Medicare: Millions Can Be Saved by Screening Claims for Overused Services
(Letter Report, 01/30/96, GAO/HEHS-96-49).
GAO provided information on Medicare payments for unnecessary medical
services, focusing on the: (1) extent to which Medicare contractors
employ medical necessity prepayment screens for procedures that are
likely to be overused; (2) potential impact of autoadjudicated
prepayment screens on Medicare spending; and (3) federal government's
role in reducing overused medical procedures billed to Medicare.
GAO found that: (1) Medicare spending for unnecessary medical services
is widespread; (2) more than half of the 17 contractors surveyed do not
use prepayment screens to check whether claimed services are necessary;
(3) 7 of the contractors paid between $29 and $150 million for
unnecessary medical services; (4) many Medicare claims are paid because
contractors' criteria for identifying unnecessary medical services vary;
and (5) the Health Care Financing Administration needs to take a more
active role in promoting local medical policies and prepayment screens
for overused medical procedures.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: HEHS-96-49
TITLE: Medicare: Millions Can Be Saved by Screening Claims for
Overused Services
DATE: 01/30/96
SUBJECT: Health insurance cost control
Questionable payments
Health care services
Insurance claims
Fraud
Documentation
Internal controls
Claims processing
Disease detection or diagnosis
Managed health care
IDENTIFIER: Medicare Program
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to the Chairman, Human Resources and Intergovernmental
Relations Subcommittee, Committee on Government Reform and Oversight,
House of Representatives
January 1996
MEDICARE - MILLIONS CAN BE SAVED
BY SCREENING CLAIMS FOR OVERUSED
SERVICES
GAO/HEHS-96-49
Screening Claims for Overused Services
(101307)
Abbreviations
=============================================================== ABBREV
FMR - focused medical review
HCFA - Health Care Financing Administration
HHS - Department of Health and Human Services
OIG - Office of the Inspector General
YAG - yttrium aluminum garnet
Letter
=============================================================== LETTER
B-258099
January 30, 1996
The Honorable Christopher Shays
Chairman, Human Resources and
Intergovernmental Relations Subcommittee
Committee on Government Reform and Oversight
House of Representatives
Dear Mr. Chairman:
Medicare is the nation's largest health care insurer. Medicare
spending totalled $162 billion in 1994, about 14 percent of the
federal budget. Without additional controls over this spending, the
Congressional Budget Office estimated in December 1995 that total
Medicare outlays will reach $336 billion in 2002.
Restraining the growth in Medicare spending has proven difficult.
This is in part because the fee-for-service payment system\1 provides
little financial incentive for physicians or patients to consider
whether diagnostic tests and some routine services are medically
necessary. Moreover, physicians paid on a fee-for-service basis may
have a financial incentive to increase their income by providing more
services than are necessary. In addition, patients often lack the
information and expertise necessary to question the medical necessity
of services ordered by physicians. As a result of these two factors,
preventing Medicare payments for unnecessary services calls for
program safeguards to check the accuracy and medical necessity of
claims.
One type of program safeguard is the use of medical policies that
define the diagnostic criteria for a service. For example, a medical
policy for echocardiography may allow payment if the patient's
diagnosis is chronic pulmonary heart disease but deny payment if the
diagnosis is indigestion. Most medical policies and diagnostic
criteria are established and applied locally by each of the 29
contractors (also called carriers) that the Health Care Financing
Administration (HCFA) uses to process and pay claims submitted by
physicians.\2 Including these diagnostic criteria in Medicare claims
processing systems can enable checking all claims for the service
against the criteria before payment. For claims that do not meet the
criteria, the claims processing systems can deny payment
automatically. Diagnostic criteria implemented this way are referred
to as autoadjudicated medical necessity prepayment screens.
Providers may resubmit claims denied by these screens with additional
or corrected information to clarify the patient's medical symptoms.
Also, providers may appeal contractors' decisions to deny their
claims.
In discussions with your staff, we agreed to examine (1) the extent
to which contractors employ medical necessity prepayment screens for
procedures that are likely to be overused nationally, (2) the
potential impact of autoadjudicated prepayment screens on Medicare
spending, and (3) the role that the federal government should play in
reducing widespread overuse of medical procedures billed to Medicare.
To address these objectives, we reviewed payments to physicians for
six groups of high-volume medical procedures,\3 which accounted for
almost $3 billion in Medicare payments in 1994.\4 These procedures
are considered to be widely overused: Evidence from the Office of
the Inspector General (OIG) in the Department of Health and Human
Services (HHS) and contractors' analyses and views indicate that
providers commonly bill for these procedures when they are not
warranted by medical symptoms.
We also surveyed 17 contractors to determine if they used medical
necessity criteria in their claims processing systems to screen
claims for the six groups of procedures in our study. For seven of
the largest contractors we also used computer programs to review a
sample of the claims they paid for the six groups of procedures.\5 If
the contractors' claims processing systems did not screen these
claims against medical necessity criteria, our programs compared the
patient diagnoses on the paid claims to diagnostic criteria used in
prepayment screens by various other Medicare contractors. We
performed our work between August 1994 and November 1995 in
accordance with generally accepted auditing standards. Appendix I
further describes our scope, data sources, and methodology.
--------------------
\1 Fee-for-service has been the traditional and predominant method of
paying for health care services in both the private and public
sectors. The Medicare fee-for-service payment system, which
currently covers more than 90 percent of all Medicare enrollees, pays
physicians a fixed amount for each service delivered. In contrast,
Medicare enrollees in managed care plans receive all services in
exchange for an annual prepaid fee.
\2 The 29 carriers that process Medicare part B payments for
physician services are referred to as contractors in this report.
Four additional carriers process Medicare claims for durable medical
equipment, but those carriers were not included in this study.
\3 We reviewed claims for the following procedures:
echocardiography, eye examinations, chest X rays, colonoscopy,
yttrium aluminum garnet (YAG) laser surgery, and duplex scan of
extracranial arteries. See table 1 for the specific procedure codes
included in our study.
