Balanced Budget Act: Implementation of Key Medicare Mandates Must Evolve
to Fulfill Congressional Objectives (Testimony, 07/16/98,
GAO/T-HEHS-98-214).

GAO discussed the Health Care Financing Administration's (HCFA)
implementation of Medicare provisions contained in the Balanced Budget
Act of 1997 (BBA), focusing on: (1) an overview of how HCFA's
implementation has progressed since GAO's earlier testimony; (2) the
efforts to inform Medicare beneficiaries about the expanded health plan
choices available to them in 1999, commonly referred to as the
information campaign; and (3) the prospective payment system (PPS) for
skilled nursing facilities (SNF), which began a 3-year phase-in in June
1998.

GAO noted that: (1) HCFA is making progress in meeting the legislatively
established implementation schedules; (2) since the passage of BBA in
August 1997, almost three-fourths of the mandates with a July 1998
deadline have been implemented; (3) however, HCFA officials have
acknowledged that many remaining BBA mandates will not be implemented on
time; (4) HCFA maintains that these delays will have a minimal impact on
anticipated Medicare program savings; (5) given the concurrent
competition for limited resources and the differing importance and
complexity of the many BBA mandates, the success or failure of HCFA's
implementation efforts should not be judged solely on meeting deadlines;
(6) rather, any assessment should consider whether the agency is meeting
congressional objectives while taking a reasoned management approach to
identifying critical BBA tasks, keeping them on track, and integrating
them with other agency priorities; (7) complying with the BBA mandate to
conduct an information campaign that provides beneficiaries with the
tools to make informed health plan choices poses significant challenges
for HCFA and participating health plans; (8) in implementing the
Medicare plus Choice program, HCFA must now assemble the necessary
comparative information about these options and find an effective means
to disseminate it to beneficiaries; (9) a parallel goal of the
information campaign is to give beneficiaries information about the
quality and performance of participating health plans to promote
quality-based competition among plans; (10) HCFA has accelerated its
goals for obtaining standardized information from plans, and GAO
believes health plan disenrollment rates provide an acceptable
short-term substitute measure of plan performance; (11) the campaign is
to be financed primarily from user fees; (12) HCFA has met the July 1,
1998, implementation date for phasing in a new payment system for SNFs;
(13) GAO is concerned, however, that payment system design flaws and
inadequate underlying data used to establish payment rates may
compromise the system's ability to meet the twin objectives of slowing
sending growth while promoting the delivery of appropriate beneficiary
care; (14) in the short term, the new payment system could be improved
if HCFA clearly stated that SNFs are responsible for insuring that the
claims they submit are for beneficiaries who meet Medicare coverage
criteria; and (15) in the longer term, further research to improve the
patient grouping methodology and new methods to monitor the accuracy of
patient assessments could substantially improve the performance of the
new payment system.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  T-HEHS-98-214
     TITLE:  Balanced Budget Act: Implementation of Key Medicare 
             Mandates Must Evolve to Fulfill Congressional Objectives
      DATE:  07/16/98
   SUBJECT:  Skilled nursing facilities
             Health insurance
             Consumer education
             Managed health care
             Claims processing
             Medical information systems
IDENTIFIER:  Medicare Choice Program
             Medicare Program
             HCFA Prospective Payment System
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


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

For Release on Delivery
Expected at 11:00 a.m.
Thursday July 16, 1998

BALANCED BUDGET ACT -
IMPLEMENTATION OF KEY MEDICARE
MANDATES MUST EVOLVE TO FULFILL
CONGRESSIONAL OBJECTIVES

Statement of William J.  Scanlon, Director
Health Financing and Systems Issues
Health, Education, and Human Services Division

GAO/T-HEHS-98-214

GAO/HEHS-98-214T


(101757)


Abbreviations
=============================================================== ABBREV

  BBA - Balanced Budget Act of 1997
  CBO - Congressional Budget Office
  HCFA - Health Care Financing Administration
  HEDIS - Healthplan Employer Data and Information Set
  HMO - health maintenance organization
  OBRA87 - Omnibus Budget Reconciliation Act of 1987
  PPS - prospective payment system
  SNF - skilled nursing facility

BALANCED BUDGET ACT: 
IMPLEMENTATION OF KEY MEDICARE
MANDATES MUST EVOLVE TO FULFILL
CONGRESSIONAL OBJECTIVES
============================================================ Chapter 0

Mr.  Chairman and Members of the Subcommittee: 

I am pleased to be here today to discuss the Health Care Financing
Administration's (HCFA) implementation of Medicare provisions
contained in the Balanced Budget Act of 1997 (BBA).\1 Taken together,
the more than 200 BBA Medicare mandates amount to what are probably
the most significant modifications to the program since its inception
30 years ago.  For example, the new Medicare+Choice provisions of BBA
will enable beneficiaries to enroll in different types of health
plans previously excluded from the Medicare program, while the
introduction of prospective payment systems will alter how
reimbursements are made to skilled nursing facilities (SNF), home
health agencies, hospital outpatient departments, and inpatient
rehabilitation facilities.  Collectively, the objective behind these
changes is to better control the growth in Medicare expenditures
while simultaneously moving the program away from its fee-for-service
orientation and toward greater acceptance of the different types of
managed care already available to those with private health
insurance.  You asked me to give an overview of how HCFA's
implementation has progressed since our testimony earlier this
year.\2 In addition to this overview, my testimony will provide more
detailed comments on two key program elements scheduled for
implementation this year that have been the subject of extensive GAO
work:  (1) the efforts to inform Medicare beneficiaries about the
expanded health plan choices available to them in 1999, commonly
referred to as the "information campaign," and (2) the prospective
payment system (PPS) for SNFs, which began a 3-year phase-in this
month. 

