Prescription Drugs: Automated Prospective Review Systems Offer
Significant Potential Benefits for Medicaid (Letter Report, 08/05/94,
GAO/AIMD-94-130).

Inappropriate drug therapy poses a significant health risk to Medicaid
patients and could be adding hundreds of millions of dollars each year
in unnecessary drug and hospitalization costs. Use of automatic systems
linked to statewide databases could provide a more thorough prospective
review than reviews based on manual or local systems.  However, the use
of automated statewide drug utilization review (DUR) is not required by
law. Although about two-thirds of the states use them, other states
believe the systems are too costly and may not provide tangible
benefits. Data provided by 5 states visited by GAO that use statewide
DUR systems resulted in the cancellation of over 128,000 prescriptions
that posed a risk to patients.  Two of these states provided data
showing that their systems led to the cancellation of over 250,000 early
refill prescriptions as potential fraud or abuse.  In addition to
patient safety, such cancellations can result in program savings to the
extent that the prescriptions are not subsequently filled or replaced
with a substitute. Prescriptions cancelled by the 5 states GAO reviewed
were valued at about $12 million.  While data were not available to show
actual savings from these cancellations, GAO believes the sheer
magnitude of cancellations indicates a significant savings potential for
the Medicaid program.

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

 REPORTNUM:  AIMD-94-130
     TITLE:  Prescription Drugs: Automated Prospective Review Systems 
             Offer Significant Potential Benefits for Medicaid
      DATE:  08/05/94
   SUBJECT:  Medicaid programs
             Fraud
             Program abuses
             Medical information systems
             Drugs
             Health care cost control
             Health hazards
             Pharmaceutical industry
             Risk management
             State-administered programs
IDENTIFIER:  Medicaid Management Information System
             Maryland
             Missouri
             Pennsylvania
             Tennessee
             West Virginia
             
**************************************************************************
* 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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to the Chairman, Committee on the District of Columbia, House
of Representatives

August 1994

PRESCRIPTION DRUGS - AUTOMATED
PROSPECTIVE REVIEW SYSTEMS OFFER
POTENTIAL BENEFITS FOR MEDICAID

GAO/AIMD-94-130

Prospective DUR Systems


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

  DUR - drug utilization review
  FDA - Food and Drug Administration
  HCFA - Health Care Financing Administration
  HHS - Department of Health and Human Services
  MMIS - Medicaid Management Information Systems

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


B-257257

August 5, 1994

The Honorable Pete Stark
Chairman, Committee on the District of Columbia
House of Representatives

Dear Mr.  Chairman: 

Concerned with the impact of adverse drug reactions on patient
safety, as well as the billions of dollars spent annually on Medicaid
prescription drug benefits, the Congress amended the Social Security
Act to require states to implement drug utilization review (DUR)
programs not later than January 1, 1993.  Under these programs,
states are required to review Medicaid prescriptions to (1) determine
whether they are appropriate, medically necessary, and not likely to
result in adverse medical reactions and (2) identify fraud, waste,
and abuse.  These reviews are to be performed before prescriptions
are filled (prospective), as well as on a quarterly basis after they
are filled (retrospective). 

The purpose of this report is to discuss states' use of information
technology in implementing their DUR programs, the effectiveness of
using automated systems to perform prospective reviews, and the
extent to which the Department of Health and Human Services' (HHS)
Health Care Financing Administration (HCFA) has encouraged the use of
automated prospective DUR systems.  This report also responds to your
September 21, 1993, request that we consider the District of Columbia
in our review of the states' implementation of DUR systems. 


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

Inappropriate drug therapy poses a significant health risk to
Medicaid patients and could be adding potentially hundreds of
millions of dollars annually in unnecessary drug and hospitalization
costs.  Pharmacies' use of automated systems linked to a statewide
database could provide a more thorough prospective review than
reviews based on manual or local systems; however, the use of
statewide automated prospective DUR systems is not statutorily
required.  Even though such systems are optional, two-thirds of the
states and the District of Columbia have or plan to have them. 
Officials from most of the remaining states said they do not plan to
acquire statewide automated prospective DUR systems because they
believe the systems are too costly to acquire and operate and may not
provide tangible benefits. 

Data provided by five states we visited that operated statewide
prospective DUR systems for portions of fiscal year 1993 show that
these systems resulted in the cancellation of over 128,000
prescriptions that posed a risk of serious adverse medical reactions. 
Also, two of these states provided data showing that their systems
led to the cancellation of over 250,000 early refill prescriptions as
potential fraud or abuse.\1 In addition to increasing patient safety,
such cancellations can result in program savings to the extent that
the prescriptions are not subsequently filled or replaced with a
substitute.  For example, prescriptions canceled by these five states
during portions of fiscal year 1993 were valued at about $12 million. 
While data were not available to show the actual savings resulting
from these cancellations, we believe the sheer magnitude of
cancellations indicates a significant savings potential for the
Medicaid program. 

With reported Medicaid hospitalization costs totaling nearly $39.6
billion for fiscal year 1993, and estimates of hospitalizations due
to inappropriate drug therapy ranging from 3 percent all the way up
to 28 percent (for the elderly population), automated prospective DUR
reviews could help prevent unnecessary hospitalizations, with
substantial potential program saving.  HCFA has taken some actions to
encourage the use of automated prospective DUR systems for the
Medicaid program, such as conducting a demonstration project with
Iowa.  However, states are looking to HCFA to provide more
information to aid them in acquiring and implementing these systems. 


--------------------
\1 Early refills include prescriptions submitted (1) for the exact
same drug, (2) for the same person, and (3) by either the same or a
different pharmacy before a predetermined amount of the drug, such as
75 percent, has been consumed.  While there may be a legitimate need
for an early refill in some situations, early refills also include
duplicate prescriptions submitted for purposes of fraud or abuse. 


