Medicare: Concerns With Physicians at Teaching Hospitals (PATH) Audits
(Letter Report, 07/23/98, GAO/HEHS-98-174).

Pursuant to a congressional request, GAO reviewed the Physicians at
Teaching Hospitals (PATH) initiative, focusing on: (1) whether the
Department of Health and Human Services' (HHS) Office of the Inspector
General (OIG) has a legal basis for conducting PATH audits; (2) whether
the OIG has followed an acceptable approach and methodology in
conducting the audits; and (3) the significance of the billing problems
identified in selected audits.

GAO noted that: (1) HHS' OIG does have a legal basis for applying the
specific criteria used in the PATH initiative; (2) the need for a
teaching physician to be physically present to bill for services
performed by residents is a longstanding requirement of Medicare; (3)
the fact that a physical presence requirement has not always been
consistently communicated or enforced does not obviate the need for
teaching physicians to document their personal involvement in services
to legitimately bill Medicare; (4) although detailed guidance for
documenting evaluation and management code was not effective until 1996,
the definitions of these codes and instructions for their use have been
available since the codes were implemented in 1992 and provided the
standard for the PATH initiative; (5) OIG's methodology on the three
audits reviewed was reasonable; (6) the criteria OIG used to assess
teaching physicians' involvement in services performed by residents were
consistent with statutory and regulatory requirements; (7) the criteria
used by OIG were no different from the information already provided to
the teaching hospitals by their carriers; (8) GAO saw no evidence that
OIG or the medical reviewers applied documentation guidance when
assessing the level of teaching physician care billed to Medicare during
the prior periods covered by the audits; (9) the results of one of the
audits examined raises questions about OIG's original intent to audit
all major teaching hospitals; (10) auditing every major teaching
hospital would be time-consuming and expensive for OIG, the carrier, and
the institutions involved; (11) a risk-based approach focusing on the
most problem-prone institutions would be a more efficient use of these
resources; (12) OIG reduced the number of institutions to be audited,
due to competing demands and other factors, but neither its original
intent nor its recent decision to reduce the number of audits used a
risk-based approach; (13) a substantial difference existed between the
billing errors identified by OIG and the amounts the institutions
ultimately agreed to repay; and (14) in essence, the Department of
Justice used OIG's audit results related to inpatient services for a
single year to estimate potential false claims for all Medicare part B
services for multiple years.

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

 REPORTNUM:  HEHS-98-174
     TITLE:  Medicare: Concerns With Physicians at Teaching Hospitals 
             (PATH) Audits
      DATE:  07/23/98
   SUBJECT:  Medical education
             Medical fees
             Patient care services
             Fraud
             Physicians
             Audits
             Hospitals
             Audit authority
IDENTIFIER:  HCFA Common Procedure Coding System
             Medicare Program
             HHS/DOJ Physicians at Teaching Hospitals Initiative
             
******************************************************************
** 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, Subcommittee on Health, Committee on Ways and
Means, House of Representatives

July 1998

MEDICARE - CONCERNS WITH
PHYSICIANS AT TEACHING HOSPITALS
(PATH) AUDITS

GAO/HEHS-98-174

Medicare PATH Audits

(101589)


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

  AAMC - Association of American Medical Colleges
  CPT - Current Procedural Terminology
  DOJ - Department of Justice
  HCFA - Health Care Financing Administration
  HHS - Department of Health and Human Services
  OIG - Office of Inspector General
  PATH - Physicians at Teaching Hospitals

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


B-278015

July 23, 1998

The Honorable Bill Thomas
Chairman, Subcommittee on Health
Committee on Ways and Means
House of Representatives

Dear Mr.  Chairman: 

About 1,200 hospitals in the United States have graduate medical
education programs for training physicians in medical specialties
after they have completed medical school.  These hospitals are known
as teaching hospitals; the physicians in training are known as
residents.  Residents receive specialized training in a particular
area of medicine and provide patient care under the supervision of a
teaching physician.  Teaching physicians are faculty members who
train and supervise residents.  Their functions may include classroom
instruction, making rounds with residents, examining specific
patients, and discussing courses of treatment. 

In December 1995, the University of Pennsylvania, without admitting
wrongdoing, entered into a voluntary settlement with the Department
of Justice (DOJ), agreeing to pay about $30 million in disputed
billings and damages for Medicare billings by teaching physicians. 
This settlement resulted from an audit performed by the Office of
Inspector General (OIG) in the Department of Health and Human
Services (HHS).  In the audit, the OIG concluded that some of the
university's teaching physicians had inappropriately billed Medicare
because medical records did not adequately document their involvement
in services provided by residents.  The audit also determined that
some teaching physicians had "upcoded" their claims--that is, billed
for more complex and, therefore, more expensive services than may
have been provided. 

Concerned that such problems might be widespread, HHS' OIG, in
cooperation with DOJ, instituted a nationwide initiative--now
commonly known as Physicians at Teaching Hospitals (PATH) audits--to
review teaching physician compliance with Medicare billing rules.\1
As of April 30, 1998, five additional PATH audits have been resolved. 
(See table 1.) In three of these cases, the institutions reached
settlements with DOJ totaling more than $37 million.  In the two
other cases, no significant errors were found; consequently, no
enforcement action was taken.  Currently, PATH audits are either
planned or under way at 37 other institutions. 



                          Table 1
          
            Resolved PATH Audits as of April 30,
                            1998

                                                Settlement
                                Date                amount
Institution                     resolved        (millions)
------------------------------  ------------  ------------
University of Pennsylvania      December             $30.0
                                 1995
Thomas Jefferson University     August 1996           12.0
Dartmouth-Hitchcock Medical     April 1997            None
 Center
Yale University                 October 1997          None
University of Virginia          November               8.6
                                 1997
University of Pittsburgh        March 1998            17.0
==========================================================
Total                                                $67.6
----------------------------------------------------------
The PATH initiative has generated considerable controversy.  The
academic medical community disagrees with HHS' OIG regarding the
billing and documentation standards that were in effect during the
time periods under review.  The medical community also contends that
DOJ is coercing settlements from teaching institutions through
threats of federal lawsuits.  On October 29, 1997, the Association of
American Medical Colleges (AAMC) and other medical associations,
specialty societies, and medical schools filed a complaint with the
U.S.  District Court for the Central District of California, seeking
to end the PATH initiative.  In addition, the Greater New York
Hospital Association, along with several New York medical schools and
teaching hospitals, filed a similar lawsuit in federal court on April
16, 1998.  AAMC's lawsuit was dismissed on April 27, 1998, for lack
of jurisdiction because no actual enforcement action was being
challenged.\2 The other lawsuit is pending. 

This letter responds to your July 14, 1997, letter and subsequent
discussions with your staff requesting that we examine the PATH
initiative.  Specifically, you asked us to determine (1) whether HHS'
OIG has a legal basis for conducting PATH audits, (2) whether the OIG
has followed an acceptable approach and methodology in conducting the
audits, and (3) the significance of the billing problems identified
in selected audits. 

To address your questions, we examined the laws, regulations, and
guidance related to teaching physician billing for Medicare services. 
To understand the OIG's approach and methodology and the nature and
extent of billing errors it found, we examined the OIG's workpapers
related to the audits conducted at the University of Pennsylvania
(Penn), Thomas Jefferson University (Jefferson), and the
Dartmouth-Hitchcock Medical Center (Dartmouth), the first three
resolved PATH audits.  We discussed the PATH initiative with staff
from OIG headquarters and field offices as well as with the Medicare
carriers who worked on these audits.\3 We also met with
representatives from the three audited institutions to obtain their
perspectives on the PATH initiative.  DOJ would not permit us to
interview key officials who negotiated the financial settlements with
Penn and Jefferson because it said certain matters related to these
institutions were still pending.  DOJ, however, did respond in
writing to questions we posed about its role in the PATH initiative
and the Penn and Jefferson settlements.  In addition, we met with
representatives from AAMC and the American Hospital Association. 

In addition to this letter, we are sending you a separate, complete
report on this subject, which we designated as "limited official use"
and which should not be further distributed.  The complete report
contains information about specific audit findings at Penn and
Jefferson that DOJ has identified as being subject to confidentiality
agreements with the audited institutions.  This letter does not
contain such information and is therefore publicly available.  Our
work was performed between August 1997 and June 1998 in accordance
with generally accepted government auditing standards.  Our scope and
methodology are discussed in more detail in appendix I. 