To prepare this testimony, we analyzed HCFA reports that track the
implementation of BBA mandates and discussed their status with HCFA
officials.  We also drew on our previous as well as ongoing work
assessing the information HCFA and health plans provide to
beneficiaries; HCFA's responsibilities under BBA for a new, annual
information campaign; and the financing for that campaign.  Finally,
we analyzed HCFA's interim final rule dated May 12, 1998, that
describes the new PPS and consolidated billing for SNFs.  Our
analysis relied on (1) discussions with HCFA officials and Medicare
contractor staff; (2) the lessons learned from implementing the PPS
for inpatient hospital services, which has been in place since the
mid-1980s; and (3) our prior work on SNF services. 

In summary, HCFA is making progress in meeting the legislatively
established implementation schedules.  Since the passage of BBA in
August 1997, almost three-fourths of the mandates with a July 1998
deadline have been implemented.  HCFA's recent publication of the
Medicare+Choice and SNF PPS implementing regulations demonstrates
that progress.  However, HCFA officials have acknowledged that many
remaining BBA mandates will not be implemented on time.  In
particular, they point out that achieving compliance with Year-2000
computer requirements, a critical agency priority, competes with BBA
mandates for computer system resources and, as a result,
implementation of a number of BBA mandates will be delayed.  HCFA
maintains that these delays will have a "minimal" impact on
anticipated Medicare program savings. 

Given the concurrent competition for limited resources and the
differing importance and complexity of the many BBA mandates, the
success or failure of HCFA's implementation efforts should not be
judged solely on meeting deadlines.  Rather, any assessment should
consider whether the agency is meeting congressional objectives while
taking a reasoned management approach to identifying critical BBA
tasks, keeping them on track, and integrating them with other agency
priorities.  Continued involvement by high-level agency officials in
this process as well as ongoing legislative oversight should enhance
the likelihood of success. 

Complying with the BBA mandate to conduct an information campaign
that provides beneficiaries with the tools to make informed health
plan choices poses significant challenges for HCFA and participating
health plans.  In the past, HCFA played almost no role in helping
beneficiaries to evaluate their health plan options--that is, in
deciding whether to remain in fee-for-service Medicare or switch to
participating HMOs.  In implementing the Medicare+Choice program,
HCFA must now assemble the necessary comparative information about
these options and find an effective means to disseminate it to
beneficiaries.  A parallel goal of the information campaign is to
give beneficiaries information about the quality and performance of
participating health plans to promote quality-based competition among
plans.  The lack of standardized information from health plans about
their benefits and the imperfect state of quality and satisfaction
measures have made HCFA's efforts to assemble this information more
difficult.  HCFA has accelerated its goals for obtaining standardized
information from plans, and we believe health plan disenrollment
rates provide an acceptable short-term substitute measure of plan
performance.  HCFA's cautious approach to implementing the
information campaign is probably warranted, given its inexperience in
such an endeavor and the campaign's important role in creating a more
competitive Medicare market. 

Questions have been raised by health plan representatives and others
about the estimated cost of the information campaign.  The campaign
is to be financed primarily from user fees--that is, through an
assessment on participating health plans.  We recently began a review
of HCFA's plans for the information campaign at your request and that
of the Senate Committee on Finance.  Since the start of our work,
HCFA modifications to its plans for the information campaign have
significantly affected the estimated costs of different components. 

Finally, HCFA has met the July 1, 1998, implementation date for
phasing in a new payment system for SNFs.  We are concerned, however,
that payment system design flaws and inadequate underlying data used
to establish payment rates may compromise the system's ability to
meet the twin objectives of slowing spending growth while promoting
the delivery of appropriate beneficiary care.  Insufficient planned
oversight of the new payment system may compound these shortcomings
and further jeopardize the potential for cost savings.  In the short
term, the new payment system could be improved if HCFA clearly stated
that SNFs are responsible for ensuring that the claims they submit
are for beneficiaries who meet Medicare coverage criteria.  In the
longer term, further research to improve the patient grouping
methodology and new methods to monitor the accuracy of patient
assessments could substantially improve the performance of the new
payment system. 


--------------------
\1 P.L.  105-33 became law on August 5, 1997. 

\2 Medicare:  HCFA Faces Multiple Challenges to Prepare for the 21st
Century (GAO/T-HEHS-98-85, Jan.  29, 1998). 


   BACKGROUND
---------------------------------------------------------- Chapter 0:1

Medicare is the nation's health care program for the elderly and
disabled, covering about 38 million people.  While the organization
and delivery of care has evolved considerably since the 1960s,
Medicare beneficiaries are still overwhelmingly enrolled in a
fee-for-service delivery system in which medical services can be
obtained from any participating provider.  Although private,
employer-based coverage shifted decisively away from fee-for-service
toward networks of providers, only a small percentage of Medicare
beneficiaries are enrolled in such networks--and, almost exclusively
in health maintenance organizations (HMO) that typically offer a more
limited choice of providers.  In contrast, many individuals with
employer-based coverage are enrolled in other types of network plans
that offer a broader choice of physicians.  While employers migrated
toward competing network-based managed care plans to help control
health care costs, Medicare focused on fee-for-service payment
innovations that moved from retrospective, cost-and-charge-based
reimbursements to prospective systems and fee schedules designed to
contain cost growth. 

The August 1997 passage of BBA dramatically changed the existing
paradigm, setting Medicare on a course toward a more competitive and
consumer-driven model.  HCFA, the agency charged with administering
the program, must accomplish this transition while continuing to
oversee the processing of about 900 million claims annually.  BBA
contained over 350 separate Medicare and Medicaid mandates, the
majority of which apply to the Medicare program.  The Medicare
mandates are of widely varying complexity.  Some, such as the
Medicare+Choice expansion of beneficiary health plan options and the
implementation of PPSs for SNFs, home health agencies, and hospital
outpatient services, are extraordinarily complex and have
considerable budgetary and payment control implications.  Others,
such as updating the conversion factor for anesthesia payments, are
relatively minor.  Although most implementation deadlines are near
term--over half had 1997 or 1998 deadlines--several are not scheduled
to be implemented until 2002. 


