Medicare: Home Health Utilization Expands While Program Controls
Deteriorate (Letter Report, 03/27/96, GAO/HEHS-96-16).

Pursuant to a congressional request, GAO examined the growth in the use
of Medicare home health benefits, focusing on the: (1) changes in the
home health industry; (2) composition of Medicare home health users; (3)
differences in utilization of home health benefits across geographic
areas; (4) incentives to overuse Medicare home health benefits; and (5)
effectiveness of payment controls in preventing payment for services not
covered by Medicare.

GAO noted that: (1) the growth in Medicare's home health benefits
resulted from less restrictive Health Care Financing Administration
(HCFA) guidelines issued in 1989; (2) 2.8 million Medicare beneficiaries
received home health services in 1993, up from 1.7 million in 1989; (3)
during the same period, the average number of home health care visits
doubled from 26 visits per year in 1989 to 57 visits per year in 1993;
(4) more than 25 percent of home health beneficiaries received at least
60 visits per year; (5) between 1989 and 1994, the number of
Medicare-certified home health agencies grew from 5,692 to 7,864; (6)
proprietary home health agencies provided beneficiaries with 78 visits
per year, while voluntary and government agencies provided beneficiaries
with 46 visits per year; (7) home health beneficiaries with the same
diagnosis received more visits from proprietary agencies than from
non-profit agencies; and (8) Medicare's home health services can be
improved by subjecting claims to medical review and audit, requiring
visits from intermediaries and physicians to beneficiaries, and
determining whether beneficiaries are qualified for such service, and
actually need or receive the service billed to Medicare.

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

 REPORTNUM:  HEHS-96-16
     TITLE:  Medicare: Home Health Utilization Expands While Program 
             Controls Deteriorate
      DATE:  03/27/96
   SUBJECT:  Home health care services
             Beneficiaries
             Program abuses
             Medical expense claims
             Medical services rates
             Health care cost control
             Claims processing
             Health resources utilization
             Internal controls
             Long-term care
IDENTIFIER:  Medicare Home Health Care Program
             Medicare Program
             Medicare Intermediary Manual
             
******************************************************************
** 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


Report to the Chairman, Special Committee on Aging, U.S.  Senate

March 1996

MEDICARE - HOME HEALTH UTILIZATION
EXPANDS WHILE PROGRAM CONTROLS
DETERIORATE

GAO/HEHS-96-16

Medicare:  Home Health Utilization

(106422)


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

  CFR - Code of Federal Regulations
  CMR - comprehensive medical review
  ESRD - end-stage renal disease
  FMR - focused medical review
  HCFA - Health Care Financing Administration
  HHA - home health agency
  HHS - Department of Health and Human Services
  OIG - Office of Inspector General
  RHHI - regional home health intermediary

Letter
=============================================================== LETTER


B-257049

March 27, 1996

The Honorable William S.  Cohen
Chairman, Special Committee on Aging
United States Senate

Dear Mr.  Chairman: 

Use of the Medicare home health benefit has seen dramatic growth in
recent years, with spending climbing from $2.7 billion in 1989 to
$12.7 billion in 1994 and projected to exceed $21.0 billion by the
year 2000.  Changes in the benefit, the home health industry, and the
characteristics of home health care users have strongly affected home
health care utilization patterns and expenditure trends. 

In several earlier reviews, we reported on lax controls over the use
of the Medicare home health benefit.  Two reports, issued in 1981 and
1986,\1 respectively, concluded that many claimed services were not
medically necessary or did not meet the coverage criteria and,
therefore, should not have been covered under the program.  We also
concluded that payment systems generally were not capable of
detecting noncovered services.  Thus, we recommended that steps be
taken to increase the capability of Medicare's claims processing and
utilization review systems to detect noncovered care. 

The recent rapid growth in Medicare home health expenditures
continues to raise questions about the extent to which abuse of the
benefit may be contributing to this growth.  Therefore, you asked us
to determine the reasons for and the nature of the growth in the use
of the Medicare home health benefit.  Specifically, you asked us to
examine

  changes in the composition of the home health industry,

  changes in the composition of Medicare home health users,

  differences in utilization patterns across geographic areas,

  incentives to overutilize services, and

  the effectiveness of payment controls in preventing payment for
     services not covered by Medicare. 

To address these issues, we reviewed pertinent laws, regulations,
court decisions, Health Care Financing Administration (HCFA)
policies, and relevant research.  We interviewed staff from HCFA, two
HCFA regional offices, and three regional home health intermediaries. 
We analyzed paid claims history data for 1989 through 1993 and
provider of service data for 1989 through 1994.  We did not, however,
examine the internal and automatic data processing controls for
automated systems from which we obtained data used in our analyses. 
With this exception, we conducted our study from July 1994 to
December 1995 in accordance with generally accepted government
auditing standards.  (See app.  I for details on our scope and
methodology.)


--------------------
\1 Medicare Home Health Services:  A Difficult Program to Control
(GAO/HRD-81-155, Sept.  25, 1981), and Medicare:  Need to Strengthen
Home Health Care Payment Controls and Address Unmet Needs
(GAO/HRD-87-9, Dec.  2, 1986). 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Recent growth in the use of Medicare's home health benefit has
largely resulted from 1989 HCFA guideline changes that made Medicare
home health coverage criteria less restrictive, resulting in an
increase in both the number of beneficiaries receiving services and
the number of services received by each beneficiary.  To illustrate,
in 1989, 1.7 million Medicare beneficiaries received home health
services; by 1993, this number had increased to 2.8 million.  During
the same time, the number of visits provided to beneficiaries
receiving home health care more than doubled, from an average of 26
visits per year in 1989 to an average of 57 visits per year in 1993. 
The number of home health beneficiaries receiving services for longer
periods of time also increased; in 1993, more than 25 percent of home
health beneficiaries were receiving 60 or more visits per year, up
from approximately 11 percent of home health beneficiaries in 1989. 

The number of Medicare-certified home health agencies (HHA) has also
grown, from 5,692 agencies in 1989 to 7,864 at the end of 1994; 83
percent of this growth has consisted of proprietary (for-profit)
agencies.  Our analyses show that proprietary agencies consistently
provide more home health visits in all areas of the country than
nonprofit agencies.  In 1993, proprietary agencies provided
beneficiaries with an average of 78 visits per year, while voluntary
and government agencies provided an average of 46 visits.  An
analysis of beneficiaries with one of four frequently occurring
diagnoses shows that proprietary agencies provide significantly more
visits than nonprofits for beneficiaries with the same primary
diagnoses.  For example, home health patients with a primary
diagnosis of diabetes received an average of 53 visits from
proprietary agencies compared with an average of 27 visits from
voluntary agencies and 24 from government agencies. 

Although we have been reporting on program weaknesses over the last
15 years, controls over the Medicare home health benefit remain
essentially nonexistent.  Few home health claims are subject to
medical review and most claims are paid without question.  Further,
because (1) few on-site coverage audits are done, (2) beneficiaries
are rarely visited by intermediaries, and (3) physicians have limited
involvement in home health care, verifying whether the beneficiaries
receiving home care truly qualify for the benefit, need the care
being delivered, or are even receiving the services billed to
Medicare is nearly impossible. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Medicare, administered by HCFA within the Department of Health and
Human Services (HHS), is a health insurance program that covers
almost all Americans 65 years old and older and certain individuals
under 65 years old who are disabled or have chronic kidney disease. 
The program, authorized under title XVIII of the Social Security Act,
provides protection under two parts.  Part A, the hospital insurance
program, covers inpatient hospital services, posthospital care in
skilled nursing homes, and care in patients' homes.  Part B, the
supplementary medical insurance program, covers primarily physician
services but also home health care for beneficiaries not covered
under part A. 


      COVERAGE CRITERIA
---------------------------------------------------------- Letter :2.1

To qualify for Medicare home health care, a person must be confined
to his or her residence (homebound); under a physician's care; and
need part-time or intermittent skilled nursing care and/or physical
therapy or speech therapy.  The services must be furnished under a
plan of care prescribed and periodically reviewed by a physician.\2

Home health benefits covered by Medicare include

  part-time or intermittent nursing care provided by or under the
     supervision of a registered nurse;

  physical, occupational, and speech therapy;

  medical social services related to the patients' health problems;
     and

  part-time or intermittent home health aide services when provided
     as an adjunct to skilled nursing or therapy care.\3

Medicare beneficiaries may receive home health care as long as it is
reasonable and necessary for the treatment of illness or injury; no
limits exist on the number of visits or length of coverage.  Medicare
does not require copayments or deductibles for home health care. 