--------------------
\1 When we refer to a specific PATH audit in this report, we will use
the name of the teaching institution.  PATH actually involves an
audit of the entities that submit teaching physician billings to
Medicare--typically physician group practice plans that are
components or affiliates of the teaching institution. 

\2 AAMC appealed the dismissal on June 23, 1998. 

\3 Carriers are insurance companies that contract with the government
to process and pay Medicare claims. 


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

HHS' OIG, in our opinion, does have a legal basis for applying the
specific criteria used in the PATH initiative.  Our analysis
indicates that the need for a teaching physician to be physically
present to bill for services performed by residents is a longstanding
requirement of the Medicare program.  The fact that a physical
presence requirement has not always been consistently communicated or
enforced does not obviate the need for teaching physicians to
document their personal involvement in services to legitimately bill
Medicare.  Furthermore, although detailed guidance for documenting
evaluation and management codes--the codes physicians use to bill
Medicare for certain services--was not effective until 1996, the
definitions of these codes and instructions for their use have been
available since the codes were implemented in 1992 and provided the
standard for the PATH initiative. 

We also found that the OIG's methodology on the three audits we
reviewed was reasonable.  The criteria the OIG used to assess
teaching physicians' involvement in services performed by residents
were consistent with statutory and regulatory requirements. 
Moreover, the criteria used by the OIG were no different from the
information already provided to the three teaching hospitals by their
carriers.  Similarly, we saw no evidence that the OIG or the medical
reviewers who assisted them retroactively applied documentation
guidance when assessing the level of teaching physician care billed
to Medicare during the prior periods covered by the audits. 

However, the results of one of the audits we examined--the audit of
the Dartmouth-Hitchcock Medical Center--raises questions about the
OIG's original intent to audit all major teaching hospitals. 
Auditing every major teaching hospital would be time-consuming and
expensive for the OIG, the carrier, and the institutions involved. 
We believe that a risk-based approach focusing on the most
problem-prone institutions would be a more efficient use of these
resources.  The OIG reduced the number of institutions to be audited,
due to competing demands and other factors, but neither its original
intent to audit all major teaching institutions nor its recent
decision to reduce the number of audits used a risk-based approach. 
The Dartmouth audit was initiated with little indication that the
institution was improperly billing Medicare.  The audit--which took
10 months and, according to Dartmouth, cost the institution about
$1.7 million in direct and indirect costs--identified billing errors
totaling $778. 

While the billing errors found at Dartmouth were immaterial, the
errors found by the OIG for the other two audits we reviewed--the
audits at the University of Pennsylvania and Thomas Jefferson
University--were more significant and resulted in referrals to DOJ. 
A substantial difference existed between the billing errors
identified by the OIG and the amounts the institutions ultimately
agreed to repay.  These amounts, however, were the outcome of
negotiations between the institutions and DOJ in an effort to avoid
litigation.  In essence, DOJ used the OIG's audit results related to
inpatient services for a single year to estimate potential false
claims for all Medicare part B services for multiple years.  While
the medical community has criticized this multiyear extrapolation, it
is not improper in the context of settlement negotiations.  Although
the institutions and DOJ did not discuss these negotiations in
detail, DOJ said it could have asked the OIG to expand the audit to
other time periods.  Representatives from the two institutions told
us that the applicable damages and penalties--if the institutions
were found liable by a court for submitting false claims--were of
great concern and influenced their decisions to agree to a settlement
with DOJ. 

The OIG found that Penn and Jefferson teaching physicians had not
always complied with Medicare billing requirements.  Based on our
review of the OIG's workpapers, however, these problems did not
appear to be as serious as the OIG has categorized them in public
statements since the settlements were reached.  The OIG has implied
that these audits found instances of billing by teaching physicians
on days they were not working and has also said that most upcoding
errors were multilevel.  The workpapers for the Penn and Jefferson
audits do not support these statements. 


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

Medicare covers almost all people aged 65 and over and certain
disabled people.\4

Administered by the Health Care Financing Administration (HCFA)
within HHS, the program has two components--hospital insurance (part
A) and supplementary medical insurance (part B).  Inpatient hospital
services, home health services, and certain other institutionally
based services are covered by part A of the program.  Part B covers
physician services, outpatient services, and various other medical
and health services. 

Medicare pays teaching hospitals for part of the costs of graduate
medical education under part A of the program.  These payments are
intended to cover a portion of teaching physicians' salaries, related
to the time they spend teaching residents.  Medicare part A also pays
a portion of the residents' salaries.  In total, Medicare paid
teaching hospitals about $8 billion in 1996 for costs associated with
the training of residents. 

Teaching physicians can also receive Medicare funds from part B of
the program when they personally provide services to Medicare
beneficiaries and, in certain circumstances, when a resident provides
services under the personal supervision of the teaching physician. 
Physicians claim part B reimbursement using five-digit codes,
developed by the American Medical Association, which indicate the
level of care provided.  For example, initial inpatient consultations
can be billed at five different levels ranging from 99251 to 99255. 
Generally, the higher the code, the higher the degree and complexity
of the service or level of care and the higher the Medicare
reimbursement. 

These two methods of paying teaching physicians have been a
longstanding concern because of the danger that Medicare will pay
twice for the same service--once as a hospital payment under part A
and again as a separately billed service under part B.  Twice
previously--in November 1971 and again in January 1986--we reported
problems with part B claims for services provided by teaching
physicians.\5 Both times we found that a significant number of the
claims we reviewed--67 percent in 1971 and 49 percent in 1986--did
not adequately document the teaching physicians' presence in services
performed by residents, raising the possibility that duplicate
payments had been made by Medicare. 

The OIG's PATH initiative stems from the continuing concern over part
B billings by physicians in a teaching setting.  Institutions
selected for a PATH audit are given the option of conducting
self-audits at their own expense, using independent external auditors
or consultants approved and supervised by the OIG.  Audits conducted
by the OIG are called PATH I, while self-audits are referred to as
PATH II.  All of the audits completed or under way since the initial
OIG audit at Penn have been PATH II reviews. 

PATH audits focus on two major areas of concern.  The first concern
is whether teaching physicians who billed part B for services
furnished by residents provided sufficient "personal direction" in
the delivery of the service.  The OIG considers that the requirement
for sufficient personal direction is met if the physician was
physically present while the service was delivered.  If the medical
records do not show evidence of the teaching physician's presence,
the OIG considers the service to be part of the teaching physician's
supervisory functions already paid under part A.  The second concern
is whether teaching physicians have inflated their part B claims by
"upcoding," that is, billing using a code that is one or more levels
higher than the level of service that was actually performed. 
Level-of-service determinations are made by carrier medical reviewers
on PATH I audits or by independent medical reviewers on PATH II
reviews.  Carrier medical review staff also assist the OIG in
monitoring the work of the external reviewers on PATH II audits. 

The cornerstone of a PATH audit is an examination of the medical
records and other documentation related to a random sample of
inpatient admissions for a 12-month period.  The results are shared
with the local U.S.  Attorney's Office, which evaluates the OIG's
findings and considers whether criminal or civil action is warranted,
including the filing of a civil lawsuit under the False Claims Act.\6
Every audit is started with the intention of reviewing the entire
sample for billing errors so as to estimate a total overpayment for
the 12-month period.  However, the audit may be terminated earlier
if, for example, the OIG concludes the errors being identified are
immaterial or if the institution decides it wants to stop the audit
and discuss a possible settlement with DOJ in an effort to limit its
False Claims Act liability.  The OIG told us that the only resolved
audit in which all sampled services were reviewed was the Penn audit. 

With the increased attention to health care fraud and abuse in recent
years, the government may now invoke the penalties and damages
prescribed in the False Claims Act for practices that in the past
might have been dealt with by seeking repayment.  The False Claims
Act has become one of the government's primary enforcement tools
because of its deterrent effect.  The act provides that anyone who
knowingly submits false claims to the government is liable for three
times the amount of damages plus a mandatory penalty of $5,000 to
$10,000 for each false claim.  The term "knowingly" is broadly
defined to mean that a person (1) has actual knowledge of the false
claim, (2) acts in deliberate ignorance of the truth or falsity of
the information, or (3) acts in reckless disregard of the truth or
falsity of the information.  In the health care setting, where
providers submit thousands of claims each year, the potential damages
and penalties provided under the False Claims Act can add up quickly. 