   PROGRESS MADE IN MEETING BBA
   MANDATES, BUT FUTURE DELAYS
   EXPECTED
---------------------------------------------------------- Chapter 0:2

Overall, BBA required HCFA to implement about 240 unique Medicare
changes.  Since August 1997, about three-quarters of the mandates
with a July 1998 deadline have been implemented.  HCFA's recent
publication of the Medicare+Choice and SNF PPS regulations are
examples of the progress HCFA has made in implementing key mandates. 
The remaining 25 percent missed the BBA implementation deadline,
including establishment of a quality-of-care medical review process
for SNFs and a required study of an alternative payment system for
certain hospitals.  It is clear that HCFA will continue to miss
implementation deadlines as it attempts to balance the resource
demands generated by BBA provisions with other competing objectives. 

Implementing BBA provisions would be daunting under the best of
circumstances, and the task is further complicated for HCFA by other,
concurrent challenges, including new antifraud provisions and other
responsibilities contained in the Health Insurance Portability and
Accountability Act of 1996\3 and BBA's creation of a new program to
reduce the number of uninsured children.  Moreover, HCFA has just
completed a major reorganization and is attempting to recruit and
train staff with the skills needed to transition the agency from a
passive purchaser of health care to an active manager of the
competitive market being created by BBA-mandated changes.  Finally,
the need to modernize its multiple automated claims processing and
other information systems, a task complicated by the Year-2000
computer challenges, is competing with other ongoing
responsibilities. 

HCFA has proposed that the Department of Health and Human Services
seek legislative relief by delaying implementation of certain BBA
provisions--those requiring major computer system changes that also
coincide with Year-2000 computer renovations.\4

According to HCFA's computer contractor, simultaneously pursuing both
BBA implementation and Year-2000 system changes risks the failure of
both activities and threatens HCFA's highest priority--uninterrupted
claims payments.  The contractor advised HCFA to seek relief from
competing requirements, which could allow the agency to focus instead
on Year-2000 computer system renovations. 

The BBA provisions to be delayed by the computer renovations include
updates to the October 1999 inpatient hospital PPS rate and the
January 2000 physician fee schedule, hospital outpatient PPS limits
on outpatient therapy services, and billing changes for SNFs.  The
appendix lists other BBA mandates that are being postponed.\5

It is difficult to assess the impact of these delays.  In some
instances, the effects are direct.  Postponing the outpatient PPS,
for instance, means that Medicare will continue to have few controls
over its outlays for these services.  Similarly, delays in
instituting per-beneficiary limits on the amount of outpatient
therapy services covered by Medicare--a rapidly expanding source of
expenditures--means that anticipated savings will be lost.  Some
delays involve mandates that are intended to complement provisions
already being implemented.  We may expect further increases in
spending for outpatient therapy services because of newly implemented
payment constraints on other therapy providers.  As another example,
consolidated billing makes SNFs responsible for virtually all
Medicare-covered services that residents receive, rather than
allowing other providers to bill directly.  The consolidated billing
provision's importance is heightened by the fact that SNFs are
starting to be paid under the new PPS rates, which cover both
services previously billed by the SNF and by certain outside
providers.  Without this provision, it may be more difficult to
adequately monitor whether bills for SNF residents are being
submitted appropriately. 


--------------------
\3 P.L.  104-191. 

\4 Any change in payment policy requires computer system changes. 
HCFA has proposed delaying system modifications required by BBA so
that resources can be focused on Year 2000 priorities.  Contractors
are required to be Year-2000 compliant by December 31, 1998.  After
compliance is achieved and remaining problems are fixed, resources
will be redirected to meeting delayed BBA requirements. 

\5 Our list is based on provisions identified in a July 8, 1998, HCFA
BBA Implementation Tracking Report. 


   ADEQUATE INFORMATION CRITICAL
   TO SUCCESS OF MEDICARE+CHOICE
---------------------------------------------------------- Chapter 0:3

BBA establishes a new Medicare+Choice program, which will
significantly expand the health care options that can be marketed to
Medicare beneficiaries beginning in the fall of 1998.  In addition to
traditional Medicare and HMOs, beneficiaries will be able to enroll
in preferred provider organizations, provider-sponsored
organizations, and private fee-for-service plans.  Medical savings
accounts will also be available to a limited number of beneficiaries
under a demonstration program.  The goal is a voluntary
transformation of Medicare via the introduction of new plan options. 
Capitalizing on changes in the delivery of health care, these new
options are intended to create a market in which different types of
health plans compete to enroll and serve Medicare beneficiaries. 
Recognizing that consumer information is an essential component of a
competitive market, BBA mandated a national information campaign with
the objective of promoting informed plan choice.  From the
beneficiary's viewpoint, information on available plans needs to be
(1) accurate, (2) comparable, (3) comprehensible, and (4) readily
accessible.  Informed beneficiary choice will be critical since BBA
phases out the beneficiary's right to disenroll from a plan on a
monthly basis and moves toward the private sector practice of annual
reconsideration of plan choice. 

The responsibility for informing beneficiaries about plan choices is
dual, falling on both HCFA and participating Medicare+Choice health
plans.  In keeping with provisions of BBA to inform beneficiaries
about new and existing health care options, HCFA is attempting to
summarize health plan coverage information and make it accessible in
a comparative format.  To ensure accessibility, BBA requires that
comparative information be available to beneficiaries via the
Internet, through a toll-free telephone number, and in printed form
by mail.  Recognizing that expanding the array of health plan choices
and organizing a top-notch information campaign is an enormous
undertaking, BBA mandates a two-step phase-in.  In 1998, HCFA is only
responsible for a "special information campaign" that gives
beneficiaries data on existing HMO options and any new
Medicare+Choice plans.  Only a few new options are expected to be
available and, though not required to do so this year, HCFA is
already providing comparative data via the Internet.  Beginning in
1999, however, the agency is charged with the orchestration of a
"nationally coordinated educational and publicity campaign" that
includes comparative data on the available health plan choices.  This
publicity campaign will support what is to become an annual event
each November--an open enrollment period in which beneficiaries may
review the options and switch to a different health plan.  As in the
past, health plans will continue to provide beneficiaries with
marketing information that includes a detailed description of covered
services.  In fact, HCFA comparative summaries will refer
beneficiaries to health plans for more detailed information. 

