73 (...continued)
newly covered groups themselves might otherwise have incurred high hospital or nursing
home bills that would have led them to spend down to Medicaid.
74 Michigan Department of Community Health, “Michigan and Vermont Implement Nation’s
First Multi-State Medicaid Pharmaceutical Pooling Program, press release, Feb. 20, 2003,
at [,1607,7-132-8347-61874 — M_2003_2,00.html] as of
Sept. 2003.
75 S. Lueck, “States’ Efforts to Cut Drug Prices Get a Boost from Medicaid Chief,” Wall
Street Journal, May 30, 2003.
76 National Conference of State Legislatures, Pharmaceutical Bulk Purchasing: Multi-state
and Inter-agency Plans, at [] as of Sept.


77 States will contribute to drug costs for dual eligibles indirectly, through required
maintenance of effort payments; see below.

include spending for nonelderly disabled dual eligibles. The shift of this spending
to Medicare may reduce states’ bargaining power.
Table 38. Number of States Making Medicaid Prescription Drug
Policy Changes, FY2003 and FY2004
FY2003Planned FY2004
AWP less greater discount178
More prescriptions under prior authorization3225
Preferred drug list1430
New or higher beneficiary copays1417
Seek supplemental rebates1121
Require generics45
Limit number of prescriptions per month54
Source: V. Smith, et al., States Respond to Fiscal Pressure: State Medicaid Spending Growth and
Cost Containment in Fiscal Years 2003 and 2004, Kaiser Commission on Medicaid and the
Uninsured, 2003.
Other Payment Requirements
Federal Rules for Specified Services
Although the services already discussed account for the vast majority of
Medicaid spending, state programs cover many other mandatory or optional services.
The following discussion is limited to services for which special payment rules are
established by federal law.
Home and Community-Based Care Option. In addition to the HCBS
waiver, Medicaid law permits states to offer home and community-based care for
functionally disabled elderly persons as an optional Medicaid service. While waiver
programs have limits on the numbers of participants, a state offering the HCBS
optional service must cover everyone meeting eligibility standards established by the
state. Perhaps for this reason, only Indiana and Texas reported payments for this
service in FY2001. Payment rates for the care must be reasonable and adequate to
meet the costs of providing the care efficiently, economically, and in conformity with
laws and quality and safety requirements. However, the federal statute provides that
payments over the course of a quarter for persons receiving the services may not
exceed 50% of what would have been paid by Medicare to treat the same average
number of patients in a nursing facility in the state.
Hospice Services. If the state elects to cover hospice services, it must follow
Medicare reimbursement rules for hospices, with minor differences. Under Medicare
rules, payment for each day of care furnished by the hospice is at fixed rates
according to the nature of the care received by the patient: a day may be classed as
routine home care, continuous home care, inpatient respite care, or general inpatient
care. Average payments per patient are subject to an annual “cap amount” updated

annually by the Secretary and applied on an aggregate basis. The hospice cap amount
for the year ending October 31, 2003 was $18,661.29. The aggregate number of
inpatient care days provided by the hospice in any 12 month period may not exceed

20% of the total number of days of hospice care provided.

Medicaid programs may or may not use the annual cap, but must impose the
limit on inpatient days; Medicaid inpatient days furnished to patients with AIDS are
not counted towards this limit. Under Medicaid, additional payment for room and
board may be made for patients who receive hospice services while residing in a
nursing facility or ICF/MR (this is not true under Medicare).
Indian Health Service. The Indian Health Service (IHS) within the
Department of Health and Human Services (DHHS) provides or purchases health
services for certain groups of Native Americans and Alaska Natives. The IHS
provides care in two different ways: directly through IHS facilities or tribally owned
and operated facilities, and on a “contract care” basis through referral to
off-reservation health care providers. The IHS requires that alternative payment
resources available to users of IHS services be exhausted before IHS will accept
financial responsibility. Native Americans may qualify for Medicaid in the same way
as any other population, by meeting categorical and financial standards.
In the case of services provided in IHS facilities to Medicaid beneficiaries, IHS78
or the facility bills Medicaid directly, at fixed rates established annually by DHHS.
Federal matching funds for services in IHS facilities are available at 100%, rather
than at the state’s usual matching rate. When services are furnished by a contract
provider, the provider is expected to collect from any “alternative resource” available
to the patient, including Medicaid, Medicare, or other health insurance, before
seeking reimbursement from IHS. If Medicaid is the responsible payer, federal
funding is available at the state’s usual matching rate; the state is liable for the
remainder, as with any other Medicaid service.
Laboratory Services. Payment for a laboratory test performed by a
physician, independent laboratory, or hospital (except tests for the hospital’s own
inpatients) may not exceed the amount that would be paid under Medicare rules for
the same test. Medicare payment is based on regional fee schedules established by
the Secretary for each type of test.
Programs of All-Inclusive Care for the Elderly (PACE). The BBA
authorized the PACE program, under which Medicare and state Medicaid programs
make integrated capitation payments for preventative, acute, and long-term care to
MCO-like organizations that furnish services to frail elderly people. (As of January
2003, 10 states had approved PACE programs.) The Medicaid component of the
capitation for PACE enrollees is negotiated by the state and the provider. It is
required to be less than would have been spent for comparable individuals in the
fee-for-service sector, taking into account the comparative frailty of PACE enrollees

