Medicare Home Health Care: Prospective Payment System Could Reverse
Recent Declines in Spending (Letter Report, 09/08/2000, GAO/HEHS-00-176).

Pursuant to a congressional request, GAO provided information on
Medicare home health care's recent declines in spending, focusing on:
(1) the declines in service use underlying the changes in spending; (2)
the extent of the changes in use across beneficiaries, home health
agencies (HHA), and locations; and (3) identify any implications these
new patterns of home health use have for the impact of the prospective
payment system (PPS).

GAO noted that: (1) the 48-percent reduction in Medicare home health
care spending following the Balanced Budget Act (BBA) of 1997 was due to
sharp declines in both the numbers of users and services used; (2) the
number of Medicare beneficiaries receiving home health services fell by
22 percent; from more than 100 users and services used; (3) during the
same period, the average number of home health visits received by each
user went down 44 percent; (4) changes in home health care varied across
agencies and types of users as well; (5) in nearly all instances,
declines were greatest for the types of agencies that had provided and
the patients who had used the most services in 1996; (6) there was a
similar pattern in the drop in usage across states; (7) declines in
rural areas were larger than in urban areas; (8) states that had the
highest levels of service use in 1996 had larger declines than states
where beneficiaries historically received fewer services; (9) the recent
changes in home health utilization occurred at least in part in response
to changes in Medicare's payment policies mandated by the BBA; (10)
because the new PPS payment rates are based on the historically high
utilization in 1998, even after adjusting for projected declines in
utilization, they likely will be generous compared with current use
patterns; (11) for this reason, home health agency responses to the PPS
could result in overpayments relative to services provide while
simultaneously raising Medicare spending; (12) under the PPS, Medicare
will make a single payment for each 60-day episode of home health care;
(13) the PPS will give agencies an incentive to increase the episodes of
care they provide; and (14) this, in turn, could cause total Medicare
home health spending to rise.

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

 REPORTNUM:  HEHS-00-176
     TITLE:  Medicare Home Health Care: Prospective Payment System
	     Could Reverse Recent Declines in Spending
      DATE:  09/08/2000
   SUBJECT:  Health care cost control
	     Internal controls
	     Medical services rates
	     Home health care services
	     Claims processing
	     Health resources utilization
	     Health insurance
IDENTIFIER:  Medicare Program
	     Medicare Prospective Payment System

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** 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.                                         **
**                                                              **
******************************************************************
GAO/HEHS-00-176
Appendix I: Scope and Methodology

32

Appendix II: Medicare Home Health Users, by State of Residence, Calendar
Years 1994, 1996, and 1999

34

Appendix III: Average Visits per Medicare Home Health User, by
State of Residence, Calendar Years 1994, 1996,
and 1999

36

Appendix IV: Comments From the Health Care Financing
Administration

38

Table 1: Average Number of Home Health Visits per User by Visit
Type, 1994, 1996, and 1999 14

Table 2: Medicare Home Health Users, by Utilization Level, 1994,
1996, and 1999 17

Table 3: Average Visits per User, by Level of Agency Service
Provision, 1994, 1996, and 1999 19

Table 4: Home Health Service Utilization Changes in Rural and
Urban Areas, 1994, 1996, and 1999 23

Table 5: History of Medicare Home Health Care Spending Growth,
1985-99 24

Figure 1: Home Health Users per 1,000 FFS Beneficiaries, 1994,
1996, and 1999 13

Figure 2: Proportion of Home Health Visits by Visit Type, 1996
and 1999 15

Figure 3: Proportion of Home Health Beneficiaries, by Utilization
Level, 1994, 1996, and 1999 18

Figure 4: Average Visits per User, by HHA Ownership, 1994, 1996,
and 1999 20

Figure 5: Average Visits per User, Highest- and Lowest-Utilization
States, 1999 22

BBA Balanced Budget Act of 1997

FFS fee-for-service

HCFA Health Care Financing Administration

HHA home health agency

HHS Department of Health and Human Services

HIPAA Health Insurance Portability and Accountability Act of 1996

IPS interim payment system

MedPAC Medicare Payment Advisory Commission

MSA metropolitan statistical area

PPS prospective payment system

SAF Standard Analytical File

Health, Education, and
Human Services Division

B-286001

September 8, 2000

The Honorable William M. Thomas
Chairman, Subcommittee on Health
Committee on Ways and Means
House of Representatives

Dear Mr. Chairman:

Until 1997, home health care was one of Medicare's fastest growing benefits,
with expenditures rising at an average annual rate of 30 percent between
1988 and 1997. Both the number of beneficiaries and the amount of services
they received increased, changing what had been a short-term,
posthospitalization benefit to one that encompassed long-term care and
vastly different service patterns across beneficiaries, agencies, and
geographic areas. In addition to concerns about Medicare expenditures, this
wide variation in use raised issues about the appropriateness of services
being provided and the lack of standards for care. To control home health
spending, in 1997 the Congress mandated that the Health Care Financing
Administration (HCFA), the agency responsible for administering the Medicare
program, implement several payment and other policy changes to the home
health benefit.1 These changes were to culminate in the implementation of a
prospective payment system (PPS) for home health services in October 1999.

Delayed until October 2000, the PPS will incorporate payment rates based on
1998 home health spending and utilization data--the latest available
information--adjusted downward to reflect projected utilization in 2001.2
However, home health spending and use have changed substantially in recent
years. Medicare home health spending dropped from a peak in 1997 of $18.3
billion to $9.5 billion in 1999. Concerned about the decline in service use
and its implications for Medicare payment policy, you asked us to (1)
examine the declines in service use underlying the changes in spending; (2)
determine the extent of the changes in use across beneficiaries, home health
agencies (HHA), and locations; and (3) identify any implications these new
patterns of home health use have for the impact of the PPS. To do this work,
we analyzed HCFA home health claims and provider data from 1994, 1996, and
1999. We did our work in accordance with generally accepted government
auditing standards between April and July 2000. (For a detailed discussion
of our scope and methodology, see app. I.)

The 48-percent reduction in Medicare home health care spending following the
BBA was due to sharp declines in both the numbers of users and services
used. The number of Medicare beneficiaries receiving home health services
fell by 22 percent, from more than 100 users per 1,000 fee-for-service (FFS)
beneficiaries in 1996 to 80 users per 1,000 beneficiaries in 1999.3 During
the same period, the average number of home health visits received by each
user went down 44 percent, to 41 visits in 1999. The decline was not uniform
across service types. Unskilled home health aide visits (including help with
personal care and simple dressing changes) dropped more than skilled service
visits (which include skilled nursing, therapies, and medical social
services). While changes in payment policies undoubtedly played a big role
in curtailing service use, other factors may have also contributed to the
decline, such as increased efforts to combat Medicare fraud and abuse.

Changes in home health use varied across agencies and types of users as
well. In nearly all instances, declines were greatest for the types of
agencies that had provided and the patients who had used the most services
in 1996. For example, the number of patients receiving more than 150 visits
fell 67 percent between 1996 and 1999, compared with a 22-percent drop
across all users. Similarly, proprietary agencies, which provided the
highest number of visits per user in 1996, reduced their service provision
by 47 percent between 1996 and 1999, compared with a 37-percent reduction
for not-for-profit and government agencies. There was a similar pattern in
the drop in usage across states. States that had the highest levels of
service use in 1996 had larger declines than states where beneficiaries
historically received fewer services. While the more than threefold
difference in utilization continued between the highest- and lowest-use
states in 1999, there were fewer states at the extremes. Declines in rural
areas were larger than in urban areas. Although variation in use across
states and agencies narrowed over this time, states with high use in 1996
continued to have rates nearly double the average visits per user of the
rest of the country in 1999, and high-visit agencies still provided half
again as many visits as the national average.