HCFA is taking a cautious approach and testing the key components of
its planned information campaign.  This caution is probably warranted
by the important role played by information in creating a more
competitive Medicare market and by the agency's inexperience in this
type of endeavor.  In March 1998, the agency introduced a database on
the Internet called "Medicare Compare," which includes summary
information on health plans' benefits and out-of-pocket costs.  The
toll-free telephone number will be piloted in five states--Arizona,
Florida, Ohio, Oregon and Washington--and gradually phased in
nationally during 1999.  Because of some concerns about its
readability, HCFA has also decided to pilot a new beneficiary
handbook in the same five states instead of mailing it to all
beneficiaries this year.  The handbook, a reference tool with about
36 pages, will describe the Medicare program in detail, providing
comparative information on both Medicare+Choice plans as well as the
traditional fee-for-service option.  For beneficiaries in all other
states, HCFA will send out a five- to six-page educational pamphlet
that explains the Medicare+Choice options but contains no comparative
information.  This schedule will allow HCFA to gather and incorporate
feedback on the effectiveness of and beneficiary satisfaction with
the different elements of the information campaign into its plans for
the 1999 open enrollment period. 


      LACK OF STANDARDIZED
      COMPARATIVE DATA HAMPERS
      HCFA AND BENEFICIARIES
-------------------------------------------------------- Chapter 0:3.1

Until BBA, Medicare lagged behind other large purchasers in helping
beneficiaries choose among plans.  The Federal Employees Health
Benefits Program, the California Public Employees' Retirement System,
Xerox Corporation, and Southern California Edison all provide their
employees with comparative information on premiums, benefits,
out-of-pocket costs, and the results from member satisfaction
surveys.  HCFA, on the other hand, has not routinely provided
plan-specific information directly to beneficiaries.  In 1996, we
reported that beneficiaries received little or no comparative
information on Medicare HMOs.\6 Among other things, we recommended
that HCFA produce plan comparison charts and require plans to use
standard formats and terminology in benefit descriptions. 

In developing comparative information for Medicare Compare, HCFA
attempted to use information submitted by health plans as part of the
contracting process.  Like beneficiaries, HCFA had difficulty
reconciling information from different HMOs because it was not
standardized across plans.  HCFA's Center for Beneficiary Services,
the new unit responsible for providing information to Medicare
enrollees, has been forced to recontact HMOs and clarify benefit
descriptions.  Recognizing that standardized contract information
would reduce the administrative burden on both health plans and
different HCFA offices that use the data, the agency has accelerated
the schedule for requiring standard formats and language in contract
benefit descriptions.  Although originally targeted by 2001, the new
timetable calls for contract standardization beginning with
submissions due in the spring of 1999.  If available on schedule,
standardized contracts should facilitate the production of
comparative information for the introduction of the annual open
enrollment period in November 1999. 

While comparative data from HCFA will provide a starting point for
selecting a health plan, beneficiaries will probably continue to rely
on marketing information and detailed benefit descriptions provided
by plans in making their ultimate choice.  Such materials may be both
difficult to use and misleading because plan marketing material is
not standardized.  In our recent review of marketing materials from
Medicare HMOs in Tampa, Florida, we found that the formats and
benefit categories varied considerably from plan to plan and
sometimes omitted key details, as in the following examples: 

  -- Marketing materials often failed to inform beneficiaries that
     they face higher out-of-pocket costs if they choose a brand-name
     drug over a generic. 

  -- HMOs differed in the terms used to describe the same benefit or
     used technical terms but did not define them.  Thus, some used
     the term "formulary" to describe the prescription drug benefit
     but did not explain that the use of nonformulary drugs may
     result in substantially higher out-of-pocket costs.\7

  -- Only five of eight Tampa plans mention mammograms in their
     benefit summaries--even though all plans covered mammograms. 
     Most plans listed mammograms under the "preventive service"
     benefit category.  One plan, however, included them under
     hospital outpatient services. 

Consistent presentation is important because beneficiaries may rely
on plans' benefit summaries when comparing coverage and out-of-pocket
cost information.  Federal employees and retirees can readily compare
benefits among health plans in the Federal Employees Health Benefits
Program because the Office of Personnel Management requires that plan
brochures follow a common format and use standard terminology.  It is
encouraging that HCFA wants to accelerate a similar requirement for
Medicare+Choice plans.  In the fall of 1999, HCFA expects to require
health plans to use standard formats and terminology to describe
covered services in the summary-of-benefits portion of the marketing
materials. 


--------------------
\6 Medicare:  HCFA Should Release Data to Aid Consumers, Prompt
Better HMO Performance (GAO/HEHS-97-23, Oct.  22, 1996). 

\7 In general, a formulary is a list of drugs that health plans
prefer their physicians to use.  The formulary includes drugs that
plans have determined to be effective and that suppliers may have
favorably priced for the plan. 


      RELIABLE PLAN PERFORMANCE
      DATA ESSENTIAL TO
      QUALITY-BASED COMPETITION
-------------------------------------------------------- Chapter 0:3.2

Comparative data on quality and performance are a key component of
the information campaign mandated by BBA and an essential
underpinning of quality-based competition.  Recognizing that the
measurement and reporting of such comparative data is a "work in
progress," the act directed broad distribution of such information as
it becomes available.  Categories of information specifically
mentioned by BBA include beneficiary health outcomes and
satisfaction, the extent to which health plans comply with Medicare
requirements, and plan disenrollment rates.  While disenrollment
rates could be prepared for publication in a matter of months, other
types of quality-related information have accuracy or reliability
problems or are still being developed. 