--------------------
\4 Medicare is authorized by title XVIII of the Social Security Act
(42 U.S.C.  sections 1395 et seq.). 

\5 See Problems in Paying for Services of Supervisory and Teaching
Physicians in Hospitals Under Medicare (B-164031(4), Nov.  17, 1971)
and Medicare:  Documenting Teaching Physician Services Still a
Problem (GAO/HRD-86-36, Jan.  21, 1986). 

\6 31 U.S.C.  sections 3729 et seq. 


   OIG HAS LEGAL BASIS FOR
   APPLYING THE CRITERIA USED IN
   PATH AUDITS
------------------------------------------------------------ Letter :3

Despite the concerns raised by representatives of the academic
medical community, HHS' OIG has legal authority to apply the
physician presence and coding criteria it is using in the PATH
initiative.  Although HCFA guidance has created some confusion,
federal Medicare law has long required that physician services be
rendered or supervised by the physician in person.  Similarly,
despite recognition that evaluation and management coding guidance
needed clarification, physicians have always been required to bill
only for services performed and to comply with billing guidance in
effect at the time. 

In our 1986 report, we found that a teaching physician's claim for
reimbursement required documentation in the patient's medical records
that the teaching physician either personally provided the service or
was present when the service was provided by a resident.  After
considering additional legislative activity and materials as well as
subsequent communications with HCFA, we continue to hold this view. 
Our report also indicated that carriers varied in their requirements
for the documentation of physician presence and that Medicare's
statement of this policy needed to be clarified.  Recognizing these
problems, the OIG has limited PATH audits to teaching hospitals that
received clear guidance from their Medicare carriers on documenting
physician presence. 

The OIG also has authority to audit the claims of physicians for
evaluation and management services from 1992 to 1995.  During these
years, Medicare required physicians to accurately code their services
in order to receive reimbursement.  Guidance in effect during that
period provided the relevant definitions for determining the
appropriateness of such coding.  The OIG has indicated that, in its
reviews, it applies only the code documentation guidance in effect
for the period being audited. 


      TEACHING PHYSICIANS MUST BE
      PHYSICALLY PRESENT TO BILL
      MEDICARE
---------------------------------------------------------- Letter :3.1

In 1966, within months of the Medicare program's inception, the
Medicare agency promulgated rules establishing principles of
reimbursement for services by hospital-based physicians.  One such
principle was that these services may be reimbursed under Medicare
part B if the physician provides "an identifiable service requiring
performance by a physician in person" (emphasis added).\7 The
following year, the agency promulgated rules specifically pertaining
to the reimbursement of attending physicians' services rendered in a
teaching setting.  These rules provided for payment where the
"physician provides personal and identifiable direction to interns or
residents who are participating in the care of his patient" (emphasis
added).\8

In 1969, Medicare issued specific guidance establishing conditions
for part B payments to supervising physicians in a teaching setting. 
This guidance--Bureau of Health Insurance, Intermediary Letter No. 
372 (IL-372)--has been central to analyses of the supervising
physician payment issue and the subject of much controversy.  Under
IL-372, to be reimbursed under part B, a teaching physician must be
the patient's "attending physician."\9 To be recognized as such, the
physician must "render sufficient personal and identifiable medical
services to the Medicare beneficiary to exercise full, personal
control over the management of the portion of the case for which a
charge can be recognized."\10

To exercise such control, the teaching physician must, among other
things, either actually perform the services required by the patient
or supervise the treatment provided by others so as to ensure that
appropriate services and quality care are provided.\11 The provision
of personal and identifiable services must be substantiated by
recordings entered by the physician in the patient's chart.\12 In
1970, further guidance was provided in Intermediary Letter No.  70-2,
indicating that a physician's status as "attending" is important
where medical or surgical services are performed in his presence.\13

In 1980, the Congress enacted a statute governing carrier
documentation requirements for part B payments for teaching physician
services.  In language similar to IL-372, it states that a carrier
shall not pay for physicians' services provided to patients under an
approved teaching program unless the physician "renders sufficient
personal and identifiable physicians' services to the patient to
exercise full, personal control over the management of the portion of
the case for which the payment is sought."\14 This provision was in
effect during the PATH initiative. 

The conference committee for this law explicitly endorsed the IL-372
guidance for documenting payment by teaching physicians.\15 Also, the
House Budget Committee (whose provision establishing criteria for
payment of teaching physicians was adopted in conference) stated that
it "strongly believes teaching physicians should personally perform
or personally supervise patient services in order to qualify for
fee-for-service payment" (emphasis added).\16

Two years later, the Congress enacted a law directing the Medicare
agency to promulgate regulations to distinguish between professional
medical services that are reimbursable under part B and those that
are not.\17 The Senate Finance Committee (whose provision calling for
regulations was subsequently adopted in conference) stated

     Under current law and regulations, services furnished by a
     physician to hospital inpatients are reimbursed on the basis of
     reasonable charges under part B only if such services are
     identifiable professional services to patients that require
     performance by physicians in person and which contribute to the
     diagnosis or treatment of individual patients.\18 (Emphasis
     added.)

The Committee's language underscores its understanding that part B
payment occurs only as a result of physicians rendering professional
services in person. 

On the basis of a review of applicable law and guidance, our 1986
report concluded that a teaching physician's claim for part B
reimbursement required documentation in a patient's medical records
that the teaching physician either personally provided the service or
was present when the service was provided by a resident.\19 However,
our 1986 report found that HCFA had failed to adequately communicate
these and other documentation requirements to providers and that the
documentation actually required by carriers varied substantially.  At
the time, we recommended that HCFA promulgate rules to clarify the
matter.  No such national rules were finalized until December 8,
1995--about 10 years after the issuance of our report. 

During those 10 years, a number of agency communications appear to
have contributed to confusion over Medicare's enforcement policy. 
For example, on December 30, 1992, the Director of HCFA's Office of
Payment Policy distributed a memorandum to HCFA regional offices
clarifying that teaching physicians must be physically present during
all procedures in order to receive part B reimbursement.  In response
to negative provider reactions to this memo, HCFA distributed an
internal memo in July 1994 stating that the instructions on the
physical presence requirement in IL-372 are "admittedly ambiguous and
have not been vigorously enforced." In April 1995, the Director of
HCFA's Bureau of Policy Development wrote to an attorney representing
Medicare providers that carriers that did not apply a "physician
presence" requirement prior to the December 1992 memo should not
institute such a policy until HCFA could issue a final rule on the
subject.  And on December 8, 1995, in the preamble to the new rules
pertaining to Medicare reimbursement in a teaching setting, HCFA
noted that while IL-372 and related issuances specifically stated
"that the attending physician had to be present when a major surgical
procedure or a complex or dangerous medical procedure was performed,"
the guidance was "vague, perhaps necessarily, on the matter of the
presence of the physician during other occasions of inpatient
service."

We view these communications as illustrating what we had concluded in
our 1986 report:  HCFA enforcement policy for reimbursing teaching
physicians under part B was not clearly communicated or consistently
enforced.  However, notwithstanding poor communication and
inconsistency in enforcement, Medicare law required documentation of
physical presence by a teaching physician for part B reimbursement. 
Accordingly, HCFA made clear in the preamble to its 1995 rules that,
despite misunderstandings resulting from IL-372, prior agency policy
had been to require teaching physician presence for all part B
billings.\20 A substantial number of carriers have correctly enforced
this requirement. 

In recognition of this confusion and its potential effect on teaching
physician compliance with Medicare billing rules, the General Counsel
of HHS sent a letter to representatives of the academic medical
community responding to concerns raised over the ongoing PATH audits. 
The July 11, 1997, letter indicated that HHS' policy for enforcing
the physician presence standard in the PATH audits would be
determined by evidence of communications between Medicare carriers
and providers, such that the OIG "will undertake PATH audits only
where carriers, before December 30, 1992, issued clear explanations
of the rules regarding reimbursement for the services of teaching
physicians."


--------------------
\7 20 C.F.R.  section 405.483(a) (1966). 

\8 20 C.F.R.  section 405.521 (1967). 

\9 IL-372 (1969), p.  1. 

\10 IL-372, p.  1. 

\11 IL-372, p.  1. 

\12 IL-372, p.  6. 