\4 We limited our review to Medicare part B payments, which generally
cover services provided by physicians and suppliers. In the part B
program, Medicare pays 80 percent of the total charge allowed and the
patient is responsible for the remainder. In this report, the total
payment allowed under Medicare is referred to as the Medicare
payment.
\5 Some contractors process Medicare claims from more than one state,
a portion of a state, or both. Our review covered claims processed
by the seven contractors in six states.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
Medicare contractors routinely pay hundreds of millions of dollars in
Medicare claims without first determining if the services provided
are medically necessary. For each of the six groups of medical
procedures we reviewed, more than half of the 17 contractors we
surveyed were not using prepayment screens to check these claims for
medical necessity. Even when evidence points to overuse nationwide,
each of the Medicare contractors usually decides on its own which
procedures to screen. For example, although HHS' OIG advised HCFA
and the contractors in 1991 to monitor the use of colonoscopies and
deny claims that were not indicated by medical symptoms or supported
with documentation, only 6 of the 17 contractors were screening
colonoscopy claims by the end of 1994.
Millions of dollars in Medicare claims for these six groups of
procedures would have been denied if all contractors had screened the
claims for medical necessity. Our review of just 7 of the 17
contractors revealed that between $29 million and $150 million was
paid for claims that may have been medically unnecessary. The range
of these estimates reflects the variation in contractors' criteria
for identifying medically unnecessary services. Because the
remaining contractors also were not using medical necessity screens
for some of these procedures, they also may have paid millions of
dollars in Medicare claims for services that should have been denied.
Problems with controlling payments for widely overused procedures
persist because HCFA lacks an effective national strategy. Although
the need for national leadership is compelling, HCFA has not
exercised its statutory authority to take an active role in promoting
more local medical policies and prepayment screens for widely
overused procedures. Instead, HCFA has relied on contractors'
abilities to focus their prepayment screens on procedures where local
use exceeds the national average. While this approach helps reduce
local overuse of some procedures, it is not designed to control
overuse of a procedure nationwide. We believe that HCFA should take
several approaches to help prevent Medicare spending for unnecessary
services.
BACKGROUND
------------------------------------------------------------ Letter :2
Medicare provides health insurance for about 37 million elderly and
disabled individuals. This insurance is available in two parts:
Part A covers inpatient hospital care and is financed exclusively
from a payroll tax. Part B coverage includes physician services,
outpatient hospital services, and durable medical equipment. Part B
services are financed from an earmarked payroll tax and from general
revenues.
The Social Security Act requires that Medicare pay only for services
that are reasonable and necessary for the diagnosis and treatment of
a medical condition.\6
HCFA contracts with private insurers such as Blue Cross and Blue
Shield plans, Aetna, and CIGNA insurance companies to process
Medicare claims and determine whether the services are reasonable and
necessary. The program was designed this way in part to protect
against undue government interference in medical practice.\7 Thus,
despite Medicare's image as a national program, each of the 29
Medicare contractors that process part B claims for physicians'
services generally establishes its own medical necessity criteria for
deciding when a service is reasonable and necessary.
Contractors do not review each of the millions of Medicare claims
they process each year to determine if the services are medically
necessary. Instead, contractors review a small percentage of claims,
trying to focus on medical procedures they consider at high risk for
excessive use. Contractor budgets limit the number of claims
contractors can review, and over the last several years, both
contractor budgets and HCFA requirements for prepayment review have
been decreasing. In 1991, HCFA required contractors to review 15
percent of all claims before payment, while in 1995, contractors are
only required to review 4.6 percent.
Since 1993, HCFA has required contractors to use a process called
focused medical review (FMR) to help them decide which claims to
review. Under the FMR process, each contractor analyzes its claims
to identify procedures where local use is aberrant from the national
average use.\8 Beginning in fiscal year 1995, HCFA has required each
contractor to select at least 10 aberrant procedures identified
through FMR and develop medical policies for those procedures. The
contractors are required to work with their local physician community
to define appropriate medical necessity criteria. This arrangement
allows contractors to take local medical practices into consideration
when establishing criteria for reviewing claims. Once physicians
have had an opportunity to comment on a medical policy, the
contractor publishes the final criteria.
Each contractor generally decides which medical procedures to target
for review and what types of corrective actions to implement to
prevent payments for unnecessary services. Contractors currently
concentrate on educating physicians about local medical policies,
hoping to decrease the number of claims submitted that do not meet
the published medical necessity criteria. Contractors also use
computerized prepayment reviews, called screens, to check claims
against the medical necessity criteria in medical policies. When
screens identify claims that do not meet the criteria, two
alternative actions are possible: first, autoadjudicated screens may
deny the claim automatically; second, all other screens may suspend
the claim for review by claims examiners, who may request additional
documentation from the physician before deciding to pay or deny the
claim.
Autoadjudicated screens usually compare the diagnosis on the claim
with the acceptable diagnostic conditions specified in the
corresponding medical policy. For example, an autoadjudicated screen
for a chest X ray would pay the claim if the patient diagnosis was
pneumonia but deny the claim if the only patient diagnosis was a
sprained ankle. Because this type of screen is entirely automated,
it can be applied to all the claims for a specific procedure at a
lesser cost than reviewing claims manually. This type of screen is
most effective for denying claims that do not meet some basic set of
medical necessity criteria. Claims denied by these screens can be
resubmitted by providers or appealed. As shown in figure 1, claims
that pass these basic criteria may be further screened against more
complex medical criteria to identify claims that warrant manual
review.
Figure 1: Overview of
Prepayment Medical Review
Process
(See figure in printed
edition.)
--------------------
\6 Medicare generally does not pay for routine screening tests such
as eye examinations, hearing tests, and routine chest X rays.
However, the Congress has enacted legislation allowing Medicare
payment for some routine services, such as a screening mammography.
\7 Section 1801 of the Social Security Act (42 U.S.C. 1395)
prohibits federal interference in the practice of medicine.