The recent changes in home health utilization occurred at least in part in
response to changes in Medicare's payment policies mandated by the BBA.
Because the new PPS payment rates are based on the historically high
utilization in 1998, even after adjusting for projected declines in
utilization, they likely will be generous compared with current use
patterns. For this reason, home health agency responses to the PPS could
result in overpayments relative to services provided while simultaneously
raising Medicare spending. Under the PPS, Medicare will make a single
payment for each 60-day episode of home health care. Some agencies may
respond to the high payments by increasing services provided to
beneficiaries. Others may maintain their reduced service levels, resulting
in overpayments relative to the services delivered within the episode. At
the same time, the PPS will give agencies an incentive to increase the
episodes of care they provide. This, in turn, could cause total Medicare
home health spending to rise.

In an earlier report, we outlined actions that HCFA should take to protect
beneficiaries, HHAs, and the Medicare program from possible negative effects
of the PPS.4 We recommended, and HCFA agreed, that the PPS will need to be
evaluated and refined periodically and that utilization monitoring and
medical review of claims will be critical to ensuring that HHAs do not stint
on care or provide unnecessary services. We also recommended that the PPS
should be modified to incorporate a risk-sharing arrangement, which would
limit aggregate HHA Medicare gains or losses, but HCFA believed that such
modification was not necessary. However, we believe that the substantial
changes in home health utilization that have occurred since the BBA lend
additional support for a risk-sharing approach. Therefore, in this report we
suggest that the Congress consider instructing HCFA to adopt risk sharing
under the home health care PPS. HCFA agreed that risk sharing is one option
to address concerns raised about the PPS but said it prefers other methods.
We continue to support implementing a risk-sharing arrangement along with
the PPS.

Medicare's home health care benefit enables certain beneficiaries with
post-acute-care needs (such as recovery from joint replacement) and chronic
conditions (such as congestive heart failure) to receive care in their homes
rather than in other settings. To qualify for home health care, a
beneficiary must be confined to his or her residence ("homebound");5 require
intermittent skilled nursing, physical therapy, or speech therapy; be under
the care of a physician; and have the services furnished under a plan of
care prescribed and periodically reviewed by a physician. If these
conditions are met, Medicare will pay for part-time or intermittent6 skilled
nursing; physical, occupational, and speech therapy; medical social
services; and home health aide visits.7 The benefit allows for an unlimited
number of visits, provided the coverage criteria are met. Beneficiaries are
not liable for any coinsurance or deductible.

Between 1990 and 1997, Medicare home health payments grew annually at a rate
of more than three times that of spending growth for the entire Medicare
program. This increase was due primarily to a steady rise in the proportion
of beneficiaries receiving home health care and in the number of visits per
person served. The number of home health users per 1,000 beneficiaries
increased from 57 to 109, and the average number of visits per user doubled
from 36 to 73 during this period. An increase in payments per visit
accounted for only a small share of the overall growth.

Originally, Medicare imposed annual limits on the number of home health care
visits covered for each beneficiary. The limitation on visits was removed by
the Omnibus Reconciliation Act of 1980,8 but utilization did not increase
appreciably because of HCFA's stringent interpretation of the coverage and
eligibility criteria. A court case challenged HCFA's interpretation, and the
decision resulted in broadened coverage guidelines for home health care,
allowing more beneficiaries to qualify for more visits.9 The benefit then
was transformed from one focused on patients needing short-term care after a
hospitalization to one that also serves patients with chronic conditions
needing longer-term care.

At the same time that much of this growth occurred, program controls were
essentially nonexistent. Few claims were subject to medical review, and
virtually all were paid. In 1986 and 1987, over 60 percent of home health
claims were reviewed, but by 1995, claims reviewed had declined to about 1
percent. As a result, utilization after 1987 is increasingly likely to
reflect a degree of inappropriate service use. Our prior investigations
found a pattern of payments for "questionable or improper" services.10 More
recently, the Department of Health and Human Services (HHS) Inspector
General also documented that some of the care provided lacked supporting
documentation required to determine medical necessity.11

Historically, most home health users received few visits, and a small
proportion of longer-term users received the majority of Medicare-funded
visits. According to the Medicare Payment Advisory Commission (MedPAC), 51
percent of home health care recipients received fewer than 30 visits and
accounted for 9 percent of all home health visits in 1996. By contrast, 15
percent of users had 150 visits or more, accounting for 59 percent of all
Medicare home health visits that year. Approximately one-third of the
beneficiaries in this latter group received over 300 visits.12 In addition,
short-term patients appeared to use a different mix of visits than did
longer-term patients. MedPAC reported that in 1996 only 6 percent of all
visits provided to short-term users--those who received nine or fewer
visits--were for aide services; skilled nursing care comprised over 75
percent of their total visits. By contrast, about 56 percent of the visits
for beneficiaries who had 100 visits or more were for home health aide
services.

There also was marked variation in home health use across geographic areas.
For example, Medicare home health users in Maryland received an average of
37 visits in 1997, with an average payment per user of $3,088. In that same
year, users in Louisiana received an average of 161 visits each, with an
average Medicare payment per user of $9,278. This wide variation in use
persisted even after controlling for patient diagnosis. Patterns of care
also differed across agency ownership and type.13 For-profit HHAs tended to
deliver more visits per beneficiary than other types of HHAs and to provide
more aide visits. For example, in 1993, for-profit HHAs provided an average
of 69 home health aide visits per beneficiary, compared with 43 and 48
visits from voluntary and government HHAs, respectively.14 Such variation
could be due to a variety of factors, including provider responses to
financial incentives, differences in patient needs, regional practice
patterns, and states' varying Medicaid coverage and eligibility policies.

Assessing whether the variation in service provision has been appropriate is
difficult. Because no agreed-upon standards exist for what constitutes
necessary or appropriate home health care, it is not clear when home health
care is warranted, how many services should be provided, or when services
should be discontinued. Many home health users have chronic and multiple
needs, so the care for a particular condition may overlap with care for
another. Furthermore, even the most basic unit of service--the visit--is not
specifically defined.

Beginning in 1995, several regulatory policies were initiated to reduce
fraud and abuse within the home health industry, which could have affected
home health use and spending. Operation Restore Trust, launched in 1995,
employed a number of approaches to uncovering fraud, including the use of
interdisciplinary teams to review individual HHAs that billed Medicare for
unusually large numbers of services. The Health Insurance Portability and
Accountability Act of 1996 (HIPAA) also contained measures to control fraud
and abuse by HHAs. For example, it stipulated that any physician who falsely
certifies a patient as eligible for home health services is liable for a
civil monetary penalty. HIPAA also provided more funding for claims review
and other safeguard activities by Medicare's claims processing contractors.
However, the proportion of claims reviewed did not increase substantially.
In January 1998, HCFA announced plans to increase the number of claims
reviewed to about 1.3 percent, far short of the peak levels in the
mid-1980s. As part of the changes included in the BBA, coverage was
eliminated for persons whose only skilled service need was venipuncture (the
drawing of blood).

Before the BBA, HHAs were paid on the basis of their costs, up to
preestablished per-visit limits. In 1996, these limits ranged from $46 for
home health aide visits to $91 for skilled nursing visits, to a high of $130
for medical social services.15 While payments varied by the type of visit,
there was no definition of what actually constituted a home health visit,
such as the time spent with the patient or the services provided. There were
no incentives to control the volume of services delivered, and as a result,
HHAs could enhance their revenues by providing more beneficiaries with more
visits.

