Medicare: Better Information Can Help Ensure That Refinements to BBA Reforms Lead to Appropriate Payments

Published by the Government Accountability Office on 1999-10-01.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                          United States General Accounting Office

-GAO                      Testimony
                          Before the Subcommittee on Health, Committee on Ways
                          and Means, House of Representatives

For Release on Delivery

Friday, October 1, 1999

                          Better Information Can
                          Help Ensure That
                          Refinements to BBA
                          Reforms Lead to
                          Appropriate Payments
                          Statement of William J. Scanlon, Director
                          Health Financing and Public Health Issues
                          Health, Education, and Human Services Division

                               ^8 GAO Accountability * Integrity * Reliability

Mr. Chairman and Members of the Subcommittee:

I am pleased to be here today as you discuss the effects of the Balanced Budget Act of
1997 (BBA) on the Medicare program. BBA set into motion significant program changes
to both modernize Medicare and rein in spending. The act's constraints on providers'
fees, increases in beneficiary payments, and structural reforms together were projected to
lower Medicare spending by $386 billion over the next 10 years. Although some BBA
provisions are in effect, data relevant to their impact are generally limited to date; other
provisions have not yet been fully phased in. As a result, the act's full effects on
providers, beneficiaries, and taxpayers will remain unknown for some time.

BBA's Medicare provisions were enacted in response to rapid program spending growth
that was neither sustainable nor readily linked to demonstrated changes in beneficiary
needs. The act's payment reforms represented bold steps to control Medicare spending
by changing the financial incentives inherent in payment methods that, prior to BBA, did
not reward providers for delivering care efficiently. To date, the Congress has remained
steadfast in the face of intense pressure to roll back certain BBA payment reforms while
waiting for evidence that demonstrates the need for modifications. Calls for BBA
changes come at a time when federal budget surpluses and lower-than-expected growth in
Medicare outlays could make it easier to accommodate higher Medicare payments.
However, as the Comptroller General cautioned last week, the surpluses are merely
projections that could fall short of expectations, and the imperative remains to find the
reforms that will make Medicare sustainable and affordable for the longer term.'

My comments today focus on payment reforms affecting certain providers in Medicare's
traditional fee-for-service program and providers in Medicare's managed care program.
Specifically, I will discuss the effects on three providers of post-acute care services-
home health agencies (HHA), skilled nursing facilities (SNF), and providers of outpatient

IMedicare Reform: Ensuring Fiscal Sustainability While Modernizing the Program Will Be Challenging
(GAOIT-HEHS/AIMD-99-294, Sept. 22, 1999).

1                                                                          GAO/T-HEHS-00-14
rehabilitation therapy-and on the health plans participating in the Medicare+Choice

In brief, some providers of post-acute care and health plans in the Medicare+Choice
program may have to rethink their business strategies as a result of BBA payment
reforms, which seek to make Medicare a more efficient and prudent purchaser.
Imperfections in the design of BBA-mandated payment systems require attention, and
better information can help policymakers distinguish between desirable and undesirable
consequences. Based on such knowledge, refinements can help ensure that payments are
not only adequate in the aggregate but are also fairly targeted to protect individual
beneficiaries and providers. Our issued and ongoing studies of various payment methods
are instructive in this regard, and a summary of our results to date follows.

*    Home health care: Our work indicates that (1) the reductions in the number of HHAs
     and changes in utilization were consistent with the objectives of the interim payment
     system to control the rapid growth that had preceded BBA and (2) appropriate access
     to Medicare's home health benefit has not been impaired. However, the prospective
     payment system (PPS) is a more appropriate tool for the long term than the interim
     payment system, because it is intended to adjust payments for differences in
     beneficiary needs. As we examine the challenges of designing a PPS, we are finding
     that that the PPS will likely require further adjustments after it is implemented as
     more information on home health costs, utilization, and users becomes available.

 *   SNF care: A PPS was implemented beginning in July 1998 with a 3-year transition to
     fully prospective rates, giving providers time to adjust to the new system. Our
     ongoing work suggests that factors in addition to the PPS have contributed to fiscal
     difficulties for some corporations operating SNFs. Nevertheless, certain
     modifications to the PPS may be appropriate to ensure that payments are targeted to
     patients who require more costly care. The potential access problems that may result
     if Medicare underpays for high-cost cases could lead to beneficiaries' staying in acute

 2                                                                    GAO)r-HEHS-00O14
    care hospitals longer, rather than foregoing care altogether. HCFA is aware of this
    potential targeting problem and is working to develop a solution.

