oversight

Medicare: Home Health Agencies With High Visit Rates Skew Averages

Published by the Government Accountability Office on 1997-06-02.

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

      United States
GAO   General Accounting
      Washington,
                            Office
                    D.C. 20648

      Health,   Education   and Human Services Division


      B-277090


      June 2, 1997

      The Honorable Fortney H. (Pete) Stark
      House of Representatives

      Subject: Medicare: Home Health Agencies With High Visit Rates Skew
                Averages

      Dear Mr. Stark:

      Medicare’s expenditures for home health care have been rising rapidly,
      increasing an average of 33 percent per year since 1989. Ln a March 1996
      report’ we concluded that a combination of factors led to this high rate of
      growth, including changes in the definition of what is covered under the benefit
      and the shift from short-term, post-hospital care to more longer-term care for
      chronic conditions. We also reported that the number of Medicare-certified
      home health agencies (HHA) has seen rapid growth in recent years, increasing
      from 5,692 agencies at the end of 1989 to 10,133 agencies as of April 1997. The
      number of proprietary (for-profit) HHAs has accounted for 83 percent of this
      growth, with an increase from 2,007 (35 percent of all HHAs) to 5,699 (56
      percent of HHAs). And, we found that proprietary agencies consistently
      provide more visits per beneficiary throughout the country than do voluntary
      and government agencies. For example, in 1993,’ proprietary agencies provided
      an average of 78 visits per year per beneficiary while voluntary and government
      agencies provided an average of 46 visits.

      The administration has proposed establishing a prospective payment system
      (PPS)3 for home health care as a way of helping control cost growth, and the

      ‘Medicare: Home Health Utilization Expands While Program Controls
      Deteriorate (GAO/HEHS-96-16, Mar. 27, 1996).
      “The most recent data available at the time of that report.
      3A PPS establishes the amount that will be paid for care in advance of the
      period to which the rate applies. Generally, if the provider has costs below the
      payment rate, it keeps the difference as a profit. If costs are higher than

                                 GAO/HEHS-9?-139R         For-Profit   Home Health   Agencies
B-277090

Congress is considering this and other proposals. You asked us whether, in
preparing our March 1996 report, we had found any reasons why proprietary
HHAs provide more visits than voluntary and governmental agencies.
Additionally, you asked whether there is any justification for the extra visits
and whether the skewing effect of the higher visit rates by proprietary
agencies could be removed when calculating the number of visits for purposes
of devising a PPS for home health.

To address these questions, we reviewed our prior work and other studies on
home health utilization. We also did some additional analyses, using the
episode of care data base for the 1996 report, to examine differences in
utilization between proprietary and nonprofit HHAs. These data were extracted
from the data bases maintained by the Health Care Financing Administration
(HCFA), which is responsible for managing the Medicare program.

In summary, our work and the work of others has consistently shown that
proprietary agencies provide more visits per beneficiary than agencies of other
types. However, while an agency could provide more visits on average than
other agencies for legitimate reasons, none of the factors we and others
explored provided an explanation related to patient need for the differences in
utilization among agency types. In developing a PPS, one way to lessen the
influence on visit rates of HHAs that consistently furnish more visits is to use
the median number of visits-the point at which half of patient cases (or
episodes of care) have fewer visits and half have more-rather than using the
average number of visits to determine payment rates for episodes of care.
Using the median could be combined with an “outlier” payment system for
exceptional cases that justifiably have high numbers of visits so that HHAs are
 not financially disadvantaged by patients who need extraordinary care.

We also have concerns about the adequacy of HCFA’s current data on home
health visit rates and costs for setting PPS rates. Our concern stems from the
low levels of medical reviews and cost report audits conducted by Medicare’s
intermediaries during the 1990s. Thorough reviews and audits should be
performed on a projectable sample of home health agencies and the results
used to adjust HCFA’s data bases before PPS rates are set.




payments, the provider incurs a loss.

 2                   GAOLHEHS-97-139R       For-Profit   Home Health   Agencies
             B-277090
             UTILIZATION PATTERNS OF PROPRIETARY HHAs

             In our 1996 report we found that in 1993, home health patients receiving care
             from a proprietary HHA had a much higher average number of visits per year
             than those receiving care from government or voluntary agencies. We found no
             explanation for this difference in use rates either in our analysis or the
             literature. Our additional analysis of beneficiary episodes of care during 1992-
             93 for the top 15 diagnoses4 also showed that proprietary agencies provided a
             higher number of visits per episode-an average of 43 visits with a median of 24
             visits for proprietary agencies versus an average of 28 visits for nonprofit
             agencies (voluntary and government agencies combined) with a median of 15.5
             While the average number of visits per episode provided by proprietary HHAs
             compared to nonprofit agencies varied somewhat by HCFA region, table I
             shows that proprietary agencies clearly provided more visits in all regions
             during 1992-93.