\13 IL-70-2 (1970), p.  B-1. 

\14 P.L.  96-499, section 9, Dec.  5, 1980; classified to 42 U.S.C. 
section 1395u(b)(7)(A)(i). 

\15 H.R.  No.  96-1479 (1980), pp.  145-46. 

\16 H.R.  No.  96-1167 (1980), pp.  69-70. 

\17 P.L.  97-248, section 109(a), Sept.  3, 1982; classified to 42
U.S.C.  section 1395xx(a).



\18 S.R.  No.  97-494 (1982), pp.  21-22. 

\19 GAO/HRD-86-36, Jan.  21, 1986. 

\20 60 Fed.  Reg.  236 (Dec.  8, 1995), p.  63139. 


      OIG MAY AUDIT EVALUATION AND
      MANAGEMENT CODES TO
      DETERMINE PROGRAM COMPLIANCE
---------------------------------------------------------- Letter :3.2

To maintain consistency in billing for physicians' services, HCFA
uses a national uniform procedure coding system known as the HCFA
Common Procedure Coding System.  Since 1983, this system has
incorporated the American Medical Association's Current Procedural
Terminology, commonly referred to as the CPT, a list of descriptive
terms and identifying codes for reporting medical services and
procedures performed by physicians. 

In 1992, the CPT was significantly affected by the scheduled
implementation of the Omnibus Budget Reconciliation Act of 1989,
which required the imposition of a Medicare fee schedule for
physicians' services based on the lesser of the actual charge for the
service or an amount determined under a resource-based relative-value
fee schedule.\21

This resulted in a complete revision of CPT codes for evaluation and
management services.\22 The 1992 CPT provided definitions or
explanations of the various levels of evaluation and management
services; ultimately, more clarity was provided by the publication of
guidelines, effective August 1995, on how to use and interpret the
codes in order to document services. 

Notwithstanding the subsequent publication of clarifying guidance,
from 1992 to 1995, Medicare required physicians to code their
services in order to receive reimbursement, and the CPTs for 1992
through 1995 provided definitions for determining the appropriateness
of such coding.  HHS' OIG has indicated that it applies code
definitions appropriate for the period being audited and provides
institutions an opportunity to review and contest findings of
suspected upcoding. 


--------------------
\21 P.L.  No.  101-239, section 6102, 103 Stat.  2106, 2169-2189
(1989); classified to 42 U.S.C.  section 1395w-4. 

\22 The 1992 CPT provides codes for six different types of services: 
evaluation and management, anesthesia, surgery, radiology, pathology,
and laboratory and medicine.  Evaluation and management services
typically involve obtaining the patient's relevant medical history, a
physical examination, and medical decisionmaking and counseling. 


   SELECTION OF INSTITUTIONS FOR
   PATH AUDITS MAY BE QUESTIONABLE
------------------------------------------------------------ Letter :4

HHS' OIG told us that when it began PATH, the intention was to audit
the major teaching hospital or faculty practice plan associated with
each of the nation's 125 medical schools.  The OIG selected these
institutions because, of the nation's 1,200 teaching hospitals, these
institutions had the greatest number of residents and received the
most Medicare revenue.  The OIG and the Medicare carriers, however,
lacked the resources to conduct all PATH audits; thus, the OIG has
offered the teaching institutions the option of hiring at their own
expense external auditors to conduct the audits.\23 Nevertheless, the
OIG and Medicare carrier staff remain involved by actively monitoring
the external auditor's work.  In July 1997, 1 year after the PATH
initiative was announced, audits were under way or planned at 49
institutions. 

This blanket approach to auditing teaching physician billing
practices may not be the most efficient use of OIG, carrier, or
teaching hospital resources.  Because audits are time-consuming and
expensive, the number of audits that can realistically be done is
limited.  While targeting the largest teaching institutions in the
country was a reasonable first step, a risk-based approach to
prioritizing PATH audits could have enabled the OIG to target
institutions most likely in violation of teaching physician billing
rules and concentrate its efforts on these institutions.\24

The OIG recently told us it intends to reduce the number of teaching
institutions that will be audited but plans to complete ongoing
audits.  The OIG attributed the reduction to competing demands for
its resources and determinations that some carriers did not provide
clear communications regarding teaching physician physical presence
rules.  Nevertheless, as the selection of the remaining ongoing
audits was not based on the risk of noncompliance with teaching
physician billing rules, the likelihood of unproductive audits, such
as the one that occurred at Dartmouth, remains. 

The OIG initiated the audit at Dartmouth with little indication that
teaching physicians at the institution were improperly billing
Medicare.  OIG field office staff told us that since they intended to
audit every major teaching institution in their region, it made
little difference to them which ones were done first.\25 Dartmouth
was the first institution to be chosen, in part, because the hospital
had only two major physician groups, which greatly simplified the
sampling of patient services.  The workpapers indicate that the OIG
told Dartmouth officials that the institution was selected for PATH
because an analysis of claims data seemed to indicate what it
referred to as "high-end" billing.  However, OIG field office staff
told us that the suspected high-end billing related only to
psychiatric services.  We found that these services represented an
insignificant number of Medicare inpatient services and were never
the focus of the actual PATH audit. 

We also found that the DOJ official who is generally regarded as the
architect of the PATH initiative previously commented that
Dartmouth's billing guidance for teaching physicians was the best he
had ever seen and, according to DOJ, had advised the OIG of the
quality of this guidance.  Quality guidance is no assurance of
compliance with Medicare billing rules.  However, without indications
of significant compliance problems and with the DOJ official's high
commendation, Dartmouth would probably not have been a good candidate
for a PATH audit under a risk-based approach. 

Ten months into the audit--after about one-half of the sampled
admissions had been reviewed--the OIG terminated the audit. 
According to a Dartmouth official, the partial audit cost the
institution about $1.7 million:  $900,000 in direct costs ($600,000
in audit expenses and $300,000 in legal fees and other costs) and
$800,000 in indirect costs attributable to a delay in a bond
financing.\26 In the end, the OIG concluded the institution had been
overpaid $778--an amount it did not deem worthy of collecting. 

While the OIG stated that the amount Dartmouth spent on its PATH
audit was far higher than it had anticipated, a Dartmouth official
told us it incurred the legal and audit costs in order to be in the
best position to defend itself should the outcome of the audit result
in litigation.  For example, Dartmouth decided to retain outside
legal counsel because at the time its audit was initiated, the
outcomes of the Penn and Jefferson settlements were well known. 
Dartmouth said it was widely believed by many in the medical
community that the government was using the threat of severe False
Claims Act penalties to compel settlements.  In addition, Dartmouth
also believed that the OIG's interpretation of IL-372 requirements
was too narrow.  Thus, it retained external auditors to expand the
scope of the audit.  By doing so, it hoped to demonstrate that it was
complying with, in its view, a more appropriate interpretation of the
requirements.  Dartmouth also wanted to expand the audit scope to
ensure that underbilled services were identified.  The institution
believed that such services would be unfairly ignored by the OIG and
thus could result in misleading conclusions about the institution's
compliance. 


--------------------
\23 In return for volunteering for a PATH II audit, the OIG advises
DOJ of the institution's level of cooperation.  DOJ may take this
cooperation into account when resolving losses the government
sustains from any claims determined to be false. 

\24 The OIG told us that in some instances the selection of
institutions was influenced by factors outside its control, such as
civil lawsuits and DOJ requests for assistance. 

\25 Within 6 months of initiating the Dartmouth audit, the OIG field
office started PATH audits at the remaining major institutions in its
region. 

\26 According to Dartmouth, investment banker and credit agency
concerns about the possible outcome of the audit delayed the bond
financing, ultimately raising the total costs of this financing. 


   METHODOLOGY FOLLOWED IN
   CONDUCTING PATH AUDITS APPEARS
   REASONABLE
------------------------------------------------------------ Letter :5

On the basis of our review of the OIG's workpapers on the Penn,
Jefferson, and Dartmouth audits, we believe that the OIG followed a
reasonable methodology in these audits in making physical presence
and level-of-service determinations--the key components of a PATH
audit.  The criteria the OIG used to assess teaching physicians'
involvement in part B services was, in our opinion, valid and
essentially the same as requirements already imposed on the teaching
physicians by local Medicare carriers during the time periods covered
by the audits.  Moreover, the OIG's workpapers show that the three
institutions were aware of the rules.  Likewise, level-of-service
determinations, which require medical background and knowledge, were
made by medical reviewers--not auditors.  We also found no evidence
of retroactive application of level-of-service documentation
guidelines. 