Medicare home health services must be furnished by Medicare-certified
HHAs or by others under arrangement with such an agency.  Agencies
participating in the program must meet specific requirements of the
Social Security Act.  HHAs are reimbursed for the reasonable costs
incurred in providing covered visits to eligible beneficiaries up to
specified cost limits established for each area of the country.\4

Medicare-certified HHAs are classified into one of three ownership
categories.  Proprietary HHAs are private, for-profit agencies. 
Voluntary agencies are private (nongovernmental), nonprofit agencies
that are exempt from federal income taxation; for example, Visiting
Nurse Associations and Easter Seal Societies.  Government agencies
are operated by a state or local government. 


--------------------
\2 The legislative authority for coverage of home health services is
contained in ï¿½ 1814, ï¿½ 1835, and
ï¿½ 1861 of the Social Security Act; governing regulations are found in
title 42 of the Code of Federal Regulations (CFR); and HCFA coverage
guidelines are found in the Medicare Home Health Agency Manual and
Medicare Intermediary Manual. 

\3 Home health aides provide hands-on personal care of beneficiaries
that must be necessary to the treatment of the beneficiary's illness
or injury.  Home health aide services include (1) personal care
services, such as assistance with eating, bathing, and toileting; (2)
simple surgical dressing changes; (3) assistance with some
medications; (4) activities to support skilled therapy services; and
(5) routine care of prosthetic and orthotic devices.  A beneficiary
whose sole need is for custodial or personal care, however, does not
qualify for home health aide services. 

\4 Under authority originally provided through ï¿½ 223 of the Social
Security Amendments of 1972 (P.L.  92-603), HCFA has established
upper limits on the amount Medicare will pay HHAs.  Based on the cost
experience of freestanding HHAs, these limits are set by type of home
health visit (such as skilled nursing or home health aide).  For each
agency, they are applied in the aggregate; that is, costs above the
limit for one type of visit can be offset by costs below the limit
for another type.  Separate limits are set for urban and rural HHAs
(because costs tend to differ between them) and adjusted to reflect
local wage rates. 


      PROGRAM ADMINISTRATION
---------------------------------------------------------- Letter :2.2

HCFA currently administers the home health care program through
nine\5 regional home health intermediaries (RHHI)--eight Blue Cross
plans and the Aetna Life and Casualty Insurance Company.\6 These
intermediaries

  serve as a communication channel between HHAs and HCFA,

  make payments to HHAs for covered services provided to Medicare
     beneficiaries, and

  establish and apply payment safeguards to prevent program abuse. 


--------------------
\5 Since conducting our review, one Blue Cross plan has dropped out
of the Medicare claims processing business; its responsibilities will
be assumed by one of the remaining eight intermediaries. 

\6 Before the consolidation of home health intermediary functions in
fiscal year 1987, 47 intermediaries administered the home health
program. 


   CHANGES IN ELIGIBILITY CRITERIA
   KEY TO HOME HEALTH GROWTH
------------------------------------------------------------ Letter :3

Changes in the legal and regulatory provisions governing the home
health benefit together with changes in HCFA's policies have played a
key role in the increase in the benefit's use.  At Medicare's
inception in 1966, the home health benefit under part A provided
limited posthospital care of up to 100 visits per year that required
a prior hospitalization of at least 3 days.  In addition, the
services could only be provided within 1 year after the patient's
discharge and had to be for the same illness.  These restrictions
were eliminated by the Omnibus Budget Reconciliation Act of 1980. 

With the implementation of the Medicare inpatient prospective payment
system in 1983, the utilization of the home health benefit was
expected to grow as patients were discharged from the hospital
earlier in their recovery period.  However, expenditures changed
little over the next 5 years\7 (see fig.  1).  The Deficit Reduction
Act of 1984 reduced the number of intermediaries processing home
health claims, and HCFA intensified education of the home health
intermediaries to promote more consistency in claims review. 
Additionally, HCFA instructed the intermediaries to increase the
number of claims receiving medical review before payment.  This
increased review in addition to a requirement for more detailed
documentation contributed to an increased claim denial rate--from 3.4
percent in 1985 to 7.9 percent in 1987.\8

   Figure 1:  Medicare Home Health
   Expenditures, 1980-94

   (See figure in printed
   edition.)

Source:  HCFA, Office of the Actuary. 

A lawsuit was filed in 1988 (Duggan v.  Bowen)\9 that struck down
HCFA's interpretation of benefit coverage requirements.  As a result
of the suit, HCFA revised the Medicare Home Health Agency and
Medicare Intermediary manuals in 1989 so that the criteria for
coverage of home health visits would be consistent with "part-time or
intermittent care," as required by statute, rather than "part-time
and intermittent care," as HCFA had been interpreting it.\10 This
change enabled HHAs to increase the frequency of visits because they
no longer had to be intermittent.  The requirements were also changed
so that patients now qualify for skilled observation by a nurse or
therapist if a reasonable potential for complications or possible
need to change treatment existed.\11 Further, the benefit now allows
maintenance therapy where therapy services are required to simply
maintain function rather than the previous criteria that patients
show improvement from such services. 

The 1989 Medicare Home Health Agency Manual changes also required
that intermediaries, in order to deny claims on the basis of medical
necessity, determine that each denied visit was not medically
necessary at the time services were ordered.  Before this change,
intermediaries were denying all visits beyond what the intermediary
judged necessary; for example, denying 10 visits out of 50 visits
claimed, if the intermediary could determine that the beneficiary
could be adequately treated with 40 visits.  The intermediary did not
need to review each visit.  This change has made it more costly for
intermediaries to determine whether services are medically necessary
and, therefore, fewer claims are denied. 

The effect of changes in Medicare law, regulations, and policy has
been that home health care is now available to more beneficiaries,
for less acute conditions, and for longer periods of time.  For
example, in 1992, approximately one-third of home health
beneficiaries entered the program without a hospital stay at one time
during the year.  Of those who had been hospitalized, only half had a
hospital stay in the 30 days before starting home health care.\12


--------------------
\7 Helbing, C., J.A.  Sangl, and H.A.  Silverman, "Home Health Agency
Benefits," Health Care Financing Review, 1992 Annual Supplement
(1992), p.  125. 

\8 Medicare:  Increased Denials of Home Health Claims During 1986 and
1987 (GAO/HRD-90-14BR, Jan.  24, 1990). 

\9 Duggan v.  Bowen, 691 F.  Supp.  1487 (D.D.C.  1988). 

\10 The manual revisions also added definitions of part-time and
intermittent (see p.  17). 

\11 This skilled observation, in turn, qualifies the beneficiary for
home health aide visits. 

\12 The increased ability of agencies to provide high-technology care
in the home has also contributed to an increase in the number of
users.  Patients who might have formerly received care in an
institution can now receive services such as infusion therapy and
ventilator care in the home. 


      MEDICARE BENEFICIARIES
      RECEIVING MORE HOME HEALTH
      SERVICES
---------------------------------------------------------- Letter :3.1

Since the Medicare Home Health Agency Manual and Medicare
Intermediary Manual changes of 1989, the percentage of Medicare
beneficiaries receiving home health services and the number of home
health visits received per year per home health user have increased
significantly.  In 1989, 1.7 million beneficiaries (5.6 percent of
the Medicare population) received home health care.  In 1993, the
number of beneficiaries receiving such care increased to 2.8 million
(8.8 percent of the Medicare population).  Beneficiaries receiving
home health services are typically female and over 75 years old;
however, the number of disabled beneficiaries under 65 years old
receiving services has been growing.  (See table II.1 in app.  II.)

The average number of visits received per home health beneficiary has
also increased dramatically since 1989.  From 1989 through 1993, the
average number of visits received per year more than doubled, from 26
to 57 visits.  Over the same period, the median number of visits
almost doubled, from 13 to 24 visits (see fig.  2).  Most of the
increase in visits has resulted from an increased use of skilled
nursing (average visits increased from 15 per year in 1989 to 26
visits per year in 1993) and home health aide visits (average visits
increased from 25 visits per year for beneficiaries who received any
aide visits in 1989 to 56 visits per year in 1993). 

   Figure 2:  Growth in Average
   Visits per Year per
   Beneficiary, 1989-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

The distribution of visits across home health beneficiaries has
become increasingly skewed toward heavy users (see fig.  3).  From
1989 to 1993, the percentage of users having more than 60 visits in a
year increased from 10.6 percent to 25.7 percent.  While
beneficiaries who had 60 or fewer visits in 1993 averaged only 20
home health visits (with a median of 15 visits), those with more than
60 visits, averaged 163 visits (with a median of 125 visits).\13 The
percentage of beneficiaries receiving more than 210 visits in 1 year
has also increased, from fewer than 1.0 percent in 1989 to 5.8
percent in 1993. 