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

The Medicaid program, authorized in 1965 under Title XIX of the
Social Security Act, is a federally aided, state-run medical
assistance program.  About 30 million people received an estimated
$84.4 billion (federal dollars) in Medicaid services during fiscal
year 1993.  At the federal level, HCFA is responsible for
administering the Medicaid program, establishing policies, developing
operating guidelines, and ensuring states' compliance with Medicaid
regulations. 

Federal regulations require states to provide certain basic medical
services, such as inpatient hospital services and physician services,
under their Medicaid programs.  Federal regulations also authorize
states to provide numerous optional medical services, such as
optometrist services and eyeglasses and prescribed drugs.  During
fiscal year 1993, all states provided prescribed drugs as part of
their Medicaid programs; the reported cost of these drugs totaled
about $6.9 billion for fiscal year 1993. 

Because of growing concern over the increased use and costs of
prescribed drugs, in 1990 the Congress amended the Social Security
Act to require states to implement drug utilization review programs
by January 1, 1993.\2 These programs must have both prospective and
retrospective drug reviews that include screening for drug therapy
problems due to therapeutic duplication, drug-disease
contraindication, drug-drug interaction, incorrect drug dosage,
incorrect duration of treatment, drug-allergy interactions, and
clinical abuse/misuse.\3 To meet the requirement for retrospective
reviews, states are required to use their mechanized claims
processing information retrieval systems--typically called Medicaid
Management Information Systems (MMIS)--or other automated systems
integrated with these systems.  The law encourages, but does not
require, states to use automated systems to conduct prospective
reviews. 

Prospective reviews are to occur before prescriptions are filled or
delivered to recipients.  Retrospective reviews are to occur at least
quarterly after prescriptions are filled.  The act states the purpose
of both reviews is to ensure that prescriptions are appropriate, are
medically necessary, and are not likely to cause adverse medical
results.  The act also directs that states' DUR programs be designed
to educate physicians and pharmacists to identify and reduce the
frequency of patterns of fraud, abuse, gross overuse, or
inappropriate or medically unnecessary care among physicians,
pharmacists, and patients or associated with specific drugs or group
of drugs. 


--------------------
\2 Public Law 101-508, November 5, 1990, (Omnibus Budget
Reconciliation Act of 1990). 

\3 Appendix I provides definitions for each drug therapy problem. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3

We contacted the Medicaid DUR coordinators in all 50 states and the
District of Columbia to determine their plans for implementing
statewide automated prospective DUR systems.  In addition, to obtain
data on system operations, costs, and potential benefits, we visited
five of the eight states that had statewide automated prospective DUR
systems in operation during 1993:  Maryland, Missouri, Pennsylvania,
Tennessee, and West Virginia. 

We contacted three vendors who design and market prospective and
retrospective systems to obtain information on the capabilities of
these systems.  In addition, we interviewed officials at HCFA's
headquarters office in Baltimore, Maryland, to determine the extent
to which HCFA has provided guidance and technical assistance to the
states on acquiring information technology for their DUR programs. 

The five states we visited provided data on the operations of their
statewide automated prospective DUR systems--data generated by the
contractors that operate these systems for the states.  These data
show the results of drug utilization reviews performed by these
systems, but differ by state both in the kinds of information
reported and the periods of operation during 1993.  Given these
variances and the number of systems, we provide these data only to
indicate the potential results of such systems; we did not attempt to
verify these data. 

We conducted our review between March 1993 and May 1994, in
accordance with generally accepted government auditing standards. 
The Department of Health and Human Services provided comments on a
draft of this report.  These comments are presented and evaluated in
the body of the report and are reprinted in appendix VI. 


   USE OF AUTOMATED PROSPECTIVE
   DUR SYSTEMS COULD INCREASE
   PATIENT SAFETY
------------------------------------------------------------ Letter :4

While considerable public policy attention has focused on the issues
of overall expenditures and price levels for pharmaceuticals, there
is a growing concern that inappropriate drug therapy may reduce the
potential benefits of available drugs or, even more importantly,
cause serious harm or even death to patients.  By using a statewide
database and automating the review of prescriptions before they are
filled, prospective DUR systems can perform more thorough reviews to
help reduce the risk of inappropriate drug therapy and increase
patient safety. 


      INAPPROPRIATE DRUG THERAPY
      THREATENS PATIENT SAFETY
---------------------------------------------------------- Letter :4.1

Patients may inadvertently be prescribed the wrong drug, the wrong
dosage, or a drug that will interact adversely with another drug they
are taking.  The effects of such inappropriate prescribing can be
serious--hospitalization may be necessary and sometimes death can
occur.  Depending on the study and the patient population considered,
estimates of the extent of hospitalizations resulting from
inappropriate drug therapy range from 3 percent for the general
population to as high as 28 percent for the elderly population.\4 In
addition, the Food and Drug Administration (FDA) operates a voluntary
system for health care professionals (physicians, nurses, and
pharmacists) to report suspected adverse drug experiences.  During
fiscal years 1990 through 1992, these health care professionals
reported 175,614 cases of adverse drug experiences.  While all
adverse drug experiences may not be reported, of these reported
cases, 24,419 (14 percent) resulted in hospitalizations and 5,684 (3
percent) resulted in death. 


--------------------
\4 Caranasos, G.  J., M.D.; R.B.  Stewart, M.S.; and L.E.  Cluff,
M.D.  "Drug-Induced Illness Leading to Hospitalization." Journal of
the American Medical Association, Vol.  228, No.  6.  (May 6, 1974);
and
Col, M.; J.E.  Fanale; and P.  Kronholm.  "The Role of Medication
Noncompliance and Adverse Drug Reactions in Hospitalizations of the
Elderly." Archives of Internal Medicine, Vol.  150, No.  4.  (Apr. 
1990) 841-845, respectively.  See appendix II for a list of studies
related to hospitalizations due to adverse drug events. 