*   Caps on coverage of outpatient rehabilitation therapy: In 1999, BBA established an
    annual $1,500 per-beneficiary cap on payments for outpatient physical therapy and
    speech/language pathology services combined and a separate $1,500 cap on
    outpatient occupational therapy. The caps reflect a legitimate need to constrain
    service use. For the vast majority of outpatient therapy users, the caps are unlikely to
    curtail access to services. Only a small share of beneficiaries receiving therapy
    services are high users. Further, most outpatient therapy users will likely have access
    to hospital outpatient departments, which are not subject to the $1,500 caps. In
    addition, owing to HCFA's partial approach to enforcing the caps,
    noninstitutionalized beneficiaries can avoid having the caps curtail service coverage
    by switching providers. Whether the caps restrict coverage for a small share of
    nursing home residents is less straightforward. A need-based payment system could
    help better target payments toward beneficiaries who genuinely require more services
    than allowed under the current dollar limits.

*   Payments to Medicare+Choice health plans: Several BBA provisions address the
    long-recognized problem of excess payments to Medicare+Choice plans. Some
    provisions have begun to be phased in, such as reducing the annual rate updates;
    others have not yet become effective, such as the use of a risk adjustment method
    based on beneficiary health status. The net effect of the implemented revisions has
    been modest and, on average, has likely removed only a portion of excess payments
    built into the base rates. Moreover, the recent and upcoming rounds of plan
    withdrawals from Medicare are not, as the industry has argued, fully attributable to
    Medicare's lowered payment rates. The evidence emerging from recent rounds of
    withdrawals suggests that market share, enrollment size, and competition from other
    health plans factor into a plan's decision to participate in Medicare. Critical to
    making Medicare+Choice payment modifications are the establishment of an
    appropriate base rate and of a risk adjustment method that pays more for serving

3                                                                    GAO/r-HEHS-00-14
      beneficiaries with serious health problems and less for serving relatively healthy


The Medicare program consists of two parts: "hospital insurance," or part A, which
covers inpatient hospital, skilled nursing facility, hospice, and certain home health care
services; and "supplementary medical insurance," or part B, which covers physician and
outpatient hospital services, outpatient rehabilitation services, home health services under
certain conditions, diagnostic tests, and ambulance and other health services and supplies.

Growth in Medicare Spending
for Home Health Care

During much of the 1990s, home health care was one of Medicare's fastest growing
benefits; between 1990 and 1997, Medicare spending for home health care rose at an
annual rate of 25.2 percent. Several factors accounted for this spending growth, most
notably the relaxation of coverage guidelines. In response to a 1988 court case, a change
in the coverage guidelines essentially transformed the benefit from one that focused on
patients needing short-term care after hospitalization to one that also serves chronic,
long-term-care patients. 2 The loosening of coverage and eligibility criteria contributed
to an increase in the number of beneficiaries receiving services and the volume of
services they received. Associated with this rise in utilization was an almost doubling in
the number of Medicare-certified HHAs to 10,524 by 1997.

Also contributing to the historical rise in home health care spending were a payment
system that provided few incentives to control how many visits beneficiaries received and
lax Medicare oversight of claims. As we noted in a previous report, even when
controlling for diagnoses, substantial geographic variation existed in the provision of
home health care, with little evidence that the differences were warranted by patient care

2   Duggan v. Bowen, 691 F. Supp. 1487 (D.D.C. 1988).

4                                                                      GAO/T-HEHS-00-14
needs. 3 Additional evidence indicates that at least some of the high use and the large
variation in practice represented inappropriate billings and unnecessary care.4 Medicare
oversight declined at the same time that spending mounted, contributing to the likelihood
that inappropriate claims would be paid. To begin to control spending, BBA
implemented an interim payment system for HHAs beginning October 1, 1997. A PPS is
scheduled to be implemented for all HHAs on October 1, 2000. 5