\8 Some contractors receive permission from HCFA to identify aberrant
procedures using alternative methods, such as trend analysis.
MANY CONTRACTORS DO NOT SCREEN
CLAIMS FOR OVERUSED SERVICES
------------------------------------------------------------ Letter :3
Most of the contractors we surveyed routinely pay claims for
procedures suspected to be widely overused without first screening
those claims against medical necessity criteria. We looked at six
groups of procedures that providers frequently perform on patients
who lack medical symptoms appropriate for the procedures. These
procedures also rank among the 200 most costly services in terms of
total Medicare payments and accounted for almost $3 billion in
Medicare payments in 1994.\9 (See table 1 below.) Four of the
procedures--echocardiography, eye examinations, chest X rays, and
duplex scans of extracranial arteries--are noninvasive diagnostic
tests. Colonoscopy can be either diagnostic or therapeutic, and YAG
laser surgery is sometimes used to correct cloudy vision following
cataract surgery.
Table 1
Medicare Services and Payments for Six
Medical Procedures (1994)
Medicare
Medicare payments\a
services (in (in
Procedure (procedure codes) thousands) millions)
------------------------------ ------------ ------------
Echocardiography (93307, 8,976 $851
93320, 93325, 93350)
Eye exams (92002, 92004, 14,400 686
92012, 92014)
Chest X rays (71010, 71020) 34,597 507
Colonoscopy (45378, 45380, 1,416 478
45385)
YAG laser surgery (66821)\ 895 325
Duplex scan of extracranial 1,513 143
arteries (93880)
==========================================================
Total 61,797 $2,990
----------------------------------------------------------
\a The total payments allowed under Medicare for each procedure code.
In the first quarter of fiscal year 1995 (Oct. 1-Dec. 31, 1994), we
surveyed 17 contractors to determine whether they were using any type
of medical necessity prepayment screens to review claims for these
six groups of procedures. As shown in table 2, the use of prepayment
screens among the contractors was not uniform, and for each of the
six procedures fewer than half the 17 contractors were using such
screens.
Table 2
Use of Medical Necessity Screens for Six
Procedures by 17 Medicare Contractors
(Oct. 1-Dec. 31, 1994)
Duplex scan
of
Echocardiogr Chest X Colonoscop YAG laser extracranial
Contractor aphy Eye exams ray y surgery arteries
------------ ------------ ------------ -------- ---------- ---------- -------------
1 X X X
2 X X X
3 X X
4 X X X
5 X
6 X X
7 X X
8 X X
9 X X X X
10
11 X X
12 X X X
13
14 X X
15 X X
16
17 X X X X X
=========================================================================================
Total 7 6 6 6 3 8
-----------------------------------------------------------------------------------------
For each group of products in our study, we found the following:
Only 7 of the 17 contractors we surveyed had prepayment screens to
review echocardiography for medical necessity, even though
echocardiography is often performed on patients with no specific
cardiovascular disorders. Ten contractors lacked such screens,
even though echocardiography is the most costly diagnostic test
in terms of total Medicare payments and despite an increase of
over 50 percent in the use of the echocardiography procedures
listed in table 1 between 1992 and 1994.
Only 6 of the 17 contractors used prepayment screens to prevent
payment for medically unnecessary eye examinations. These
contractors have medical necessity criteria to deny claims for
routine eye examinations and to allow payments only for certain
conditions, such as cataracts, diabetes, and hypertension.
Only 6 of the 17 contractors had prepayment screens to review chest
X ray claims for medical necessity, although HCFA had alerted
Medicare contractors that providers frequently bill for chest X
rays that are not warranted by medical symptoms and are thus
medically unnecessary.
Only 6 of the 17 contractors had medical necessity prepayment
screens to review colonoscopy claims. In 1991, HHS' OIG
reported that nationwide almost 8 percent of colonoscopies paid
by Medicare were not indicated by diagnosis or medical
documentation.
Only 3 of the 17 contractors had prepayment screens for YAG laser
surgery even though federal guidelines exist that indicate the
diagnostic conditions for performing this surgery. Also, at a
national meeting of Medicare contractors in 1994, HCFA officials
discussed the need to avoid paying for unnecessary YAG laser
surgery following cataract removal.\10
Only 8 of the 17 contractors had implemented prepayment screens for
duplex scans even though HCFA had alerted Medicare contractors
that providers commonly bill for noninvasive vascular tests such
as duplex scans without adequately documenting the patient's
medical symptoms.
A primary reason all contractors do not screen claims for nationally
overused procedures is that, following HCFA's instructions for FMR,
contractors have been targeting procedures that are overused locally,
based on comparisons with national average use. The shortcomings of
this approach are discussed later in this report.
Our survey of the 17 contractors represents a snapshot of the use of
prepayment screens for these procedures in the first quarter of
fiscal year 1995. Typically, contractors turn screens on and off
depending on their local circumstances. For example, one contractor
began using a screen for echocardiography in March 1995, and another
contractor implemented screens for chest X rays and eye examinations
in January 1995 because these procedures were overused locally. By
contrast, one contractor discontinued using an autoadjudicated screen
for eye examinations in February 1995 because the diagnostic criteria
for payment in the screen were considered too narrow.\11 Nonetheless,
these fluctuations in contractors' use of screens do not reflect a
coordinated approach to screening nationally overused procedures.
--------------------
\9 All procedure codes for these six medical services ranked among
the top 200 most costly services in 1994, except code 92002 (eye
examination for a new patient).
\10 On October 6, 1995, HCFA published a draft national medical
policy that specifies the Medicare medical criteria for payment of
YAG laser surgery claims.
\11 The discontinued autoadjudicated screen was not among those used
in our tests.