      THE AUTOMATED PROSPECTIVE
      DUR PROCESS
---------------------------------------------------------- Letter :4.2

A statewide automated prospective DUR system, integrated with a
state's MMIS, can assist pharmacists in submitting Medicaid
prescription drug benefit claims and in assessing the full range of
medications a patient is receiving through the Medicaid program.  It
also supports expanded patient counseling.  The automated prospective
DUR process is described below and illustrated in appendix III. 

For automated prospective DUR systems, when a Medicaid patient
submits a prescription to be filled, the pharmacy transmits patient
identification and prescription information to the statewide MMIS
database.\5 The information is transmitted via computer terminal and
point-of-sale equipment (similar to equipment currently used to
authorize credit card purchases).  In an on-line, real-time mode, the
system will edit the benefit claim, verify the recipient's
eligibility, determine whether the claim is payable (adjudication),
and automatically screen the prescription against the patient's known
Medicaid medical and prescription history and drug utilization review
criteria.  According to contractor data, in usually just seconds, the
system responds to the pharmacy indicating whether the recipient is
eligible and the claim payable, and whether there is a potential drug
therapy problem such as drug-to-drug interaction. 

If the recipient is ineligible or the claim is not payable, the claim
is denied.  If there is a potential drug therapy problem, the
pharmacist consults with the recipient and/or the recipient's
physician according to the seriousness of the problem.  Depending on
the outcome of this consultation, the pharmacy may then fill the
original prescription, resubmit the claim for a new drug prescribed
by the physician, or submit a reversal to cancel the claim. 

Figure 1 provides an example of the information a pharmacy receives
on its computer terminal when a potential drug therapy problem is
identified.  In this example, on April 1, 1993, the recipient
submitted a prescription for an acne treatment drug that can cause
birth defects.  The information provided to the pharmacist by the
prospective DUR system showed that in March 1993 the recipient
obtained a prescription for a prenatal vitamin, indicating that the
recipient was pregnant.  This information was available even though,
as the physician and pharmacy codes indicate (code 3), the recipient
obtained the vitamin prescription from a different physician and a
different pharmacy than for the acne treatment drug. 

   Figure 1:  Example of
   Prospective DUR Computer Screen

   (See figure in printed
   edition.)


For pharmacies in states that do not use an automated prospective DUR
system, pharmacists are generally limited to comparing the
prescription to medical history and prescription information
maintained at their specific pharmacy or chain of pharmacies. 
Automation may help screen for potential adverse drug reactions, but
a local system would not have the benefit of the patient's complete
Medicaid medical and prescription history as contained in a state
MMIS.  Moreover, there would be no on-line eligibility verification
or claim submission/adjudication. 


--------------------
\5 Given the sensitive nature of Medicaid data files in the state's
MMIS, the automated prospective DUR system restricts access to
authorized users.  We did not assess the adequacy of data security
for any state system, but according to one contractor that operates
systems for two states, such systems use commercially available
software security systems to limit unauthorized access.  According to
this contractor, these security systems require the use of passwords
to access MMIS and track such access by password, transaction, data
set, and individual.  They also produce audit trail reports that show
both successful and unsuccessful access attempts. 


      STATE SYSTEMS IDENTIFY
      INAPPROPRIATE DRUG THERAPY
---------------------------------------------------------- Letter :4.3

Contractor-reported results from operations of statewide automated
prospective DUR systems in Maryland, Missouri, Pennsylvania,
Tennessee, and West Virginia indicate that these systems can reduce
the threat of inappropriate drug therapy for Medicaid patients. 
During portions of 1993, these states' systems identified over 1
million prescriptions that posed a potentially serious risk to
patient safety, with pharmacies canceling over 128,000 of these
prescriptions.  For example, during the period May through October
1993, the Tennessee system showed pharmacies canceled a reported
6,263 prescriptions in which there was a risk of severe drug-drug
interactions and 299 prescriptions with a risk of serious birth
defects.  Similarly, during the period January through October 1993,
the Maryland system showed pharmacies canceled a reported 14,516
prescriptions where there was a risk of severe drug-drug interactions
and 132 prescriptions with a risk of serious birth defects. 
(Appendix IV contains detailed contractor data on DUR system
operations for four of the five states contacted in our review.  The
fifth state--West Virginia--changed DUR contractors during 1993 and
did not have consistent data available for operations during the
year.)

Comparison data were not readily available for prospective reviews
performed by pharmacies without an automated system or by pharmacies
with a local system for the individual pharmacy or chain.  However,
officials from several states said they do not believe states can
effectively meet requirements to conduct prospective drug utilization
reviews without the use of statewide automated prospective DUR
systems because individual pharmacies do not have access to a
Medicaid recipient's complete medical and pharmacy history unless the
recipient obtains all his/her prescriptions from the same pharmacy,
which is unlikely for many recipients.  For example, Tennessee's data
show that more than 47 percent of all its Medicaid recipients used
more than one pharmacy during a 12-month period.  In addition,
Maryland's Medicaid data showed that 1,525 recipients visited three
or more different pharmacies during a 1-month period, and 5,199
recipients visited three or more physicians during the same month. 


   PROSPECTIVE DUR SYSTEMS MAY
   RESULT IN SUBSTANTIAL SAVINGS
------------------------------------------------------------ Letter :5

Automated prospective DUR systems may result in actual savings when
pharmacies cancel prescriptions identified as inappropriate drug
therapy.  Such savings may occur from the value of those
prescriptions canceled and not subsequently filled with another
prescription.  However, even more substantial program savings may
occur when a hospitalization due to an adverse drug reaction is
prevented.  Other savings may also be attributed to the value of
cancellations because of potential fraud, waste, and abuse.  When
costs to acquire and operate automated prospective DUR systems are
compared to these potential savings, the systems appear to be very
cost-effective. 