      OIG REQUIRED DOCUMENTATION
      OF PHYSICAL PRESENCE
---------------------------------------------------------- Letter :5.1

The workpapers for all three audits show that in examining sampled
services, the OIG required documentation--such as written comments,
notes, or reports in the patients' medical records--that demonstrated
the teaching physician either provided the service or was physically
present while the resident provided the service.  Countersignatures
by teaching physicians on residents' notes were not considered
acceptable evidence by the OIG because, with countersignatures alone,
it was not possible to ascertain whether physicians were personally
involved in these services or were acknowledging a later review of
the residents' notes as part of their routine teaching
responsibilities.  The OIG did accept a countersignature, however, if
other information in the medical record demonstrated that the
teaching physician was with the patient when the service was
provided.  This is essentially the same criterion we used in
evaluating teaching physician services in 1986; and, in our judgment,
this criterion is compatible with Medicare requirements for
reimbursing teaching physicians for their services. 


      CARRIERS NOTIFIED
      INSTITUTIONS OF PHYSICAL
      PRESENCE REQUIREMENTS
---------------------------------------------------------- Letter :5.2

The OIG's workpapers show that Xact Medicare Services--the Medicare
carrier that processed part B claims for Penn and Jefferson--had for
many years clearly interpreted the personal and identifiable services
requirement of IL-372 to mean that a teaching physician had to be
physically present to bill.  In 1982, for example, the carrier issued
a manual to doctors, hospital administrators, and medical records
personnel that specified that teaching physicians had to be
physically present to bill for services provided by residents.  The
manual also said that a physician's countersignature of a note
entered by a resident or nurse was not evidence that a part B covered
service was provided unless the note indicated that the physician was
present.  In 1988, the carrier distributed a newsletter to providers
that reiterated its documentation requirements and emphasized that
teaching physicians needed to document their presence in patients'
medical records in order to bill part B.  For example, the newsletter
warned that Medicare considered a part B payment to be an overpayment
if a physician was reimbursed for a physical examination performed by
a resident but was not present during the examination. 

C&S Administrative Services, the carrier that processed part B claims
for Dartmouth, had interpreted the physical presence requirement
similarly, although its interpretation was not as longstanding.\27 In
1993, the carrier began to publicize its expectation that teaching
physicians needed to be present to bill for services provided by
residents.\28 For example, a May newsletter to providers stated, in
part, that a physician's fee was payable in a teaching hospital if
(1) the physician personally performed the service or (2) was
physically present when the resident performed the service for which
payment was sought.  This guidance was repeated in a December
newsletter.  In addition, the carrier's Associate Medical Director
for Government Programs wrote to several medical societies in
September that the carrier expected each patient's medical record to
document the teaching physician's involvement in the service billed. 


--------------------
\27 National Heritage Insurance Company is the current Medicare
carrier for Dartmouth. 

\28 Prior to 1993, the carrier accepted countersignatures to
residents' notes as evidence of a teaching physician's presence and
involvement in the billed service.  The carrier also only required a
notation in patients' medical records once every 3 days in order to
bill for daily visits. 


      TEACHING INSTITUTIONS WERE
      AWARE OF PHYSICAL PRESENCE
      REQUIREMENTS
---------------------------------------------------------- Letter :5.3

The OIG's workpapers also show that all three of the audited
institutions were aware that teaching physicians were required to be
present in order to bill Medicare. 

  -- In the 1980s, Xact conducted numerous reviews of physician
     billings at teaching hospitals in the state of Pennsylvania to
     ensure the institutions were complying with Medicare physical
     presence requirements.  At least three audits occurred at
     Jefferson--in 1981, 1986, and 1988--with the carrier reporting
     rates of noncompliance of 15, 45, and 69 percent, respectively. 
     Feedback to the institution from the carrier, in our view, left
     no doubt as to what was expected.  The carrier's report on the
     results of its 1986 review, for example, mentioned that
     emergency room visits were discrepant because supervising
     physicians had failed to indicate whether they had personally
     performed services or were present while residents performed the
     services.  In response to these findings, Jefferson attempted to
     educate its teaching physicians about IL-372 requirements and
     advised them that to bill Medicare for their services, Jefferson
     teaching physicians had to either perform the services
     themselves or supervise treatment "at the elbow" of the
     resident. 

  -- Although we were unable to determine if Penn had been subject to
     similar audits by the carrier, evidence obtained by the OIG
     demonstrated that Penn was also aware of what the carrier
     required.  In a 1986 memorandum to physicians and residents, a
     department chairman discussed the need for attending physicians
     to be present in order to bill for services performed by
     residents.  Mentioning our 1986 report, the chairman stated,
     "Medicare auditors have been instructed to enforce the published
     guidelines, and refunds will be required for undocumented
     services." Other evidence in the OIG's workpapers indicated that
     Penn became concerned about its billing practices in 1992 and
     took steps--including development of new billing instructions,
     physician training, and internal review--to improve compliance
     with IL-372 requirements.  The billing guidelines largely
     mirrored what the carrier already required. 

  -- In 1991, Dartmouth issued a uniform billing policy for its
     teaching physicians.  The institution's guidance stated, in
     part, that medical records must contain a notation indicating
     that the attending physician personally performed the service or
     was physically present while the resident performed the service. 
     A countersignature, the guidance stated, verifies only that the
     attending physician reviewed the note and does not imply that
     the attending physician was present or that the attending
     physician personally rendered a service. 


      MEDICAL REVIEWERS MADE
      LEVEL-OF-SERVICE
      DETERMINATIONS
---------------------------------------------------------- Letter :5.4

Medical reviewers, not auditors, determined whether the codes used by
teaching physicians to claim reimbursement accurately reflected the
service provided.  In our view, this was a reasonable approach. 
Determining the appropriateness of the level of service billed
involves examining related medical records and other information. 
Such determinations require medical knowledge as well as familiarity
with the codes used to bill Medicare.  If auditors do not possess
such expertise, then auditing standards require that they seek the
assistance of specialists who have the appropriate qualifications and
experience.  The carriers' medical reviewers who assisted the OIG in
these PATH audits were registered nurses, most with many years of
experience in conducting postpayment medical reviews of Medicare
claims.\29 The medical reviewers we interviewed told us that their
work on PATH audits was essentially the same as what they had
routinely done on other postpayment reviews. 


--------------------
\29 Reviews of paid claims, known as postpayment reviews, are
routinely performed by Medicare carriers and are the primary means
for systematically identifying which providers are inappropriately
billing part B. 


      NO EVIDENCE OF RETROACTIVE
      APPLICATION OF EVALUATION
      AND MANAGEMENT DOCUMENTATION
      GUIDELINES
---------------------------------------------------------- Letter :5.5

Although some have alleged that the PATH audits involve a retroactive
application of evaluation and management documentation guidance, we
found no evidence to support such allegations in the audits we
reviewed.  The carrier medical reviewers we interviewed told us that
in evaluating the appropriateness of the levels of service billed,
they applied the criteria in effect for the period under audit.  We
did not have the expertise to evaluate the medical reviewers'
determinations and, as a result, did not attempt to do so. 
Nevertheless, we found no evidence in the workpapers that the medical
reviewers had used inappropriate criteria.  For example, we observed
that when the carrier questioned the appropriateness of the code used
for a particular service, the institutions were given the opportunity
to comment and provide additional information, and in some instances,
the medical reviewers revised their determinations on the basis of
the additional information.  Although we saw many instances where the
institutions and the carriers' medical reviewers ultimately disagreed
on the appropriate code, we found no evidence that the institutions'
basis for disagreeing was that the carriers' medical reviewers had
applied documentation guidance retroactively.  Moreover, the
officials we interviewed from the teaching institutions never raised
this as an issue.  Officials from Penn, for example, told us they did
not think the OIG had applied evaluation and management documentation
guidance retroactively at their institution.  They believed, however,
that the guidance for documenting evaluation and management services
during the period covered by their PATH audit was vague and,
therefore, auditing these services was inappropriate. 