The best-known quality-of-care measures available focus on a health
plan's history in delivering preventive services such as mammography,
flu shots, and eye exams for diabetics.  These indicators, referred
to by the acronym HEDIS (Health Plan Employer Data and Information
Set), were jointly developed by a group of large purchasers,
including HCFA, and health plans.  HCFA has already collected data on
many HEDIS measures from Medicare HMOs with the intent of publishing
them.  Since the HEDIS data are self-reported by plans, HCFA
contracted for an audit to verify the accuracy.  HCFA recently
reported serious accuracy problems that it attributed to immature
health plan information systems and ambiguities in the HEDIS
measurement specifications.  Though committed to making the HEDIS
information available as quickly as possible, HCFA emphasized that
its premature release would be unfair to both plans and
beneficiaries.  Finally, efforts have been under way for some time to
develop measures that actually demonstrate the quality of the care
delivered--often referred to as "outcome" measures.  As noted, the
current HEDIS measures look at how frequently a health plan delivers
specific services, such as immunizations, not at outcomes.  The
development and dissemination of reliable health outcome measures is
a much more complicated task and remains a longer-term goal.\8

Before passage of BBA, HCFA had funded a survey to measure and report
beneficiaries' satisfaction with their HMOs.  For example, Medicare
enrollees were asked how easy it was to gain access to appropriate
care and how well their physicians communicated with them about their
health status and treatment options.  HCFA plans to make the survey
results available on its Medicare Compare Internet site this fall and
to include the data in mailings to beneficiaries during the fall 1999
information campaign.  We believe that the usefulness of HCFA's
initial satisfaction survey for identifying poor performing plans is
limited because it surveyed only those individuals satisfied enough
with their plan to remain enrolled for at least 12 months.  HCFA is
planning a survey of those who disenrolled, which could help
distinguish among the potential causes of high disenrollment rates in
some plans, such as quality and access issues or beneficiary
dissatisfaction with the benefit package. 

For the short term, disenrollment rates for health plans provide a
broad indicator of satisfaction that has long been available through
HCFA's enrollment database.  Only since passage of BBA has HCFA begun
to develop formats to make these data useful for public consumption. 
We have urged the dissemination of disenrollment rates in reports to
the Congress over the past 3 years, and we have published comparative
rates for individual markets to illustrate the wide variability in
HMOs' ability to satisfy and retain enrollees.  Our most recent
report shows that many HMOs had relatively high voluntary
disenrollment rates.\9 In many markets, the highest disenrollment
rates exceeded the lowest by more than fourfold.  In a few markets,
the range in rates was even wider.  For example, in Houston, Texas,
the highest disenrollment rate was nearly 56 percent, while the
lowest was 8 percent.  The large range in disenrollment rates among
HMOs suggests that this single variable could be a powerful tool in
alerting beneficiaries about potentially significant differences
among plans and the need to seek additional information before making
a plan choice. 


--------------------
\8 HCFA's current HEDIS initiative contains a single outcome measure
that will require data collection and analysis over several years. 

\9 Medicare:  Many HMOs Experience High Rates of Beneficiary
Disenrollment (GAO/HEHS-98-142, Apr.  30, 1998). 


      ESTIMATING COST OF
      INFORMATION CAMPAIGN
      COMPLICATED BY LACK OF
      EXPERIENCE
-------------------------------------------------------- Chapter 0:3.3

Questions have been raised by health plan representatives and others
about the estimated cost of the information campaign.  The campaign
is to be financed primarily from user fees--that is, an assessment on
participating health plans.  We are conducting a review of HCFA's
information campaign plans at your request and that of the Senate
Committee on Finance.  Our work began recently, and since then HCFA
has modified its plans significantly, affecting the estimated costs
of different components.  While we cannot yet make an overall
assessment, it is clear that the operation of the toll-free number is
the most expensive component and, because of a lack of prior
experience, is the most difficult cost to estimate. 

The cost of the toll-free number comprises 44 percent of the total
information campaign budget.  HCFA projects fiscal year 1998 costs of
$50.2 million to support set up as well as operations during fiscal
year 1999.  All but $4 million will come from user fees collected
from existing Medicare HMOs.  For fiscal year 2000, operations costs
are projected to grow to $68 million.\10

As noted earlier, HCFA will gradually make the toll-free number
available nationwide between October 1998 and August 1999.  HCFA's
approach to establishing the toll-free number appears to be geared
toward controlling costs.  Customer service representatives will
attempt to handle straightforward information requests on the spot
but will refer beneficiaries with more complicated or detailed
questions to the Medicare+Choice plans or to state and local
counselors.\11 This referral concept should limit the duration of
calls and hence their cost.  It is important that the toll-free
number meet beneficiaries' reasonable needs or expectations. 
However, until HCFA actually gains experience with the toll-free
number, it has no firm basis to judge either the duration of the
calls or the type of information beneficiaries will find useful.  The
phased implementation of the toll-free numbers should give HCFA a
better idea of what beneficiaries want and may necessitate
adjustments to current plans. 

Ultimately, the design of this and other aspects of the information
campaign should be driven less by cost and more by how effective they
are in meeting beneficiary needs and contributing to the intended
transformation of the Medicare program.  Consequently, we will be
looking at (1) whether the estimated cost of the planned activities
is appropriate and efficient in the near term, and (2) whether, over
the longer term, the impact and effectiveness of these activities
might be increased. 


--------------------
\10 During its first operational year, the primary costs associated
with the toll-free number will be for a contract to provide trained
customer service representatives.  Smaller contracts will support the
leasing of phone lines and the provision of recorded messages; the
mailing of requested printed materials and the processing of
disenrollment requests; and referrals of complex questions to HCFA. 

\11 The toll-free number will offer prerecorded information 24 hours
a day with customer service representatives available from 8:00 a.m. 
to 4:30 p.m.  weekdays.  These representatives will answer basic
questions about Medicare+Choice rules and the types of plans
available in specific areas.  In answering benefit questions, the
service representatives will rely on the Medicare Compare summary
data that are also available on the Internet.  Finally, the service
representatives will refer requests for printed comparative
information, disenrollment, or difficult policy issues to a separate
contractor. 