   Figure 3:  Distribution of
   Medicare Beneficiary Visits,
   1989 and 1993

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 


--------------------
\13 Home health users with more than 60 visits during the year were
more likely to have a primary diagnosis of the chronic diseases of
diabetes and hypertension than those receiving fewer than 60 visits. 
On the other hand, home health users with fewer than 60 visits were
more likely to be diagnosed with osteoarthritis or hip fracture than
those receiving more than 60 visits. 


      HOME HEALTH INDUSTRY
      EXPANDING RAPIDLY
---------------------------------------------------------- Letter :3.2

The home health industry has experienced rapid growth since 1989. 
The number of Medicare-certified HHAs increased from 5,692 in 1989 to
7,864 at the end of 1994.  Growth has occurred mainly in HCFA's
Dallas, San Francisco, and Chicago regions.\14 (See fig.  II.1 in
app.  II for individual state growth data.)

Recent HHA growth has primarily taken place in proprietary agencies,
while the percentage of more traditional nonprofit home health
providers--visiting nurse associations and government agencies--has
declined (table 1).  In 1989, approximately 35 percent of all
Medicare-certified HHAs were proprietary.  In 1994, close to 50
percent of all HHAs were in this category.  (See fig.  II.1 in app. 
II for state breakdowns.) This increased percentage of proprietary
agencies was responsible for 83 percent of the growth in the number
of HHAs between 1989 and 1994.\15



                                Table 1
                
                   Growth in HHAs Providing Medicare
                           Services, 1989-94


                                                Percen          Percen
HHA type                                Number     t\a  Number     t\a
--------------------------------------  ------  ------  ------  ------
Government                               1,443   25.35   1,353   17.20
Proprietary                              2,007   35.26   3,815   48.51
Voluntary                                2,242   39.39   2,696   34.28
======================================================================
Total                                    5,692  100.00   7,864  100.00
                                                                     \
----------------------------------------------------------------------
\a Percentages may not add to 100 due to rounding. 

Source:  GAO analysis of HCFA's Provider of Service File. 


--------------------
\14 The Dallas region includes Arkansas, Louisiana, New Mexico,
Oklahoma, and Texas; the San Francisco region includes Arizona,
California, Hawaii, and Nevada; and the Chicago region includes
Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin. 

\15 In 1989, 29.5 percent of home health beneficiaries received care
from proprietary agencies.  In 1993, this number had increased to
33.3 percent of home health beneficiaries.  At the same time, total
visits provided by proprietary agencies increased from 38.3 percent
of total home health visits in 1989 to 45.8 percent of all visits in
1993. 


   UTILIZATION VARIES BY
   GEOGRAPHIC AREA AND TYPE OF HHA
------------------------------------------------------------ Letter :4

A comparison of average visits per beneficiary receiving home health
services in 1993 indicates that beneficiaries in certain HCFA
regions--most notably in the Atlanta, Boston, and Dallas
regions--receive considerably more services on average than
beneficiaries in other areas (see table 2).  (Refer to fig.  II.3 in
app.  II for data on total home health visits per Medicare
beneficiary by state.) A further breakdown of these figures by
ownership category indicates that in all regions, proprietary HHAs
provide many more services per case than voluntary or government-run
agencies.\16 (See fig.  II.2 in app.  II for state breakdowns.)



                                         Table 2
                         
                         Average and Median Home Health Visits by
                          HCFA Region\a and Ownership Type, 1993


HCFA
region      Average    Median   Average    Median   Average    Median   Average    Median
---------  --------  --------  --------  --------  --------  --------  --------  --------
Boston\a       66.2        26      89.8        37      57.4        24      62.9        24
New            38.5        19      52.4        25      29.9        13      38.3        19
 York\b
Philadelp      40.1        19      53.0        23      38.4        16      35.8        18
 hia\c
Atlanta\d      79.3        37      91.6        44      60.8        28      68.4        32
Chicago\e      44.2        20      60.5        28      38.9        17      38.8        18
Dallas\f       77.1        33      92.8        43      55.5        25      56.8        24
Kansas         43.2        20      59.4        27      38.1        17      39.8        18
 City\g
Denver\h       55.6        22      85.0        33      42.3        18      47.7        21
San            39.7        17      53.0        23      29.8        14      31.4        15
 Francisc
 o\i
Seattle\j      36.3        17      55.2        23      33.8        16      32.0        16
=========================================================================================
Total          56.7        24      78.0        34      45.9        19      46.1        20
-----------------------------------------------------------------------------------------
\a Includes Connecticut, Maine, Massachusetts, New Hampshire, Rhode
Island, and Vermont. 

\b Includes New Jersey, New York, Puerto Rico, and the Virgin
Islands. 

\c Includes Delaware; Washington, D.C.; Maryland; Pennsylvania;
Virginia; and West Virginia. 

\d Includes Alabama, Florida, Georgia, Kentucky, Mississippi, North
Carolina, South Carolina, and Tennessee. 

\e Includes Illinois, Indiana, Michigan, Minnesota, Ohio, and
Wisconsin. 

\f Includes Arkansas, Louisiana, New Mexico, Oklahoma, and Texas. 

\g Includes Iowa, Kansas, Missouri, and Nebraska. 

\h Includes Colorado, Montana, North Dakota, South Dakota, Utah, and
Wyoming. 

\i Includes Arizona, California, Hawaii, and Nevada. 

\j Includes Alaska, Idaho, Oregon, and Washington. 

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

On the national level, proprietary agencies have provided a
significantly higher number of average visits per home health
beneficiary since 1989 (see fig.  4). 

   Figure 4:  Growth in Medicare
   Home Health Visits per
   Beneficiary by Type of Agency,
   1989-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

A recent study\17 noted that some of the regional variation in
services may reflect differences in the availability of substitute
services.  Additionally, the study reported some regional differences
in patient characteristics; however, these differences did not seem
to have a clear pattern that might partially explain variations in
utilization.  Another study\18 indicated that regional variation
could in part be explained by patient characteristics.  For instance,
the study found that compared with Medicare home health users
nationally, beneficiaries in the East South Central\19 region were
more likely to be frail, chronically ill, and in poorer health.  The
study also noted that home health care in the East South Central
region tended to be delivered outside large metropolitan counties and
in counties that had unusually high percentages of elderly persons
living in poverty (both characteristics associated with higher than
average home health use). 

While evidence might suggest that the availability of substitute
services and beneficiary case-mix may explain some of the regional
variation in utilization of home health services, why proprietary
agencies consistently provide more visits in all regions is not
clear.  To learn more about the differences between care provided by
proprietary and other types of HHAs, we conducted an episode-of-care
analysis\20 for four diagnoses:  diabetes, heart failure,
hypertension, and hip fracture.\21 (See app.  I for our methodology
and app.  III for detailed results.)

For these diagnoses, proprietary agencies, on average, provided care
for the longest period of time and provided the most visits per
episode during the period studied (see table 3).  Although
government-run agencies provided care for similar lengths of time as
proprietary agencies, government-run HHAs provided 32 to 45 percent
fewer visits to beneficiaries with these four diagnoses.  Voluntary
agencies, in general, provided care for the shortest period of time
for all four diagnoses, but they provided slightly more visits per
episode than government-run agencies.  Variations in utilization
between the different types of HHAs were most notable in cases of
diabetes, which is regarded as a chronic problem, and less notable in
cases of hip fracture, which is more of an acute problem. 



                                         Table 3
                         
                          Average Episode Length and Visits per
                             Episode--Four Diagnoses, 1992-93


            Average   Average   Average   Average   Average   Average   Average   Average
Diagnosis  length\a    visits  length\a    visits  length\a    visits  length\a    visits
---------  --------  --------  --------  --------  --------  --------  --------  --------
Diabetes       59.0      38.2      61.3      28.7      63.7      52.6      55.3      30.5
Heart          54.9      32.1      56.1      25.8      59.7      43.4      52.1      27.5
 failure
Hypertens      57.4      34.9      58.5      26.4      62.0      44.0      52.8      27.8
 ion
Hip            43.3      29.3      42.9      24.1      44.6      35.3      42.8      27.3
 fracture
-----------------------------------------------------------------------------------------
\a In days. 