Growth in Medicare Spending
For SNF Care

As required by BBA, on July 1, 1998, SNFs began a 3-year transition to a PPS, under
 which providers are paid a prospective rate for each day of care. Previously, SNFs were
 paid the reasonable costs they incurred in providing Medicare-covered services.
 Although there were limits on the payments for the routine portion of care (that is,
 general nursing, room and board, and administrative overhead), payments for ancillary
 services, such as rehabilitative therapy, were virtually unlimited. Because higher
 ancillary service costs triggered higher payments, facilities had no incentive to provide
 these services efficiently or only when necessary. Thus, between 1992 and 1995, daily
 ancillary costs grew 18.5 percent a year, compared to 6.4 percent for routine service
 costs. Moreover, new providers were exempt from the caps on routine care payments for
 up to their first 4 years of operation, which encouraged greater participation in Medicare.

 3 Medicare: Home Health Utilization Expands While Program Controls Deteriorate (GAO/HEHS-96-16,
 Mar. 27, 1996).
 4Medicare: Improper Activities by Mid-Delta Home Health (GAO/T-OSI-98-6) and Office of the Inspector
 General, Department of Health and Human Services, Variation Among Home Health Agencies in Medicare
 Payment for Home Health Services (July 1995). Our 1997 analysis of a small sample of high-dollar claims
 found that over 40 percent of these claims should not have been paid by the program. See Medicare: Need
 to Hold Home Health Agencies More Accountable for Inappropriate Billings (GAO/HEHS-97-108, June
  13, 1997).
 5BBA required the HHA PPS to be in place on October 1, 1999. Subsequent legislation delayed the
 implementation by 1 year, eliminating any transition period.

  5                                                                          GAO/T-HEHS-00-14
Growth in Medicare Spending for Outpatient
Rehabilitation Therapy Services

Rehabilitation therapy comprises a substantial portion of the post-acute-care services
provided by SNFs and other providers, such as rehabilitation therapy agencies and
comprehensive outpatient rehabilitation facilities. Between 1990 and 1996, payments for
outpatient rehabilitation therapy alone rose at an average rate of 18 percent a year,
compared to 9.7 percent average growth rate for the same period for overall Medicare
spending. BBA reforms were designed to control both the price and volume of therapy
services provided in outpatient settings-the former by a fee schedule and the latter by
per-beneficiary coverage caps.6 Specifically, BBA limits coverage for outpatient therapy
to $1,500 per beneficiary for physical therapy and speech/language pathology services,
with a separate $1,500 per-beneficiary limit for occupational therapy. Hospital outpatient
departments are exempt from these coverage limits.

Historical Overpayvments to
Medicare Health Plans

BBA sought to moderate Medicare's payments to managed care plans because
beneficiaries who joined Medicare managed care cost-not saved-the government
money. That is, the government was paying more to cover beneficiaries in managed
care-an estimated several billion dollars more-than it would have if these individuals
had remained in the traditional fee-for-service program. Medicare payments to managed
care plans have been estimated to be too high by as much as 16 percent.7 Beginning in
1998, BBA made several changes to the method used to set Medicare+Choice plan
payments, not all of which will reduce excess payments. Among other things, BBA

6Payments    for inpatient rehabilitation therapy services, such as those provided by SNFs, HHAs, and
rehabilitation facilities, are not subject to the fee schedule and are paid under other rules. In addition,
outpatient therapy provided by critical access hospitals is not subject to the fee schedule.
7In a 1996 study, HCFA estimated that payments were too high by 8 percent in 1994. [See Gerald Riley
and others, "Health Status of Medicare Enrollees in HMOs and the Fee-for-Service Sector in 1994. Health
Care FinancingReview, vol. 17, no 4 (Summer 1996)]. In a 1997 study, we estimated that aggregate
payments to California plans were too high by 16 percent. [See Medicare HMOs: HCFA Can Promptly
Eliminate Hundreds of Millions in Excess Payments (GAO/HEHS-97-16, Apr. 25, 1997)].

