Medicare: Data Limitations Impede Measuring Quality of Care in Medicare ESRD Program

Published by the Government Accountability Office on 1997-07-11.

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

      United Statea
3A0   General Accoanting Off‘ice
      Wafhhgto~ D.C. 20648

      Health, Ednadon    and Human Services Division

      July 11, 1997
      The Honorable Pete Stark
      Ranking Minority    Member
      Subcommittee on Health
      Committee on Ways and Means
      House of Representatives
      Dear Mr. Stark:
      Subject: Medicare: Data Limitations tiDede Measuring QuaIiw
               of Care in Medicare ESRD Program

      Severalresearch studies indicate th&t care furnished to Medicare beneficiaries
      with end-stagerenal disease (ESRD) does not meet established standards.
      Patients undergoing dialysis, the most common form of therapy, are often
      anemic, undernourished, experience complications, or die prematurely. At least
      one of these studies concluded that differences in patient mortality among
      dialysis providers is related tb differences in the characteristics of dialysis
      facilities rather than those of patients.’ Some providers and renal disease
      experts are also concerned about the influence of various provider ownership
      and financial arrangements on the quality of care furnished to ESRD patients.
      For example, in 1991the Institute of Medicine (IOM), an advisory body that
      examinesissues related to public health, expressed concern that the federal
      government was not monitoring the health care quality implications of the
      growth in for-profit dialysis facilities and those owned by large ~hains.~ More
      recently, some dialysis providers have alleged that ESRD patients belonging to
      health maintenance organizations (HMO) receive poorer care than patients
      belonging to the standard Medicare ESRD program. These latter concerns have

      ‘Wfiam McClellan and J. Michael Soucie, “Facility Mortality Rates for New
      End-StageRenal Disease Patients, American Journal of Kidney Diseases Vol24,
      No. 2 (Aug. 1994),pp. 28089.
      ?Ol$ Kidney Failure and the Federal Government, Richard Rettig and Norman
      Levinsky, eds. (Washington, D-C.: National Academy Press, X391),pp. 169-63.
                                          GAOLKEHS-97-1318   ESBD Quality   of Care

become more immediate because the Congress is considering legislation to lift
the current restriction against ESRD patients joining Medicare HMOs with risk

Because of your concerns about the quality of care furnished to Medicare
ESRD patients, you asked us to (1) identify accepted performance standards
for measuring quality of care provided ESRD patients; and (2) using these
performance standards, compare the quality of care furnished to ESRD patients
between providers such as chain-aftiliated and unaBil.iated dialysis facilities,
and between HMOs and providers paid through the standard Medicare ESRD

To attempt to answer these questions, we interviewed ESRD experts and
reviewed relevant literature about measuring the quality of care furnished to
ESRD patients. We also investigated several data sources that potentially could
be used to analyze the differences between various provider types. These
sources included the ESRD Core Indicators Project results for 1993, 1994, and
1995 conducted by the Health Care Financing Administration (HCF’A), which
administers the Medicare program, and special studies conducted by the United
States Renal Data System (USRDS) Coordinating Center, which maintains
extensive data on patient characteristics and treatment for all ESRD patients4
For this study, we elected to use the Core Indicators Project data because they
included the most recent available data (1995) containing indicators that
measure clinical factors that most experts considered to be related to
outcomes for ESRD patients. We also obtained data fiorn HCFA files
identifying all ESRD beneficiaries enrolled in Medicare risk HMOs. We did our
work between May 1996 and June 1997 in accordance with generally accepted
 government auditing standards.

In summary, we found that most experts we interviewed and applicable
literature we reviewed agree that clinical indicators measuring dialysis

3HMOs with Medicare risk contracts are paid an amount fixed in advance by
Medicare for each beneficiary to provide all Medicare-covered services. Thus,
the HMO assumes the Gnancial risk of providing all necessary care in return for
the fixed payment amount

4The USRDS was created and is maintained by the Coordinating Center under a
contract with National Institute of Diabetes and Digestive and Kidney Diseases.
This contract is presently held by the University of Michigan Kidney
Epidemiology and Cost Center, which also conducts specific ESRD-related
research studies for other sponsors..

 2                                GAO/HEHS-97-137R       ESRD Quality      of Care

effectiveness, anemia, and nutritional status-urea reduction ratio, hematocrit
levels, and serum albumin levels, respectively-are valid performance indicators
for measuring the quality of care ESRD patients receive.5 These indicators are
currently used by HCFA to evaluate the care furnished to Medicare
beneficiaries with ESRD. Almost all experts we interviewed and applicable
literature we reviewed also agreed that these indicators were correlated with
morbidity and mortality, the ultimate outcome measures.