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

Several HCFA and intermediary officials expressed concern that the
growing number of proprietary agencies may be generating increased
utilization of home health services.  They believe that because the
beneficiary incurs no cost and little data exist on the effectiveness
of different plans of care, HHAs primarily compete by offering
greater numbers of services to beneficiaries.  Some HHS Office of
Inspector General and intermediary officials further believe that the
nonprofit HHAs are being forced to offer increasingly more services
in order to stay in business. 


--------------------
\16 For example in 1993, beneficiaries receiving home health aide
visits from proprietary agencies received an average of 69 of such
visits, while those receiving home health aide visits from voluntary
and government agencies received an average of 43 and 48 visits,
respectively. 

\17 Mauser, E., and N.A.  Miller, "A Profile of Home Health Users in
1992," Health Care Financing Review, Fall (1994), p.  17. 

\18 Schore, J., "Patient, Agency, and Area Characteristics Associated
with Regional Variation in the Use of Medicare Home Health Services,"
Mathematica Policy Research, Inc.  (1994), reference number 7899-400. 

\19 Consists of Alabama, Kentucky, Mississippi, and Tennessee. 

\20 Because home health episodes are not clearly defined by admission
and discharge dates, we defined episodes of care as a series of home
health visits preceded and followed by a 60-day period with no
visits. 

\21 Approximately 22 percent of home health beneficiaries have one of
these four primary diagnoses. 


   BENEFIT CONTROLS WEAKENED AS
   UTILIZATION EXPANDS
------------------------------------------------------------ Letter :5

In two reports issued in 1981 and 1986, respectively,\22 we
criticized HCFA's administration of the Medicare home health benefit. 
We reported that about 27 percent of the visits reviewed at 37
agencies and paid for under the benefit were questionable or
improper.  We attributed those problems to the vagueness of the
coverage criteria (particularly uncertainty over the exact meaning of
terms such as homebound and intermittent care), insufficient
information being submitted with the claims upon which to base a
coverage decision, and poor performance of the intermediaries in
reviewing claims.  We also noted that other control problems were
adversely affecting proper utilization of the home health benefit,
including insufficient physician involvement and inadequate
monitoring of beneficiary status.  In revisiting these issues, we
found that while controls had improved during the mid- and late
1980s, they have largely deteriorated since then. 


--------------------
\22 Medicare Home Health Services:  A Difficult Program to Control
(GAO/HRD-81-155, Sept.  25, 1981), and Medicare:  Need to Strengthen
Home Health Care Payment Controls and Address Unmet Needs
(GAO/HRD-87-9, Dec.  2, 1986). 


      CONSIDERABLE ROOM FOR
      INTERPRETATION OF COVERAGE
      CRITERIA REMAINS
---------------------------------------------------------- Letter :5.1


         HOMEBOUND STATUS
-------------------------------------------------------- Letter :5.1.1

The Social Security Act requires that a beneficiary be "confined to
the home" (homebound) to be eligible for Medicare home health care. 
In our 1981 report, we found that determining whether beneficiaries
are homebound is difficult due to the inadequacy of the definition
provided by HCFA.  The report recommended that HCFA's criteria for
determining homebound status be clarified and made more specific. 
The Omnibus Reconciliation Act of 1987 added a definition of
homebound to the Social Security Act\23

using the same wording as the HCFA Home Health Agency Manual
definition.  Therefore, the definition of homebound remains
essentially unchanged and considerable discretion remains in
interpreting and applying the homebound definition. 

As stated in the Medicare Home Health Agency Manual, homebound means
that

     "the condition of these patients should be such that there
     exists a normal inability to leave home and, consequently,
     leaving their homes would require a considerable and taxing
     effort. 

     "[Further,] if the patient does in fact leave the home, the
     patient may nevertheless be considered homebound if the absences
     from the home are infrequent or for periods of relatively short
     duration, or are attributable to the need to receive medical
     treatment." (See app.  IV for a full definition.)

Several HCFA and intermediary officials said that few denials are
made on the basis that the beneficiary was not homebound.\24 One
intermediary official said that the RHHI made fewer than 10 denials a
year based on the homebound criteria.  An HCFA official further noted
that although the RHHIs tend to interpret the homebound criteria
fairly consistently, the criteria are so broad that very few claims
are denied on the basis that the coverage criteria have not been met. 
Finally, even if intermediaries do make a denial based on the
homebound criteria, so much room for interpretation still exists in
the infrequent or short duration requirements that such denials may
end up being reversed at the reconsideration or appeals level.\25

A recent study conducted by one of the RHHIs\26 identified some of
the types of abuses that are difficult for the intermediary to
prevent because of the range of interpretations possible for the
homebound criterion.  For example, the study identified an instance
where a physician called the RHHI to complain that some of his
patients were being told by an HHA that they were homebound because
they did not own a car.  The survey also revealed an example of a
home health beneficiary who would put her home health care on hold so
that she could go fishing for a week or two.  She would then come
back and resume her care. 


--------------------
\23 Social Security Act ï¿½ 1814(a), 42 U.S.C.  1395f(a). 

\24 Denials were made only in obvious cases; for example, when the
aide frequently noted in the patient's record that she was unable to
supply care because the patient was not home at the time. 

\25 In 1991, HCFA attempted to develop numerical parameters to better
define the terms:  infrequent, short duration, and confined to the
home.  However, HCFA's proposal received so many negative responses
during the comment period, from intermediaries as well as home health
agencies, that this proposal was never implemented.  Many HHAs
expressed concerns that such absolute limits would rob them of
flexibility in interpretation of the benefit. 

\26 In 1993, Aetna of Florida did a pilot study that involved sending
a sample of physicians and beneficiaries detailed lists of claims
filed on behalf of their patients or themselves, respectively. 


         INTERMITTENT CARE
-------------------------------------------------------- Letter :5.1.2

The Medicare Home Health Manual sets the parameters of the term
intermittent in two ways.  The first pertains to beneficiary
eligibility requirements; to meet the requirement for intermittent
skilled nursing care, an individual must have a "medically
predictable recurring need for skilled nursing services."\27 In most
instances, the definition will be met if a patient requires a skilled
nursing service at least once every 60 days.\28 In contrast, a person
expected to need more or less full-time skilled nursing care over an
extended period of time would usually not qualify for home health
benefits because he or she needs a higher level of care. 

The second parameter of intermittent pertains to the frequency of
visits allowed by Medicare in a given time frame and is usually used
together with the term part-time.\29 According to the Medicare Home
Health Agency Manual,\30 intermittent care is defined as

  up to and including 28 hours per week of skilled nursing and home
     health aide services combined provided on a less than daily
     basis;

  up to 35 hours per week of skilled nursing and home health aide
     service combined that are provided on a less than daily basis,
     subject to review by fiscal intermediaries on a case-by-case
     basis, and determined on the basis of documentation justifying
     the need for and reasonableness of such additional care; or

  up to and including full-time (that is, 8 hours per day) skilled
     nursing and home health aide services combined that are provided
     and needed 7 days per week for temporary but not indefinite
     periods of time of up to 21 days with allowances for extensions
     in exceptional circumstances where the need for care in excess
     of 21 days is finite and predictable.\31

Because a range of interpretations is possible for intermittent, the
requirement is difficult to enforce.  For example, individuals can be
provided intermittent services (for example, blood tests or periodic
skilled observation) every 60 days simply to qualify for aide
services on a long-term basis.\32 Under the part-time or intermittent
coverage rules, determining whether someone who needs daily care for
an extended period meets the intermittent requirement or might
require institutional care is difficult.  Moreover, without further
review, to determine whether daily care itself is really necessary is
not possible.  During our recent investigation of a large home health
organization, for example,\33 employees alleged instances where
managers instructed nurses to visit new patients daily for the first
14 or 21 days of care regardless of condition--intermediaries usually
do not question daily visits during the first 21 days of care. 


--------------------
\27 Social Security Act, ï¿½ 1814(a)(2)(C); Medicare Home Health Agency
Manual, ï¿½ 204.4 and ï¿½ 205.1C. 

\28 The manual further states that since the need for intermittent
skilled nursing care makes the individual eligible for other covered
home health services, the intermediary should evaluate each claim
involving skilled nursing services furnished less frequently than
once every 60 days.  Thus, it is possible that a beneficiary may
receive skilled care less than once every 60 days. 

\29 Part-time means any number of days per week up to and including
28 hours per week of skilled nursing and home health aide services
combined for fewer than 8 hours per day, or up to 35 hours per week
of skilled nursing and home health aide services combined for fewer
than 8 hours per day subject to review by fiscal intermediaries on a
case-by-case basis, based upon documentation justifying the need for
and reasonableness of such additional care. 

\30 Revised in 1989 to implement the decision of the District Court
of the District of Columbia in the Duggan v.  Bowen case. 