The BBA mandated substantial changes to Medicare's method of paying for home
health services. Beginning October 1, 1997, HHAs were paid under an interim
payment system (IPS), which incorporated tighter per-visit cost limits than
previously in place and subjected each agency to an annual Medicare revenue
cap, which is the product of a per-beneficiary amount and the number of
patients it served. The per-beneficiary amount is a blend of each agency's
historical average payments for treating a Medicare beneficiary and a
regional or national average amount.16 To ensure that Medicare payments
under the IPS cover its costs, an HHA needs to keep the average cost of its
visits below the per-visit limits and keep its average cost per Medicare
beneficiary below its per-beneficiary amount. For agencies with previously
higher per-visit costs or that provided more visits per user, adjustments to
the IPS may involve delivering visits more efficiently, changing the mix or
reducing the number of visits provided to each user, increasing the
proportion of lower-cost patients it treats, or some combination of these
strategies.

Beginning in October 2000, HHAs will be paid under the PPS. An agency will
receive a single payment for each 60-day episode of care for a Medicare
beneficiary, regardless of the services actually delivered during the
period.17 There is no limit on the number of episodes a beneficiary may
receive. A base payment will be adjusted to reflect patient characteristics
that have been shown to affect service use. Payments for patients expected
to use the most services in an episode will be over 5 times the payment for
patients expected to use the fewest services. Each episode payment also will
be adjusted for differences in labor costs across geographic areas. HCFA
will make outlier payments for certain extremely high cost episodes. The BBA
required HCFA to set payment levels so that Medicare home health
expenditures would be equivalent to what would have been spent under the
IPS, with those limits reduced by 15 percent. This 15-percent reduction has
been delayed until October 1, 2001, and the Secretary of Health and Human
Services must report to the Congress within 6 months of implementation of
the PPS on the need for the 15-percent or other reduction.18

In previous work on the home health PPS, we noted several concerns about
HCFA's proposed, and now final, design.19 Given the wide variation in
service use, the 60-day unit of payment may not be suitable for all
patients. Furthermore, the adjustments to the episode payment may not
adequately account for differences in patient needs and, because the
adjustments rely heavily on what services are provided to patients, they may
be open to manipulation by agencies. Because of uncertainties about the
effects of the PPS on beneficiaries, agencies, and the program, we
recommended that a

risk-sharing arrangement, which limits the losses and gains a provider can
experience over a period of time, be added to the PPS.20 HCFA did not agree
with this recommendation, indicating that risk sharing was not needed, given
the adjustments included in the PPS, and that risk sharing would make the
PPS difficult to implement. While we are sympathetic to HCFA's concerns and
do not believe that the PPS should be delayed in order to implement risk
sharing, we nevertheless remain convinced that the magnitude of potential
excessive payments to some HHAs and large losses for others warrants this
added complexity. We also recommended, and HCFA concurred, that the PPS be
modified as appropriate as experience is gained under the PPS. To address
concerns about the appropriateness of potential service reductions within
episodes and whether each episode of care a beneficiary receives is
medically necessary, we recommended that adequate resources be devoted to
utilization monitoring and medical review. In agreeing with this
recommendation, HCFA outlined the various activities it has planned to
ensure that the data agencies submit are accurate, that its payments to
agencies are appropriate, and that timely utilization data is readily
available for possible PPS refinements.

Since peaking in 1997, Medicare home health expenditures have declined
rapidly so that by 1999 spending was about the same as it was in 1993. The
drop in spending reflected a decrease in home health service use, both in
the number of beneficiaries using home health care and the number of visits
provided to each user. The patterns of decline have resulted in a benefit
that involves a larger proportion of skilled services (skilled nursing and
therapies) and considerably fewer home health aide services. The fall in
visits per user is consistent with the objectives of the IPS but exceeds the
reduction necessary for some agencies to stay within the limits of the IPS.
The reduction in the number of home health users may be due in part to
initiatives to combat fraud and abuse and to some agencies' overreaction to
the IPS, which may have led them to avoid certain types of high-cost
patients.

Services in 1999

After having been a major driver in home health spending growth from the
early 1980s through 1997, the number of FFS beneficiaries receiving home
health visits has decreased. The percentage of FFS beneficiaries getting
home health care fell 22 percent between 1996 and 1999. In 1996, more than
100 of every 1,000 FFS beneficiaries received home health care, compared
with 80 in 1999 (see fig. 1). This decline, which followed a 15-percent
increase in home health users between 1994 and 1996, brought the number of
users in 1999 to below 1994 levels.

Source: GAO analysis of HCFA home health claims data.

The number of visits per home health user also dropped substantially over
this period. In 1999, the average home health user received 41 visits,
compared with 73 visits in 1996 (see table 1). The average number of visits
per user decreased for all visit types, although the amount of the decline
varied significantly. The most notable drop was in home health aide use. In
1999, home health aide users received, on average, about half the number of
home health aide visits that they received in 1996, 37 compared with 73
visits. Users of skilled nursing services in 1999 received almost one-third
fewer skilled nursing visits than they did in 1996. Reductions in therapy
visits were more modest than home health aide or overall average declines.

                          Average visits per
                                                         Percentage change
                          visit-type usera
                          1994      1996     1999        1994-96   1996-99
 All visits               65        73       41          12%       -44%
 Skilled nursing          29        32       22          10%       -31%
 Home health aide         64        73       37          14%       -49%
 Physical therapy         13        14       11          8%        -21%
 Speech therapy           12        11       8           -8%       -27%
 Occupational therapy     8         8        7           0         -13%
 Medical social
 services                 3         3        2           0         -33%

aNumbers in these columns reflect the number of visits of a particular type
received by beneficiaries who used any of that visit type. Thus, the
averages by visit type do not sum to the average of all visits.

Source: GAO analysis of HCFA's home health claims data for 1994, 1996, and
1999.

Because of the disproportionate reduction in aide visits and overall drop in
use, post-acute-care services are becoming a more important component of the
Medicare home health benefit. Compared with previous years, the average user
in 1999 is more likely to receive therapy services, and less likely to
receive home health aide services. Nearly one-half of all home health users
received physical therapy visits in 1999, up more than 20 percent over 1996.
By contrast, 38 percent of users received home health aide services in 1999,
which is 22 percent below 1996 levels. Furthermore, aide visits in 1999
comprised a smaller share of all visits (34 percent), which is similar to
the share of aide visits in 1987 (see fig. 2). Skilled nursing visits have
become a larger share of all visits, comprising nearly half of total visits
in 1999, and therapy services have increased their proportion of total
visits as well. Combined, skilled services made up two-thirds of all visits
in 1999, compared with half of all visits in prior years. These shifts are
consistent with care that reflects more short-term, post-acute use rather
than care for longer-term chronic conditions.

Source: GAO analysis of HCFA's home health claims data for 1996 and 1999.

Changes

The reduction in the visits per user is consistent with agency incentives
under the IPS to keep average per-user costs below the per-beneficiary
amount, yet it appears that some HHAs may have overreacted to the IPS. Some
agencies reduced the number of visits provided to beneficiaries. In
addition, some agencies modified their admitting practices to lower the
number of beneficiaries likely to need longer-term and more costly services.
Our previous work found that HHAs said they had increased their efforts to
identify the anticipated service needs of prospective patients; were more
reluctant to accept longer-term, expensive patients; and stepped up their
monitoring of patients' needs for timely discharge.21 These results are
consistent with a MedPAC-sponsored survey in which some HHAs reported that
because of the IPS, they were no longer taking Medicare patients they
previously would have admitted.22 The types of patients HHAs were most
likely to report they no longer admitted or discharged sooner included
longer-term, chronic, and diabetic patients, all of whom are generally
associated with longer-term utilization and heavy use of aide services.