6                                                                               GAO/T-HEHS-00-14
required a new risk adjustment method-a mechanism for adjusting payment rates on the
basis of a beneficiary's expected annual health care costs. It will be implemented in two
stages. Beginning in 2000, HCFA plans to phase in an interim method based on inpatient
hospital data; in 2004 it plans to implement a more comprehensive method incorporating
additional medical data from other settings. The interim risk adjustment, if fully phased
in, would reduce payments by 7 percent. BBA also reduced updates to health plan
payment rates for a 5-year period ending 2002, for a cumulative rate reduction of less
than 3 percent. However, the effect of these reductions is substantially moderated
because BBA used 1997 payment rates as the foundation for rates in 1998 and future
years. According to HCFA actuaries, a forecast error caused the 1997 rates to be an
estimated 4.2 percent too high and, consequently, aggregate plan payments in 1998 were
$1.3 billion too high. The excess payments resulting from this forecast error will increase
over time with managed care enrollment because it is built into the base rate.8


BBA's new payment policies addressing rapid spending growth for home health care
included the establishment of an interim payment system, which is currently in effect,
 and a requirement to replace that system with a PPS by October 2000. Our published and
 ongoing studies discuss the effects of these BBA payment reforms and concerns about
 their design and implementation.

 Concerns have been raised about the effect of the interim system, but, as we reported in
 May 1999, there was little evidence that appropriate access to Medicare's home health
 benefit has been impaired. 9 The pre-BBA payment system had controls for payments per
 visit but left volume unchecked. Since enactment of BBA, home health agencies have
 been paid under the interim payment system, which attempts to control the costs and total

 8 BBA   did not allow HCFA to adjust the 1997 rates for forecast errors, although such adjustments had been
 a critical component of the pre-BBA rate-setting process. BBA permits HCFA to correct forecasts in future
 years but did not include a provision to allow a correction of its 1997 forecast.

 7                                                                              GAO/T-HEHS-00-14
volume of services. Indeed, our work indicates that overall home health utilization in the
first 3 months of 1998 was below that in 1996 when Medicare spending for home health
services nearly peaked. Moreover, the sizeable variation in utilization across counties has
narrowed, a change consistent with the incentives of the interim payment system.
Although these changes occurred at the time that about 14 percent of HHAs closed their
doors to Medicare business, we found little evidence that beneficiary access to services
was inappropriately curtailed.

Nevertheless, a home health PPS is a more appropriate payment tool because it can align
payments with patient needs. Under PPS, payments will reflect the needs of the
agencies' current beneficiaries rather than historical spending patterns. However, our
ongoing work on this subject shows that a number of design issues remain, and the
payment system will likely require continued adjustments even after implementation next
year. It appears that HCFA intends to pay HHAs a per-episode rate for each 60-day
period during which a patient receives services. Such per-episode payments are designed
to balance competing goals of controlling service provision while giving HHAs
flexibility to vary the intensity or mix of services delivered during the episode. Evidence
indicates that HHAs do lower their costs in response to prospective payments for an
episode of care. Whether they will inappropriately cut care remains to be seen. Under
this prospective payment approach, HHAs also have incentives to increase the number of
episodes of care provided, which could escalate, rather than constrain, Medicare
spending. HCFA will need to adequately monitor service provision to ensure that
beneficiaries receive the care they need and the number of episodes are not
inappropriately increased.

The design of the case-mix adjustment mechanism is critical to adequately pay for
patients with high service needs, yet not overpay for others with lower needs. Designing
this mechanism requires detailed information about services and beneficiary
characteristics, and such information is currently available only for a sample of users.

 Medicare Home Health Agencies: Closures Continue With Little Evidence Beneficiary Access Is Impaired
(GAOIHEHS-99-120, May 26, 1999).

8                                                                          GAO/T-HEHS-00-14
Furthermore, the wide geographic and agency-level variation in service use indicates that
standards of care are not well-defined, nor are the criteria for who should use the benefit.
As a result, the factors that will be used under PPS for grouping patients with similar
resource needs may not adequately distinguish among types of home health patients, and
the PPS payment adjuster that will be associated with each patient group may not reflect
appropriate cost differences. Systematic errors could result in overpayments for some
beneficiaries and underpayments for others. Underpayments could lead to impaired

Large variations in historic spending patterns mean that a PPS, which will be based on
average payment amounts, will undoubtedly cause payment levels to rise for certain
HHAs and fall for others. Although the PPS may incorporate an outlier policy-that is,
extra payments for extremely costly cases-additional mechanisms to moderate payment
changes may be appropriate. For example, an "inlier" policy to reduce the payment for a
patient who receives few services may be warranted, particularly given the fact that
multiple episode payments may be made for a single beneficiary. Policies addressing
both extremes of service use could protect the access of beneficiaries with high needs and
protect Medicare from overpaying for low-cost cases. A risk-sharing method, to account
 for cost differences across agencies, could provide further protection against
 underpayments or overpayments. Given the heterogeneous use of this benefit and the
 unresolved PPS design issues, moderating payments through risk-sharing might be
 warranted, even though such a mechanism would weaken HHAs' incentives to provide
 care more efficiently.