\31 ï¿½ 206.7(B). 

\32 For example, Aetna's 1993 survey identified cases where patients
received unneeded skilled services in order to qualify for aide
services. 

\33 See Medicare:  Allegations Against ABC Home Health Care
(GAO/OSI-95-17, July 19, 1995). 


      LESS INFORMATION IS
      AVAILABLE TO INTERMEDIARIES
      FOR MAKING COVERAGE
      DECISIONS
---------------------------------------------------------- Letter :5.2

In August 1985, HCFA implemented standardized medical information
forms\34 for HHAs to use in requesting payment from intermediaries. 
These plan-of-care and update forms, which were to be submitted with
the initial claim and the claim closest to the recertification date
60 days later, gave medical reviewers more detailed information on
each beneficiary's general physiological condition, homebound status,
functional limitations, nutritional requirements, services
prescribed, and services received.  The additional information was
intended to increase the accuracy and consistency of coverage
decisions. 

In our 1990 report,\35 we noted that the regional intermediaries
generally agreed that denials associated with the implementation of
the new forms were a contributor to increases in denials in fiscal
year 1986.\36 HCFA has, however, dropped the requirement for routine
submittal.  In a September 1994 revision, the Medicare Intermediary
Manual was changed to state: 

     "These forms [485/486] are no longer submitted routinely with
     the initial claim or other subsequent claim.  The completed
     HCFA-485, signed by the physician, is retained in the HHA files
     and a copy of the signed form is submitted [to the intermediary]
     when requested for medical review.  The HCFA-486 is completed
     only when required for medical review."

An HCFA official explained that the primary reason for dropping this
requirement was that over time HHAs learned how to fill out the forms
in a manner that would most likely result in the services being
approved for payment; the completed forms all started to look alike
and were less useful.  Currently, the only information the
intermediary routinely receives is the bill from the HHA.  A notation
in the annotated intermediary version of the Medicare Home Health
Agency Manual states that RHHIs will

     "assume that the type and frequency of services ordered are
     reasonable and necessary unless objective clinical evidence
     clearly indicates otherwise, or there is a lack of clinical
     evidence to support coverage."

Because the current billing form alone does not supply adequate
information to make this type of determination, most bills are paid
without question. 


--------------------
\34 HCFA forms 485, Home Health Certification and Plan of Care, and
486, Medical Update and Patient Information.  Other forms implemented
at the same time include an addendum to the plan of treatment and
patient information form (the 487 form) and an intermediary medical
information request form (the 488 form). 

\35 Medicare:  Increased Denials of Home Health Claims During 1986
and 1987 (GAO/HRD-90-14BR, Jan.  24, 1990). 

\36 We concluded that this initiative contributed to more claims
denials because (1) medical reviewers had more information on which
to make coverage decisions and (2) some intermediaries denied claims
because certain information was missing, instead of requesting the
required data. 


      LITTLE MEDICAL REVIEW IS
      DONE
---------------------------------------------------------- Letter :5.3

The regional home health intermediaries are responsible for
procedures to assure that they only make payments for home health
services that are covered by Medicare and avoid paying for services
that are (1) provided to beneficiaries who do not meet Medicare home
health criteria, (2) not reasonable or medically necessary, or (3) in
excess of the services called for by the approved plan of treatment. 
Currently, the RHHI's primary procedure for detecting noncovered
services is medical review of
claims.\37


--------------------
\37 Medical review involves reviewing additional information
requested from and submitted by the HHA, such as the 485, Plan of
Care Form, and the beneficiary's medical records. 


         PREPAYMENT REVIEWS
-------------------------------------------------------- Letter :5.3.1

The Consolidated Omnibus Budget Reconciliation Act of 1985 more than
doubled the funds available for medical review and audit of home
health and other Medicare claims.  This allowed intermediaries to
increase the number of medical reviews performed; they conducted
medical reviews on 62 percent of home health claims processed in
fiscal years 1986 and 1987.  The increased number of claims subjected
to medical review resulted in more denials and higher denial rates
even though the percentage of claims being denied during medical
review did not increase significantly.  For example, in both 1985 and
1987, intermediaries denied about 10 percent of the claims subjected
to medical review.  However, because over twice as many claims were
subjected to medical review in 1987, there were over twice as many
denials.  As a result, the HCFA-reported denial rate was 7.9 percent
in 1987 compared with 3.4 percent in 1985. 

Due to budget cuts since 1989,\38 however, intermediaries are now
required to conduct medical reviews (pre- and postpayment) on a
target of 3.2 percent of all claims, including home health claims.\39
At the same time, home health claims volume increased from 5.5
million claims in 1989 to 16.6 million claims in 1994.  Of the 3.7
percent of home health claims denied in fiscal year 1994, only 0.6
percent were denied because the services were determined, through
medical review, to not be medically necessary or because the
beneficiary did not meet the qualifying coverage criteria.\40

As a result of decreased review, HHAs are less likely to be caught if
they abuse the home health benefit.  An HCFA official noted that HHAs
are aware that the intermediary only reviews a small number of claims
and, therefore, can take chances billing for noncovered services.  As
long as they do not trigger the criteria that would cause the claim
to be flagged,\41 HHAs can submit abusive claims that will never be
reviewed. 

Besides covering so few claims, prepayment medical review is limited
in its ability to detect noncovered care in that it is simply a paper
review done at the offices of the RHHI.  According to HCFA and
intermediary officials, it is often not possible to obtain enough
information from a paper review alone no matter how complete the
medical records submitted, to determine whether a provider is abusing
the benefit or committing fraud.  If the codes are valid, the forms
filled out properly, and no unusual patterns are identified during
the FMR process, the claim goes through.  For example, our
investigation of a large home health organization turned up
allegations that staff were directed to alter or falsify medical
records to ensure continued or prolonged visits, including recording
visits that were never made or noting that patients were homebound
even after they were no longer confined to the home.\42 To further
illustrate, an intermediary official noted that sometimes the wrong
diagnosis is put on the claim form to make beneficiaries appear
sicker than they really are and, thus, in need of more care. 


--------------------
\38 In 1989, total part A contractor funding for medical
review/utilization was $61 million.  Due to subsequent budget cuts,
by 1992 this funding dropped to $31 million--an almost 50-percent
cut.  Fiscal year 1995 funding is $33.1 million.  Payment safeguard
funding for the home health benefit is based on the number of home
health claims processed by the intermediary in relation to other
types of claims processed. 

\39 The contractor Budget and Performance Requirements, 1995
stipulate that "(t)he target review level is 3.2 percent, however,
intermediaries may reduce the review level based on resources
available.  The minimum acceptable review level is 1 percent although
it is expected that intermediaries review as many claims as
possible."

\40 The remaining denials stemmed primarily from Common Working File
edit checks; for example, the patient was not eligible to receive
Medicare benefits, Medicare was the secondary payer, the bill was for
noncovered services (such as some supplies), or the bill was a
duplicate. 

\41 To identify noncovered services, intermediaries currently
evaluate claims through a focused medical review process (FMR). 
According to FMR procedures, the intermediaries are to target the
review of claims where there is the greatest risk of inappropriate
payment.  Intermediaries analyze utilization data and develop
measures to identify predictors of aberrant utilization among their
providers.  The intermediaries then put edits in their claim
processing systems that flag the claims that exceed the chosen
criteria; for example, high levels of utilization.  A small
percentage of the flagged claims are then reviewed.  Even if a claim
exceeds the screening criteria, it may still not be selected for the
small percentage that are actually reviewed. 

\42 See Medicare:  Allegations Against ABC Home Health Care
(GAO/OSI-95-17, July 19, 1995). 


         POSTPAYMENT REVIEW
-------------------------------------------------------- Letter :5.3.2

Postpayment utilization review differs from prepayment review in that
its principal focus is on identifying HHAs that are providing
significant amounts of noncovered care rather than on identifying
services provided to specific beneficiaries.  In 1982, HCFA
implemented a selective postpayment utilization review program that
has cost effectively identified extensive noncovered services paid
for by Medicare.  The essential component of postpayment review,
comprehensive medical review (CMR), is a thorough postpayment
evaluation of claims and medical documentation that may involve an
audit at the provider's site.\43 On-site audits give the reviewer
access to the information in the provider's records, including plans
of care and documentation of visits. 