Some agencies responded to the IPS by reducing per-beneficiary costs more
than would have been necessary to remain under the per-beneficiary amounts.
The per-beneficiary amounts, which were based on 1994 cost data and updated
annually, essentially used service levels in that year as the standard.
However, home health service use in 1999 dipped below 1994 levels. The
average home health user received 41 visits in 1999, compared with 65 visits
in 1994. Moreover, the IPS had no limitations on the number of beneficiaries
that an agency could serve and be paid by Medicare. Yet, the proportion of
FFS beneficiaries receiving home health services in 1999 was 10 percent
lower than in 1994. Other policy initiatives, such as Operation Restore
Trust, which increased scrutiny of claims, and stronger physician
certification requirements, may have prompted HHAs to be more vigilant in
their admissions and discharge processes. In our previous work on agency
closures, we found that the caseload of agencies that had stopped serving
Medicare beneficiaries included patients who were ineligible for Medicare
home health care.23 In a study of four states, HHS' Office of Inspector
General found that improper or highly questionable home health services
dropped from 40 percent of the total in 1995 to 19 percent of services in
1998.24 In the MedPAC survey, 77 percent of agencies reported an increased
reluctance on the part of physicians to refer Medicare patients for
services. HHAs told us that a drop in physician referrals and the
elimination of venipuncture as a qualifying service for home health care
reduced agency caseloads.

Experienced Largest Declines

The historically wide variation in home health service use across
beneficiaries, types of providers, and geographic areas has narrowed
substantially because of disproportionate declines in utilization among the
highest users in these categories. The number of longer-term beneficiaries
receiving 150 or more home health visits per year dropped by two-thirds,
compared with a 22-percent reduction in all users. High-visit HHAs accounted
for a disproportionate share of the overall utilization decline after 1996,
as well as a greater share of the increase before 1996. Among HHAs that
historically delivered the most services, the average number of visits per
user decreased more than for all agencies. And the states with the highest
utilization experienced greater declines after 1996 compared with the rest
of the country, although wide variation in use persists. While rural areas
experienced greater reductions compared to urban areas in the proportion of
beneficiaries using services, rural users continue to receive more visits.

Short-Term Use

Long-term home health users, those receiving 150 or more home health visits
per year, declined dramatically between 1996 and 1999, both in absolute
numbers and as a proportion of all home health users. After substantial
increases, the number of high-use beneficiaries per 1,000 FFS enrollees
dropped 67 percent from 1996 to 1999, three times the decline among all
users (see table 2). As a result, high-use beneficiaries as a proportion of
total users fell by half over this period (see fig. 3). Conversely, the
number of beneficiaries receiving fewer than 10 visits increased, and their
share of all home health users rose from 22 to 31 percent.

                            Home health users per
 Utilization level          1,000 FFS enrollees          Percentage change
 (visits/year)
                            1994     1996     1999       1994-96   1996-99
 Low (1-9)                  20       23       24         15%       4%
 Medium (10-149)            58       65       51         12%       -22%
 High (150+)                11       15       5          36%       -67%
 Total                      89       102      80         15%       -22%

Source: GAO analysis of HCFA home health claims data.

1996, and 1999

Source: GAO analysis of HCFA home health claims data.

The difference in utilization across HHAs has declined since 1996, but
substantial variation continues. High-visit HHAs, the 20 percent of HHAs
with the highest average number of visits per user in 1996, experienced
greater early increases, followed by larger declines than other agencies. In
1996, these HHAs provided an average of 151 visits per user, a 30-percent
increase over 1994 levels, but this fell by over half to 67 visits in 1999
(see table 3). By contrast, historically low-visit HHAs continued to reduce
service provision between 1996 and 1999 by 15 percent. Given their steeper
rate of decline, high-visit HHAs accounted for a disproportionate share of
the total drop in visits, even after controlling for the mix of HHAs
participating in the Medicare program.25 Among HHAs serving Medicare
beneficiaries in 1994, 1996, and 1999, over one-third of the recent
reduction in visits was attributable to high-visit HHAs.

                             Average visits per
 HHA utilization level in    patient                     Percentage change
 1996
                             1994     1996    1999       1994-96   1996-99
 Low                         28       27      23         -4%       -15%
 Medium                      54       60      35         11%       -42%
 High                        116      151     67         30%       -56%

Note: Only HHAs open all 3 years (1994, 1996, and 1999) were included in
this analysis. The 20 percent of HHAs with the highest visits per user in
1996 were categorized in the high group, with the lowest 20 percent defined
as low, and the middle 60 percent defined as medium.

Source: GAO analysis of HCFA home health claims data.

We found similar patterns of changes in utilization across agency ownership
categories (see fig. 4). Between 1994 and 1996, for-profit HHAs increased
their service provision more than other agencies, and then reduced visits by
almost half between 1996 and 1999.

Source: GAO analysis of HCFA home health claims data.

States Persists

The wide range of utilization among HHAs is likewise seen across states, as
are the substantial changes in use over time. The difference in visits per
user between the highest- and lowest-utilization states has increased since
1994 (see app. II). In 1999, there was over a fourfold difference in average
visits per user between the lowest-utilization state (Oregon) and the
highest (Louisiana) (see fig. 5), and over a threefold difference among
states in the number of home health users per 1,000 FFS Medicare
beneficiaries (see app. III). Although the range in utilization remains
large across states, there were fewer states with extremely high use levels
in 1999 than there were in 1996. From 1994 to 1996, utilization in the eight
states with the highest usage rates in 1996 grew at double the rate of other
states.26 By 1999, visits per user in these states had fallen by 47 percent
from 1996 levels, compared with a 39-percent decrease for the rest of the
country. These same eight states also had a greater reduction in the number
of users per 1,000 FFS Medicare beneficiaries (33 percent, compared with a
19-percent reduction for the rest of the United States).

1999

Source: GAO analysis of HCFA home health claims data.

Higher Than for Urban Users

In 1999, 75 out of 1,000 Medicare beneficiaries in rural areas received home
health services, compared with 82 beneficiaries per 1,000 who lived in urban
areas.27 The number of home health users in rural areas declined more than
in urban areas between 1996 and 1999, although the number of visits rural
users received remained higher (see table 4). Rural beneficiaries on average
received 15 percent more visits than their urban counterparts, primarily
because of more home health aide visits.

                       Users per 1,000 Medicare      Average visits per
 Beneficiary           FFS enrollees                 user
 residence
                       1994     1996     1999        1994   1996    1999
 Urban                 88       101      82          62     70      40
 Rural                 92       105      75          74     83      46

Source: GAO analysis of HCFA home health claims data.

Increase Program Spending

In the past, home health service provision has fluctuated in response to
changes in Medicare's payment and coverage policies. The PPS incorporates
new incentives, and agencies are likely to respond by modifying how they
care for Medicare beneficiaries in both the services provided within an
episode of care and the number of episodes provided to each patient. HHA
behavior could result in substantial overpayments relative to the level of
services actually delivered and huge increases in Medicare home health
spending. Adequate controls are necessary to mitigate these risks.

Incentives

HHAs appear to have responded to previous Medicare payment incentives by
changing their patterns of service delivery (see table 5).28 In 1985,
legislation more than doubled HCFA's funding for home health claims review
after which Medicare outlays grew only 1 percent annually through 1988.
Restrictions on coverage were relaxed as a result of the Duggan v. Bowen
lawsuit decision in 1989, followed by spending growth at an annual rate of
30 percent. Utilization peaked in 1997 when BBA changes were implemented.
Under the IPS, agencies have faced strong financial incentives to control
the average number of visits and the average cost of care delivered to their
patients. Once again, Medicare policies appear to have affected the delivery
of services, as spending decreased 32 percent between 1998 and 1999.