78 The Alaska Native and American Indian Direct Reimbursement Act of 2000 (P.L. 106-
417) authorized direct Medicaid billing by tribes, tribal organizations, or Alaska Native
health organizations that are operating IHS owned or leased facilities.

and “such other factors as the Secretary determines to be appropriate.” (The
Secretary has not so far specified any additional factors.)
Volume Purchasing. States may arrange for “volume purchasing” of
laboratory services (other than those provided by hospitals or rural health clinics) or
medical devices, such as durable medical equipment or eyeglasses. One or more
providers of the specified service may be selected by the state, through competitive
bidding or other means, as the sole source of the items covered in an area or
statewide. Some states will permit other providers to furnish the item or service, but
only if they are prepared to meet the price of the approved source. The state must
ensure that services remain accessible to beneficiaries. If the arrangement is for
laboratory services, the laboratory must be state-licensed and/or meet other
requirements established by the Secretary, and no more than 75% of the laboratory’s
charges may be for Medicare and Medicaid beneficiaries.
Coordination with Medicare
While coverage under Medicare Part A (hospital insurance) is automatic for
persons meeting eligibility standards, coverage under Part B (supplemental medical
insurance) requires payment of a monthly premium by the beneficiary. Some persons
not automatically eligible for Part A coverage may also obtain that coverage by
paying a premium. In addition, Medicare beneficiaries are liable for cost-sharing:
deductible and coinsurance payments imposed for most Medicare-covered services.
State Medicaid programs are required to help defined populations of low-income
Medicare beneficiaries with Medicare-required premiums and sometimes
As Table 39 shows, there are five groups of Medicare beneficiaries eligible for
Medicaid assistance. The amount of help they receive generally decreases with rising
income. In addition to meeting income tests, beneficiaries must meet resource tests.
For dual eligibles, assets are limited to the SSI standard ($2,000 for an individual and79

$3,000 for a couple.); for the other groups, the limits are twice these amounts.
79 For the QMB amd SLMB groups, states can also use more liberal methodologies for
counting resources, meaning that they could disregard some amount of excess assets.

Table 39. Medicaid Benefits for
Low-Income Medicare Beneficiaries
Family incomeMedicaid pays
Full Medicaid beneficiariesVaries by stateAll Medicaid-covered
(dual eligibles)services; Medicaid secondary
to Medicare
Qualified MedicareAbove state MedicaidAll Medicare premiums and
beneficiaries (QMBs)eligibility level, no greatercost-sharing
than 100% of FPG
Specified low-incomeBetween 100% and 120% ofMedicare Part B premium
Medicare beneficiariesFPG
Qualified individuals (QIs)Between 120% and 135% ofMedicare Part B premium
provision expires NovemberFPG, but enrollment limited
20, 2004through fixed dollar funding
cap for each state
Qualified disabled andNo greater than 200% of FPGMedicare Part A premium
working individuals
Source: Title XIX of the Social Security Act.
Note: FPG = Federal poverty guideline.
Dual eligibles are elderly or disabled people who are eligible for Medicare and
who also qualify for full Medicaid benefits, such as SSI recipients, the medically
needy, and people meeting special eligibility standards for long-term care coverage.
States pay Medicare premiums and cost-sharing for these beneficiaries and cover
services included in the state plan and not covered by Medicare — most notably,80
long-term nursing facility care and outpatient prescription drugs. Note that, when
the new Medicare prescription drug benefit takes effect in 2006, state Medicaid
programs will no longer be permitted to cover drugs that are included in the Medicare
benefit. Instead, dual eligibles will receive the low-income subsidies provided for81
under the Medicare law.
Qualified Medicare beneficiaries (QMBs) do not qualify for full Medicaid
coverage, but have incomes no greater than 100% of the federal poverty guideline
(FPG): $8,980 for an individual and $12,120 for a couple in 2003. Medicaid
programs must pay both Part A and Part B premiums and required Medicare
cost-sharing for QMBs. Specified low-income Medicare beneficiaries (SLMBs) have