According to records obtained from HCFA, only 51 on-site audits were
conducted by the nine RHHIs combined in fiscal year 1994.\44 Thus,
fewer than 1 percent of all Medicare- certified HHAs were audited. 
Intermediaries are required to perform 10 on-site CMRs each year for
all provider types, including, for example, outpatient, skilled
nursing, and rehabilitation facilities.  An HCFA representative noted
that CMRs are so resource intensive that they may be done only in
instances where a high level of return is expected.  Because HHA
claims may comprise a relatively small portion of an intermediary's
total claims volume, the intermediary may not do any home health
CMRs. 

One of the best ways to verify information provided by the HHA is to
visit beneficiaries at home.  Beginning in 1984, intermediaries were
required to make visits to a sample of five beneficiaries at targeted
agencies to assess coverage status; however, this requirement was
subsequently dropped due to cuts in contractor funding.  In March
1995, HCFA revised the Medicare Intermediary Manual to say that
intermediaries may perform visits to selected beneficiary homes but
they are not required to do so.  According to officials at the
intermediaries visited, only one of the three was doing any
beneficiary visits as part of its CMRs.\45

A proposed sampling procedure for CMRs involves selecting a valid
statistical sample of claims from agencies suspected of abusive
practices and extrapolating the denial rate (and therefore payment
recovery rate) in the sample to similar claim types during the same
period.  In our 1986 report, we suggested that by using statistically
valid sampling techniques, such as those being used to estimate
physician overpayments under Medicare part B, overpayments to HHAs
for noncovered services could be projected to all claims submitted by
the agency during the sampling period and could result in millions of
dollars in additional recoveries.  In addition, we recommended that
HCFA require intermediaries to use such procedures.  However, RHHIs
are currently not required to use a projectable sample of home health
visits to extend recoveries--recoveries are, therefore, limited to
the cost of actual services reviewed and denied.  HCFA is circulating
a new draft sampling plan that delineates the methodology for
selecting a representative sample.  However, previous attempts to
implement statistically valid postpayment sampling have not been
successful, primarily due to opposition from the home health industry
and other health care providers. 


--------------------
\43 Providers are selected for a CMR based on performance patterns
identified in prepayment review; for example, the provider submits
noticeably altered documentation, has a pattern of not complying with
physician orders, or is suspected of fraud.  In this type of CMR,
overpayments are only collected on claims that are reviewed. 

\44 The number of on-site audits ranged from none to 15 for each
RHHI. 

\45 This intermediary noted, however, that HHAs sometimes coach the
beneficiaries on what to say and do to ensure that they would
continue to get home health coverage. 


      PHYSICIANS NOT ACTIVELY
      INVOLVED IN MONITORING
      PATIENT CARE
---------------------------------------------------------- Letter :5.4

With the enactment of the Medicare program, it was expected that the
physician would play an important role in determining utilization of
services.  Medicare law and regulations, therefore, require that home
health items and services must be furnished under a plan of treatment
established and periodically reviewed by a physician.  HCFA requires
that the plan is to be reviewed and recertified in writing by the
attending physician at least every 62 days.  The physician is
expected but not required to see the patient. 

Few data exist about the current nature of physician involvement in
home health care.  Concerns have been raised, based on audits of
certain HHAs and anecdotal reports, that physicians are not
appropriately involved in planning and coordinating home health
services.  For example, both HCFA and intermediary officials
expressed concern that HHAs were preparing the plans of treatment and
the physicians were signing them with little or no review. 

A recent report issued by HHS' Office of the Inspector General
(OIG)\46 that was based on a survey of physicians and HHAs around the
country found that physicians generally have a relationship with
patients for whom they sign plans of care.  Physicians usually
reported initiating referrals for home care and reviewing the plans
of care that they sign; however, most do not prepare the plans of
care themselves.  The report also found that physicians were most
involved when caring for patients with complex medical problems and
were less involved when caring for patients with chronic or less
complex conditions.  Thus, physicians frequently are not aware of the
ongoing HHA services being provided to patients and billed to the
Medicare program.  HHS' Inspector General pointed out\47 the
importance of recognizing that physicians usually do not make home
visits themselves to monitor the HHA services provided and do not
directly manage the care that a patient receives from an HHA.  An
intermediary official noted that some physicians feel that because
they are ordering nonmedical services, which will generally not harm
the patient, not much review is required. 

The 1993 Aetna of Florida pilot study revealed examples of different
levels of physician involvement.  In one instance, a physician wrote
that he took every Friday off to spend the whole day reviewing home
health plans of care.  Another physician, who received 100 plans of
care a week, wrote a letter to his intermediary reprimanding it for
asking him to read the plans of care.  Our investigation of a large
home health organization found that physicians typically rely on
nurses' verbal recommendations, written recommendations, or both.  We
also noted allegations that physicians' signatures were forged and
plans of care were altered after certification without the
physicians' knowledge. 

To compensate physicians for the time spent on preparing and
reviewing home health plans-of-care forms, HCFA issued a new
regulation in 1994 providing separate payment for physician care plan
oversight services.  As of January 1995, HCFA began allowing
participating physicians to be paid for oversight requiring at least
30 minutes.  Currently, the payment rate is approximately $81 per
patient. 


--------------------
\46 See "The Physician's Role in Home Health Care," HHS, OIG,
OEI-02-94-00170 (Washington, D.C.:  1995). 

\47 Testimony of June Gibbs Brown, Inspector General, HHS, before the
Senate Special Committee on Aging (Mar.  21, 1995). 


      PHYSICIANS AND BENEFICIARIES
      NOT AWARE OF SERVICES BILLED
---------------------------------------------------------- Letter :5.5

Neither the beneficiaries receiving nor the physicians ordering home
health services are sent information about which services Medicare
has paid.  Beneficiaries do not receive an explanation of benefits
because they are not billed for in-home services.  Therefore, neither
the physician nor the beneficiary has any way of knowing whether
Medicare is paying the HHA for services not rendered or whether the
home health services are provided according to the plan of care. 


      DENIED CLAIMS LIKELY TO BE
      PAID UNDER WAIVER OF
      LIABILITY
---------------------------------------------------------- Letter :5.6

Under the waiver-of-liability provision of the Social Security Act
(ï¿½ 1879) Medicare will pay for denied services if the beneficiaries
and providers did not know and had no reason to know that the
services were not medically reasonable and necessary or were based on
the need for custodial rather than skilled care.  In implementing
this provision, HCFA generally presumed that HHAs did not know
services were not covered as long as their number of denials did not
exceed 2.5 percent of total visits billed.  When a provider exceeded
the 2.5-percent rate in a calendar quarter, Medicare would not
reimburse the provider for denied services, usually for the next
3-month period.\48

According to statistics obtained from HCFA, in fiscal year 1994
approximately half of all claims denied for lack of medical necessity
or for not meeting the coverage criteria were eligible for waiver. 
Of those eligible for waiver, 73 percent were ultimately paid.  In
fiscal year 1994, the total amount reimbursed under waiver was
approximately $45.5 million. 

Because so few claims are reviewed and so few technical and medical
necessity denials are made, most providers, especially those who
submit large numbers of claims, would never exceed the 2.5-percent
rate threshold.\49 In an earlier report,\50 we noted that savings
could be realized by changing the waiver-of-liability rules and
recommended that HCFA establish more stringent eligibility
requirements for the application of waiver of liability for health
care providers under part A of Medicare.\51


--------------------
\48 The Omnibus Budget Reconciliation Act of 1986 created a second
waiver-of-liability category under which the beneficiary is not
liable when services are denied for technical reasons; that is,
because the beneficiary was not homebound or did not require
intermittent skilled nursing care.  HCFA pays providers for services
denied for technical reasons using the same 2.5-percent rate
criterion that applies to medical necessity denials. 

\49 Providers who do exceed the threshold are usually targeted by
intermediaries for high levels of review. 

\50 Savings Possible by Modifying Medicare's Waiver of Liability
Rules (GAO/HRD-83-38, Mar.  4, 1983). 

\51 We suggested several ways to achieve savings:  (1) eliminating
the presumption that providers did not know or could not reasonably
be expected to know that certain services were not covered and
applying the waiver provision on a case-by-case basis; (2) tightening
the denial rate criteria used to determine presumed eligibility (that
is, reducing the threshold); and (3) changing the waiver-of-liability
procedure so that after providers have participated in Medicare for
some period of time, there would no longer be a presumption of
eligibility.  The legal provision requiring HCFA to use the
presumptive level expired in January 1996.  HCFA instructed the
intermediaries to discontinue using presumption of eligibility for
waiver of liability and make waiver decisions for each denied claim. 