 Period           Annual spending    Significant change shaping service use
                  rate
                                     Tightened interpretation of coverage
 1985−88    1%                 criteria; increased emphasis on
                                     medical review of home health claims.
                                     Loosening of coverage criteria allowed
 1989−97    30%                more beneficiaries to receive more
                                     services.
                                     IPS limited per visit payments and
                                     limited aggregate agency payments;
 1998−99    -32%               heightened scrutiny of claims; changed
                                     qualifying criteria for "skilled"
                                     services.

Sources: HCFA, A Profile of Medicare Home Health Chart Book (Baltimore, Md.:
Nov. 1999), and GAO analysis of Medicare home health spending data from
HCFA's Office of the Actuary.

The PPS, to be implemented October 1, 2000, will incorporate further major
policy changes for Medicare that could have a profound effect on home health
service use. Instead of per-visit limits and controls on the average costs
of treating Medicare patients, HHAs will receive one payment for each 60-day
episode of care, regardless of the actual services provided. Agencies will
be rewarded financially for keeping their per-episode costs below the
payment rate and thus will have a strong incentive to reduce the number of
visits provided during an episode and to shift to a less costly mix of
visits. Historical responses to policy changes suggest that agencies are
likely to respond to the incentive to reduce services provided within an
episode and to increase the number of episodes they deliver.

Be Excessive for Others

The PPS will use payment rates based on 1998 home health spending and
utilization data. Although by 1998 home health care utilization had already
started falling from its peak in 1997, the PPS rates will still be based on
an average experience that is higher than current usage. Thus, the episode
payments could present an ample cushion for many agencies. Not only could
the episode amounts allow for more visits during a 60-day period than the
average agency is now providing, but because there is no limit on the number
of episodes an HHA may provide to a patient, agencies may revert to treating
beneficiaries for longer periods.

The adjustments HHAs may make to adapt to the episode-based PPS will depend
on their current service patterns. Agencies that have continued to incur
expenses above the national average will be pressured to lower their episode
costs, which is likely to require decreasing the number of visits provided
or shortening their duration. Agencies with below-average costs, probably
reflecting fewer average visits for a given episode, will be rewarded
financially under the PPS. Some of these agencies may increase service
provision. Others, however, may choose to maintain their relatively low
expenses (and probably low visit levels) or reduce services even further,
thereby increasing profits. In such cases, the PPS likely will pay too much
relative to the services delivered in each episode.

We noted in our April 2000 report that the adjusters to the basic payment
rate to reflect patients' needs are more sensitive to differences in the
amount of therapy services provided than to differences in patients'
clinical indicators.29 We remain concerned that the financial benefit of
providing more therapy services to receive higher payments may interfere
with the goal of the PPS to provide payments that support efficiently
delivered care that meets patients' needs.

Spending

Agencies can enhance their revenues by serving more longer-term users and
extending the length of time they serve patients in order to be paid for
additional episodes. For some patients, the scheduling of visits could
determine whether an agency is paid for one episode or two. In addition, the
design of the PPS allows agencies to receive a full episode payment for a
small number of visits. While the episode payment is based on an average of
27 visits, agencies can receive an episode payment if they provide as few as
5 visits.30 As 16 percent of episodes in 1998 consisted of one to four
visits, adding only a few visits would allow the agency to receive the full
episode payment.

The budgetary implications of growth in the number of episodes are
considerable. HCFA has projected 5.3 million full episodes for 2001, almost
13 percent fewer than in 1998. Because the industry has historically
responded to changes in payment policy in ways that enhanced agency
revenues, this projection may not adequately anticipate potential service
growth in response to the PPS' strong incentives. If the number of episodes
in 2001 exceeds HCFA's projection by as little as 5 percent, program
expenditures could be roughly half a billion dollars more than projected.

HCFA has included three mechanisms under the PPS to counter the incentives
to stint on services and generate additional episodes. First, HCFA will
curtail gross overpayments for very low-service episodes by paying on a
per-visit basis (through the low-utilization payment adjustment) when fewer
than five services are provided in a 60-day period. Second, adjustments to
the payment for an episode can be made if a significant change in patient
condition occurs. The episode payment can be raised on a prorated basis if a
patient's condition deteriorates or if therapy service provision increases
after the beginning of an episode; the payment can be decreased on a
prorated basis if the home health agency reports significant improvement in
a patient's condition during the course of care that changed the required
services. The third mechanism is a requirement for medical review of a
portion of claims to detect underservice and unnecessary episodes. For
fiscal year 2001, HCFA has targeted just over 2 percent of home health
claims for review, even though provider incentives will be different than
under previous payment methods. HCFA has characterized its planned
utilization monitoring and medical review activities as similar to reviews
conducted before the implementation of the PPS, when the payment incentives
were different.

It is unclear whether these controls, in combination with HCFA's planned
activities and continued anti-fraud-and-abuse activities, will be sufficient
to counter incentives to provide fewer services within an episode and to
generate additional episodes, especially given agencies' historical ability
to quickly respond to such incentives. Furthermore, the lack of standard
definitions of appropriate home health care will confound efforts to
identify instances of excessive use or inadequate care.

The fluctuations in Medicare home health use suggest that agencies will
continue to respond to their payment and policy environments by changing the
volume and mix of the services they provide to Medicare beneficiaries.
Indeed, the PPS is based on the premise that appropriate financial
incentives cause HHAs to deliver services more efficiently. Previously, we
expressed concern that the wide, unexplained variation in service use and
inadequate patient-level payment adjusters could result in substantial
underpayments to some agencies and for some types of patients and
overpayments for others under a PPS based on national average costs. After
examining HHA responses to the IPS and the basis for Medicare's PPS, we
continue to believe that additional protections for beneficiaries, agencies,
and the program need to be incorporated into the payment mechanism through a
risk-sharing arrangement that limits the aggregate losses or gains for each
agency. Risk sharing would insulate agencies from extreme financial losses,
protect beneficiaries from impaired access or inadequate care, and shield
Medicare from burgeoning expenditures. HCFA disagreed with our
recommendation that it implement a risk-sharing mechanism in conjunction
with the PPS. HCFA argued that doing so would complicate the administration
of the payment system and that the mechanism was not needed because certain
features of the PPS, such as the case mix adjustment mechanism and the
potential for unlimited episodes, would adequately protect beneficiaries and
the program. We acknowledge HCFA's concerns that a risk-sharing arrangement
adds administrative complexity to the PPS, but believe that the
uncertainties about appropriate payment levels, as well as the lack of
consensus regarding what constitutes adequate treatment, require this
payment system modification. Further, we continue to have reservations about
the adequacy of some of the features of the PPS that HCFA believes will
offer protections from any unintended consequences of the new payment
system. A risk-sharing arrangement would minimize excessive payments to some
agencies and extreme losses for others, and it would moderate incentives to
underserve beneficiaries and inappropriately change treatment patterns.
Given the number of agencies and beneficiaries affected, and the potential
effect on Medicare expenditures, we believe the added complexity engendered
by risk sharing is warranted.

As service use changes in response to the PPS, we and HCFA agree that it
will be important to refine the payment system. The rates will need to be
evaluated to ensure that HHAs are not overpaid relative to the services
provided. The Secretary of Health and Human Services' report, due by April
1, 2001, that will evaluate the need for a 15-percent payment reduction will
be an important first step in assessing the adequacy of current payment
rates. Ongoing refinements of the payment system?including reconsideration
of the episode length, the average payment rate, and the patient-level
payment adjusters?will continue to be needed, to account for changes in HHA
service delivery and beneficiary needs.

Even as the system is improved, however, payment mechanisms alone may not be
adequate to ensure appropriate service use. As we previously recommended and
as was agreed to by HCFA, sufficient resources must be devoted to ensuring
that any service reduction within episodes is appropriate and that each
episode of care a beneficiary receives is medically necessary.