80 Technically, a state could choose not to pay premiums (“buy in”) for dual eligibles whose
incomes exceed 120% of poverty (chiefly persons in institutions who are spending down).
However, states are given a financial incentive to obtain Part B coverage for all Medicaid
beneficiaries who could qualify for it. If the state fails to buy in for a beneficiary and then
makes Medicaid payments for services that could have been covered by Medicare, it may
not claim federal matching for the resulting expenditures. (There is an exception for
services furnished prior to the date of the beneficiary’s Medicaid application and covered
as a result of a retroactive grant of Medicaid eligibility.)
81 States will make maintenance of effort payments to the federal government. These will
begin in 2006 at 90% of states’ estimated per capita drug spending (less dispensing fees and
rebates) for dual eligibles and will phase down to 75% in 2014 and later years.

incomes between 100% and 120% of the FPG. Medicaid pays only the Part B
premium and does not pay cost-sharing for SLMBs.
Qualifying individuals (QIs) have incomes between 120% and 135% of the
FPG. States pay the Part B premium for these individuals with 100% federal
funding, subject to a fixed-dollar annual cap for each state; this can mean that not all
eligible applicants will receive assistance. The provision for QIs expires November

20, 2004.

Finally, states are required to pay the Part A premium, but not the Part B
premium or cost-sharing, for qualified disabled and working individuals. These are
certain persons whose social security disability insurance benefits cease after they
return to work but who are permitted to continue to receive Medicare by paying the
Part A premium.
When states pay Medicare deductibles and coinsurance for dual eligibles and
QMBs, they have two options for computing the Medicaid payment amount. They
may choose to pay the full cost-sharing amounts determined by Medicare. Or they
may pay the difference, if any, between what Medicare paid and what the state would
have paid under its usual Medicaid rules for the same service. For example, suppose
the Medicare allowed fee for a given physician service is $100 and the state’s
Medicaid fee schedule allows $75 for the same service. Medicare pays 80%, or $80,
leaving $20 to be paid as coinsurance. The state may pay the full $20, or it may
determine that the practitioner has already been paid in full and pay nothing. (This
option for states was clarified by the BBA after a number of court rulings requiring
payment of full cost-sharing.)
Table 40 shows which states have chosen to pay full Medicare cost-sharing
(MR) and which pay up to the ordinary payment limits under the state plan (SP). As
the table indicates, some states pay differently for Part A and Part B cost-sharing,
while others have exceptions for specific services. Only major exceptions are noted
in the table.
Table 40. Medicaid Payment Policies for Medicare Cost-Sharing
StatePayment methodNotes
AlaskaMRHospital inpatient/outpatient SP
AlabamaSPPart A deductibles MR
Ar ka nsa s M R
ArizonaMRNon-AHCCCS services only
CaliforniaSPSkilled nursing facility MR
Co lo r a d o SP
ConnecticutMR (Part A)/SP (Part B)
District of ColumbiaSPPart A deductibles MR
Delawa re SP
Flo r id a SP
Geo r gia SP
Hawa ii MR
Iowa MR
IdahoMR (Part A)/SP (Part B)

StatePayment methodNotes
IllinoisSPSkilled nursing facility MR
Indiana SP
Kansas SP
K e nt uc ky M R
Lo ui s i a n a S P
MassachusettsSP (Part A), MR (Part B)
Maryland MR
Maine SP
Michigan SP
Minne so ta MR
MissouriMRHospital inpatient SP
M i ssissip p i SP
MontanaMR (Part A)/SP (Part B)
North CarolinaSP
North DakotaMRSkilled nursing facility SP
Neb r aska MR
New HampshireSP
New JerseySPHospital inpatient/outpatient MR
New MexicoMR
Neva d a SP
New YorkMR
Ohio MR
OklahomaPart A MR, Part B: 94% of
deduct and 75% of coins
Orego n SP
P e nnsyl va ni a S P
Rhode IslandSP
South CarolinaMR
South DakotaMR
TennesseeSPMost dual eligibles in TennCare
TexasMRHospital inpatient SP
UtahMRHospital inpatient/outpatient, other
specified services SP
Vir ginia SP
Vermo nt MR
W a shi ngt o n SP
W i sc o nsin S P
West VirginiaMR
Wyoming MR
Source: Medicaid state plans and amendments approved as of Nov. 7, 2002, except as follows:
Florida Medicaid program, Summary of Services 2002, [],
July 25, 2003. New Jersey Administrative Code, 10:49. North Carolina Medicaid Special Bulletin
VI, Sept. 2002, [], as of
July 2003.
Note: MR = pays full Medicare cost-sharing; SP = pays up to state plan limit.