CONTRACTORS WITHOUT SCREENS PAY
MILLIONS FOR SERVICES THAT MAY
BE UNNECESSARY
------------------------------------------------------------ Letter :4
Seven large Medicare contractors\12 paid millions of dollars in
claims for services that may have been unnecessary. These
contractors did not use diagnostic medical criteria to screen claims
for some of the six groups of procedures in our study. The claims
paid for these services included a range of patient diagnoses that
did not meet the criteria established by other contractors. For
example, a chest X ray was paid for a patient with a diagnosis of
injuries to the hand and wrist, an echocardiogram was paid for a
patient with a diagnosis of chronic conjunctivitis, and a therapeutic
colonoscopy examination was paid for a patient with a mental health
diagnosis of hysteria. If the seven contractors had used
autoadjudicated diagnostic screens for the six groups of procedures,
they would have denied between $38 million and $200 million in claims
for services in 1993, as shown in table 3.
Table 3
Medicare Payments That Would Have Been
Denied by Autoadjudicated Screens\a
(Dollars in thousands)
Lowest Highest
Procedure (procedure code) estimate estimate
-------------------------- -------------- --------------
Echocardiography (93307, $10,475 $74,632
93320, 93325, 93350)\b
Eye exams (92002, 92004, 611 931
92012, 92014)
Chest X rays (71010, 819 37,166
71020)
Colonoscopy (45378, 45380, 5,798 61,886
45385)
YAG laser surgery (66821) 14,934\c 14,934\c
Duplex scan of 5,971 10,710
extracranial arteries
(93880)
==========================================================
Total $38,608 $200,259
----------------------------------------------------------
\a Medicare payments were calculated by multiplying the estimated
number of denied services by the average allowance under Medicare for
the procedure at the contractor. The estimates are based on a
5-percent sample of beneficiaries at each of seven contractors. We
used claims paid for services provided in 1993.
\b We combined the results from two echocardiography screens to
estimate payments that would have been denied for all four
echocardiography procedure codes.
\c We used only one prepayment screen for YAG laser surgery.
Therefore, the lowest and highest estimates are the same for payments
that would have been denied.
The range of estimated payments for claims that would have been
denied reflects differences among contractors' criteria for
identifying medically unnecessary services. Although different
contractors had screens for the same procedure, they used different
diagnoses to determine medical necessity. For example, a colonoscopy
screen we used from one contractor paid claims with a diagnosis of
gastritis, while another contractor's screen denied such claims.
Because of these differences among the contractors' screens, we
applied screens from two or three different contractors for each
group of procedures, except for YAG laser surgery.\13 Thus, our test
results show a range of estimated payments for claims that would have
been denied, depending on the medical necessity criteria used. The
tables in appendix II list the estimated payments for claims that
would have been denied by each of the tested screens.
The seven contractors we reviewed were among the largest in terms of
the number of claims processed, accounting for about 37 percent of
all Medicare part B claims, and almost 38 percent of all the claims
for the six groups of procedures in our study. To estimate the paid
claims that would have been denied, we applied autoadjudicated
screens developed by several contractors in our survey to a sample of
the 1993 claims paid by the seven contractors.\14 We only applied
these screens if the tested contractor did not have a medical
necessity diagnostic screen of its own in place in 1993 for the
specific procedure tested. We used autoadjudicated screens because
decisions to pay and deny claims based on medical necessity criteria
are automated and, therefore, do not require additional medical
judgment. Appendix I provides additional details on our methodology.
When claims are denied by prepayment screens, the billing physician
can (1) resubmit the claim with additional or corrected information
or (2) appeal the denial. In either case, the contractors may
ultimately pay claims that they have initially denied. Contractors'
claims processing systems generally do not track the claims denied by
autoadjudicated prepayment screens to determine if they are
resubmitted or appealed and then paid. However, based on a limited
analysis of claims denied by contractors with autoadjudicated
screens, we estimate that about 25 percent of the denied claims were
ultimately paid.\15 Assuming that the 25-percent rate is typical for
autoadjudicated screens, about 75 percent of the payments in table 3,
or between $29 million and $150 million, were for services that would
be considered unnecessary using the criteria established by various
contractors.
Our estimates of payments for unnecessary services involve only six
groups of procedures and cannot be statistically generalized beyond
the 7 contractors included in our analysis. However, all 29
contractors--not just the 7 whose claims we reviewed--operate under
FMR requirements designed to correct local rather than national
overutilization problems. Therefore, the other 22 contractors also
may lack screens for some of these procedures and, hence, may have
paid millions of dollars in claims for services that should have been
denied.
For widely overused procedures such as the six we tested,
autoadjudicated screens can be a low-cost, efficient way to screen
millions of claims against basic medical necessity criteria.
Contractor officials said that these screens are much less expensive
to implement than screens that suspend claims for manual review.
Consequently, as funding for program safeguards declines,
autoadjudicated screens can be used to maintain or even increase the
number of claims reviewed. Moreover, for procedures where the
medical review decisions can be automated, autoadjudicated screens
can quickly identify and deny claims where the patient diagnosis is
inconsistent with the procedure performed. In contrast, when claims
examiners manually review claims, the risk exists that the medical
necessity criteria may be misinterpreted and applied inconsistently.
However, for certain procedures or medical policies, autoadjudicated
screens may not be appropriate. For example, some medical policies
are not easily defined with diagnostic codes and require manual
review of documentation, such as medical records, to determine if a
service is medically necessary.
Denying claims using autoadjudicated or other prepayment screens can
increase administrative costs if providers frequently resubmit denied
claims or appeal the denials. Contractor officials said that these
costs can be minimized if providers are educated to bill
appropriately in the first place. By combining direct provider
education with screens that enforce agreed upon medical criteria,
contractors can, over time, reduce the number of claims submitted for
unnecessary services.
--------------------
\12 Of the 17 contractors we surveyed, we selected the 7 contractors
that processed the most Medicare claims. In table 2, the 7
contractors we selected are numbered 1 through 7.
\13 For YAG laser surgery we only applied the one screen that we had
identified at the time we began our analysis.
\14 We obtained all the tested screens from some of the 17
contractors in our initial survey. Some of the screens were obtained
from one of the seven contractors included in our tests.