   SIGNIFICANCE OF THE PATH
   RESULTS AT THE UNIVERSITY OF
   PENNSYLVANIA AND THOMAS
   JEFFERSON UNIVERSITY
------------------------------------------------------------ Letter :6

While inpatient billing errors found by HHS' OIG at Dartmouth were
immaterial, those identified at both Penn and Jefferson were more
significant and were used by DOJ as a basis for negotiating
settlements with the institutions.  Although we had access to the
OIG's workpapers, confidentiality agreements between DOJ and these
two institutions preclude us from disclosing the specific details of
these findings.  The amounts the institutions ultimately agreed to
repay were substantially higher than the overpayments found by the
auditors.  However, these settlement amounts--about $10 million in
disputed billings plus about $20 million in damages at Penn and
almost $6 million for disputed billings plus an equal amount in
damages at Jefferson--were based on DOJ's extrapolation of the OIG's
findings.  The extrapolations covered time periods and types of
services that were not audited by the OIG.  While these
extrapolations have been criticized by the medical community, they
were arrived at during settlement negotiations between the
institutions and DOJ in an effort to avoid False Claims Act
litigation.  Such negotiations are not bound by rules of evidence or
methodological constraints. 


      BILLING ERRORS AT PENN AND
      JEFFERSON
---------------------------------------------------------- Letter :6.1

The workpapers for these audits show evidence that some teaching
physicians had not documented their compliance with Medicare billing
requirements and, therefore, may not have been entitled to payment. 
Because it is not practical or efficient to audit all teaching
physicians' claims as part of a PATH audit, a random sample of 100
Medicare inpatient admissions, typically involving 1,500 to 2,000
individual physician services for a 1-year period, is selected.  All
100 of the 1993 sampled admissions were reviewed during the Penn
audit, enabling the OIG to project the results to Medicare inpatient
admissions for the year audited.  In contrast, only 50 of the 100
1994 sampled admissions were completely reviewed at Jefferson--only
one-half of the number the OIG considered the minimum to make an
extrapolation. 

According to the OIG's Penn workpapers, the auditors found
overpayment errors in many of the services they examined.  By
extrapolating these sample results, the OIG concluded that teaching
physicians had been significantly overpaid by Medicare for inpatient
part B services in 1993.  Similar extrapolations could not be made
for the review of 1994 services at Jefferson because the audit was
stopped before the entire sample could be reviewed.  Nevertheless,
the results of this partially completed sample were used to negotiate
Jefferson's settlement. 


      AUDIT RESULTS USED TO
      EXTRAPOLATE SETTLEMENT
      AMOUNTS
---------------------------------------------------------- Letter :6.2

As a result of settlement negotiations with DOJ, the institutions
agreed to return millions of dollars for disputed billings plus
millions more in damages.  Penn agreed to repay about $10 million for
disputed billings related to inpatient and outpatient Medicare part B
services over a 6-year period--significantly more than what the OIG
estimated its teaching physicians had been overpaid in 1993 for
inpatient services only.  Penn also agreed to pay another $20 million
in damages.  Jefferson agreed to repay about $6 million for disputed
billings covering a 5-year period plus an equal amount in damages. 
These amounts were based on DOJ's extrapolations of the auditor's
findings. 

The academic medical community has expressed concern and criticism
that the estimates used by DOJ in settlement negotiations were not
statistically valid.  Indeed, these estimates were not based on
statistically valid calculations.  We found that DOJ used the results
of the Penn and Jefferson audits covering inpatient services for 1
year to estimate potential false claims for both inpatient and
outpatient services for multiple, unaudited years:  5 additional
years at Penn and 4 additional years at Jefferson.  We also found
that because the review of services was not completed at Jefferson,
the sample was not sufficient to make statistically valid estimates
of the total overpayment to the year audited, much less multiple
years.  In addition, we found that at both institutions, DOJ
projected evaluation and management coding errors to time periods
that preceded implementation of the current codes. 

In the context of settlement negotiations, however, such
extrapolations are not improper.  If this matter had gone to court,
DOJ might not have been able to use such extrapolations to establish
the existence of False Claims Act violations; the extrapolations
could have been challenged on the basis that they were not
statistically sound.  However, the settlement negotiations were
undertaken between DOJ and the institutions in an effort to avoid
litigation, and the settlement process is not governed by rules of
evidence or methodological constraints. 

We are not in a position to know exactly why the parties agreed to
the settlements because, consistent with the agreement to treat the
settlement negotiations as confidential, the institutions and DOJ did
not discuss the negotiations with us in detail.  Officials from all
three institutions, however, told us that the severe fines and
penalties applicable under the False Claims Act were of great concern
to them and influenced all of their decisions regarding their PATH
audits.  Attorneys from both Penn and Jefferson told us that these
fines and penalties could have been financially devastating to their
institutions had the institutions been found liable in court for
submitting even relatively few false claims.  Penn told us that
during the course of negotiations, DOJ took the position that the
institution had submitted 1.4 million claims during the period
covered by their settlement.  Thus, the institution believed that if
a court determined that only 2 percent, or about 28,000, of these
claims were false, it would have faced false claims penalties of
approximately $280 million even before the calculation and tripling
of damages.  This possibility, they said, foreclosed any realistic
recourse that the institution had to litigate rather than settle this
matter.  Similarly, Jefferson officials emphasized that they would
not have settled with DOJ had the potential damages and penalties
been less onerous.  Instead, the institution saw itself faced with
False Claims Act provisions that had "almost no intent standard"\30
as well as triple damages, severe monetary penalties, and potential
exclusion from the Medicare program.  Given such a threat, they said,
their fiduciary obligations left them no choice other than to settle. 

DOJ told us that, instead of relying on the OIG's 1-year results to
estimate the overpayments for the other years, it could have asked
the OIG to audit the other services and time periods.  In fact, DOJ
said it offered the institutions such an option.  According to DOJ,
the institutions declined and agreed to accept the settlement to
avoid the cost and disruption such additional audit work would have
entailed. 


--------------------
\30 Although one must "knowingly" submit a false claim to be liable
under the False Claims Act, no proof of specific intent to defraud
the government is required.  "Knowingly" is defined to include actual
knowledge or deliberate ignorance or reckless disregard of the truth
or falsity of the information submitted. 


   NATURE OF THE FINDINGS MAY BE
   OVERSTATED
------------------------------------------------------------ Letter :7

While the audit results at Penn and Jefferson indicate noncompliance,
based on our review of the workpapers, the problems in these two PATH
audits do not appear to be as serious as publicly portrayed by HHS'
OIG since the settlements.  In October 1997 testimony, the OIG
reported that both settled and ongoing PATH audits had identified
significant instances of noncompliance with the physical presence
standard; the four examples cited involved physicians who had billed
Medicare for services on days when they were on leave or out of
town.\31 In an April 1997 response to a congressional inquiry, HHS'
Inspector General said that serious upcoding errors had been
identified in the first two PATH reviews and that the huge majority
of these errors were related to multilevel upcoding.  At the time of
these statements, the Penn and Jefferson audits were the only PATH
audits that had resulted in settlements with DOJ. 

The OIG's workpapers for Penn and Jefferson do not contain convincing
evidence that teaching physicians were not working on days they
billed Medicare.\32 Rather, the workpapers show that teaching
physicians did not always document their presence when services were
rendered by a resident.  While this lack of documentation may be
insufficient to obtain Medicare reimbursement, it does not
necessarily mean that the teaching physicians were not at work when
the services were rendered.  Although the Penn workpapers show that
the auditors suspected that a few teaching physicians might have
billed Medicare on days they were not working, the auditors told us
that a settlement was reached before they could determine if this was
the case.  Similarly, the Jefferson workpapers show no evidence of
such abuses, and the OIG auditors told us there was no time to look
for such evidence before the audit was terminated. 

Likewise, the OIG's statements that the PATH audits have identified
serious upcoding errors--the huge majority of which were related to
multilevel upcoding errors--are not, in our opinion, substantiated by
the workpapers.  Rather, the workpapers show the overwhelming
majority of the upcoding errors that were found by the auditors at
both institutions involved one-level discrepancies.  At one of these
institutions, not only were few multilevel errors found by the OIG,
but one physician accounted for about 70 percent of these multilevel
errors. 

Assuming medical reviewers examined services for both undercoding and
overcoding at Penn and Jefferson, errors that reveal one-level
differences consistently favoring teaching physicians may indicate
abuse.  In fact, the workpapers indicated that carrier medical
reviewers found few instances of undercoding errors at Penn and none
at Jefferson.\33

Although the medical reviewers who assisted the OIG on these two
audits told us that they looked for both undercoded and overcoded
services during their reviews, we lacked the expertise to verify
their statements. 