 Despite industry charges to the contrary, SNF payment rates under BBA are likely to
 provide sufficient, or even generous, compensation for providers. Nevertheless, the
 distribution of these payments may be out of balance, because the current case-mix
 adjustment method may not adequately ensure that providers serving high-cost

 9                                                                    GAO/T-HEHS-00-14
beneficiaries are paid enough and that those serving low-cost beneficiaries are not paid
too much.

Under the new PPS, SNFs receive a payment for each day of covered care provided to a
Medicare-eligible beneficiary. By establishing fixed payments and including all services
provided to beneficiaries under the per diem amount, the PPS attempts to provide
incentives for SNFs to deliver care more efficiently. Under the PPS, SNFs that
previously boosted their Medicare ancillary payments-either through higher use rates or
higher costs-will need to modify their practices more than others. Scaling back the use
of these services, however, may not necessarily affect the quality of care. There is little
evidence to indicate that the rapid growth in Medicare spending was due to a
commensurate increase in Medicare beneficiaries' need for services.

Recent industry reports have questioned the ability of some organizations that operate
SNF chains to adapt to the new PPS. Indeed, Medicare payment changes have been
blamed for one corporation's filing for protection under bankruptcy law and the potential
for another to similarly file. However, our ongoing work suggests that the PPS should
not have an untoward impact on most SNFs and is only one of many factors contributing
to the poor financial performance of these corporations. For most SNFs Medicare
patients constitute a relatively small share of their business. In addition, the PPS rates are
being phased in, to allow time for facilities to adapt to the new payment system, and most
of the payments are still tied to each facility's historical costs. However, heavy
investments in the nursing home and ancillary service businesses in the years
immediately before the enactment of BBA, both to expand their acquisitions and upgrade
facilities to provide higher-intensity services, has created difficulties for some
corporations. Now under tighter payment constraints for both their SNF and ancillary
 service operations, these debt-laden enterprises will not be able to rely on overly
 generous Medicare payments. Thus, while PPS does represent a constraint on Medicare
revenue and SNFs will have to adapt, the performance of some large post-acute providers
is a reflection of many Medicare payment policy changes and strategic decisions made
during a period when Medicare was exercising too little control over its payments. We

 10                                                                   GAO/T-HEHS-00-14
are gathering additional information and will report soon on the effect of the PPS on SNF
solvency and beneficiary access to care.

We believe that overall payments to SNFs are adequate. In fact, we and the Department
of Health and Human Services Inspector General (HHS IG) are concerned that the PPS
rates Medicare pays may be too generous. Most of the data used to establish these
rates-from 1995 cost reports-have not been audited and are likely to include excessive
ancillary costs due to the previous system's incentives and the lack of appropriate
program oversights

We are also concerned that payments for individual beneficiaries could be
inappropriately too high or low because of certain PPS design problems. The first of
these involves the patient classification system. The classification system was based on a
small sample of patients and, because of the age of the data, may not reflect current
treatment patterns. As a result, it may aggregate patients with widely differing needs into
too few payment groups that do not distinguish adequately among patients' resource
needs. In addition, the variation in non-therapy ancillary services costs does not appear
to have been adequately accounted for in the payment rates, which may inappropriately
compress the range in payments. Accordingly, access problems or inadequate care could
result for some high-cost beneficiaries. Hospitals have reported an increase in placement
problems due to the reluctance of some facilities to admit certain beneficiaries with high
expected treatment costs, which will increase hospital lengths of stay for these patients.
HCFA is aware of the limitations of the patient classification system and is working to
refine the system to more accurately reflect patient differences.

 Another concern is that the current patient classification system preserves the opportunity
 for SNFs to increase their compensation by supplying unnecessary services. A SNF can
 benefit by manipulating the services provided to beneficiaries, rather than increasing
 efficiency. For example, by providing certain patients an extra minute of therapy over a

 11                                                                  GAO/r-HEHS-00-14
defined threshold, a facility could substantially increase its Medicare payments without a
commensurate increase in its costs.