\15 As described in appendix I, we estimated the percentage of denied
claims that would be ultimately paid by analyzing claims for
echocardiography processed by one contractor and claims for duplex
scans processed by another contractor. In each case, the contractors
used autoadjudicated screens for these services.
GREATER HCFA LEADERSHIP IS
NEEDED
------------------------------------------------------------ Letter :5
HCFA does not have a national strategy for using prepayment screens
to deny payments for unnecessary services among Medicare's most
highly overused procedures. HCFA does periodically alert contractors
about some of these procedures at semiannual national contractor
meetings and through occasional bulletins. However, the agency does
not identify widely overused procedures in a systematic manner.
Moreover, the agency does not ensure that contractors implement
prepayment screens or other corrective actions for these procedures.
Medicare legislation does not preclude HCFA from requiring its
contractors to screen claims for nationally overused procedures.
However, HCFA has chosen to avoid the appearance of interfering in
local medical practice. HCFA usually does not establish medical
policies or tell the contractors which procedures warrant medical
policies or prepayment screens.\16 Instead, HCFA relies primarily on
the contractors' local FMR efforts to identify and prevent Medicare
payments for unnecessary services. This process, according to HCFA
officials, allows contractors to take medical practice into
consideration when making medical necessity determinations.
Although FMR can work well for overutilization problems that are
truly local, the process is not designed to address nationwide
overutilization of a medical procedure. The national average use of
a procedure generally serves as a benchmark for identifying local
overutilization problems, but the benchmark itself may already be
inflated by millions of dollars in payments for unnecessary services.
For example, in several states the use of echocardiograms greatly
exceeded the 1992 national average of 101 services per 1,000
beneficiaries.\17 Some of the contractors servicing those states have
designed and implemented prepayment screens for this procedure.
Meanwhile, other contractors targeted different procedures and
allowed unconstrained use of echocardiograms. This focus on local
overuse may be one of the factors that led to a national 12-percent
increase in echocardiography use by 1994--and a new benchmark of 113
echocardiograms per 1,000 beneficiaries.
HCFA can take a more active role in controlling spending for widely
overused procedures without intruding on the contractors'
responsibilities to establish their own prepayment screens. HCFA has
an oversight responsibility to monitor and evaluate contractors'
screens and other efforts to prevent payments for unnecessary
services. Yet HCFA does not know (1) which contractors have
diagnostic screens for which medical procedures, (2) the medical
necessity criteria used in these screens, or (3) the effectiveness of
the screens in denying claims for unnecessary services. Furthermore,
without this information HCFA cannot identify best practices and
promote approaches such as autoadjudicated medical necessity screens
where they can be a cost-effective alternative or complement to
screens that flag claims for manual review.
HCFA funded a central database on local medical policies, but this
resource is not being effectively used. HCFA has encouraged the
contractors to use the database to research other contractors'
medical policies before drafting their own. However, according to
some contractors, the usefulness of the database is limited because
it is not regularly updated. Moreover, HCFA has not taken the
initiative to use the database to evaluate the contractors' medical
policies and identify those worthy of consideration by all
contractors for controlling widely overused procedures.
HCFA can also encourage greater use of medical necessity criteria for
widely overused procedures by providing contractors with more model
medical policies. About 2 years ago, HCFA established clinical
workgroups composed of contractor medical directors to develop model
medical policies that the contractors can adapt for local use.
Specifically, contractors can work with their local medical community
to review model policies, adapt them to reflect local medical
practice, and implement them in prepayment screens. This has been an
important step in promoting greater efficiency in developing local
medical policies. However, since the workgroups' inception, only one
model policy has been published.\18 According to HCFA and contractor
officials, progress has been limited in part because HCFA often takes
longer to review draft model policies than its goal of 45 days.
HCFA officials said that they are considering provisions for greater
use of autoadjudicated screens in a new, national claims processing
system. However, full implementation of that system is scheduled for
late in 1999. In addition, what types of screens will be included in
the system remains unclear, as well as how the contractors will chose
which screens to modify, implement, and use and how HCFA will monitor
and evaluate the effectiveness of the screens. Meanwhile, HCFA
continues to allow contractors to pay millions of dollars for
services that may be unnecessary.
--------------------
\16 HCFA has mandated that contractors use medical necessity
prepayment screens for four procedures (routine foot care, mycotic
nails, chiropractic visits, inpatient rehabilitation medicine visits)
and an injection (Epoetin Alpha). Contractors can request a waiver
to alter or eliminate mandated screens, except screens for mycotic
nails and inpatient rehabilitation medicine visits.
\17 This example is based on echocardiography procedure code 93307,
complete real-time echocardiography with two dimensional image
documentation, with or without M-mode recording (a form of
ultrasound).
\18 The model policy covers noninvasive vascular studies.
CONCLUSIONS
------------------------------------------------------------ Letter :6
While the rapid increase in Medicare costs threatens the long-term
viability of the Medicare program, many Medicare part B contractors
continue to routinely pay claims for widely overused services,
without first determining if the services are reasonable and
necessary. Even when evidence indicates that problems with payments
for specific medical procedures are widespread, HCFA has not ensured
that contractors help correct national problems as well as local
aberrancies. More specifically, HCFA policies do not encourage
contractors to reduce a national norm already inflated by millions of
dollars in payments for unnecessary services.
Our tests of paid claims against criteria used by some of the
contractors show that millions of dollars are being paid for services
that do not meet basic medical necessity criteria. Although our
tests were limited to seven contractors, our survey of 17 contractors
indicates that nationally, additional millions of Medicare dollars
may have been paid for claims that should have been denied.
Prepayment screens are an important tool in preventing payments for
unnecessary services. Funding for program safeguards, such as
medical policies and prepayment screens, has been declining, however,
while the volume of Medicare claims is increasing. In this
environment, autoadjudicated diagnostic screens offer a low-cost way
to ensure that all claims for selected procedures pass a basic
medical necessity test before payment. Greater use of
autoadjudicated screens could complement, rather than replace, the
contractors' efforts to use FMR and other types of prepayment screens
to address local overutilization problems.