      VALUE AND POTENTIAL SAVINGS
      IMPLICATIONS OF
      INAPPROPRIATE DRUG THERAPY
      CANCELLATIONS
---------------------------------------------------------- Letter :5.1

Table 1 shows contractor data on (1) the number of Medicaid
prescriptions canceled after being identified as inappropriate drug
therapy and (2) the associated value of these cancellations for five
states' automated prospective DUR systems during varying periods of
operation for 1993.  As this information demonstrates, the value of
these cancellations can be substantial--at least $3.5 million for
these states.  We recognize the value of such cancellations may not
always result in actual savings to the program.  For example,
prescriptions identified as inappropriate drug therapy may be
replaced with another prescription.  However, we believe the value of
cancellations do indicate potential program savings. 



                           Table 1
           
           Summary of Contractor Data on Number and
              Value of Prescriptions Canceled as
                  Inappropriate Drug Therapy

                                     Number of      Value of
                    Period of     prescription  prescription
State               operation       s canceled    s canceled
------------------  ------------  ------------  ------------
Maryland            Jan. 1,             51,519    $1,614,753
                     1993-
                     Oct. 31,
                     1993
Missouri            Mar. 1,             30,271       860,495
                     1993-
                     Oct. 31,
                     1993
Pennsylvania        Sep. 1,             17,414         N/A\a
                     1993-
                     Oct. 31,
                     1993
Tennessee           May. 1,             25,824       932,726
                     1993-
                     Oct. 31,
                     1993
West Virginia       Jan. 1,              3,296      $104,634
                     1993-
                     Feb. 28,
                     1993
============================================================
Totals:                                128,324    $3,512,608
------------------------------------------------------------
\a At the time of our review, Pennsylvania was not preparing reports
showing the value of prescription cancellations. 

In addition to these potential direct savings from canceled
prescriptions, numerous studies indicate that a percentage of all
hospitalizations are due to inappropriate drug therapy, such as
adverse drug-drug interactions (see appendix II).  For example, in
September 1990 congressional hearings on the Medicaid budget
initiative, the chief executive officer of the American
Pharmaceutical Association, National Professional Society of
Pharmacists, estimated that nearly 20 percent of all hospital
admissions can be traced to some kind of drug misadventure.\6 We
recognize that hospitalizations from inappropriate drug therapy may
involve a host of different factors that prospective drug utilization
review may not identify, such as drug use in combination with alcohol
or over-the-counter medication.  However, with Medicaid
hospitalization costs of nearly $39.6 billion for fiscal year 1993,
we believe automated prospective DUR systems could help prevent some
hospitalizations due to inappropriate drug therapy, which could, in
turn, result in substantial potential program savings. 


--------------------
\6 Medicaid Budget Initiatives:  Hearings before the Subcommittee on
Health and the Environment, Committee on Energy and Commerce, House
of Representatives, September 10 and 14, 1990. 


      POTENTIAL SAVINGS BY
      PREVENTING FRAUD, WASTE, AND
      ABUSE
---------------------------------------------------------- Letter :5.2

Automated prospective DUR systems can also provide savings by
preventing potential fraud, waste, and abuse.  While some early
refill claims may be legitimate and ultimately refilled, others can
involve potential fraud or abuse.  In our August 1993 report on drug
fraud in the Medicaid program,\7 we cited an instance in which a
recipient submitted six prescriptions for the same drug within a
4-day period at six different pharmacies.  This recipient also
obtained a total of 85 prescriptions within an 18-day period that
cost the Medicaid program $1,314. 

An automated statewide prospective DUR system could prevent such
fraud and abuse by denying many claims as early refills.  For
example, the Maryland and Tennessee systems automatically deny early
refill claims.  For 10 months of operation during 1993, contractor
reports show that the Maryland system denied 144,880 early refill
prescriptions valued at about $5.2 million.  Similarly, for 6 months
during 1993, contractor reports show that the Tennessee system denied
108,456 early refill prescriptions valued at $3.1 million. 


--------------------
\7 Medicaid Drug Fraud:  Federal Leadership Needed to Reduce Program
Vulnerabilities (GAO/HRD-93-118, Aug.  2, 1993). 


      PRELIMINARY DATA SHOW
      SYSTEMS CAN BE
      COST-EFFECTIVE
---------------------------------------------------------- Letter :5.3

Preliminary cost data gathered from states during our review indicate
that automated prospective DUR systems are relatively inexpensive and
are cost-effective to implement and operate when compared to
potential cost savings.\8 Moreover, the actual cost to the states can
be even lower because federal funding for the Medicaid program can
reimburse up to 90 percent of system development/implementation costs
and 75 percent of system operating costs. 

According to information contained in the advance planning documents
submitted to HCFA by the five states we visited, their total one-time
system development and implementation costs ranged from $105,000 to
$675,000.  Most of these states also provided actual or estimated
system operation costs for 1993.  Comparing such costs to the
potential benefits realized by the systems indicate the potential
cost-effectiveness of automated prospective DUR systems.  For
example, the total one-time system installation cost for Tennessee's
system was about $420,000, and the total operational cost for its
initial 6 months of operation was about $578,000.  In contrast,
contractor data show the value of prescriptions canceled during this
6-month period exceeded $4 million.  For Maryland, the total one-time
system acquisition cost was about $165,000, and the total operational
cost for the system's initial 10 months of operation was about
$472,000.  Contractor data show the value of prescriptions canceled
during this period exceeded $6.7 million. 