One-level differences, however, may indicate legitimate differences
in judgment.  HCFA, OIG, and carrier staff with whom we spoke
acknowledged that coding discrepancies can be subjective and do not
necessarily reflect fraud or abuse.  A HCFA official told us, for
example, that because the time periods covered by these two audits
predated HCFA's 1996 documentation guidelines, legitimate one-level
disagreements could have occurred between providers and payers as to
what constituted appropriate documentation.  Indeed, the OIG's
workpapers show that many one-level discrepancies found at Penn were
dropped by DOJ during settlement negotiations. 


--------------------
\31 The Physicians at Teaching Hospitals (PATH) Audits, OIG testimony
before the Senate Committee on Appropriations, Subcommittee on Labor,
HHS, Education, and Related Agencies, Oct.  21, 1997. 

\32 An OIG official told us that one of the four examples cited in
the October 1997 testimony involved a physician from one of these two
institutions.  The other three examples involved physicians from an
institution that was not part of the PATH initiative at the time of
the testimony.  We did not review the work related to this audit. 

\33 The workpapers showed, however, that some undercoding errors were
found at Jefferson by its external auditors. 


   CONCLUSIONS
------------------------------------------------------------ Letter :8

The method of paying teaching physicians for their services creates
the potential for Medicare to pay for some services twice--once
through part A and again through part B.  To prevent inappropriate
payments, Medicare has issued guidance addressing when teaching
physicians may bill part B for their services.  Although this
guidance was not always clear and, as a result, has been interpreted
differently over the years by Medicare carriers, federal law has long
required that teaching physicians billing part B either provide the
service themselves or be physically present while a resident provides
the service.  In addition, when teaching physicians submit part B
claims, Medicare requires that they accurately code their services in
accordance with applicable guidance.  For these reasons, the OIG's
PATH initiative, which involves auditing teaching physicians' part B
claims for compliance with these requirements, is consistent with
law. 

The OIG also followed a reasonable methodology in conducting its
work.  In assessing compliance, the OIG used the same criteria that
the carriers already expected the teaching physicians to follow when
submitting part B claims.  Furthermore, when medical knowledge was
required, the OIG relied on carrier medical reviewers to make these
assessments.  We are concerned, however, that the OIG's initial
intent to audit teaching physicians affiliated with all 125 of the
nation's largest medical schools was not sufficiently risk-based,
resulting in potentially unproductive audits and an inefficient use
of resources, as the Dartmouth audit suggests.  Instead, the OIG
should have considered a risk-based approach, which may have more
clearly identified institutions with suspected billing problems and
then targeted its efforts accordingly.  Because PATH audits can be
time-consuming and expensive for both the government and the
institutions, we believe that the OIG should have had a sound basis
for asking the institutions to incur these costs. 

Penn and Jefferson agreed to repay the federal government amounts
that were substantially higher than the errors identified by the OIG. 
While the medical community has been critical of the use of
extrapolations to calculate these amounts, it is important to
recognize that these settlement amounts were the outcomes of
discussions that occurred between the institutions and DOJ, and such
negotiations are not bound by rules of evidence or methodological
constraints.  The institutions agreed to the settlement amounts
rather than subjecting themselves to additional audit work and
possibly defending themselves in court against a False Claims Act
lawsuit. 

Finally, the results of the Penn and Jefferson audits show that the
OIG identified instances of teaching physician noncompliance with
Medicare billing rules.  Regardless of whether the noncompliance is
due to a mistake, carelessness, or outright fraud, teaching
physicians are not entitled to reimbursement if they fail to comply
with Medicare requirements.  However, the OIG has characterized the
problems found at these institutions as more serious than its
workpapers establish. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :9

We provided a draft of this report to the HHS Inspector General and
the Department of Justice.  We also provided excerpts of the draft
report to the University of Pennsylvania, Thomas Jefferson
University, and the Dartmouth-Hitchcock Medical Center.  The excerpt
each institution received consisted only of factual material
pertaining to that institution.  We received written responses from
all five organizations. 

The draft we provided to the HHS Inspector General and DOJ and the
draft excerpts we provided to Penn, Jefferson, and Dartmouth
contained specific details about the audit findings at these
institutions.  In commenting on our draft, both Penn and Jefferson
claimed they had verbal agreements with DOJ to keep the information
related to their settlements confidential.  DOJ subsequently
confirmed that such agreements were made and objected to inclusion of
confidential information in our report.  DOJ also said that
publication of the findings underlying the settlements would have a
detrimental effect on its ongoing PATH efforts.  In addition, DOJ
told us that it is currently involved in litigation to prevent the
release of this information pursuant to a Freedom of Information Act
request from a major newspaper.  Consequently, DOJ asked us to
prepare this version of the report, which does not include certain
details DOJ identified as subject to its confidentiality agreements. 
This redacted version will be made available to the public.  Other
than its concerns about the confidentiality of the Penn and Jefferson
audit results, DOJ said that it generally concurred with the
substance of our draft report. 

We received comments from HHS' Inspector General and the three
institutions discussed in this report.  Summaries of these comments
are provided below.  In addition, DOJ, the HHS Inspector General, and
the institutions provided technical changes, which we incorporated as
appropriate. 


      COMMENTS FROM THE HHS
      INSPECTOR GENERAL
---------------------------------------------------------- Letter :9.1

The HHS Inspector General's comments revolved around three issues: 
(1) the overstatement of the findings, (2) the selection of
institutions for audit, and (3) the costs of the Dartmouth PATH
review. 

Concerning the first issue, the Inspector General acknowledged that
the OIG had misstated the extent of multilevel upcoding found at Penn
and Jefferson.  She added that these misstatements were of no
consequence in these two, or any other, PATH reviews.  As we stated
in our report, regardless of the reason, if teaching physicians fail
to comply with Medicare requirements, they are not entitled to
reimbursement.  In our view, however, overstating the seriousness of
the findings is unnecessary and unfair to the audited institutions. 

The Inspector General also stated that the physical presence problems
noted in the OIG's October 1997 testimony pertained to teaching
institutions across the country, not just Penn and Jefferson.  In
addition, she pointed out that at both Penn and Jefferson, physical
presence errors found by the OIG were significant.  We noted,
however, that one of the four examples cited involved a physician
from one of these two institutions, while the remaining three
examples involved an institution that was not considered by the OIG
to be a PATH audit at the time of the testimony.  Moreover, the
examples, in our opinion, leave the impression that the OIG found
many instances in which teaching physicians billed Medicare on days
that they were not working or were out of town.  While we cannot
comment on three of the examples, on the basis of our review of the
Penn and Jefferson workpapers, we found that the OIG had no
convincing evidence of the type of abuses cited in the testimony. 

Concerning the second issue, the Inspector General agreed that a
risk-based approach was an excellent method for selecting
organizations to audit.  However, she said that the OIG could not use
a risk-based approach in PATH because it had no techniques for
narrowing the selection process to the most problem-prone
institutions.  Moreover, with regard to our criticism of the OIG's
selection of Dartmouth for a PATH audit despite a DOJ official's
positive characterization of that institution's billing guidance, the
Inspector General pointed out that the billing guidance at both Penn
and Jefferson was similar to Dartmouth's. 

We recognize that it may be difficult to consistently apply a
risk-based approach in auditing, but we are unconvinced that such an
approach was not possible for PATH.  For instance, while we agree
that the billing guidance at the three institutions was similar, we
believe indications of improper billing at Dartmouth were weak.  In
contrast, indications suggesting the possibility of improper billing
by Penn and Jefferson teaching physicians were stronger.  We believe
that these indications, which were available prior to the initiation
of these audits, demonstrate that it is possible to target
problem-prone institutions.  To illustrate, DOJ told us it had
received information pertaining to significant violations of teaching
physician billing rules by some Penn physicians and that the
institution had not taken effective action to address the problem. 
DOJ then approached the OIG for assistance in its investigation. 
Similarly, as we noted in our report, carrier audits showed that
Jefferson had a history of significant compliance problems with
teaching physician billing rules, making it an appropriate target for
an audit. 