To forestall widespread overuse of specific medical procedures, HCFA
can help the contractors much more than it has. HCFA has begun to
capitalize on the knowledge and skills of the contractor medical
directors by using contractor workgroups to develop model medical
policies. More model policies can help contractors control spending
for nationally overused procedures by providing them with generally
accepted criteria for identifying and denying claims for unnecessary
services. However, HCFA needs to support the efforts of the
workgroups and review model policies on a more timely basis so that
these efforts can succeed. Also, to exercise stronger leadership by
promoting best practices, HCFA needs to collect and evaluate
information on the medical criteria and prepayment screens now being
used by the contractors.
RECOMMENDATIONS
------------------------------------------------------------ Letter :7
To help prevent Medicare payments for unnecessary services, we
recommend that the Secretary of HHS direct the Administrator of HCFA
to
systematically analyze national Medicare claims data and use
analyses conducted by HHS' OIG and Medicare contractors to
identify medical procedures that are subject to overuse
nationwide;
gather information on all contractors' local medical policies and
prepayment screens for widely overused procedures, evaluate
their cost and effectiveness, and disseminate information on
model policies and effective prepayment screens to all the
contractors; and
hold the contractors accountable for implementing local policies,
prepayment screens (including autoadjudicated screens), or other
corrective actions to control payments for procedures that are
highly overused nationwide.
AGENCY COMMENTS
------------------------------------------------------------ Letter :8
We provided HHS an opportunity to comment on our draft report, but it
did not provide comments in time to be included in the final report.
However, we did discuss the contents of this report with HCFA
officials from the Bureau of Program Operations, including the
Director of Medical Review and the Medical Officer. In general, they
agreed with our findings.
We obtained written comments on our draft report from several part B
contractor medical directors who serve on the Contractor Medical
Director Steering Committee. We selected this committee as a focal
point for obtaining contractor comments because of its role as a
liaison between the contractors and HCFA and the communication
network for the contractor medical directors. Their comments support
our conclusions (see app. III). In summary, they suggested the
development of contractor workgroups to rapidly produce model medical
policies for the six groups of procedures in our study.
---------------------------------------------------------- Letter :8.1
As agreed with your office, unless you release its contents earlier,
we plan no further distribution of this report for 30 days. At that
time, we will send copies to other congressional committees and
members with an interest in this matter, the Secretary of Health and
Human Services, and the Administrator of the Health Care Financing
Administration. We will also make copies available to others upon
request.
This report was prepared by William Reis, Assistant Director; Teruni
Rosengren; Stephen Licari; Michelle St. Pierre; and Vanessa Taylor
under the direction of Jonathan Ratner, Associate Director. Please
call me on (202) 512-7119 or Mr. Reis on (617) 565-7488 if you or
your staff have any questions about this report.
Sincerely yours,
Sarah F. Jaggar
Director, Health Financing
and Public Health Issues
SCOPE, DATA SOURCES, AND
METHODOLOGY
=========================================================== Appendix I
We reviewed HCFA's statutory authority and responsibilities for
administering the Medicare program and HCFA's regulations and
guidance to contractors on the development of local medical policies
and the implementation of prepayment screens. We also discussed
HCFA's oversight of these functions with officials at its Bureau of
Program Operations.
Before selecting the six groups of medical procedures included in our
study, we reviewed previous GAO and HHS OIG reports, HCFA guidance,
and other studies on overused medical services. We also reviewed
HCFA's list of 200 medical procedure codes, ranked by total
Medicare-allowed charges, and obtained Medicare contractors' views on
procedures that are likely to be overused. Based on the information
gathered from these sources, we selected six groups of procedures
generally considered widely overused.
Because little centralized information exists on Medicare
contractors' use of prepayment screens or the medical necessity
criteria included in those screens, we contacted 17 of the 29
contractors that process Medicare part B claims for physician
services. We also visited three of the Medicare contractors and
attended two of the semiannual contractor medical director
conferences. In the course of these contacts, we decided to limit
our collection of detailed information on medical necessity criteria
and prepayment screens to 17 contractors who could provide us the
information we needed.
To estimate the Medicare payments for unnecessary services that could
be prevented by broader use of prepayment screens, we tested
autoadjudicated prepayment screens on claims paid by seven
contractors in six states. The seven contractors in our analysis
were among the largest contractors in terms of the number of claims
processed in 1993 and they did not use a medical necessity prepayment
screen for some of the six groups of procedures in our study.
We based our tests on data from the Medicare Physician Supplier
Component of the 1993 HCFA 5 Percent Sample Beneficiary Standard
Analytic File. The Physician Supplier Component contains all
Medicare part B claims for a random sample of beneficiaries. Our
analysis is based on all paid claims in the database for the seven
contractors and the six groups of procedures in our review.
For each screen and tested contractor, we estimated the services and
payments that would have been denied by
simulating the screen using a computer algorithm to determine the
number of services in the sample that would have been denied by
the screen,\19
weighing this number to reflect the universe of services, and
multiplying this weighted number by the average Medicare allowance
for the procedure at the contractor.
The average Medicare-allowed amount for each procedure code at each
contractor in 1993 was calculated based on data from HCFA's part B
Extract Summary System.
For five of the procedures, we applied two or three different
autoadjudicated diagnostic screens currently used by other
contractors in order to illustrate the impact of using different
screens. By applying multiple screens, we were able to examine the
range of services that would have been denied depending on the
medical necessity criteria used. For example, one of the colonoscopy
screens paid claims with a diagnosis of gastritis, while another did
not. For YAG laser surgery, however, we only applied the one screen
that we had identified at the time we began our analysis. We only
applied a particular screen to a contractor's claims if that
contractor did not have a medical necessity diagnostic screen in
place in 1993 for the specific procedure being tested. We obtained
our tested screens from several of the 17 contractors in our initial
survey. Some of the screens we used were obtained from one of the
seven contractors that we subsequently tested.