The development and operational costs shown above for the Tennessee
and Maryland prospective DUR systems also include costs to install
and operate on-line, real-time Medicaid eligibility verification and
claims submission and adjudication.  Both of these capabilities can
provide substantial additional benefits.  For example, according to a
May 1992 HHS Office of Inspector General report, Massachusetts'
on-line Medicaid eligibility verification system saved the state $8.5
million during its first year of operation.\9 These savings were
attributable to identifying individuals who were not eligible for
Medicaid benefits before they obtained Medicaid services.  In
addition, on-line claims submission and adjudication eliminates costs
states incur to keypunch paper Medicaid drug claims.  According to a
senior programmer/analyst for the Missouri prospective DUR
contractor, on-line claims submission/adjudication allowed the firm
to eliminate five keypunch operator positions. 


--------------------
\8 We did not independently verify these cost data. 

\9 Point-of-Service Claims Management Systems for Medicaid,
Department of Health and Human Services, Office of Inspector General,
May 1992. 


   HCFA COULD INFLUENCE STATE
   IMPLEMENTATION OF AUTOMATED
   PROSPECTIVE DUR SYSTEMS
------------------------------------------------------------ Letter :6

The Social Security Act does not require states to use automated
systems to perform prospective drug utilization reviews.  At the time
of our review, about one-third of the states did not have plans to
acquire automated systems for this purpose.  While HCFA has provided
some information to the states on the use of automated prospective
DUR systems, it has generally provided little guidance.  Additional
information, especially on costs and benefits, could encourage more
states to adopt these systems.  Additional guidance would also help
standardize procedures for implementing the systems. 

As of December 31, 1993, eight states were operating automated
prospective DUR systems.  Medicaid DUR coordinators from 22 other
states plus the District of Columbia\10 told us they plan to begin
operating such systems--most sometime during 1994 and 1995.  DUR
coordinators from the remaining 20 states told us their states did
not have any plans to acquire and operate these systems.\11 Appendix
V lists the status of states that are implementing automated
prospective DUR systems. 

Despite the significant potential benefits and relatively low cost of
automated prospective DUR systems, DUR coordinators from 11 of the 16
states not planning to acquire these systems said their decision was
based on the belief that the systems would not provide tangible
benefits to offset system costs.  For fiscal year 1992, these 16
states' Medicaid programs dispensed about 82 million prescriptions to
over 5.6 million Medicaid patients.  DUR coordinators from five of
these states said that HCFA had not provided them information on the
costs and benefits of automated prospective DUR systems.  These
officials added that if they had data showing the systems were
cost-effective, they might reconsider their decision. 

DUR coordinators from four of the states we contacted specifically
said that information on the capabilities and desirable features of
automated prospective DUR systems would be beneficial.  Since
automated prospective DUR systems are not statutorily required, there
are no federal standards or guidelines for implementation.  As a
result, system functions can vary by state.  For example, data for
the five state systems we reviewed showed that four systems screen
for pregnancy conflicts; the remaining state does not.  Also, one of
the five state systems automatically suspends a claim if a drug
therapy problem is identified.  The system requires the pharmacist to
enter a code into the system to override the alert and adjudicate the
claim for payment.  If the alert is not overridden within 48 hours,
the system automatically cancels the claim.  In contrast, the other
four systems automatically adjudicate a claim for payment even though
there may be a drug therapy alert.  These systems require pharmacists
to enter a code to reverse (cancel) the claim only if they believe a
serious problem exists. 

HCFA has taken some steps to provide information on the use of
automated prospective DUR systems.  For example, HCFA cosponsored a
2-day seminar on the use of these systems and conducted a
demonstration project with the state of Iowa.  However, according to
the HCFA official responsible for overseeing states' DUR programs,
HCFA has not had the staff needed to take a more active role in
encouraging states to acquire automated prospective DUR systems or to
issue standards or guidelines on their use. 


--------------------
\10 According to the DUR Coordinator for the District of Columbia's
Office of Medical Assistance, the District's City Council has
approved funding for an automated prospective DUR system.  Work has
begun for the District's MMIS contractor to integrate the prospective
DUR system with the MMIS.  This official estimates the system will
begin operations in September 1994. 

\11 Four of these states are not required to perform drug utilization
reviews because most of their Medicaid patients are covered under
some type of a health maintenance organization plan.  The Social
Security Act does not require states to conduct drug utilization
reviews for covered outpatient drugs that are dispensed by health
maintenance organizations. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

Although data are not yet available to conclusively show the costs
and benefits of using statewide automated prospective DUR systems for
Medicaid prescriptions, these systems offer a significant opportunity
to increase patient safety.  Limited operation of automated
prospective DUR systems by several states demonstrates this and
indicates the significant cost savings potential of these systems. 
More than two-thirds of the states and the District of Columbia agree
that these systems are beneficial and either operate or plan to
implement automated prospective DUR systems within the next few
years.  Additional information about system use and benefits would
encourage more states to consider these systems.  In addition, all
states could benefit from federal system guidance to help ensure
standard implementation of effective systems. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :8

To increase patient safety and the potential for significant program
savings, we recommend that the Secretary of the Department of Health
and Human Services direct the Administrator of the Health Care
Financing Administration to (1) gather information on the costs and
benefits of automated prospective DUR systems, (2) develop guidance
on desirable features and capabilities for these systems, and (3)
provide this information and guidance to all states. 


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

In commenting on a draft of our report, the Department of Health and
Human Services generally agreed with our recommendations and the
facts presented.  The Department agreed that data are not yet
available to conclusively show the benefits and costs of using
statewide automated DUR systems, that these systems offer a
tremendous opportunity to improve patient safety, and that all states
could benefit from federal system guidance. 