   HCFA STRIVING TO ADDRESS
   PROBLEMS
------------------------------------------------------------ Letter :6

In response to the changing climate surrounding home health care, the
Administrator of HCFA convened an internal task force in the spring
of 1994 (the Medicare Home Health Initiative) to examine the home
health benefit from both a policy and an operations perspective.  As
of September 1995, the task force has held four open workgroup
meetings at which HCFA officials solicited ideas and suggestions for
benefit improvement from physician organizations, representatives of
beneficiary groups, the home health industry, state governments and
their Medicaid agencies, and others.  The task force has also issued
a draft revision of the conditions of participation, developed a
pamphlet to better inform beneficiaries of what services are covered,
and developed draft sampling instructions for postpayment utilization
review.  Further, the task force has implemented a four-state pilot
program to investigate providing home health beneficiaries with
claims information, begun pilots of team on-site medical review of
HHAs,\52 revised the Medicare Intermediary Manual to allow
unannounced on-site audits,\53 and implemented a two-state pilot
program involving training state surveyors to assess patient
eligibility as a part of HHA annual surveys. 

These efforts by HCFA are commendable and should help somewhat in
gaining control of the use of the home health benefit.  However, as
discussed earlier in this report, HCFA cannot address many of the
major problems, such as the changes in the manuals made in response
to a court decision, that make it harder to control use of services
and the shortage of funds to perform program safeguard activities. 


--------------------
\52 The March 1995 Medicare Intermediary Manual revisions added the
following statement under Review Options:  "Team Reviews have been
found to be very effective and are to be conducted whenever
appropriate.  The team may consist of medical review, and/or audit
and fraud and abuse staff, state surveyors, carrier and/or Medicaid
staff depending upon the issues identified.  At minimum, prior to
conducting CMRs you are to consult and share information with other
internal and external (as appropriate) staff to determine if there
are issues that you should be aware of or if a team review is
needed."

\53 Until a March 1995 revision to the Medicare Intermediary Manual,
RHHIs were required to give HHAs 10 days' notice before doing an
on-site audit (none if fraud was suspected).  The manual now leaves
the decision of whether to give notice and how much to the RHHI. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

The Medicare home health program is judged by HCFA as being very
difficult to control.  While quantifying how much of the recent
growth in home health care is due to abuse of the benefit is not
possible, lax benefit controls leave the door open for abuses such as
overutilization to occur.  While HCFA has made some notable attempts
to remedy several specific problems, a number of fundamental issues
remain.  For example: 

  In response to a court decision, HCFA revised its requirements for
     determining Medicare home health eligibility.  The revisions
     made it possible for more beneficiaries to qualify for Medicare
     home health services and more HHAs to receive payment for higher
     numbers of visits and for longer periods of care.  Historically,
     part A of Medicare's home health benefit was directed at acute
     conditions after hospitalization.  While many beneficiaries
     still use the benefit in this way, an increasing number of
     beneficiaries are receiving visits that are more directed at
     long-term care for chronic conditions. 

  Physicians tend to depend on HHAs to design plans of care,
     especially for less complex cases, and agencies as a rule have
     incentives to furnish as many visits as possible.  This
     combination can lead to the overprovision of services. 

  Medicare has reduced on-site audits and reviews so that HHAs have
     less incentive to follow Medicare rules.  The percentage of
     claims that are reviewed has decreased from over 60 percent in
     1987 to approximately 3 percent in 1994.  We have testified on a
     number of occasions that program safeguard activities are cost
     effective, returning close to $14 in savings for each $1
     invested in 1994, and cuts in payment safeguard areas translate
     into increased program losses from fraud, waste, and abuse. 
     When claims volume increases and medical review of claims
     declines, intermediaries' ability to detect and prevent
     erroneous payments is substantially lessened.  Further, even
     when claims are denied, they were often paid because the HHA
     qualified for a waiver of liability. 

  It is nearly impossible for intermediaries to assess from paper
     review alone whether a beneficiary meets the eligibility
     criteria, whether the services received are appropriate given
     the beneficiary's current condition, and whether the beneficiary
     is actually receiving the services billed to Medicare.  Coverage
     criteria, such as confined to the home or intermittent, are not
     meaningful when the HHAs are in effect the only ones monitoring
     beneficiaries. 


   MATTERS FOR CONSIDERATION BY
   THE CONGRESS
------------------------------------------------------------ Letter :8

The emphasis of Medicare's home health benefit program has recently
shifted from primarily posthospital acute care to more long-term
care.  At the same time, HCFA's ability to manage the program has
been severely weakened by coverage changes mandated by court
decisions and a decrease in the funds available to review HHAs and
the care they provide.  The Congress may wish to consider whether the
Medicare home health benefit should continue to become more of a
long-term care benefit or if it should be limited primarily to a
posthospital acute care benefit.  The Congress should also consider
providing additional resources so that controls against abuse of the
home health benefit can be better enforced. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :9

We provided HHS an opportunity to comment on our draft report, but it
did not provide comments in time for them to be included in the final
report. 


---------------------------------------------------------- Letter :9.1

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 7 days
after its issue date.  At that time, we will send copies to the
Secretary of HHS, the Administrator of HCFA, interested congressional
committees, officials who assisted our investigation, and other
interested parties.  Copies also will be made available to others on
request. 

Please contact me on (202) 512-6808 if you have any questions about
this report.  Other GAO contacts and staff acknowledgments are listed
in appendix V. 

Sincerely yours,

Sarah F.  Jaggar
Director, Health Financing and
 Policy Issues


SCOPE AND METHODOLOGY
=========================================================== Appendix I

Our work was done primarily at HCFA headquarters.  We also visited
three regional home health intermediaries (in Chicago, Illinois;
Milwaukee, Wisconsin; and Clearwater, Florida) and two HCFA regional
offices (Chicago and Atlanta) to obtain workload and performance
data, information concerning RHHI claims review operations, and an
update on HCFA's implemented and planned program changes.  We also
interviewed officials at HHS' Office of the Inspector General in
Baltimore and Atlanta. 

We reviewed pertinent laws, regulations, court decisions, and HCFA
policies to identify changes in eligibility determination and medical
review practices.  And we reviewed studies related to home health
benefit utilization and control issues. 

To identify home health growth patterns and variations in
utilization, we analyzed data from Medicare's Provider of Service and
Home Health National Claims History files.  These data include
information on all paid claims for the period 1989 through 1993.\54

We used data from the Provider of Service file to determine agency
growth through time and across geographic regions and to identify
provider ownership type.  And we used the Medicare claims data to
calculate mean and median home health visits, by total and by each
type of service, broken out by geographic area\55 and HHA ownership
types. 

While the average visits per year provides a general indication of
variations in utilization of home health services, it does not
indicate the length of each individual's episode of care nor does it
provide a picture of the intensity of services provided during this
time.  To obtain a more in-depth look at variations in practice
patterns, both across regions and among various types of HHAs, we
conducted an episode-of-care analysis for four diagnoses:  diabetes,
heart failure, hypertension, and hip fracture.\56 The first three
diagnoses were selected because they are among the most common
primary diagnoses associated with home health care.\57 Hip fracture
was selected because it is generally regarded as having a more
predictable pattern of treatment with a more finite end point.  We
selected beneficiaries with one of the above primary diagnoses who
began receiving home health services in 1992.  We then tracked
beneficiaries' visits up to 210 days after their episode start
date.\58

The principal sources of our automated data were Medicare paid claims
data systems, which are subject to periodic HCFA reviews and
examinations.  HCFA relies on the data obtained from these systems as
evidence of Medicare-covered services and expenditures and to support
its management and budgetary decisions.  For this reason, we did not
independently examine the internal and automatic data processing
controls for automated systems from which we obtained data used in
our analyses.  With this exception, we conducted our work in
accordance with generally accepted government auditing standards
between July 1994 and December 1995. 


--------------------
\54 Most recent data available. 

\55 Geographic breakdowns were based on the state of beneficiary
residence. 

\56 Because home health episodes are not clearly defined by admission
and discharge dates, we defined episodes of care as a series of home
health visits preceded and followed by a 60-day period with no
visits. 

\57 Approximately 22 percent of Medicare home health patients have
one of these four primary diagnoses. 

\58 Because some episodes of care continued indefinitely, that is,
without a 60-day gap in claims for visits, we selected a cutoff point
that allowed us to analyze utilization of most of the patients.  A
cutoff point of 210 days allowed us to look at full episodes for
99.22 percent of hip fracture beneficiaries, 94.60 percent of heart
failure beneficiaries, 94.59 percent of hypertension beneficiaries,
and 91.54 percent of diabetes beneficiaries. 