Questions have been raised about a BBA coverage restriction for a third group of post-
acute-care services-outpatient rehabilitation therapy. Together with a fee schedule that
replaces reasonable cost reimbursement for these services, BBA established an annual
$1,500 per-beneficiary cap on payments for outpatient physical therapy and
speech/language pathology services combined and a separate $1,500 per-beneficiary cap
on outpatient occupational therapy." Services provided by hospital outpatient
departments are exempt from the per-beneficiary caps.

Rehabilitation therapy providers have raised concerns that the $1,500 limits will
arbitrarily curtail necessary treatments for Medicare beneficiaries, particularly victims of
stroke, hip injuries, or multiple medical incidents within a single year. These concerns
have led to several legislative proposals to include various exceptions to the caps or
eliminate them altogether.

 Our ongoing work on this topic suggests that eliminating the caps without substituting
 other controls could undermine BBA's comprehensive strategy for restricting payments
 for outpatient therapy services. Controlling the price for each unit of service-as is done
 with the new requirement that outpatient therapy providers be paid using Medicare's
 physician fee schedule-may not necessarily control Medicare expenditures if utilization

 '1The HH-IS IG recently reported on the inappropriateness of the base year costs. See Physical And
 Occupational Therapy in Nursing Homes: Cost of Improper Billings to Medicare (HHS IG, OEI-09-97-
 00122, Aug. 1999).
 11Physical therapy includes treatments-such as whirlpool baths, ultrasound, and therapeutic exercises-to
 relieve pain, improve mobility, maintain cardiopulmonary functioning, and limit the disability from an
 injury or disease. Speech/language pathology services include the diagnosis and treatment of
 communication, swallowing, oral motor and related cognitive functions and their disorders. Occupational
 therapy helps patients learn the skills necessary to perform daily tasks, diminish or correct pathology, and
 promote health.

 12                                                                              GAO/T-HEHS-00-14
rises. This is particularly likely, given the price and utilization controls established
through PPSs on other providers of rehabilitation therapy. Thus, the per-beneficiary caps
serve to limit the volume of services provided.

For the vast majority of beneficiaries, the coverage caps are unlikely to curtail access to
needed services. An analysis by the Medicare Payment Advisory Commission shows
that, in 1996, most users (86 percent) did not exceed $1,500 in payments for physical
therapy and speech/language pathology services or for occupational therapy.' 2 Moreover,
as the fee schedule likely reduces payments for many providers, the proportion of
beneficiaries that are unaffected by the caps could be even higher in 1999 because
beneficiaries could receive more services before reaching the per-beneficiary caps than
under the former cost-based system.

Even for beneficiaries exceeding $1,500 in payments under the fee schedule, mitigating
factors exist. First, under the BBA exemption, Medicare beneficiaries have no limits on
coverage for rehabilitation therapy provided by hospital outpatient departments, which
are widely available nationwide. In addition, the caps will initially not be applied as
specified in BBA. Implementing the caps involves many programming changes to
Medicare's automated information systems that HCFA is unable to undertake concurrent
with its year 2000 preparation efforts. As a result, HCFA's claims processing contractors
will be unable to track therapy payments on a per-beneficiary basis. Instead, effective
January 1, 1999, HCFA employed a transitional approach to implementing the caps.
Under this approach, each provider of therapy services is responsible for tracking its
billings for each Medicare patient and stopping them at the $1,500 threshold. The
consequence of this partial implementation is that noninstitutionalized beneficiaries may
 switch to a new provider when they have reached the $1,500 limit under their current

12A July 1998 report sponsored by the National Association for the Support of Long-Term Care and
NovaCare, a rehabilitation services company, projects that 87 percent of beneficiaries will not exceed the
per-beneficiary cap.