Given the uncertainties for beneficiaries, HHAs, and the Medicare program
associated with the home health agency PPS, we believe that the Congress
should consider requiring HCFA to implement a risk-sharing arrangement under
the PPS to moderate excessive HHA gains or losses as soon as practicable. We
believe that a risk-sharing arrangement would offer protection to Medicare
beneficiaries, home health agencies, and the Medicare program from any
unintended consequences of the home health PPS.

In commenting on a draft of this report, HCFA found our analysis useful in
understanding trends in home health utilization and payment trends under the
IPS. HCFA concurred that many home health agencies may have over-reacted to
the IPS by curtailing service provision after 1997 more than was necessary.
HCFA also agreed that refinements to the PPS will be an ongoing activity
based on HHA behavior and reiterated its commitment to monitor provider
responses under the new system to ensure beneficiary access to needed
services. While HCFA agrees with us that risk sharing in conjunction with
the PPS is one option to moderate inappropriate behavior, it continues to
have reservations about implementing such a provision. HCFA also provided
technical comments, which we incorporated in the final report as
appropriate.

HCFA raised concerns about a risk-sharing provision. First, it believes that
a risk-sharing arrangement that limits HHA profits or losses through a
comparison of Medicare payments with Medicare costs undermines the
incentives of the PPS. HCFA said that this would encourage HHAs to increase
their costs--potentially in ways unrelated to patient care--thus rewarding
provider inefficiency. HCFA also said that costs are not the best measure of
whether patients' service needs are being met. Further, it is concerned that
relying on costs in the payment system perpetuates the need for an elaborate
cost settlement reconciliation system. Because of these concerns, HCFA
prefers a visit-based measure of utilization to correct inappropriate
behavior. HCFA is also concerned that HHAs need time to adapt to the new
payment system and therefore that it would be premature to immediately
implement risk sharing before HHA responses to the PPS can be evaluated and
before PPS adjustments, if any, are made on the basis of observed behavior.
Further, HCFA believes that HHAs compete for patients on the basis of
service delivery and that competition among HHAs will be a primary driver of
agency behavior and performance under the PPS.

We agree with HCFA that a visit-based approach to moderating inappropriate
behavior would improve the current PPS, but we continue to believe that a
risk-sharing arrangement based on a comparison of Medicare payments and
costs is preferable. First, it offers HHAs more flexibility than a
visit-based approach with respect to the services they provide under the
PPS, because HHAs could balance visit costs, mix, and volume in meeting
beneficiary care needs and keeping their costs in line with Medicare
payments. Second, because cost-based risk sharing depends on HHA cost data,
using this information in conjunction with the PPS could improve cost
reporting data, which will be critical to evaluating the PPS. We acknowledge
that a risk-sharing arrangement based on agency costs lessens the incentive
for an HHA to cut its costs, but we believe that it could be designed in a
way that would offset any incentive to maintain high costs. For example, if
risk sharing always required HHAs to incur some portion of their losses,
agencies would continue to have an incentive to lower their costs. Further,
HCFA does not acknowledge the protection afforded by a risk-sharing approach
against Medicare overpayments for episodes or Medicare expenditure growth
due to increased numbers of episodes, which we believe are important
justifications for this payment modification. Finally, we believe that risk
sharing is an important tool in moderating the incentive HHAs can have under
the PPS to stint on services and to protect Medicare patients from
underservice.

We believe that risk sharing should be implemented as soon as practicable
because our analyses of recent and historical utilization and spending data
indicate that agencies respond dramatically and quickly, but not necessarily
appropriately, to changes in Medicare payment policies. In its comments,
HCFA noted the rapid growth in utilization between 1990 and 1997, and
agencies' overreaction to the IPS between 1997 and 1999. Similarly, we
believe that agencies may immediately respond to the incentives of the PPS
in ways that may jeopardize beneficiary access to services or quality of
care and increase program expenditures.

We agree with HCFA that agency competition for patients and medical review
and monitoring efforts may deter HHAs from underserving beneficiaries.
However, we remain concerned that these features may be insufficient to
counter the financial incentives to stint on services within an episode and
to provide unnecessary episodes. Further, relying on competition to enforce
appropriate agency behavior may place unrealistic expectations on a
vulnerable population to have information about agencies' provision of
services and assumes that beneficiaries have choices in selecting a
provider, which is not necessarily true for all beneficiaries, particularly
those located in rural areas. Given the potential limitations of competition
and medical review in guarding against potential underservice, risk sharing
could provide HCFA with an additional tool to protect Medicare's
beneficiaries.

HCFA's comments are included as appendix IV.

We are sending copies of this report to the Honorable Nancy-Ann Min DeParle,
Administrator of HCFA, and interested congressional committees. We will also
make copies available to others upon request.

If you have any questions about this report, please call me or Laura Dummit,
Associate Director, at (202) 512-7119. Major contributors included Carol
Carter, Jean Chung, James E. Mathews, Kara Sokol, and Wayne Turowski.

Sincerely yours,
William J. Scanlon
Director, Health Financing and
Public Health Issues

Scope and Methodology

We conducted our analyses using Medicare provider, claims, and beneficiary
files for calendar years 1994, 1996, and 1999. We chose 1994 as our starting
point because its patterns of utilization and spending were used to set the
interim payment system (IPS) payment limits. We analyzed 1996 data because
the 1997 home health claims data include both pre-IPS and IPS claims. We did
not analyze 1998 data, since HCFA had well-documented problems constructing
this claims file. We thus selected 1999 claims data to reflect utilization
patterns under the IPS.

Agency ownership and location were extracted from HCFA's end-of-year
Provider of Services files for 1994, 1996, and 1999. We included in our
analysis only those providers that were listed as active in each year.

We used 100 percent of Medicare claims from HCFA's home health Standard
Analytical Files (SAF), final action claims, for 1994, 1996, and 1999 to
analyze patterns and trends in home health utilization. These files were
edited in three ways. First, the claims file for each year was compared with
the Medicare corresponding Denominator File to exclude claims for
beneficiaries who had enrolled in a Medicare managed care plan at any point
in the year. Second, the HHAs included in the claims data were compared with
the Provider of Service files for each year, and only claims from agencies
participating in Medicare were included in our analyses. Last, we excluded
aberrant values for service counts.

We used a 1999 claims file that was generated in May 2000, although HCFA's
SAFs are usually not complete until June of the year following the claims
year. After analyzing the distribution of claims by month, we concluded that
the file was roughly 95 percent complete. Subsequent comparisons with HCFA's
projections indicated that our estimate of the number of beneficiaries
receiving home health services in 1999 was 4 percent lower than HCFA's final
total. As a result, numbers presented in this report are likely to slightly
understate actual utilization in 1999 and may slightly overstate the
declines reported between 1996 and 1999.

In 1999, HCFA implemented a policy change that affected how home health
agencies reported the units of service when submitting claims for payment to
Medicare. Until July 1, 1999, units represented the number of visits;
starting October 1, units represented the number of 15-minute increments
making up the visit; and between July 1 and September 30, both counting
methods were used on the claims. We incorporated these policy changes in our
calculation of units from the claims files and verified our calculations by
analyzing the monthly distribution of visits during 1999.

For our analysis of changes in the number of Medicare beneficiaries using
home health services, we controlled for changes in Medicare enrollment by
using home health users per 1,000 Medicare fee-for-service (FFS)
beneficiaries. Our analysis only reflects Medicare FFS enrollees because
HCFA data on service use exclude those enrolled in managed care plans and
because the payment methods of interest in our analysis only apply to those
receiving home health care under FFS. In analyzing geographic
characteristics, we used the beneficiary's residence, reflecting HCFA's
decision to pay agencies on the basis of where the patient resides, not
where the agency is located. Because beneficiaries may receive care from
multiple agencies, which could be of different types, we counted each unique
beneficiary/agency combination as a separate home health user when analyzing
service use by agency characteristics. As a result, the user counts included
in our analyses of HHA characteristics are roughly 10 percent higher than
those included in the beneficiary-level data.