We were unable, however, to evaluate the differences between the quality of
ESRD care furnished in chainafFiliated and unaftiliated dialysis facilities or the
care provided by HMOs and providers in the standard Medicare ESRD program
because of limitations with data availability. Existing HCFA data about chain
affiliation of dialysis facilities is unreliable. When we matched ESRD
beneficiaries in HCFA’s Core Indicators files with HCFA data on ESRD
beneficiaries who belong to HMOs, we found too few beneficiaries belonging to
HMOs in each annual sample to give us confidence in the results. Even after
we combined the three annual tiles, the sample size was too small to permit us
to make reliable inferences about differences in quality of care between the
HMO and non-HMO ESRD populations when comparing beneficiaries with
similar characteristics such as age, gender, race, socioeconomic status, and
health conditions. If HCFA maintained up-to-date information about the chain
&iliations of diaJysis facilities and included a larger sample of HMO enrollees
in its Core 3ndicators Project, a comparison could be made of different types of
providers and delivery systems that would give us greater confidence in the
results. HCFA program officials agreed and said they would consider collecting
data to perform these analyses.


ESRD is chronic failure of kidney function. Persons with this degree of kidney
disease will die within a short time without long-term kidney dialysis or a
kidney transplant. Treatment for chronic kidney failure is very costly. In 1994,
Medicare paid a total of about $8.2 billion to treat approximately 242,000 ESRD
beneficiaries who received covered services-almost $34,000 per patient. In

the urea reduction ratio is the ratio of the reduction of the level of blood urea
nitrogen, a metabolic toxin, resulting from the dialysis treatment. Hematocrit is
a measure of the percentage of total blood volume, which consists of oxygen
carrying mature red blood cells. Serum albumin is a measure of the amount of
albumin, a simple protein, found in the blood.

3                                 GAOIFIEHS-97-137R      ESRD Quality    of Care
contrast, in 1994, the average Medicare expenditure was $4,637 per enrollee
without ESRD.

In 1972, 7 years after Medicare was enacted to cover many of the he&h care
expenses incurred by persons 65 years old and older, the Congress enacted
legislation that extended Medicare eligibility to persons under age 65 with
ESRD.’ This categorical eligibility was enacted because many persons who
could have benefited from dialysis or a transplant died because they could not
afford to pay for their care and payments from other sources were inadequate.
Under the 1972 provision, most U.S. residents not yet entitled to Medicare who
develop ESRD become Medicare eligible 3 months after beginning kidney
dialysis7 Medicare’s ESRD program has grown rapidly since its early years.
Between 1978 and 1995, the number of ESRD patients covered by Medicare
grew from 45,000 to an estimated 248,000-an average annual increase of more
than 10 percent per year. However, very few of these patients are enrolled in
Medicare HMOs.

Under current law, a beneficiary diagnosed with ESRD is prohibited from
enrolling in a Medicare HMO. However, persons already enrolled in such an
HMO when diagnosed with ESRD may remain enrolled in it.* By the end of
1995, only about 6,400-2.6 percent of ESRD beneficiaries-were enrolled in
Medicare risk HMOs. This percentage is only one-fourth of the percentage of
all Medicare beneficiaries enrolled in risk HMO.+10.1 percent at the end of
1995. However, the administration has proposed legislation to end the
prohibition. While it does not appear that this proposal will be enacted in
1997, the pending House reconcihation bill includes a requirement for the
Department of Health and Human Services to conduct a study and make
recommendations. If the administration’s proposal is enacted in the future, we

6P.L. 92-603, Social Security Amendments of 1972.
7Medicare eligibility can begin sooner if the patient receives a kidney transplant
or enters into a course of training for home dialysis.
*ESRD beneficiaries, like other Medicare beneficiaries, may elect to disenroll
from a Medicare HMO effective the beginning of the next month. However,
they then become liable for Medicare coinsurance and copayments, which are
substantial for these patients. This can be a signiticant burden for beneficiaries
without some form of secondary health insurance. We do not have estimates
of the number of ESRD patients who disenroll from Medicare risk HMOs once
diagnosed with ESRD.

4                                 GAOAEHS-97-137R         ESRD Quality     of Care

believe that the numbers of ESRD patients in HMOs could grow rapidly
because the beneficiaries would incur fewer out-of-pocket costs.