   ANTICIPATED SAVINGS AT RISK
   WITH NEW SNF PAYMENT SYSTEM
---------------------------------------------------------- Chapter 0:4

On July 1, 1998, HCFA began phasing in a Medicare PPS for SNFs, as
directed by BBA.\12 Under the new system, facilities receive a
payment for each day of care provided to a Medicare-eligible
beneficiary (known as the per diem rate).  This rate is based on the
average daily cost of providing all Medicare-covered SNF services, as
reflected in facilities' 1995 costs.  Since not all patients require
the same amount of care, the per diem rate is "case-mix" adjusted to
take into account the nature of each patient's condition and expected
care needs. 

Previously, SNFs were paid the reasonable costs they incurred in
providing Medicare-allowed services.  There were limits on the costs
that were reimbursed for the routine portion of care, that is,
general nursing, room and board, and administrative overhead. 
Payments for capital costs and ancillary services, such as
rehabilitation therapy, however, were virtually unlimited. 
Cost-based reimbursement is one of the main reasons the SNF benefit
has grown faster than most components of the Medicare program. 
Because providing more services generally triggered higher payments,
facilities have had no incentive to restrict services to those
necessary or to improve their efficiency. 

Prospective payment is intended to slow spending growth by
controlling the increase in Medicare payments per day of SNF care. 
Facilities that can care for beneficiaries for less than the case-mix
adjusted payment will benefit financially.  Those with costs higher
than the per diem amount will be at risk for the difference between
costs and payments.  The PPS for hospitals is credited with
controlling outlays for inpatient hospital care.  Similarly, the
Congressional Budget Office (CBO) estimates that over 5 years the SNF
PPS could save $9.5 billion compared with what Medicare would have
paid for covered services. 

Although HCFA met the deadline for issuing the implementing
regulations for the new SNF per diem payment system, features of the
system and inadequate data used to establish rates could compromise
the anticipated savings.  As noted in previous testimony, design
choices and data reliability are key to implementing a successful
payment methodology.\13 We are concerned that the system's design
preserves the opportunity for providers to increase their
compensation by supplying potentially unnecessary services. 
Furthermore, the per diem rates were computed using data that
overstate the reasonable cost of providing care and may not
appropriately reflect the differences in costs for patients with
different care needs.  In addition, as a part of the system, HCFA's
regulation appears to have initiated an automatic eligibility
process--that is, a new means of determining eligibility for the
Medicare SNF benefit, that could expand the number of beneficiaries
who will be covered and the length of covered stays.  The planned
oversight is insufficient, increasing the potential for these aspects
of the regulations to compromise expected savings.  Immediate
modifications to the regulations and efforts to refine the system and
monitor its performance could ameliorate our concerns. 


--------------------
\12 HCFA has had problems with computer system changes to implement
the system.  As a result, providers with cost reporting periods
beginning July 1 through September 30, 1998, will receive interim
payments based on the old payment system that will be adjusted
retroactively on or about October 1.  To avoid major disruptions to
the industry, the PPS will be phased in.  For the first 3 years, SNFs
will receive a blended payment of old and new rates. 

\13 Medicare Post-Acute Care:  Home Health and Skilled Nursing
Facility Cost Growth and Proposals for Prospective Payment
(GAO-T-HEHS-97-90, Mar.  4, 1997); Medicare Post-Acute Care:  Cost
Growth and Proposals to Manage It Through Prospective Payment and
Other Controls (GAO/T-HEHS-97-106, Apr.  9, 1997); Medicare:  Recent
Legislation to Minimize Fraud and Abuse Requires Effective
Implementation (GAO/T-HEHS-98-9, Oct.  9, 1997). 


      RATES PAID FOR MANY PATIENTS
      BASED ON SERVICE USE INSTEAD
      OF NEED
-------------------------------------------------------- Chapter 0:4.1

To reflect differences in patient needs that affect the cost of care,
the SNF PPS divides beneficiaries into 44 case-mix groups.  Each
group is intended to define clinically similar patients who are
expected to incur similar costs.\14 An adjustment is associated with
each group to account for these cost differences.  A facility then
receives the same daily payment for all of its patients in each
group.  The case-mix classification method used in this PPS relies
heavily on service use, particularly rehabilitation therapy
(physical, occupational, or speech therapy), to assign patients to
the different groups.  Categorizing patients on the basis of expected
service use conflicts with a major objective of a PPS--to break the
direct link between providing services and receiving additional
payment. 

A SNF has incentives to reduce the costs of the patients in each
case-mix group.  Because the groups are largely defined by the
services the patient is to receive, a facility could do this by
providing the minimum level of services that characterize patients in
that group (see table 1).  This would reduce the average cost for the
SNF's patients in that case-mix group, but not lower Medicare
payments for these patients.  For patients needing close to the
maximum amount of therapy services in a case-mix group, facilities
could maximize their payments relative to their costs by adding more
therapy so that the beneficiary was categorized in the next higher
group.  An increase in daily therapy from 140 to 144 minutes, for
example, would change the case-mix category of a patient with
moderate assistance needs from the "very high" to the "ultra high"
group, resulting in a per diem payment that was about $60 higher.  By
thus manipulating the minutes of therapy provided to its
rehabilitation patients, a facility could lower the costs associated
with each case-mix category and increase its Medicare payments. 
Rather than improve efficiency and patient care, this might only
raise Medicare outlays. 



                          Table 1
          
            SNF Prospective per Diem System for
                   Rehabilitation Groups

                                          Per diem payment
                    Average daily                 (federal
                    therapy                unadjusted rate
Rehabilitation      (for 5 days per              for urban
groups              week)                      facilities)
------------------  ------------------  ------------------
Ultra high          144+ minutes                   $345.90
Very high           100 to 143 minutes              286.30
High                65 to 99 minutes                249.64
Medium              30 to 64 minutes                238.87
----------------------------------------------------------
Notes:  Rates listed are for patients receiving this amount of
therapy who also need moderate assistance with personal care, such as
getting in and out of bed, toileting, moving from a chair to a bed,
and eating. 

Source:  GAO analysis of data from HCFA's May 12, 1998, interim final
rule. 