Concerning the third issue, the Inspector General was critical of the
amount Dartmouth spent on its PATH audit.  For example, she said that
while the Dartmouth audit was terminated after a review of about half
the sample, the audit costs were significantly higher than what it
cost the OIG to do the entire Penn audit.  Moreover, she questioned
the need for Dartmouth to spend $200,000 in legal fees, given the
absence of a formal legal dispute.  Our report, while presenting
Dartmouth's view of why its costs were so high, does not take a
position on the appropriateness of these expenses. 


      COMMENTS FROM PENN AND
      JEFFERSON
---------------------------------------------------------- Letter :9.2

In their written responses, Penn and Jefferson took issue with
comments we made in the draft about their carrier's interpretation of
IL-372 requirements.  The institutions contended that the carrier's
interpretation of IL-372 conflicted with the Medicare statute,
regulations, and other HCFA correspondence.  The institutions,
however, were not provided the portion of our draft that discussed
our analysis of IL-372 requirements.  In this analysis, we recognized
that HCFA guidance had created some confusion.  However, we concluded
that federal Medicare law had long required that physician services
be rendered or supervised by the physician in person. 

Both Penn and Jefferson also objected to the section of our report
where we discussed their audit results collectively.  Penn, for
example, claimed its audit was unique because, among other things, it
predated the PATH initiative and was the only audit that targeted
services from 1993, when evaluation and management codes were barely
a year old.  Penn also contended that its audit was unique because it
was the only institution that was not given the opportunity to
conduct a self-audit and was the only institution for which the
entire audit sample was reviewed.  Similarly, Jefferson stated that
it believed the two institutions were very different and that
conclusions based on information from the Penn audit should not be
applied to Jefferson.  While we agree that there were many
differences between the two audits, we believe our report makes a
clear distinction between the audit results that were found at Penn
and Jefferson.  Moreover, our discussions of the audit results from
the two institutions is not intended to compare them to each other
but rather to describe the significance and seriousness of the
findings found by the OIG.  This information is based on our
examination of the OIG's workpapers for both audits as well as our
discussions with OIG and carrier staff who conducted the work. 

Penn also objected to the section of our report that discusses its
awareness of the physical presence requirement prior to its audit. 
Penn emphasized that it had made substantial efforts to improve its
billing practices before the audit started and believed these efforts
showed that the institution was not involved in a scheme to defraud
the government.  However, Penn indicated that the OIG did not credit
these activities.  Instead, Penn believed the OIG penalized the
institution by contending that these efforts demonstrated that the
institution was aware of the problem but had not taken sufficient
corrective action.  Our inclusion of this material presents factual
information about Penn's awareness of the physical presence
requirements prior to its audit.  We cannot comment on how the OIG or
DOJ may have treated this information in negotiating a settlement
with the institution. 

Penn suggested we clarify the statement in our report that Penn
officials did not believe that the OIG had retroactively applied
evaluation and management documentation guidelines in its audit. 
Penn said that the guidance was retroactively applied in the sense
that the OIG and DOJ extrapolated these findings to years which
predated the establishment of these evaluation and management codes. 
Penn also asserted that not only did it view the guidance as
inadequate, but the carrier and the OIG did as well.  As noted
elsewhere in our report, we pointed out that DOJ extrapolated
evaluation and management coding errors to time periods that preceded
implementation of the codes.  Our report also notes that both OIG and
carrier staff acknowledged that one-level coding differences could be
subjective. 


      COMMENTS FROM DARTMOUTH
---------------------------------------------------------- Letter :9.3

Dartmouth raised objections about the way we addressed it and its
carrier's interpretation of IL-372 requirements.  While not disputing
that its billing guidance requires "physical presence," Dartmouth
said it interpreted physical presence to mean that the teaching
physician met this requirement by being "on the premises," not
necessarily "at the elbow" of the resident, while the service was
being provided.  It also contended that the newsletters its carrier
issued in 1993 interpreted the meaning of physical presence in the
same way.  However, neither Dartmouth's billing guidance nor the
carrier newsletters specify that a teaching physician has met the
physical presence requirement merely by being on the premises. 

In its comments, Dartmouth also said we did not fully explain its
reasons for retaining outside counsel, hiring independent auditors,
and expanding the scope of its audit.  We have expanded the
discussion of this issue.  In addition, Dartmouth contended that the
aggregate financial impact of its PATH audit exceeded $3 million, not
$1.7 million as we have reported.  The cost figures we used were
given to us by a Dartmouth official when we met with him to discuss
PATH and were the same cost figures presented by Dartmouth to Senator
Patrick Leahy in an October 1997 briefing document.  Finally,
Dartmouth said that our statement that it had been overpaid $778 was
inaccurate and misleading because the audit results actually showed
that overpayments were offset by underpayments.  While this is
essentially true for the results of the external auditor's work
(which determined that the government owed Dartmouth about $5), the
OIG's verification of this work determined that, after netting
overbilling and underbilling errors, Dartmouth had been overpaid
$778, as we noted in our report. 


---------------------------------------------------------- Letter :9.4

As arranged with your office, unless you announce its contents
earlier, we plan no further distribution of this report until 30 days
after the date of this letter.  At that time, we will send copies of
the limited official use version of this report to the Inspector
General of HHS and the Attorney General.  We will provide this
redacted version to officials from the organizations we visited and
other interested parties.  We also will make copies of the redacted
version available to others upon request.  Please call me at (202)
512-7114 or Leslie G.  Aronovitz at (312) 220-7600 if you or your
staff have any questions about this report.  Other major contributors
to this report include Paul D.  Alcocer, Barry R.  Bedrick, George H. 
Bogart, Robert T.  Ferschl, and Geraldine Redican-Bigott. 

Sincerely yours,

William J.  Scanlon
Director, Health Financing and
 Systems Issues


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

To determine whether HHS' OIG has a legal basis for conducting PATH
audits, we examined pertinent laws and regulations and HCFA guidance
and correspondence related to teaching physician billing for Medicare
part B services.  We also examined related guidance issued by the
Medicare carriers that processed part B claims for Pennsylvania and
New Hampshire teaching physicians.  In addition, we reviewed our 1971
and 1986 reports that described problems in documenting teaching
physician services.  We discussed the PATH initiative and teaching
physician billing rules with representatives from the OIG, HCFA, the
Association of American Medical Colleges, and the American Hospital
Association. 

To understand the OIG's approach and methodology in carrying out PATH
audits and to determine the nature and significance of the billing
problems being identified in the audits, we reviewed the OIG's work
related to the first three resolved audits.  These were audits of the
University of Pennsylvania and Thomas Jefferson University--both
located in Philadelphia, Pennsylvania--and the Dartmouth-Hitchcock
Medical Center located in Lebanon, New Hampshire.  Our review
involved examining the OIG's workpapers and discussing our questions
and observations with OIG staff from the Philadelphia and Boston
field offices who were involved in the work.  We also met with
medical review staff from the carriers who assisted the OIG on these
three audits, and we interviewed representatives from each of the
institutions to obtain their perspectives. 

We attempted to interview officials from the U.S.  Attorney's Office
for the Eastern District of Pennsylvania because they played a key
role in the development and expansion of the PATH initiative and
because they negotiated the Penn and Jefferson settlements.  DOJ
would not permit these officials to meet with us because it said
certain matters related to Penn and Jefferson were still pending. 
DOJ, however, did respond in writing to our questions regarding its
role in PATH in general and the Penn and Jefferson settlements in
particular. 

Our review of the OIG's workpapers for the three audits focused on
understanding how the work was carried out--particularly the work
related to physical presence and level-of-service determinations, the
key components of PATH.  In carrying out this work, we did not
attempt to assess the OIG's compliance with auditing standards, nor
did we redo the audits to verify the validity of the OIG's findings. 
Our review was also limited to the workpapers in the OIG's
possession.  The OIG did not have all of the workpapers prepared by
external auditors for the two PATH II audits--Jefferson and
Dartmouth.  The OIG's workpapers for these audits, however, included
the results of its verification reviews of the external auditors'
work and other information, which, in our judgment, were sufficient
to enable us to understand what was done. 

We performed our work between August 1997 and June 1998 in accordance
with generally accepted government auditing standards. 


KEY EVENTS RELATED TO THE PATH
INITIATIVE
========================================================== Appendix II



   (See figure in printed
   edition.)



   (See figure in printed
   edition.)


*** End of document. ***