Kidney dialysis is the most common form of treatment for chronic kidney
failure. The most frequent form of dialysis is hemodialysis, a process that
involves passing the patient’s blood through a device that removes metabolic
poisons and excess fluids. In 1993, 59 percent of ESRD patients received
hemodialysis, usually provided in a hemodialysis facility three times per week.g

Several studies evaluating the quality of dialysis care for ESRD patients are
currently under way. The HCFA Core Indicators Project is a multiyear study
intended to improve dialysis care in the United States. For each year since
1993, project staff have selected a random sample of patients &om each of
HCFA’s 18 ESRD network geographic areas.” The sample is designed so that
the results will support conclusions apphcable to each network area and to the
nation. For each randomly selected patient, clinical data necessary to measure
certain quality indicators are obtained fiorn medical records. These data are
analyzed by HCFA and the networks and then given to dialysis centers to be
used as a benchmark against which improvements in care can be measured.
Data from 1993 through 1995 are now available, and during 1997, the project is
collecting data from 1996. Results fiorn this project showed that in 1993-
before the adoption of a formal standard-only 43 percent of a nationally
representative sample of adult hemodialysis patients received adequate
dialysis. IL Results for 1995, although considerably better, were still poor-only
59 percent of a similar sample of patients received adequate dialysis.

the other therapies include kidney transplantation (27.3 percent of ESRD
patients), peritoneal dialysis (11.3 percent), and home hemodialysis (0.2
percent). Peritoneal dialysis is usually performed by patients in their homes
and involves introducing dialysis fluids directly into the patient’s abdominal
“?t’he 18 network areas, designated by the Secretary of HHS, cover the nation.
In each network area, a network organization under contract to HCFA
conducts oversight activities for the appropriateness of services and patient
safety for ESRD patients. Two of the 18 networks did not participate in the
first year of the project. All networks have participated since then.
“Adequate dialysis is defined as a urea reduction ratio of less than or equal to
0.65. This standard was recommended in November 1993 by a consensus
conference convened by the National Institutes of Health.

5                                GAOLHEHS-97-137R       ESRD Quality     of Care
Other studies of ESRD care are also being conducted, The USRDS
Coordinating Center is collecting data over time on a l-arge sample of dialysis
patients. One of its goals is to characterize the total renal patient population
and another is to develop and analyze data on the effects of various treatment
modalities by disease and patient group categories. Significant amounts of
information have been produced through this effort, some of which
corroborates that ESRD patients generally receive poor care. Another national
study, also being conducted by the Coordinating Center and funded by Amgen
Corporation, is dete rmining the impact of hemodialysis practice patterns on
patient outcomes.12 Data from this study are not yet available.


Morbidity and mortality are among the often used indicators for outcomes
associated with quality care. As a result, when evaluating the quality of care
for ESRD patients, many experts rely on indicators that studies have shown to
be closely associated with morbidity and premature mortality among dialysis
patients. We found that commonly agreed upon indicators of quality are
measures of dialysis effectiveness, level of anemia, and nutritional status (urea
reduction ratio, hematocrit level, and serum albumin level, respectively).13
Since 1994, as part of its ESRD Core Indicators Project, HCFA has collected
data annually on these indicators to provide a basis for improving care
furnished to Medicare ESRD dialysis patients-l4

“Amgen Corporation is a manufacturer of genetically engineered drugs,
including recombinant human erythropoietin (EPO), which is used to treat
anemia in dialysis patients.
13Analternative, and according to many experts, superior, measure of dialysis
effectiveness to the urea reduction ratio is K#V. KUV, however, requires
additional information and must be estimated using a complex formula. A
HCFA official told us that in an effort to keep the data collection instrument
short, the Core Indicators Project had decided not to collect all the necessary
information to calculate this measure. In its comments on this
correspondence, however, HCFA said that it will collect all information needed
to calculate KW for the 1997 data collection effort.
14Awork group composed of renal community representatives established to
provide guidance to HCFA identified a total of four indicators at the start of
the Core Indicators Project. HCFA has dropped the fourth-blood pressure-as
a measure of quality because no clinical standard exists for blood pressure
                                  GAO/HEHS-97-137R        ESRD Quality     of Care


     Because HCFA reimburses dialysis facilities at a fixed rate per dialysis session
     for patients in the standard Medicare ESRD program, facilities have a strong
     incentive to control c~sts.‘~ The rate is not increased for inflation and, in fact,
     has declined in absolute dollars since the current reimbursement method was
     implemented in 1983. This cost-control incentive may be especially strong in
     for-profit facilities, which constitute over 60 percent of the total, because they
     need to pay a return on equity to investors as well as other expenses
     associated with operating a dialysis facility. Many for-profit dialysis facilities
     belong to multifacility chains. HMOs have a similar incentive to control costs
     because they are fully capitated for all care provided to an ESRD beneficiary.