The Department said that it had contracted for the development of
methodological guidance on how to evaluate the cost-effectiveness of
conducting prospective DUR through electronic systems, and that it
would distribute the resulting report to all state Medicaid programs
when available.  The Department also noted that it expects to receive
shortly each State's annually required report on their DUR program. 
It agreed that these reports should provide better information on
states' experiences with on-line prospective DUR, and that the
Department would share its analysis of these reports with all of the
states as soon as possible. 

The Department did not address our recommendation concerning
development and distribution of guidance regarding desirable features
and capabilities of automated DUR systems.  Because patient safety is
a primary concern, it is important that the Department expeditiously
develop and transmit such guidance to all states. 


---------------------------------------------------------- Letter :9.1

As agreed with your office, unless you publicly announce the contents
of this report earlier, we plan no further distribution until 30 days
from the date of this letter.  We will then send copies to the
Secretary of Health and Human Services; the Director, Office of
Management and Budget; other interested congressional committees; and
state Medicaid directors and DUR coordinators.  Copies will also be
made available to others upon request. 

Please contact me at (202) 512-6252 if you or your staff have any
questions concerning this report.  Other major contributors to this
report are listed in appendix VII. 

Sincerely yours,

Frank W.  Reilly
Director, Information Resources Management/
 Health, Education, and Human Services


GLOSSARY OF DRUG UTILIZATION
REVIEW TERMS
=========================================================== Appendix I


      ADVERSE DRUG-DRUG
      INTERACTION
------------------------------------------------------- Appendix I:0.1

The potential for, or the occurrence of, an adverse medical effect as
a result of the recipient using two or more drugs together. 


      ADVERSE MEDICAL RESULT
------------------------------------------------------- Appendix I:0.2

A clinically significant undesirable effect experienced by a patient
due to the course of drug therapy. 


      DRUG-AGE CONTRAINDICATION
------------------------------------------------------- Appendix I:0.3

Use of a drug that is not recommended for the age group of the
patient.  This can occur when the patient is too old or young for the
given medication. 


      DRUG-ALLERGY INTERACTIONS
------------------------------------------------------- Appendix I:0.4

The significant potential for, or the occurrence of, an allergic
reaction as a result of drug therapy. 


      DRUG-DISEASE
      CONTRAINDICATION
------------------------------------------------------- Appendix I:0.5

The potential for, or occurrence of, an undesirable alteration of the
therapeutic effect of a given prescription because of the presence,
in the patient for whom it is prescribed, of a disease condition or
the potential for, or the occurrence of, an adverse effect of the
drug on the patient's disease condition. 


      GROSS OVERUSE
------------------------------------------------------- Appendix I:0.6

Repetitive overutilization without therapeutic benefit. 


      INCORRECT DRUG DOSAGE
------------------------------------------------------- Appendix I:0.7

Dosage that lies outside the daily dosage range specified in
predetermined standards as necessary to achieve therapeutic benefit. 
Dosage range is the strength multiplied by the quantity dispensed
divided by days of supply. 


      INCORRECT DURATION OF DRUG
      TREATMENT
------------------------------------------------------- Appendix I:0.8

The number of days of prescribed therapy exceeds or falls short of
the recommendations contained in the predetermined standards. 


      OVERUTILIZATION
------------------------------------------------------- Appendix I:0.9

Use of a drug in quantities or for durations that put the recipient
at risk of an adverse medical result. 


      PREGNANCY CONFLICT
------------------------------------------------------ Appendix I:0.10

Use of prescribed drug that is not recommended during pregnancy. 


      THERAPEUTIC DUPLICATION
------------------------------------------------------ Appendix I:0.11

The prescribing and dispensing of two or more drugs from the same
therapeutic class such that the combined daily dose puts the
recipient at risk of an adverse medical effect or incurs additional
program costs without additional therapeutic benefit. 


      UNDERUTILIZATION
------------------------------------------------------ Appendix I:0.12

A drug used by a recipient in insufficient quantity to achieve a
desired therapeutic goal. 


STUDIES RELATED TO
HOSPITALIZATIONS DUE TO ADVERSE
DRUG EVENTS
========================================================== Appendix II

Beard, K.  "Adverse (Drug) Reactions as a Cause of Hospital Admission
in the Aged." Drugs & Aging (1992). 

Bergman, U.  and B.  E.  Wilholm.  "Drug-Related Problems Causing
Admission to a Medical Clinic." European Journal of Clinical
Pharmacology (1981). 

Bero, L.  A., Ph.D.; H.  L.  Lipton, Ph.D.; and J.  A.  Bird, Ph.D. 
"Characterization of Geriatric Drug-Related Hospital Readmissions."
Medical Care, Vol.  29, No.  10.  (May 1991). 

Black, A.  J., M.B., and K.  Somers, M.B.  "Drug-Related Illness
Resulting in Hospital Admission." Journal of the Royal College of
Physicians of London, Vol.  18, No.  1.  (Jan.  1984). 

Caranasos, G.  J., M.D.; R.B.  Stewart, M.S.; and L.E.  Cluff, M.D. 
"Drug-Induced Illness Leading to Hospitalization." Journal of the
American Medical Association, Vol.  228, No.  6.  (May 6, 1974). 

Col, M.; J.E.  Fanale; and P.  Kronholm.  "The Role of Medication
Noncompliance and Adverse Drug Reactions in Hospitalizations of the
Elderly." Archives of Internal Medicine, Vol.  150, No.  4.  (Apr. 
1990) 841-845. 

Colt, Henri G., M.D., and Alvin P.  Shapiro, M.D.  "Drug-Induced
Illness as a Cause for Admission to a Community Hospital." Journal of
American Geriatrics Society (Apr.  1989). 