HCFA needs to continue research efforts to move away from a patient
classification system so closely linked to service use.  If the
case-mix categories were more dependent on patient characteristics,
the facility would have to improve the efficiency with which it
provides care to maximize its Medicare payments relative to costs. 
We recognize that this will be a challenging task.  It is difficult
to group patients by the amounts of care needed using methods that
are less susceptible to manipulation by a SNF.  Nevertheless, being
able to classify patients appropriately is critical to ensuring that
Medicare can control its SNF payments and that SNFs are adequately
compensated for their mix of patients. 


--------------------
\14 The groups are defined by a classification system developed by
HCFA contractors.  The categories in this system are known as
Resource Utilization Groups.  For the Medicare SNF PPS, version III
of the classification system, commonly called RUGS-III, is being
used. 


      INADEQUATE DATA LIKELY
      INFLATE AND DISTORT PAYMENT
      RATES
-------------------------------------------------------- Chapter 0:4.2

We are also concerned that the data underlying the SNF rates
overstate the reasonable costs of providing services and may not
appropriately reflect costs for patients with different care needs. 
The rates to be paid SNFs are computed in two steps.  First, a base
rate reflecting the average per diem costs of all Medicare SNF
patients is calculated from 1995 Medicare SNF cost report data.  This
base rate may be too high, because the reported costs are not
adequately adjusted to remove unnecessary or excessive costs. 
Second, a set of adjustors for the 44 case-mix groups is computed
using information on the costs of services used by about 4,000
patients.  This sample may simply be too small to reliably estimate
these adjustors. 

Most of the cost data used to set the SNF prospective per diem rates
were not audited.  At most, 10 percent of the base year--1995--cost
reports underwent a focused audit in which a portion of the SNFs'
expenses were reviewed.  Of particular concern are therapy costs,
which are likely inflated because there have been no limits on
cost-based payments.\15 HCFA staff report that Medicare has been
paying up to $300 per therapy session.  These high therapy costs were
incorporated in the PPS base rates.  Even if additional audits were
to uncover significant inappropriate costs, HCFA maintains that it
has no authority to adjust the base rates after the July 1, 1998,
implementation of the new payment system. 

The adjustors for each category of patients are based on data from
two 1-day studies of the amount of nursing and therapy care received
by fewer than 4,000 patients in 154 SNFs in 12 states.  Almost all
Medicare patients will be in 26 of the 44 case-mix groups.  For about
one-third of these 26 groups, the adjustors are based on fewer than
50 patients.  Given the variation in treatment patterns among SNFs,
such a small sample may not be adequate to estimate the average
resource costs for each group.  As a result, the case-mix adjusted
rates may not vary appropriately to account for the services
facilities are expected to provide--rates will be too high for some
types of patients and too low for others. 


--------------------
\15 Medicare:  Tighter Rules Needed to Curtail Overcharges for
Therapy in Nursing Homes (GAO/HEHS-95-23, Mar.  30, 1995); Medicare: 
Early Resolution of Overcharges for Therapy in Nursing Homes Is
Unlikely (GAO/HEHS-96-145, Aug.  16, 1996). 


      AUTOMATIC ELIGIBILITY
      PROCESS COULD EXPAND
      MEDICARE COVERAGE
-------------------------------------------------------- Chapter 0:4.3

Medicare's SNF benefit is for enrollees who need daily skilled care
on an inpatient basis following a minimum 3-day hospitalization. 
Before implementation of the prospective per diem system, SNFs were
required to certify that each beneficiary met these criteria.\16 With
the new payment system, the method for establishing eligibility for
coverage will also change.  Facilities will assign each patient to
one of the case-mix groups on the basis of an assessment of the
patient's condition and expected service use, and the facility will
certify that each patient is appropriately classified.  Beneficiaries
in the top 26 of the 44 case-mix groups will automatically be deemed
eligible for SNF coverage.  If facilities do not continue to assess
whether beneficiaries meet Medicare's coverage criteria, "deeming"
could represent a considerable new cost to the program. 

Some individuals who are in one of these 26 deemed categories may
only require custodial or intermittent skilled care, but HCFA's
regulations appear to indicate that they could still receive Medicare
coverage.  Medical review nurses who work with HCFA payment
contractors indicated in interviews that some patients included in
the 26 groups would not necessarily need daily skilled care.  This
may be particularly true at a later point in the SNF stay, since SNF
coverage can only begin after a 3-day hospitalization.  Individuals
with certain forms of paralysis or multiple sclerosis who need
extensive personal assistance may also need daily skilled care
immediately following a hospital stay for pneumonia, for example. 
After a certain period, however, their need for daily skilled care
may end, but their Medicare coverage will continue because of
deeming.  Similarly, certain patients with minor skin ulcers will be
deemed eligible for Medicare coverage, whereas previously only those
with more serious ulcers believed to require daily care were covered. 
Thus, more people could be eligible and Medicare could be responsible
for longer stays unless HCFA is clear that Medicare coverage criteria
have not been changed. 

Deeming eligibility would not be a problem if all patients in a
case-mix group met Medicare's coverage criteria.  To redefine the
patient groups in this way would require additional research and
analysis.  However, an immediate improvement would be for HCFA to
clarify that Medicare will only pay for those patients that the
facility certifies meet Medicare SNF coverage criteria. 


--------------------
\16 Medicare coverage criteria are that the patient require skilled
nursing services or skilled rehabilitation services, that is,
services that must be performed by or under the supervision of
professional or technical personnel; that these skilled services be
required on a daily basis; and, as a practical matter, considering
economy and efficiency, that the services can be provided only on an
inpatient basis in a SNF. 


      LACK OF STRINGENT OVERSIGHT
      COULD FURTHER DIMINISH
      SAVINGS
-------------------------------------------------------- Chapter 0:4.4

Whether a SNF patient is eligible for Medicare coverage and how much
will be paid are based on a facility's assessment of its patients. 
Yet, HCFA has no plans to monitor those assessments to ensure they
are appropriate and accurate.  In contrast, when Texas implemented a
similar reimbursement system for Medicaid, the state instituted
on-site reviews to monitor the accuracy of patient assessments and to
determine the need for training assessors.  In 1989, the first year
of its system's operation, Texas found widespread over-assessment. 
Through continued on-site monitoring, the error rate has dropped from
about 40 percent, but it still remains at about 20 percent. 