To examine the response of users, agencies, and areas of high utilization to
policy changes, we categorized beneficiaries, HHAs, and states as low-use,
medium-use, and high-use according to the average number of visits per user
in 1996. The low- and high-use cutoff points for beneficiaries and agencies
were set such that roughly 20 percent of the observations in 1996 fell into
each category, with the remaining group defined as medium-use. High-use
states were defined as those with utilization 20 percent or more above the
national mean.

To control for agencies opening and closing between 1994 and 1999, we
created a cohort of agencies open in all 3 years and examined them
separately. Their utilization trends were similar to those included in this
report.

Our analysis of HCFA's proposed PPS was based on the Federal Register final
rule31 and briefings with HCFA officials.

Medicare Home Health Users, by State of Residence, Calendar Years 1994,
1996, and 1999

                         Medicare home health
                         users per 1,000 FFS          Percentage change
       State                  enrollees

                       1994     1996     1999     1994-96  1996-99 1994-99
 Alabama              113      123     77         9.1      -37.9   -32.2
 Alaska               50       67      48         34.6     -29.0   -4.4
 Arizona              64       75      54         17.1     -28.3   -16.0
 Arkansas             93       103     70         10.8     -32.6   -25.3
 California           93       107     86         15.0     -20.2   -8.3
 Colorado             86       95      75         11.4     -21.4   -12.5
 Connecticut          102      119     106        16.4     -10.9   3.6
 Delaware             82       90      70         10.7     -22.3   -14.0
 District of
 Columbia             71       86      72         20.1     -15.6   1.4
 Florida              112      121     92         7.7      -24.1   -18.2
 Georgia              104      108     72         4.2      -33.1   -30.2
 Hawaii               38       42      34         10.9     -18.7   -9.9
 Idaho                78       95      64         22.3     -32.7   -17.7
 Illinois             87       100     78         15.8     -22.2   -9.9
 Indiana              73       85      60         16.8     -29.6   -17.7
 Iowa                 63       76      56         19.4     -26.3   -11.9
 Kansas               66       83      54         25.5     -34.1   -17.2
 Kentucky             87       105     81         20.5     -22.8   -7.0
 Louisiana            131      151     109        15.2     -28.1   -17.1
 Maine                94       111     94         18.4     -15.3   0.3
 Maryland             75       85      75         13.2     -11.9   -0.3
 Massachusetts        121      142     115        17.0     -18.5   -4.7
 Michigan             84       96      86         13.7     -10.0   2.4
 Minnesota            45       55      41         22.3     -25.5   -8.8
 Mississippi          135      152     103        12.3     -32.2   -23.9
 Missouri             100      113     84         12.7     -25.8   -16.4
 Montana              68       81      63         20.6     -22.9   -7.0
 Nebraska             61       74      60         21.0     -19.5   -2.6
 Nevada               71       81      63         15.6     -22.3   -10.2
 New Hampshire        98       111     89         13.2     -19.9   -9.3
 New Jersey           74       89      84         20.9     -5.3    14.5
 New Mexico           78       93      71         18.5     -23.1   -8.9
 New York             72       85      81         18.8     -4.7    13.2
 North Carolina       82       96      78         17.4     -19.0   -4.9
 North Dakota         66       76      63         15.9     -17.5   -4.3
 Ohio                 76       90      71         19.1     -20.7   -5.6
 Oklahoma             104      129     88         23.9     -31.3   -14.9
 Oregon               75       79      65         5.4      -17.5   -13.0
 Pennsylvania         98       113     101        16.0     -11.2   3.1
 Rhode Island         104      125     109        20.7     -12.9   5.2
 South Carolina       82       97      75         17.9     -22.7   -8.8
 South Dakota         55       71      54         30.1     -24.6   -1.9
 Tennessee            127      131     86         2.9      -34.4   -32.5
 Texas                102      125     84         23.4     -32.7   -17.0
 Utah                 88       96      71         9.7      -25.8   -18.6
 Vermont              130      138     114        6.2      -17.6   -12.5
 Virginia             76       90      76         18.2     -16.0   -0.7
 Washington           70       77      63         10.1     -17.7   -9.4
 West Virginia        72       88      64         20.8     -27.4   -12.2
 Wisconsin            55       61      51         10.7     -17.0   -8.1
 Wyoming              75       86      58         15.2     -33.4   -23.2
 Nationwide           89       102     80         15.0     -22.0   -10.3

Source: GAO analysis of HCFA's home health claims data and beneficiary
enrollment data for 1994, 1996, and 1999.

Average Visits per Medicare Home Health User, by State of Residence,
Calendar Years 1994, 1996, and 1999

                           Visits per person
                                                      Percentage change
        State                   served
                          1994   1996    1999     1994-96  1996-99 1994-99
 Alabama                 112    121     63        8.0      -48.5   -44.4
 Alaska                  43     48      23        11.3     -51.5   -46.0
 Arizona                 55     60      29        8.0      -51.7   -47.8
 Arkansas                75     79      45        5.6      -43.5   -40.4
 California              47     53      31        11.5     -41.1   -34.4
 Colorado                60     70      39        16.4     -44.9   -35.9
 Connecticut             46     55      40        19.3     -27.7   -13.8
 Delaware                46     50      31        9.2      -38.7   -33.0
 District of Columbia    42     51      42        20.7     -16.3   1.0
 Florida                 75     79      43        4.8      -45.7   -43.1
 Georgia                 103    105     52        2.5      -50.6   -49.4
 Hawaii                  44     47      25        6.0      -45.9   -42.6
 Idaho                   55     64      30        17.6     -54.1   -46.0
 Illinois                51     54      32        5.0      -40.2   -37.2
 Indiana                 72     77      37        6.6      -52.5   -49.4
 Iowa                    45     49      27        7.0      -44.8   -40.9
 Kansas                  56     63      32        12.5     -49.5   -43.2
 Kentucky                65     71      48        9.3      -32.7   -26.4
 Louisiana               126    162     95        28.6     -41.0   -24.1
 Maine                   59     65      39        10.3     -39.2   -32.9
 Maryland                37     38      29        2.9      -24.2   -22.0
 Massachusetts           75     85      46        13.4     -45.1   -37.8
 Michigan                45     51      34        13.5     -34.3   -25.5
 Minnesota               40     47      27        19.4     -43.0   -31.9
 Mississippi             113    127     80        13.1     -37.6   -29.4
 Missouri                50     55      32        10.1     -40.8   -34.8
 Montana                 50     53      31        5.8      -42.1   -38.7
 Nebraska                42     45      28        7.5      -38.6   -34.0
 Nevada                  68     66      39        -2.6     -40.9   -42.4
 New Hampshire           51     59      38        16.2     -34.9   -24.4
 New Jersey              40     44      31        10.9     -29.8   -22.1
 New Mexico              57     75      36        30.0     -51.3   -36.7
 New York                44     51      36        15.6     -30.2   -19.4
 North Carolina          57     56      37        -3.3     -34.3   -36.5
 North Dakota            44     44      27        1.4      -39.0   -38.1
 Ohio                    51     53      32        4.4      -38.7   -36.0
 Oklahoma                105    144     72        36.9     -50.0   -31.5
 Oregon                  39     37      22        -5.7     -41.1   -44.5
 Pennsylvania            43     47      32        10.0     -32.2   -25.3
 Rhode Island            43     52      34        20.2     -34.5   -21.3
 South Carolina          67     66      41        -1.5     -38.6   -39.6
 South Dakota            39     45      26        15.9     -43.0   -33.9
 Tennessee               115    114     67        -1.4     -41.1   -41.9
 Texas                   98     132     64        35.2     -51.8   -34.8
 Utah                    96     113     55        17.8     -51.0   -42.3
 Vermont                 56     67      44        20.3     -34.7   -21.5
 Virginia                49     56      39        15.2     -30.5   -20.0
 Washington              39     35      23        -9.5     -34.4   -40.7
 West Virginia           50     58      34        15.7     -40.9   -31.6
 Wisconsin               41     44      29        7.2      -33.0   -28.2
 Wyoming                 73     75      37        2.3      -50.5   -49.4
 Nationwide              65     73      41        12.7     -43.6   -36.5

Source: GAO analysis of HCFA's home health claims data and beneficiary
enrollment data for 1994, 1996, and 1999.