     Jn its 1991 report on the Medicare ESRD program, the IOM pointed out that the
     proportion of for-profit and chain-affjhated dialysis facilities compared with
     nonprofit facilities was increasing. The IOM report expressed concern that no
     one was monitoring the implications of this change for quality of care to
     patients.‘” In addition, some providers and consumer advocates we interviewed
     alleged that they had observed that care provided to patients belonging to some
     Hh4Os was of lesser quality than that provided to beneficiaries in the standard
     Medicare ESRD program. Therefore, we had planned, using HCFA data, to
..   determine whether statistically signikant quality differences existed (1)
     between the care furnished to beneficiaries served by chain-mated        dialysis
     facilities and unmated     facilities and (2) between the care furnished to
     beneficiaries enrolled in HMOs and those receiving care paid for through the
     standard Medicare program.

     We were unable to analyze differences in the quality of care furnished by chain-
     affiliated dialysis centers and those without such an affiliation because HCFA’s
     data on chain affiliation are unreliable. A HCFA official told us that these data
     have not been consistently updated to reflect the rapid changes in the industry.
     For example, chain A’s purchase of a dialysis facility from chain B might not

     levels in patients with ESRD. Experts we interviewed agreed that blood
     pressure is not a good quality indicator for dialysis patients.
     ‘?his so-called “composite rate” covers all services normally associated with a
     dialysis treatment. Other services, such as some diagnostic tests and
     administration of some drugs, notably EPO, are compensated separately.
     “Kidnev Failure and the Federal Government, pp. 162-63.

     7                                 GAOBIEIHS-97-137R       ESRD Quality     of Care

have been noted in HCFA databases. Or a new dialysis chain may have been
formed as a result of a merger of several unaffiliated centers, but these changes
may not have been recorded. When we found we were unable to use HCFA
data, we attempted to obtain information on chain affiliation of dialysis
facilities directly from the larger chains we could identify. However, not aU
those we contacted provided the data.

We also could not determine whether differences existed between the quality
of care furnished to Medicare beneficiaries with ESRD in HMOs and those
covered by the standard Medicare program. The number of ESRD patients in
the Core Indicators Project samples found to be enrolled in HMOs was very
small. The number of HMO cases in any one of the 3 years that data were
available ranged from 80 to 150. Even after we combined the date for the 3
years, the maximum number of HMO cases available for analysis was 335, with
available cases for some analyses falling below that number because of missing
data elements.17 We were particularly concerned that the small number of
cases limited our ability to compare beneficiaries with similar characteristics
such as age, gender, socioeconomic condition, and race.

Because of the differing financial incentives inherent in different care delivery
systems, we believe it is as important to monitor the quality of care furnished
to patients by different provider types as it is to monitor the quality of care
furnished to all patients regardless of provider. In discussing our work with
HCFA officials, we suggested that the Administrator of HCFA (1) maintain
current information about members of chain-affiliated dialysis facilities, and (2)
modify the ESRD Core Indicators Project to collect sufficient data on ESRD
patients enrolled in risk HMOs to permit a valid comparison of the quality of
care of hemodialysis patients in HMOs with those in the standard Medicare
ESRD program. The HCFA officials we talked with agreed with our
suggestions and said they would keep us informed of what they did in this


HCFA chose not to comment formally on this letter, although it did provide
technical comments, which we incorporated where appropriate. In these
technical comments, HCFA stated that it will collect the core indicators data
from the last quarter of 1996 on all ESRD patients enrolled in Medicare

“In contrast, the maximum number of non-HMO cases available for analysis
was just over 18,000.

 8                                 GAO/HEHS-97-137R       ESRD Quality    of Care

managed care plans. The core indicators data from ESRD patients in these
plans will be compared with the core indicators data fkom the patients in the
standard Medicare ESRD program. These data will be collected during the
summer of 1997, and HCFA anticipates that the comparative analysis will be
available by the end of 1997.

We will make copies of this correspondence available to interested parties on

If you have any questions about this correspondence, please call me at (202)
512-6543or Assistant Director Sandra K. Isaacson at (202) 512-7174. Other
major contributors to this study included Peter Schmidt and George Lorenzen.

Sincerely yours,

Bernice Steirkardt
Director, Health Services Quality
 and Public Health Issues


9                                   GAO/HEHS-97-137R   ESRD Quality    of Care
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 37050
Washington, DC 20013

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 (202) 512-6061, or TDD (202)       512-2537.

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 caB (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:


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

United States
General Accounting    Office                   BulkRati.
Washington,   D.C. 20548-0001            Postage & Fees Paid
                                     I     Permit No. GIOO
Official   Business
Penalty    for Private   Use $300

Address    Correction    Requested