The current plans for collecting patient assessment information
actually discourage rather than facilitate oversight.  A SNF will
transmit assessment data on all its patients, not just those eligible
for Medicare coverage, to a state agency that will subsequently send
copies to HCFA.\17 However, the claim identifying the patient's
category for Medicare payment is sent to the HCFA claims contractor
that pays the bill.  At the time it is processing the bill, the
claims contractor will not have access to data that would allow
confirmation that the patient's classification matches the
assessment. 

To some extent, the implementation of the SNF prospective per diem
system reduces the opportunities for fraud in the form of duplicate
billings or billing for services not provided.  Since a SNF is paid a
fixed per diem rate for most services, it would be fraudulent to bill
separately for services included in the SNF per diem.  Yet, the new
system opens opportunities to mischaracterize patients or to assign
them to an inappropriate case-mix category.  Also, as was the case
with the former system, methods to ensure that beneficiaries actually
receive required services could be strengthened.  As with the
implementation of any major payment policy change, HCFA should
increase its vigilance to ensure that fraudulent practices discovered
in nursing homes, similar to problems noted in our prior work, do not
resurface.\18


--------------------
\17 The assessment of all SNF patients is actually a requirement of
the Omnibus Budget Reconciliation Act of 1987 (OBRA 87), which
established requirements for SNFs participating in Medicare and
Medicaid.  The SNF prospective payment system rule added a
requirement that these assessment data be given to the appropriate
state agency.  Previously, they had remained with the SNF. 

\18 Nursing Homes:  Too Early to Assess New Efforts to Control Fraud
and Abuse (GAO/T-HEHS-97-114, Apr.  16, 1997); Fraud and Abuse: 
Providers Target Medicare Patients in Nursing Facilities
(GAO/HEHS-96-18, Jan.  24, 1996). 


   CONCLUSIONS
---------------------------------------------------------- Chapter 0:5

HCFA faces numerous concurrent challenges, many of them the result of
significant changes to the Medicare program mandated by BBA.  Given
the BBA workload alone, implementation delays were probably
inevitable.  And now, HCFA has been advised by its contractor that
its highest priority--uninterrupted claims processing through the
timely completion of Year-2000 computer renovations--may be
jeopardized by some BBA mandates that also require computer system
changes.  Though HCFA is implementing what will become an annual
information campaign associated with Medicare+Choice, it has little
experience in planning and coordinating such an undertaking.  The
ability of the campaign to provide accurate, comparable,
comprehensive, and readily accessible information will help to
determine the success of the hoped for voluntary movement of Medicare
beneficiaries into less costly, more efficient health care delivery
systems.  While BBA computer system-related delays may jeopardize
some anticipated program savings, slower Medicare expenditure growth
is also at risk because of weaknesses in the implementation of other
mandates.  HCFA could take short-term steps to correct deficiencies
in the new SNF PPS.  However, longer-term research is needed to
implement a payment system that fully realizes the almost $10 billion
in savings projected by CBO. 


-------------------------------------------------------- Chapter 0:5.1

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


BBA PROVISIONS DELAYED BY YEAR
2000 COMPUTER RENOVATIONS
=================================================== Appendix Apppendix

                                                Required
BBA section             Provision               implementation date
----------------------  ----------------------  ----------------------
4001                    Collection of non-      No date specified
                        inpatient encounter
                        data from plans

4011, 4012              Medicare+Choice         1/1/99
                        competitive pricing
                        demonstration

4014                    SHMO: Plan for          1/1/99
                        integration of part C
                        and SHMO

4015                    Medicare subvention:    1/1/98\a
                        Project for military
                        retirees

4103                    Prostate cancer         1/1/00
                        screening

4313                    Reporting and           No date specified
                        verification of
                        provider
                        identification numbers
                        (employer
                        identification numbers
                        and Social Security
                        numbers)

4402                    Maintaining savings     10/1/97\b
                        from temporary
                        reductions in capital
                        payments for
                        PPS hospitals

4403                    Disproportionate share  10/1/97\c
                        payment adjustment

4432                    PPS rates for SNFs      7/1/98\d

4432                    SNF consolidated        7/1/98
                        billing for part B
                        services

4441                    Payment update for      10/1/97\e
                        hospice services

4502                    Update to conversion    1/1/99\f
                        factor

4505                    Implementation of       3/1/98\g
                        resource-based
                        practice expense RVUs

4505                    Implementation of       1/1/00
                        resource-based
                        malpractice RVUs

4523                    Hospital outpatient     1/1/99
                        PPS

4531                    Prospective payment     1/1/00
                        fee schedule for
                        ambulance services

4541                    Application of $1,500   1/1/99
                        annual limit to
                        outpatient
                        rehabilitation therapy
                        services

4551                    DME payment provisions  1/1/98\h

4555                    Ambulatory surgical     10/1/97\i
                        center update

4602                    Interim payment for     10/1/97\j
                        home health services:
                        per beneficiary limit

4603                    Prospective payment     10/1/99
                        for home health
                        services

4603                    Requirements for home   10/1/98
                        health payment
                        information

4624                    Payments to hospitals   1/1/98\k
                        for direct costs of
                        graduate medical
                        education of
                        Medicare+Choice
                        enrollees
----------------------------------------------------------------------
\a Collection of encounter data may be delayed. 

\b October 1, 1999, updates and changes may be delayed. 

\c October 1, 1999, updates and changes may be delayed. 

\d October 1, 1999, updates may be delayed. 

\e October 1, 1999, updates may be delayed. 

\f January 1, 2000, update may be delayed. 

\g January 1, 2000, transition may be delayed. 

\h January 1, 2000, update for orthotics and prosthetics may be
delayed. 

\i October 1, 1999, update may be delayed. 

\j Implementation of proration provision delayed. 

\k January change may be delayed. 


*** End of document. ***