Davidson, K.  W., M.D.; A.  Kahn, M.D.; and R.  D.  Price, Ph.D. 
"Reduction of Adverse Drug Reactions by Computerized Drug Interaction
Screening." The Journal of Family Practice, Vol.  25, No.  4.  (1987)
371-375. 

Hallas, J., M.D., O.  Davidsen; E.  Grodum; N.  Damsbo; and L.  Gram. 
"Drug-Related Illness as a Cause of Admission to a Department of
Respiratory Medicine." Respiration, Vol.  59.  (1992) 30-34. 

Levy, M.; L.  Mermelstein; and D.  Hemo.  "Medical Admissions Due to
Noncompliance With Drug Therapy." International Journal of Clinical
Pharmacology, Therapy, and Toxicology, Vol.  20, No.  12.  (1982)
600-604. 

Levy, M.; M.  Lipshitz; and M.  Eliakim.  "Hospital Admissions Due to
Adverse Drug Reactions." The American Journal of the Medical Sciences
(Jan.-Feb.  1979). 

McKenney, J.  M.  and W.L.  Harrison.  "Drug-Related Hospital
Admissions." American Journal of Hospital Pharmacy, Vol.  33.  (Aug. 
1976). 

Miller, R.  R., Pharm.D., Ph.D.  "Hospital Admissions Due to Adverse
Drug Reactions." Archives of Internal Medicine, Vol.  134.  (Aug. 
1974). 

Public Citizen Health Group.  Worst Pills, Best Pills:  The Older
Adults Guide to Avoiding Drug-Induced Death or Illness (New York: 
Pantheon, 1988). 

Stewart, R.  B., M.S.; P.  K.  Springer, M.D.; and J.  E.  Adams,
M.D.  "Drug-Related Admissions to an Inpatient Psychiatric Unit."
American Journal of Psychiatry (September 1980). 

A Study of the Impact of Inappropriate Ambulatory Drug Therapy on
Hospitalization, Health Information Designs, Inc., Washington, D.C.: 
May 16, 1978. 

Sullivan, S.  D., D.  H.  Kreling, and T.  K.  Hazlet. 
"Noncompliance With Medications Regimens and Subsequent
Hospitalizations:  A Literature Analysis and Cost of Hospitalization
Estimate." Journal of Research in Pharmaceutical Economics, Vol.  2,
No.  2.  (1990). 

Trunet, P.; I.T.  Borda; A.  V.  Rouget; M.  Rapin; and F.  Lhoste. 
"The Role of Drug-Induced Illness in Admissions to an Intensive Care
Unit." Intensive Care Medicine (1986). 


PROCESSING A MEDICAID PRESCRIPTION
DRUG BENEFIT CLAIM USING A
STATEWIDE AUTOMATED PROSPECTIVE
DUR SYSTEM
========================================================= Appendix III



   (See figure in printed
   edition.)


WARNINGS AND CLAIM CANCELLATIONS
RESULTING FROM STATES' PROSPECTIVE
DUR SYSTEMS
========================================================== Appendix IV

The following tables provide detailed contractor data on the results
of automated prospective DURs as provided by four of the five states
we visited that operated such systems during 1993.  West Virginia
data are not presented because it changed DUR contractors during 1993
and did not have consistent data available for operations during the
year.  These data include (1) the number and types of drug therapy
alert messages sent via states' prospective DUR systems, (2) the
number of claims reversed and the value of these reversals, and (3)
the number and value of early refill prescriptions denied and not
resubmitted.  Table content may vary between states due to
differences in the types of data reported by states' systems.  Also,
the periods of operation during 1993 differ by state.  Given these
variances and the number of systems, we did not attempt to verify
these data. 



                          Table IV.1
           
           Maryland Prospective DUR System Warnings
           and Cancellations From January 1, 1993,
                     to October 31, 1993

------------------------------  --------------  ------------
Total claims processed                             3,365,929
Total cost of claims processed                   $95,025,298
                                  Total claims      Value of
                                      reversed     reversals
Drug therapy alerts
Drug-drug interactions                  14,516      $456,534
Pregnancy conflict                         132         1,346
Therapeutic duplication                  5,480       258,510
Underutilization                         4,486       141,940
Excessive daily dose                     7,082       168,058
Excessive daily dose over age            7,338       385,252
 65
Excessive daily dose children              225         4,640
Excessive quantity dispensed             3,337        86,347
Insufficient daily dose for              8,654       105,377
 age
Drug-disease interaction                   269         6,749
Total                                   51,519     1,614,753
Early refills denied                   144,880     5,173,146
============================================================
Total claims reversed/value            196,399    $6,787,899
Reversals/value of reversals              5.83          7.14
 as a percent
------------------------------------------------------------


                          Table IV.2
           
           Missouri Prospective DUR System Warnings
            and Cancellations For the Period March
                 1, 1993, to October 31, 1993

------------------------  ----------  ----------  ----------
Total claims processed                             3,340,917
                           Number of      Number       Total
                            warnings   of claims    value of
                                sent    reversed   reversals
Drug therapy problem
Drug disease conflict          3,773         138      $3,449
Drug indicated disease        41,814       1,316      35,515
 conflict
Below minimum dose range       7,294         244       6,186
Above maximum dose range       3,473          96       7,464
Below minimum daily dose     151,700       5,406      92,431
Above maximum daily dose     136,470       5,585     183,010
Significant side effect        7,681          91       1,767
Additive side effect           1,340          29         867
Side effect medical              261           1           4
 condition
Indicated for prior              279           3          21
 drug's side
 effect
Drug-drug interaction         46,205       1,320      30,619
Duplicative therapy same     241,421      11,386     294,041
 drug
Therapeutic duplication       77,106       2,908      76,684
RX initiates 90-day           11,916         268      19,118
 therapy
RX applies to 90-day          22,516         473      34,906
 therapy
Current RX exceeds 90-        37,509         675      59,422
 day
 therapy
Maintenance dose              20,211         332      14,991
============================================================
Total                        810,969      30,271    $860,495
------------------------------------------------------------