 13                                                                             GAO/T-HEHS-00-14
The effect of the per-beneficiary caps on nursing home residents is less clear. HCFA's
policy explicitly states that the hospital outpatient department exemption does not apply
to those therapy services furnished to nursing facility residents. Moreover, the ability of
beneficiaries to switch outpatient providers under HCFA's partial implementation
approach is, practically speaking, not available to nursing facility residents. Under new
billing requirements, the nursing facility in which the beneficiary resides is required to
bill for outpatient therapy provided to the resident, regardless of the entity that actually
delivered the service. Therefore, unlike their noninstitutionalized counterparts, nursing
facility residents cannot switch providers to restart the $1,500 coverage allowance.
Under these circumstances, some nursing home residents-like those needing extensive
rehabilitation therapy resulting from such conditions as stroke or hip fractures-could be
vulnerable to out-of-pocket costs for therapy.

Even the risk for these more vulnerable beneficiaries may be moderated, however,
because nursing home residents seeking therapy for such conditions would likely receive
a complement of rehabilitation services as a SNF inpatient-before the outpatient therapy
coverage limit begins to apply. For example, individuals suffering a stroke or undergoing
hip replacement would likely spend at least 3 days in an acute care hospital, which,
combined with the need for daily skilled nursing care or therapy, would make them
eligible for a Medicare-covered SNF stay of up to 100 days, during which they would
likely receive therapy services. After their Medicare coverage period ends, nursing
facility residents can continue to receive outpatient therapy services under Medicare part
B, subject to the coverage limits. BBA mandates that HCFA develop a classification
 system based on diagnosis to determine differences in patients' therapy needs and
 propose possible alternatives to the caps in a report due January 1, 2001. This report will
 be significant in that a need-based system could help ensure adequate coverage for those
 beneficiaries requiring an extraordinary level of services and prevent overprovision to
 those requiring only limited amounts.

 14                                                                    GAOIT-HEHS-00-14

Developing appropriate refinements to BBA reforms affecting Medicare+Choice requires
consideration of several aspects of Medicare's managed care program. At the moment,
plan withdrawals from Medicare+Choice in 1999, and recent announcements that
additional plans will withdraw in 2000, have prompted debate about whether BBA
reforms have resulted in inadequate payment rates. At the same time, our published and
ongoing work indicates that Medicare managed care payments to health plans likely
continue to exceed the cost of providing Medicare-covered benefits

Our analysis of the 1999 withdrawals showed that payment rates alone could not explain
plans' participation decisions. Withdrawals were not limited to low payment rate
counties. The data suggested that local market conditions affected plans' participation in
a county. A plan was more likely to withdraw from counties it had recently entered,
where its enrollment was low, or where its market share was small relative to other plans
serving the same county. Although our final analysis will not be available for a few
weeks, our preliminary assessment suggests that similar factors help explain the pattern
of the 2000 withdrawals.

Plan withdrawals may well reflect a normal market correction spurred by a changing
business environment. Prior to BBA, health plans could expand into new areas with
relatively little risk because overgenerous Medicare rates provided protection from the ill
consequences of small enrollment or large competitors. Between 1993 and 1998, the
number of plans and enrollees tripled. However, as BBA slowed payment growth, health
plans may have reevaluated their expansion decisions, making such factors as potential
enrollment, market share, and competition a key part of plans' decisions to withdraw
from certain geographic areas.

15                                                                  GAOir-HEHS-00-14
A local example illustrates the importance of nonpayment factors in plans' participation
decisions. In 1996, Blue Cross' Free State health plan in Maryland-which until that
time had served only some of the state's large urban counties--extended service
statewide. Free State recently announced, however, that beginning in 2000 it would
substantially reduce its geographic service area. The plan is withdrawing from 17 rural
and small urban counties even though BBA will increase the average base rate in those
counties by nearly 6 percent next year. In contrast, the large urban counties will receive
only a 2.4 percent average rate increase, but Free State will continue its Medicare
participation in these counties.

According to industry representatives, it is difficult for health plans to serve counties with
few providers and enrollees because providers have little incentive to discount their fees
and plans cannot spread risk over a large enrollment base. Although we cannot know
with certainty, these factors may have influenced Free State's decision to discontinue
service in Caroline County and 16 other rural and small urban Maryland counties. For
example, in Caroline County Free State faced no competitors, had enrolled 19 percent of
the beneficiaries, and would have received a 7.5 percent Medicare rate increase in 2000.
However, less than 4,700 beneficiaries live in the county and Free State's 19 percent
market share represented an enrollment of less than 900 beneficiaries. In contrast, the
plan will continue to serve seven counties where the number of beneficiaries ranges from
about 15,200 to 116,600.