Table 1: Average and Median Visits Per Care Enisode-Ton 15 Diapnoses. 1992-93




             ‘Approximately half of all home health beneficiaries have 1 of these 15
             diagnoses.
             5These numbers are lower than those for visits per beneficiary per year, as a
             beneficiary may have more than one episode in a year.

3                                GAOMEHS-97-139R        For-Profit   Home Health   Agencies
              B-277090
aProprietary to nonprofit.

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

              An HHA may provide more visits on average than another for legitimate
              reasons. For example, one study found that regional variation in utilization
              could be explained, in part, by patient characteristics6 Beneficiaries in the
              East South Central Census region, for instance, were more likely to live in
              poverty and be in poorer health than beneficiaries in other parts of the country,
              and these characteristics are associated with higher than average home health
              use. However, this does not explain why one type of agency would provide
              more visits than another type in the same geographic area. A multivariate
              model of visits rendered per episode indicated that a proprietary status of an
              HHA is a highly significant predictor of utilization and that proprietary agencies
              provided an average of 30 percent more visits per episode than did nonprofit
              agencies, regardless of region.

              Further, in our work, we found that within the same state or region,
              proprietary agencies provided more visits for beneficiaries with the same
              primary diagnosis. 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 30 visits from nonprofit agencies during episodes
              occurring in 1992-93. Tables 2 through 5 show comparisons of mean and
              median home health visits for four diagnoses in the 10 HCFA regions during
              1992-93. These tables show that although visit rates varied among the HCFA
              regions, proprietary agencies consistently furnished more visits for each of the
              diagnoses in every region of the country.




               ‘Jennifer Schore, Patient, Agencv. and Area Characteristics Associated with
               Regional Variation in the Use of Medicare Home Health Services, prepared by
               Mathematics Policy Research, Inc., for HCFA, Contract No. HCFA-500-89-0047,
               Sept. 30, 1994. This study looked at reasons for regional variations in home
               health utilization using home health episodes starting in 1990, 1991, and 1992.

 4                                 GAO/HEHS-97-139R        For-Profit   Home Health    Agencies
                  B-277090

Table 2: Average and Median Visits Per EDisode for Diabetes bv HCFA Region. 1992-93

    HCFA region              Proprietary                 Nonprofit                         Ratio
                         Mean         Median         Mean        Median            Mean            Median




    All              I       52.6              26       30.2               16             1.7           1.6
Source: GAO analysis of data from the Medicare Standard Analytical File: Home Health
Claims History Database.


Table 3: Average and Median Visits Per EDisode for Hwertension                    bv HCFA Region. 1992-93




                             30.5              16       20.9               12             1.5           1.3
    All                      44.0              25       27.5               15             1.6           1.7
Source: GAO analysis of data from the Medicare Standard Analytical File: Home Health
Claims History Database.

5                                          GAO/HEHS-97-139R          For-Profit     Home Health        Agencies
             B-277090

Table 4: Average and Median Visits Per EDisode for Heart Failure bv HCFA Region. 1992-93




&i.ms History Database.

Table 5: Average and Median Visits Per EDisode for HiD Fracture bv HCFA Region. 1992-93




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

6                                GAO/HEHS-97-139R      For-Profit   Home Health   Agencies
B-277090
The Department of Health and Human Services’ Office of the Inspector General
(OIG), has also examined differences in visit rates by type of agency. A 1997
OIG study7 looked at the operating practices of low average utilization and high
average utilization HHAs by surveying a random sample of 150 of each of these
two types of agencies. OIG found that HHAs at the high utilization end tended
to be for-profit and freestanding organizations. OIG found, however, that
program operations were similar in high- and low-utilization agencies and did
not explain the variation. For example, both types of agencies provided a
similar mix of skilled nursing, physical therapy, occupational therapy, speech
therapy, medical social services, and aide services. Additionally, OIG did not
find any difference between the two types of agencies in terms of beneficiary
age, race, gender, deaths while in care, qualifying conditions, and principal
diagnostic codes. OIG also did not find any differences in the quality of care
provided as measured by the number of deficiencies and complaints recorded
by HCFA’s survey and certification program for HHAs or by their accreditation
status. OIG concluded that nothing in its findings would suggest that
beneficiaries in the high-utilization, proprietary agency-dominated group were
any sicker or in any greater need of services than those beneficiaries in the
lower-utilization groups.