                          Table IV.3
           
             Pennsylvania Prospective DUR System
              Warnings and Cancellations For the
           Period September 1, 1993, to October 31,
                             1993

                               Total       Total     Percent
                              alerts      claims      alerts
                                sent    reversed    reversed
--------------------------  --------  ----------  ----------
Drug therapy alerts
Drug-drug interaction          3,412         643       18.85
Low dose alert                21,827       4,810       22.04
High dose alert               25,177       6,924       27.50
Drug age precaution              135          22       16.30
Drug pregnancy alert              19           9       47.37
Therapeutic duplication       24,810       5,006       20.18
Total                         75,380      17,414       23.10
Early refills                 14,042       5,151       36.68
============================================================
Total                         89,422      22,565       25.23
------------------------------------------------------------
Note:  At the time of our review, Pennsylvania was not preparing
reports showing value of prescription cancellations. 



                          Table IV.4
           
               Tennessee Prospective DUR System
              Warnings and Cancellations for the
           Period May 1, 1993, to October 31, 1993

--------------------------------  ------------  ------------
Total claims processed                             4,659,263
Total cost of claims processed                  $103,821,733
                                         Total      Value of
                                        claims     reversals
                                      reversed
Drug therapy alerts
Drug-drug interactions                   6,263      $198,507
Pregnancy conflict                         299         2,683
Therapeutic duplication                  5,716       244,625
Underutilization                         1,684        45,937
Excessive daily dose                     6,312       158,571
Excessive daily dose over age 65           319        19,755
Excessive daily dose children              154         2,654
Excessive quantity dispensed             3,194       238,772
Insufficient daily dose over age         1,641        16,239
 65
Drug-disease interaction                   242         4,983
Total                                   25,824       932,726
Early refills denied                   108,456     3,106,450
============================================================
Total claims reversed/value            134,280    $4,039,176
 Reversals/value of                  2.88          3.89
 reversals as a percent
------------------------------------------------------------

STATUS OF STATE IMPLEMENTATION OF
PROSPECTIVE DUR SYSTEMS
=========================================================== Appendix V


                                             Date operations
State                                                  began
----------------------------------------  ------------------
Illinois                                        January 1993
 Maryland                                       January 1993
 New Mexico                                     October 1993
 Missouri                                      February 1993
 Pennsylvania                                      June 1993
 Tennessee                                        April 1993
 Vermont                                       November 1993
 West Virginia                                     July 1992
------------------------------------------------------------

                                              Date operation
State                                       planned to begin
----------------------------------------  ------------------
Alabama                                         Date not set
 Arkansas                                       Date not set
 California                                             1994
 Colorado                                               1994
 Connecticut                                            1995
 Delaware                                               1994
 District of Columbia                                   1994
 Florida                                                1994
 Indiana                                                1994
 Idaho                                                  1995
 Iowa                                                   1994
 Kentucky                                               1994
 Maine                                                  1994
 Minnesota                                              1994
 Nebraska                                       Date not set
 New Hampshire                                          1994
 New Jersey                                             1995
 New York                                               1994
 Oklahoma                                               1994
 Oregon                                                 1994
 Texas                                                  1994
 Utah                                                   1994
 Virginia                                               1994
------------------------------------------------------------

State
------------------------------------------------------------
Alaska
Arizona\a
Georgia
Hawaii\a
Kansas
Louisiana
Massachusetts
Michigan
Mississippi
Montana
Nevada\a
North Carolina
North Dakota
Ohio
Rhode Island
South Carolina
South Dakota
Washington\a
Wisconsin
Wyoming
------------------------------------------------------------
\a These states do not plan to acquire automated prospective DUR
systems because most of their Medicaid patients will be covered under
some type of a health maintenance organization plan.  The Social
Security Act does not require states to conduct drug utilization
reviews for covered outpatient drugs that are dispensed by health
maintenance organizations. 




(See figure in printed edition.)Appendix VI
COMMENTS FROM THE DEPARTMENT OF
HEALTH AND HUMAN SERVICES
=========================================================== Appendix V



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII

ACCOUNTING AND INFORMATION
MANAGEMENT DIVISION, WASHINGTON,
D.C. 

William B.  Ritt, Assistant Director
Shane D.  Hartzler, Reports Analyst

KANSAS CITY REGIONAL OFFICE

John B.  Mollet, Evaluator-in-Charge
George L.  Jones, Adviser
John L.  Womble, Senior Evaluator



RELATED GAO PRODUCTS
=========================================================== Appendix 0

Medicaid:  A Program Highly Vulnerable to Fraud (GAO/T-HEHS-94-106,
Feb.  25, 1994). 

Medicaid Drug Fraud:  Federal Leadership Needed to Reduce Program
Vulnerabilities (GAO/HRD-93-118, Aug.  2, 1993). 

Medicaid Prescription Drug Diversion:  A Major Problem, But State
Approaches Offer Some Promise (GAO/T-HRD-92-48, July 29, 1992). 

Prescription Drug Monitoring:  States Can Readily Identify Illegal
Sales and Use of Controlled Substances (GAO/HRD-92-115, July 21,
1992). 

Prescription Drugs:  HCFA's Proposed Drug Utilization Review System
Ignores Quality of Care Issues (GAO/PEMD-89-26BR, July 13, 1989). 

Prescription Drugs:  Information on Selected Drug Utilization Review
Systems (GAO/PEMD-89-18, May 24, 1989).