Comments From the Health Care Financing Administration

(201054)

Table 1: Average Number of Home Health Visits per User by Visit
Type, 1994, 1996, and 1999 14

Table 2: Medicare Home Health Users, by Utilization Level, 1994,
1996, and 1999 17

Table 3: Average Visits per User, by Level of Agency Service
Provision, 1994, 1996, and 1999 19

Table 4: Home Health Service Utilization Changes in Rural and
Urban Areas, 1994, 1996, and 1999 23

Table 5: History of Medicare Home Health Care Spending Growth,
1985-99 24

Figure 1: Home Health Users per 1,000 FFS Beneficiaries, 1994,
1996, and 1999 13

Figure 2: Proportion of Home Health Visits by Visit Type, 1996
and 1999 15

Figure 3: Proportion of Home Health Beneficiaries, by Utilization
Level, 1994, 1996, and 1999 18

Figure 4: Average Visits per User, by HHA Ownership, 1994, 1996,
and 1999 20

Figure 5: Average Visits per User, Highest- and Lowest-Utilization
States, 1999 22
  

1. The Balanced Budget Act of 1997 (BBA) (P.L. 105-33, title IV, chapter I,
111 Stat. 251, 466) mandated these changes.

2. The Omnibus Consolidated and Emergency Supplemental Appropriations Act of
1999 (P.L. 105-277) delayed the PPS implementation by 1 year to October 1,
2000.

3. Medicare payment policies discussed in this report do not apply to HHAs
providing services to beneficiaries enrolled in managed care plans. These
beneficiaries and their service use have not been included in this report's
analyses.

4. Medicare Home Health Care: Prospective Payment System Will Need
Refinement as Data Become Available (GAO-HEHS-00-9, Apr. 7, 2000).

5. A beneficiary is considered homebound when he or she has a condition that
results in a normal inability to leave home except with considerable and
taxing effort, and absences from home are infrequent or of relatively short
duration or are attributable to receiving medical treatment (section 204.1,
HCFA's Home Health Agency Manual, June 12, 2000).

6. "Part-time or intermittent" means skilled nursing and home health aide
services furnished any number of days per week as long as they are furnished
fewer than 8 hours each day and for 28 or fewer hours each week. Subject to
review on a case-by-case basis, a beneficiary may receive up to 35 hours of
home care per week, or up to and including 8 hours per day (full-time), 7
days per week, for temporary periods up to 21 days or longer in exceptional
circumstances (section 206.7, Home Health Agency Manual).

7. 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 certain medications; (4) activities to support
skilled therapy services; and (5) routine care of prosthetic and orthotic
devices (section 206.2, Home Health Agency Manual).

8. P.L. 96-499, sec. 930, 94 Stat. 2599, 2631.

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

10. Medicare: Home Health Utilization Expands While Program Controls
Deteriorate (GAO/HEHS-96-16, Mar. 27, 1996). More recent work has continued
to find billing abuses and improper claims: Medicare: Need to Hold Home
Health Agencies More Accountable for Inappropriate Billings
(GAO-HEHS-97-108, June 13, 1997) and Medicare: Improper Activities by
Mid-Delta Home Health, ( GAO/OSI-98-6, Mar. 12, 1998).

11. HHS Office of Inspector General, Review of Medicare Home Health Services
in California, Illinois, New York and Texas, A-04-99-01194 (Washington,
D.C.: HHS, Nov. 1999).

12. MedPAC, Report to the Congress: Context for a Changing Medicare Program
(Washington, D.C.: MedPAC, June 1998).

13. Agencies may be not-for-profit, for-profit (or proprietary), or
government-owned.

14. GAO/HEHS-96-16, Mar. 27, 1996.

15. The per-visit limits are for urban, freestanding HHAs. The per-visit
cost limits for rural agencies are higher than the urban limits.

16. The IPS per-beneficiary payment limits were based on data for cost
reporting periods ending in fiscal year 1994, reflecting 1993 costs and
utilization, and are updated annually by the home health market basket
index.

17. Payments would be adjusted if the episode of care is interrupted, such
as when a beneficiary elects to transfer to another HHA, when a beneficiary
is discharged because treatment goals are attained but then returns to the
same HHA, or when the beneficiary experiences a significant change in
condition. Episodes with extremely low service use (four or fewer visits)
will receive a low-utilization payment adjustment based on per-visit costs.

18. The Medicare, Medicaid and SCHIP Refinement Act of 1999 (P.L. 106-113,
app. F, title I) delayed the 15-percent reduction in the payments required
under the PPS until 12 months after implementation of the PPS. As a result,
the rates for fiscal year 2001 will be set so that spending would be the
same as spending under the IPS. The 15-percent reduction would be applied in
fiscal year 2002.

19. GAO-HEHS-00-9, Apr. 7, 2000.

20. For a more complete discussion of risk sharing, see our previous report,
GAO-HEHS-00-9, Apr. 7, 2000.

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

22. Abt Associates, Survey of Home Health Agencies (Cambridge, Mass.: Sept.
1999).

23. Since the IPS went into effect, almost 3,000 HHAs have stopped
participating in the Medicare program.

24. Improper services included services not documented, services to
beneficiaries not homebound, services without a valid physician order, and
services not reasonable or necessary. HHS Office of Inspector General,
Review of Medicare Home Health Services in California, Illinois, New York
and Texas, A-04-99-01194 (Washington, D.C.: HHS, Nov. 1999).

25. Between 1996 and 1999, there was considerable fluctuation in the number
of HHAs serving Medicare beneficiaries. Almost 1,400 HHAs began serving
beneficiaries and almost 1,900 stopped, for a net reduction of over 500
HHAs. In addition, previous GAO analysis found that the number of HHAs under
for-profit ownership, as opposed to not-for-profit or government ownership,
changed from 1996 to 1999 (see GAO/HEHS-99-120, May 26, 1999). To control
for the changing mix of agencies, for this analysis we examined only HHAs
that were open in 1994, 1996, and 1999.

26. The eight states are Louisiana, Oklahoma, Texas, Mississippi, Alabama,
Tennessee, Utah, and Georgia.

27. Urban and rural counts reflect beneficiary residence. Urban areas are
defined as those within a metropolitan statistical area (MSA), and rural
areas are those outside an MSA.

28. Because per-visit payments under prior law were relatively constant over
time, growth in payment rates composed only a small share of Medicare
spending growth for home health services. The changes in spending over time
are, therefore, mostly attributable to changes in utilization.

29. GAO/HEHS-00-9, Apr. 7, 2000.

30. Under the PPS, HHAs can receive a full episode payment for a patient
receiving 5 or more visits within a 60-day period.

31. "Medicare Program: Prospective Payment System for Home Health Agencies,"
final rule, Federal Register, Vol. 65, No. 128 (July 3, 2000).
*** End of document. ***