Because our estimates were based on a sample of claims, our estimates
are subject to sampling error. We calculated 95-percent confidence
intervals for each of our estimated payments for services that would
have been denied by the tested screens. This means the chances are
about 19 out of 20 that the actual payments for services that would
have been denied at each of the tested contractors would fall within
the range covered by our estimate, plus or minus the sampling error.
Sampling errors for our estimates are included in appendix II.
Some of the payments that would have been denied by the tested
screens would eventually be paid if they were resubmitted with
corrected or additional information or successfully appealed.
Because contractors' claims processing systems generally do not track
claims denied by autoadjudicated screens to determine how many are
ultimately paid, we developed our own estimates. Using the 1993 HCFA
5 Percent Sample Beneficiary Standard Analytic File, we analyzed
echocardiography claims processed by one contractor and duplex scan
claims processed by another contractor. In each case, the
contractors used autoadjudicated screens for these services. For
each contractor, we used computer programs to identify claims for the
services that were denied for medical necessity in a 3- month period
in 1993. We then determined whether another claim was submitted and
paid for the same service, provided on the same day, for the same
beneficiary, and by the same provider. Our analysis showed that 23
to 25 percent of the echocardiography and duplex scan claims denied
for medical necessity were subsequently paid. Based on these results
we used 25 percent as our estimate of claims denied that would
ultimately be paid. The actual percentage will likely vary by type
of medical procedure and the diagnostic criteria used in the screen.
However, because of the costs and inefficiencies associated with
denying a large percentage of services and then later reprocessing
and paying those services, we believe that contractors would not be
likely to continue using a prepayment screen that inappropriately
denies more than 25 percent of the services.
--------------------
\19 We consulted staff at the contractors whose screens we tested to
ensure that we applied the screens correctly. Also, we manually
reviewed printouts of the claims that were denied by the tested
screens to ensure that only nonpayable diagnoses were denied.
ESTIMATED PAYMENTS FOR VARIOUS
PROCEDURES DENIED BY SELECTED
PREPAYMENT SCREENS (1993)
========================================================== Appendix II
Table II.1
Estimated Payments for Echocardiography
Error
Prepayment range
screens\a and Payments for
tested Paid Denied for denied payments
contractors\b services services services\c \d
---------------- -------- -------- ---------- --------
Screen 1
----------------------------------------------------------
Contractor A 388,200 40,600 $3,608,000 �$146,30
0
Contractor B 363,900 271,900 18,949,700 �163,600
Contractor D 104,800 14,600 900,400 �57,700
Contractor F 161,200 114,700 9,883,000 �139,700
Contractor G 213,500 149,600 11,836,200 �153,000
==========================================================
Total 1,231,60 591,400 $45,177,30 �$660,30
0 0 0
Screen 2
----------------------------------------------------------
Contractor A 749,900 34,400 $5,885,600 �$268,60
0
Contractor B 619,800 153,200 17,218,700 �362,600
Contractor D 269,900 31,200 2,976,900 �134,500
Contractor F 296,000 67,000 8,583,400 �275,500
Contractor G 392,600 72,600 8,097,500 �244,200
==========================================================
Total 2,328,20 358,400 $42,762,10 �$1,285,
0 0 400
Screen 3
----------------------------------------------------------
Contractor A 749,900 3,800 $511,900 �$70,800
Contractor B 619,800 73,200 8,051,300 �265,700
Contractor D 269,900 3,300 271,900 �40,000
Contractor F 296,000 2,600 321,600 �53,500
Contractor G 392,600 11,800 1,317,900 �107,100
==========================================================
Total 2,328,20 94,700 $10,474,60 �$537,10
0 0 0
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a The prepayment screens presented in the table apply to different
echocardiography codes. Screen 1 was used for echocardiography codes
93320 and 93325. Screens 2 and 3 were used for echocardiography
codes 93307, 93320, 93325, and 93350. Therefore, screen 1 was
applied to a smaller number of paid services than the other screens.
\b Two of the seven contractors in our study had medical necessity
screens to identify unnecessary echocardiography tests, therefore,
those two contractors were not included in this analysis.
\c Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\d The error range for estimated payments was based on a 95-percent
confidence level.
Table II.2
Estimated Payments for Eye Examinations
Error
Prepayment range
screens\a and Payments for
tested Paid Denied for denied payments
contractor\b services services services\c \d
---------------- -------- -------- ---------- --------
Screen 1
----------------------------------------------------------
Contractor A 1,271,60 1,100 $51,000 �$13,400
0
Contractor D 499,400 3,700 164,500 �22,900
Contractor E 527,300 5,500 305,800 �35,000
Contractor G 838,700 8,300 409,700 �38,700
==========================================================
Total 3,137,00 18,600 $931,000 �$110,00
0 0
Screen 2
----------------------------------------------------------
Contractor A 1,271,60 600 $28,800 �$10,100
0
Contractor D 499,400 300 15,000 �6,900
Contractor E 527,300 3,800 212,400 �29,300
Contractor G 838,700 7,200 355,100 �35,700
==========================================================
Total 3,137,00 11,900 $611,300 �$82,000
0
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a The prepayment screens presented in the table were used for eye
examination codes 92002, 92004, 92012, and 92014.
\b Three of the seven contractors in our study had medical necessity
screens to identify unnecessary eye examinations, therefore, those
three contractors were not included in this analysis.
\c Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\d The error range for the estimated payments was based on a 95-
percent confidence level.