In addition to our work on plan withdrawals, our assessment of BBA payment changes
indicates that, relative to the cost of providing the package of traditional Medicare
benefits, payments to health plans remain excessive. For one thing, plans annually
receive a billion-plus-dollar overpayment in aggregate as a consequence of BBA's terms
for setting the base payment rate. This problem, owing to an uncorrected forecast error,
will be built into future base rates because BBA has not provided explicit authority for
HCFA to correct the forecast error. In our June 1999 report, we suggested that the
Congress consider certain modifications to Medicare's base payment rates to health plans

16                                                                    GAOrT-HEHS-00-14
to eliminate, among other things, the excess payments resulting from the 1997
uncorrected forecast error.

Moreover, payments continue to exceed plans' costs of providing Medicare-covered
services. In 1999, the average plan was required to provide $54 in extra benefits per
member per month so that projected Medicare payments would not exceed the plan's
projected costs and normal profits. In addition, the average plan voluntarily provided
another $54 in benefits per member per month. The additional benefits can be reflected
not only in coverage for services, but also in reduced beneficiary cost sharing. For
example, in 1999 most plans did not charge a monthly premium and charged only a small
copayment for outpatient services.

In 2000, enrollment in a Medicare+Choice plan will remain a relatively inexpensive way
for a beneficiary to obtain prescription drug coverage in many areas. On average, plans
will charge beneficiaries $16 per month in premiums and most will offer prescription
drug coverage. Beneficiaries will be charged a copay for prescription drugs that will
average about $17 for brand name drugs and $7 for generic drugs. In contrast, the
average monthly premium for private supplemental insurance policies (Medigap)
offering, among other things, prescription drug coverage ranges from $136 to $194 per
month in 1999. Moreover, those Medigap policies require a $250 deductible with a 50-
percent copayment.

 Given that Medicare has spent more for the generally healthier beneficiaries enrolled in
 Medicare+Choice plans than for the generally sicker beneficiaries in traditional
 Medicare, the need to have payments better reflect beneficiaries' expected health care
 costs is critical. HCFA's new risk adjustment method, based on certain health status
 measures, is scheduled for phased implementation in 2000 and represents a major
 improvement over the current method. For the first time, Medicare managed care plans
 can expect to be paid more for serving beneficiaries with serious health problems and less
 for serving relatively healthy ones. The method scheduled for implementation in 2004
 will be an improvement over the method used in 2000 because it is intended to include

 17                                                                 GAO/T-HEHS-00-14
better health status measures derived from more comprehensive data not currently

HCFA's plan to phase in the 2000 risk adjustment method slowly is designed to balance
the needs of taxpayers and beneficiaries. In 2000, only 10 percent of health plans'
payments will be adjusted using the new method. This proportion will be increased each
year until 2003, when 80 percent of plans' payments will be adjusted using the interim
system. Although a gradual phase-in of the interim risk adjuster delays the full realization
of Medicare savings, it also minimizes potential disruptions for both health plans and
beneficiaries. In 2004, HCFA intends to implement a more finely tuned risk adjuster that
uses medical data from physician offices, outpatient departments, and other health care
settings and providers-in addition to the inpatient hospital data on which the interim
adjuster is based. This more comprehensive risk adjustment system cannot be
implemented currently because many plans say they do not have the capability to report
such comprehensive information.


In conclusion, BBA payment reforms seek to curb unnecessary Medicare spending. As
the reforms begin to have their intended effects, pressure is building to return to more
generous payment policies. Evidence to date shows that BBA is moving Medicare in the
right direction but that adjustments will be needed along the way. These adjustments
should be based on thorough, quantitative assessments so that misdiagnosed problems do
not lead to misguided solutions. With the health care of seniors and the tax dollars of all
Americans at stake, it will be prudent to uphold new payment policies that exact .
efficiencies but make adaptations when substantiated evidence supports the need to do so.

Mr. Chairman, this concludes my prepared statement. I will be happy to answer any
questions you or other Members of the Subcommittee might have.

18                                                                  GAO/T-HEHS-00-14

For future contacts regarding this testimony, please contact William J. Scanlon at (202)
512-7114. Individuals making key contributions to this testimony included James C.
Cosgrove, Hannah F. Fein, and Deborah Spielberg.


 19                                                                 GAO/T-HEHS-00-14