Another study’ also found a marked difference in the amount of care provided
by proprietary and nonprofit home health agencies. Even after controlling for
the health and functional status of the beneficiary, as well as age, sex, and
living situation, those beneficiaries getting care from a nonprofit agency
received, on average, 21 fewer visits than those receiving care from a for-profit
one.

CONSIDERATIONS IN DESIGNING A PPS FOR MEDICARE HOME HEALTH

Cost-based reimbursement for home health has been criticized by health
financing experts as providing few incentives for controlling volume of
services, operating efficiently, or controlling costs. It is cited as one of the
reasons for the significant growth in home health spending since 1989.
However, the growth in cost per visit has been relatively modest during the

‘Office of the Inspector General, Onerating Practices of High-Cost and Low-
Cost Home Health Agencies, 031-04-93-00261 (Washington, D.C.: HHS, Feb.
1997).
‘Elizabeth Mauser, Does Organizational Form Matter: Imnlications for the Home
Health Care Industrv, paper presented at the American Public Health
Association Meeting, San Diego, CA, Oct. 30-Nov. 2, 1995.

7                     GAOLHEHS-9?-139R        For-Profit   Home Health    Agencies
B-277090

1990s averaging 3.6 percent per year. As we have noted, expenditure growth
has resulted from an increasing volume of services covered under the program
-both in terms of increases in the numbers of users and in the number of
covered visits per user.

The administration has recently proposed moving home health payments from
cost reimbursement to a PPS, beginning in 1999. The goal in designing a PPS is
to ensure that providers have incentives to control costs and that, at the same
time, payments are adequate for efficient providers to furnish needed services
and at least recover their costs. If payments are set too high, for example,
Medicare will not save money and cost-control incentives can be weak.

As we mentioned in our recent testimony to the Subcommittee on Health of the
House Ways and Means Committee,g we are concerned that the unit of service
selected for such a system should not lead to increased costs or lower quality
of care. A per visit PPS would give agencies incentives not only to hold down
the cost per visit but also to maximize the number of visits and could result in
higher overall Medicare costs. For this reason, a per episode PPS may be
attractive, but there are also potential problems with this choice. For example,
agencies could gain by increasing their caseloads or by reducing the number of
visits provided during an episode, thus potentially lowering the quality of care.
If an episode of care is chosen as the unit of service, HCFA would need a
method to ensure that beneficiaries receive adequate services and that any
reduction in services that can be accounted for by past over-provision of care
will not result in windfall profits for agencies.

We also have concerns about the appropriateness of using current data on visit
rates to determine payments under a PPS for episodes of care. As we reported
in March 1996, controls over the use of home health care are virtually
nonexistent. Operation Restore Trust, a joint effort by federal and state
agencies in several states to identify fraud and abuse in Medicare and Medicaid,
also found very high rates of noncompliance with Medicare’s coverage
conditions in targeted agencies. Audits found that from 19 to 64 percent of
home health visits paid for by Medicare did not meet Medicare guidelines for
reasons such as the patients were not homebound, visits were not medically
necessary, and visits were not provided. Because these audits were done at
 agencies suspected of having problems, the results are not projectable to all


‘Medicare Post-Acute Care: Home Health and Skilled Nursing Facilitv Cost
Growth and ProDosals for Prospective Pavment (GAO/T-HEHS-97-90, Mar. 4,
 1997).

 8                   GAOEIEHS-97-139R       For-Profit   Home Health    Agencies
B-277090

HHAs. Nevertheless, this project’s results and the fact that intermediaries do
only a very small number of on-site compliance audits each year” indicate that
substantial amounts of noncovered care are likely to be reflected in HCFA’s
home heath care utilization data. For these reasons, we suggested that HCFA
conduct thorough on-site medical reviews of a projectable sample of HHAs to
give it a basis to adjust utilization rates for purposes of establishing a PPS.

Additionally, in our March 1997 testimony, we discussed concerns about the
quality of HCFA’s home health care cost data for PPS rate-setting purposes.
Because only a relatively small portion of HHAs have had on-site cost report
audits performed by the intermediaries, there is little assurance that reported
costs are reasonable and/or related to patient care. Further, an examination of
caregiver compensation in 1995 as a proportion of the Medicare cost limits
indicated that there was considerable opportunity to inflate overhead expenses
(see table 6). The average caregiver compensation per visit ranged from a high
of 49 percent of the Medicare cost limits for physical therapy services to less
than 26 percent of the cost limits for home health aide services. Because of
our concerns about HCFA’s home health cost data bases, we suggested that
HCFA conduct audits of a projectable sample of cost reports to help ensure
that inflated costs are not used as the base for PPS rate setting.