DETAILED DATA TABLES
========================================================== Appendix II



                                                                      Table II.1
                                                       
                                                       Characteristics of Medicare Home Health
                                                                    Beneficiaries


               Number of               Number of
              beneficiar            beneficiarie                    Number of                     Number of                     Number of
                     ies   Percent             s   Percent      beneficiaries     Percent     beneficiaries     Percent     beneficiaries     Percent
------------  ----------  --------  ------------  --------  -----------------  ----------  ----------------  ----------  ----------------  ----------
               1,682,139       100     1,955,170       100          2,228,701         100         2,526,978         100         2,836,912         100

Sex
-----------------------------------------------------------------------------------------------------------------------------------------------------
Female         1,073,674     63.83     1,253,685     64.12          1,433,111       64.30         1,628,771       64.46         1,834,899       64.68
Male             608,169     36.15       701,451     35.88             795,89       35.70           898,207       35.54         1,002,008       35.32

Race
-----------------------------------------------------------------------------------------------------------------------------------------------------
White          1,422,569     84.57     1,650,857     84.44          1,876,723       84.21         2,118,644       83.84         2,369,476       83.52
Black            185,959     11.05       213,174     10.90            243,891       10.94           278,888       11.04           314,829       11.10
Other             25,448      1.51        32,676      1.67             42,558        1.91            54,208        2.15            67,328        2.37

Age
-----------------------------------------------------------------------------------------------------------------------------------------------------
Less than 65      96,103      5.71       114,578      5.86            132,153        5.93           156,143        6.18           186,840        6.59
65-70            263,740     15.68       304,211     15.56            338,514       15.19           373,912       14.80           407,177       14.35
71-80            692,258     41.15       793,679     40.59            897,289       40.26         1,009,229       39.94         1,111,308       39.17
81 or older      630,038     37.45       742,702     37.99            860,745       38.62           987,694       39.09         1,131,587       39.89

Medicare eligible
-----------------------------------------------------------------------------------------------------------------------------------------------------
Aged without   1,573,317     93.53     1,824,379     93.31          2,076,599       93.18         2,345,992       92.84         2,621,323       92.40
 ESRD\a
Aged with          7,022      0.42         9,540      0.49             12,388        0.56            16,321        0.65            19.382        0.68
 ESRD
Disability        94,728      5.63       112,010      5.73            128,320        5.76           150,133        5.94           178,340        6.29
 without
 ESRD
Disability         3,616      0.21         4,659      0.24              5,647        0.25             7,550        0.30             9.632        0.34
 with ESRD
ESRD only          3,456      0.21         4,582      0.23              5,747        0.26             6,982        0.28             8,235        0.29
-----------------------------------------------------------------------------------------------------------------------------------------------------
\a End-stage renal disease. 

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

   Figure II.1:  Growth in
   Medicare Home Health Agencies,
   1989-94

   (See figure in printed
   edition.)

Source:  GAO analysis of the Medicare Provider of Service File. 

   Figure II.2:  Average and
   Median Number of Visits per
   Beneficiary per Year, 1989 and
   1993

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

   Figure II.3:  Home Health
   Visits per Medicare
   Beneficiary, 1993, in
   Descending Order

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database and the Social Security
Bulletin, Annual Statistical Supplement, 1994. 


EPISODE-OF-CARE ANALYSIS
========================================================= Appendix III

The following figures present tables (figs.  III.1 to III.4) that
show the average length of episode and the average number of visits
per episode for patients with a primary diagnosis of diabetes, heart
failure, hypertension, and hip fracture for the different types of
HHAs.  Length of episode refers to the average period of time during
which a beneficiary receives care,\59 and visits per episode refers
to the average number of home health services a beneficiary receives
during that time.  We examined episodes of care beginning during
1992.  For these episodes we tracked care throughout 1992 and 1993. 

Much variation in both lengths of episode and average number of
visits per episode can be seen among the different types of agencies
for these four diagnoses.  For example, on a national level,
proprietary agencies provided an average of 53 visits to
beneficiaries with diabetes over an average period of 64 days. 
Government agencies, on the other hand, provided an average of 29
visits to diabetic beneficiaries over a similar period of time.  The
variation in utilization between the different types of agencies is
less pronounced in cases of hip fracture, which may be regarded as an
acute condition, than in cases of diabetes, heart failure, and
hypertension, which may be regarded as more chronic conditions. 

Variations in utilization are also seen across geographic regions. 
For example, beneficiaries diagnosed with hypertension receiving care
in the Atlanta or Dallas regions received more care for longer
periods of time than beneficiaries in other regions with the same
diagnosis.  (See fig.  III.3.) HHAs in these two regions, on average,
consistently provided more care for cases of diabetes, heart failure,
and hypertension, while HHAs in the Boston region provided the most
care to beneficiaries with hip fracture.  Some of the variation
between regions may be explained by case-mix differences and
availability of alternative sources of care.  And some of the
differences are probably due to geographic variations in practice
patterns. 

Table III.5 shows the average number of two types of visits provided
to beneficiaries--skilled nursing visits and home health aide visits. 
Again, proprietary agencies provided more of these types of services
for all diagnoses.  For example, in cases of hypertension,
proprietary agencies provided almost twice as many skilled nursing
visits as voluntary agencies during a beneficiary's episode of care. 

   Figure III.1:  Average Episode
   Length and Visits per
   Episode--Diabetes, 1992-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from Medicare Standard Analytical File: 
Home Health Claims History Database. 

   Figure III.2:  Average Episode
   Length and Visits per
   Episode--Heart Failure, 1992-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

   Figure III.3:  Average Episode
   Length and Visits per
   Episode--Hypertension, 1992-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 

   Figure III.4:  Average Episode
   Length and Visits per
   Episode--Hip Fracture, 1992-93

   (See figure in printed
   edition.)

Source:  GAO analysis of data from the Medicare Standard Analytical
File:  Home Health Claims History Database. 



                              Table III.1
                
                Average Home Health Services per Episode
                      by Type of Service, 1992-93

                                                Govern  Propri  Volunt
                                                  ment   etary     ary
----------------------------------------------  ------  ------  ------
Skilled nursing visits
----------------------------------------------------------------------
Diabetes                                         16.70   30.79   19.14
Heart failure                                    12.74   19.53   14.54
Hypertension                                     12.14   18.41   13.54
Hip fracture                                      8.24   11.79    9.37

Home health aide visits
----------------------------------------------------------------------
Diabetes                                         36.11   48.71   32.86
Heart failure                                    27.51   38.81   26.13
Hypertension                                     32.48   43.28   29.48
Hip fracture                                     18.78   24.81   19.78
----------------------------------------------------------------------
Source:  GAO analysis of the Medicare Standard Analytical File,
National Claims History Data. 



(See figure in printed edition.)Appendix IV

--------------------
\59 Capped at 210 days.  Refer to app.  I for methodology. 


HCFA DEFINITION OF CONFINED TO THE
HOME
========================================================= Appendix III



(See figure in printed edition.)



(See figure in printed edition.)


GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
=========================================================== Appendix V

GAO CONTACTS

Thomas G.  Dowdal, Assistant Director, (202) 512-6588
Patricia A.  Davis, Evaluator-in-Charge, (202) 512-3011

STAFF ACKNOWLEDGMENTS

The following team members also contributed to this report:  Adrienne
S.  Friedman, Senior Evaluator; MaryEllen Fleischman, Computer
Specialist; and Mary W.  Reich, Attorney Advisor. 


RELATED GAO PRODUCTS
============================================================ Chapter 0

Medicare:  Allegations Against ABC Home Health Care (GAO/OSI-95-17,
July 19, 1995). 

Medicare:  Increased Denials of Home Health Claims During 1986 and
1987 (GAO/HRD-90-14BR, Jan.  24, 1990). 

Medicare:  Need to Strengthen Home Health Care Payment Controls and
Address Unmet Needs (GAO/HRD-87-9, Dec.  2, 1986). 

Savings Possible by Modifying Medicare's Waiver of Liability Rules
(GAO/HRD-83-38, Mar.  4, 1983). 

The Elderly Should Benefit From Expanded Home Health Care but
Increasing These Services Will Not Insure Cost Reductions
(GAO/IPE-83-1, Dec.  7, 1982). 

Response to the Senate Permanent Subcommittee on Investigations'
Queries on Abuses in the Home Health Care Industry (GAO/HRD-81-84,
Apr.  24, 1981). 

Medicare Home Health Services:  A Difficult Program to Control
(GAO/HRD-81-155, Sept.  25, 1981). 

Home Health Care Services--Tighter Fiscal Controls Needed
(GAO/HRD-79-17, May 15, 1979). 


*** End of document. ***