Table II.3
Estimated Payments for Chest X Rays
Error
Prepayment range
screens\a and Payments for
tested Paid Denied for denied payments
contractors\b services services services\c \d
---------------- -------- -------- ---------- --------
Screen 1
----------------------------------------------------------
Contractor B 2,146,00 597,400 $7,708,300 �$73,100
0
Contractor C 2,147,20 530,100 7,914,000 �83,000
0
Contractor D 1,734,50 693,400 9,288,100 �75,300
0
Contractor G 1,436,10 617,100 10,701,400 �91,900
0
==========================================================
Total 7,463,80 2,438,00 $35,611,80 �$323,30
0 0 0 0
Screen 2
----------------------------------------------------------
Contractor B 2,146,00 654,900 $8,449,700 �$75,100
0
Contractor C 2,147,20 521,400 7,743,600 �82,200
0
Contractor D 1,734,50 441,900 6,017,500 �67,400
0
Contractor G 1,436,10 669,500 11,514,200 �92,600
0
==========================================================
Total 7,463,80 2,287,70 $33,725,00 �$317,30
0 0 0 0
Screen 3
----------------------------------------------------------
Contractor B 2,146,00 10,200 $136,400 �$11,600
0
Contractor C 2,147,20 8,900 132,000 �11,900
0
Contractor D 1,734,50 4,200 55,900 �7,400
0
Contractor G 1,436,10 23,300 494,300 �27,500
0
==========================================================
Total 7,463,80 46,600 $818,600 �$58,400
0
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a The prepayment screens presented in the table were used for chest
X ray codes 71010 and 71020.
\b Three of the seven contractors in our study had medical necessity
screens for chest X rays, therefore, those three contractors were not
included in this analysis.
\c Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\d The error range for estimated payments was based on a 95-percent
confidence level.
Table II.4
Estimated Payments for Colonoscopy
Error
Prepayment range
screens\a and Payments for
tested Paid Denied for denied payments
contractors\ services services services\b \c
---------------- -------- -------- ---------- --------
Screen 1
----------------------------------------------------------
Contractor A 119,700 46,100 $13,291,70 �$418,10
0 0
Contractor B 83,600 37,600 11,172,800 �368,600
Contractor C 70,400 28,300 7,613,500 �298,900
Contractor D 59,200 24,900 7,812,400 �325,100
Contractor E 46,500 21,400 7,488,400 �353,800
Contractor F 47,400 17,800 5,913,900 �313,800
Contractor G 64,800 24,100 8,593,700 �375,100
==========================================================
Total 491,600 200,200 $61,886,40 �$2,453,
0 400
Screen 2
----------------------------------------------------------
Contractor A 119,700 25,600 $7,878,400 �$396,30
0
Contractor B 83,600 23,100 7,050,400 �348,000
Contractor C 70,400 17,500 4,875,700 �282,100
Contractor D 59,200 15,000 4,829,200 �298,500
Contractor E 46,500 14,500 5,231,800 �336,600
Contractor F 47,400 12,300 4,218,300 �294,100
Contractor G 64,800 15,300 5,658,500 �353,200
==========================================================
Total 491,600 123,300 $39,742,30 �$2,308,
0 800
Screen 3
----------------------------------------------------------
Contractor A 119,700 3,700 $1,332,200 �$184,10
0
Contractor B 83,600 3,700 1,227,600 �167,900
Contractor C 70,400 4,700 1,344,100 �167,200
Contractor D 59,200 800 283,300 �84,000
Contractor E 46,500 1,600 647,000 �137,100
Contractor F 47,400 900 333,500 �94,100
Contractor G 64,800 1,600 630,500 �134,800
==========================================================
Total 491,600 17,000 $5,798,200 �$969,30
0
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a The prepayment screens presented in the table were used for
colonoscopy codes 45378, 45380, and 45385.
\b Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\c The error range for estimated payments was based on a 95-percent
confidence level.
Table II.5
Estimated Payments for YAG Laser Surgery
Error
range
Payments for
Tested Paid Denied for denied payments
contractors\a services services services\b \c
---------------- -------- -------- ---------- --------
Contractor A 57,700 9,000 $2,813,200 �$239,70
0
Contractor B 31,100 2,900 1,054,900 �159,000
Contractor C 67,100 23,600 4,355,000 �200,900
Contractor D 25,300 2,400 1,094,700 �183,400
Contractor E 26,100 3,300 2,134,300 �304,700
Contractor F 23,900 2,900 1,314,500 �200,000
Contractor G 19,100 3,200 2,167,800 �302,200
==========================================================
Total 250,300 47,300 $14,934,40 �$1,589,
0 900
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a Only one prepayment screen was used.
\b Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\c The error range for estimated payments was based on a 95-percent
confidence level.
Table II.6
Estimated Payments for Duplex Scans of
Extracranial Arteries
Error
Prepayment range
screens\a and Payments for
tested Paid Denied for denied payments
contractors\b services services services\c \d
---------------- -------- -------- ---------- --------
Screen 1
----------------------------------------------------------
Contractor B 101,200 53,000 $7,536,700 �$193,70
0
Contractor F 55,800 25,400 3,173,000 �126,700
==========================================================
Total 157,000 78,400 $10,709,70 �$320,40
0 0
Screen 2
----------------------------------------------------------
Contractor B 101,200 28,300 $4,021,600 �$173,70
0
Contractor F 55,800 15,600 1,949,300 �113,700
==========================================================
Total 157,000 43,900 $5,970,900 �$287,40
0
----------------------------------------------------------
Notes: All numbers are rounded to hundreds.
The estimated number of and payments for denied services were derived
from a 5-percent beneficiary sample of 1993 claims for each
contractor.
\a The prepayment screens in the table were used for duplex scan code
93880.
\b Five of the seven contractors in our study had medical necessity
screens to identify unnecessary duplex scans, therefore, those five
contractors were not included in this analysis.
\c Estimated payments for denied services were calculated by
multiplying the estimated number of denied services by the average
Medicare allowance for the procedure at the contractor.
\d The error range for estimated payments was based on a 95-percent
confidence level.
(See figure in printed edition.)Appendix III
COMMENTS FROM THE CONTRACTOR
MEDICAL DIRECTOR STEERING
COMMITTEE
========================================================== Appendix II
(See figure in printed edition.)
*** End of document. ***