“Our 1996 report found that only a fraction of 1 percent of Medicare-certified
HHAs received an on-site comprehensive medical review in 1994 and that
without on-site reviews, it was unlikely that visits not meeting Medicare
requirements would be identified.

                    GAO/HEHS-97-139R        For-Profit   Home Health Agencies
                     B-277090

    Table 6: Home Health Caregiver Comnensation as a Proportion of Medicare Cost Limits,
    1995

                                Per visit compensation,by                Per visit cost knits        Ratio of median
                                percentile                                                           compensation to
                                                                                                     cost limit
  Type of visit                     25th         Median       75th        Urban           Rural       Urban     Rural

IISkilled nursing           I        $25.00 1      $26.93 1   $31.80 1     $91.16 1         $99.83 1 29.54% 1 26.98%
  Occupational therapy




  Home health aide          I          11.00 1      11.75 1    13.74 I      45.98 1          46.30 1 25.55% 1 25.38%

    Source: GAO analysis of data from National Association for Home Care, Basic Statistics
    About Home Care 1996, and Schedule of Limits on Home Health Anencv Costs Per Visit, 60
    FR 8389 (Washington, D.C.: Feb. 14, 1995).

                     USE OF MEDIAN NUMBER OF VISITS WOULD HELP REMOVE SKEWING
                     EFFECT

                     The distribution of number of visits for home health care is skewed because of
                     the episodes of care with high volumes of visits. This, in turn, has the effect of
                     increasing the mean (average) number of visits per episode of care. One way
                     to statistically address skewing is to use the median number of visits, the point
                     where half of the cases have fewer and half have more visits. For example,
                     while the mean number of visits per case for episodes of care for the top 15
                     diagnoses in 1992-93 was 33, the median was 18. Thus, using the median to
                     establish PPS rates would provide greater incentives to control visit volume
                     than would use of the mean.

                     Regardless of the measure of visits used to establish a per episode PPS, it
                     would also be appropriate to provide for an outlier payment mechanism.
                     Outlier payments could be made for cases where the number of visits
                     significantly exceed the median (or whatever is used to establish payment
                     rates) and the excess visits are demonstrated to be medically appropriate. This
                     would help overcome the incentive to refuse to take cases that may require
                     higher than normal numbers of visits. Another potential problem with a per


                      10                         GAO/HEHS-97-139R            For-Profit         Home Health    Agencies
B-277090
episode PPS is that it gives agencies incentives to increase caseload. A review
mechanism would also be needed to help ensure that the patients treated by
HHAs actually qualify for the benefit. Finally, a per episode PPS could result in
HHAs’ underserving patients because of the incentive to hold down the number
of visits furnished. A quality of care monitoring process would be needed to
help prevent this effect.

When faced with similar utilization and quality concerns about Medicare’s
proposed inpatient hospital PPS, the Congress directed the utilization and
quality control peer review organizations to assess the necessity of admissions
and quality of care of PPS hospital cases.

AGENCY COMMENTS

We discussed the contents of this letter with HCFA officials and reflected their
comments where appropriate.



We are sending copies of this correspondence to the Administrator of HCFA
and interested congressional committees. We will make copies available to
others on request.

Should you or your staff have any questions, please call me at (202) 512-7114,
or Tom Dowdal, Senior Assistant Director, at (202) 512-6588. Tricia Davis and
Robert DeRoy also contributed to this letter.

Sincerely yours,




William J. Scanlon
Director, Health Financing and Systems Issues




(101548)



11                  GAOLKEHS-97-139R       For-Profit   Home   Health   Agencies
Ordering    Information

The first copy of each GAO report and testimony is free.
Additional   copies are $2 each. Orders should be sent to the
following address, accompanied    by a check or money order
made out to the Superintendent     of Documents, when
necessary. VISA and Mastercard      credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 6015
Gaithersburg, MD 20884-6015

or visit:

Room 1100
700 4th St. NW (corner    of 4th and G Sts. NW)
U.S. General Accounting     Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (301) 2534066, or TDD (301)        413-0006.

Each day, GAO issues a list of newly available reports and
testimony.   To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information         on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with “info” in the body to:

info@www.gao.gov

or visit GAO’s World Wide Web Home Page at:

http://www.gao.gov
United States
General Accounting  Offke
Washington, D.C. 20548-0001


Official   Business
Penalty    for Private   Use $300

Address    Correction    Requested