VA Health Care: Exploring Options to Improve Veterans' Access to VA
Facilities (Letter Report, 02/06/96, GAO/HEHS-96-52).

Pursuant to a congressional request, GAO provided information on
veterans' use of Department of Veterans Affairs' (VA) medical
facilities, focusing on: (1) users' characteristics; (2) the geographic
accessibility of VA and private medical facilities that provide standard
benefits; and (3) options to improve accessibility of VA health care.

GAO found that: (1) in the early 1990s, over 80 percent of veterans who
received health care services obtained them from non-VA sources; (2)
veterans who used VA medical facilities generally had lower incomes and
were less likely to have private health insurance than veterans who
obtained health care from non-VA facilities: (3) veterans with
service-connected disabilities utilized VA facilities more often than
other veterans; (4) about 50 percent of all veterans lived over 25 miles
from a VA facility and 11 percent of veterans lived within 5 miles of a
VA hospital; (5) although VA hospitals and outpatient clinics were
geographically less accessible to veterans than private medical
facilities, veterans had better access to certain specialty services
through VA facilities; (6) options to improve veterans' access to VA
health care include determining whether to improve access for current
users, all veterans, or selected veterans groups, and comparing the
costs of VA-provided services and contractor-provided services; and (7)
although VA facilities are more costly to operate, they lessen the
chances of program abuse by giving VA more control over resources.

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

 REPORTNUM:  HEHS-96-52
     TITLE:  VA Health Care: Exploring Options to Improve Veterans' 
             Access to VA Facilities
      DATE:  02/06/96
   SUBJECT:  Veterans hospitals
             Health care services
             Veterans benefits
             Income statistics
             Health resources utilization
             Handicapped persons
             Demographic data
             Health insurance
             Health care planning
IDENTIFIER:  Chicago (IL)
             Rockford (IL)
             Salem (OR)
             Dublin (GA)
             
******************************************************************
** 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 Ranking Minority Member, Subcommittee on Compensation,
Pension, Insurance, and Memorial Affairs, Committee on Veterans'
Affairs, House of Representatives

February 1996

VA HEALTH CARE - EXPLORING OPTIONS
TO IMPROVE VETERANS' ACCESS TO VA
FACILITIES

GAO/HEHS-96-52

Accessibility of VA Health Care

(406077)


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

  CHAMPUS - Civilian Health and Medical Program of the Uniformed
     Services
  HMO - health maintenance organization
  VA - Department of Veterans Affairs

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


B-256197

February 6, 1996

The Honorable Lane Evans
Ranking Minority Member
Subcommittee on Compensation, Pension,
 Insurance, and Memorial Affairs
Committee on Veterans' Affairs
House of Representatives

Dear Mr.  Evans: 

The Department of Veterans Affairs (VA) runs one of the nation's
largest health care systems.  As a major health care provider, it
competes with private health care providers to serve veterans.  In
fiscal year 1994, VA served about 2.4 million of the nation's 27
million veterans.  The other 24.6 million veterans either received
care from private providers or did not seek health care services. 

When the VA health care system was established in 1930, public and
private health insurance were virtually nonexistent.  VA developed
its system as a direct delivery system, with the government owning
and operating its own health care facilities.  Since then, the VA
system has become one of the nation's largest networks of direct
delivery health care providers, with 173 hospitals and 376 outpatient
clinics nationwide.  Because public and private health insurance
programs have also grown, most veterans now have one or more
alternatives to VA health care.  Many veterans indicate that they use
private providers because they live too far from a VA hospital or
outpatient clinic.  VA has recently encouraged its facilities to
improve veterans' access to VA health care. 

This report responds to your request for information on veterans' use
of VA medical facilities.  It discusses (1) characteristics of recent
users of VA medical facilities; (2) the geographic accessibility of
VA and private medical facilities that provide standard benefits; and
(3) options that VA facilities might explore to improve accessibility
of VA health care, such as where to locate new medical facilities and
whether to establish new VA-operated facilities or contract with
private providers. 

We reviewed VA studies and previous GAO reports for information on
users of VA medical facilities.  To assess medical facilities'
accessibility to veterans, we interviewed VA and private-sector
officials, analyzed VA and private-sector provider documents, and
compared sites of VA and private facilities in four locations that
represent the types of markets typically served by VA: 

  Chicago, Illinois--an urban area with a VA medical center,

  Rockford, Illinois--an urban area with a freestanding VA outpatient
     clinic,

  Salem, Oregon--an urban area with no VA medical facility, and

  Dublin, Georgia--a rural area with a VA medical center. 

To identify the options that VA facilities might consider to improve
accessibility of VA health care, we reviewed VA policy directives and
other guidance and interviewed VA headquarters and selected medical
center officials (see app.  I for a detailed discussion of our scope
and methodology).  We did our work between January 1994 and November
1995 in accordance with generally accepted government auditing
standards. 


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

Veterans who used VA medical facilities during the early 1990s had
lower incomes and were less likely to have private insurance than
veterans who obtained their health care from non-VA facilities. 
Also, almost half the veterans who used VA facilities had conditions
incurred during or aggravated by military service.  Most veterans
lived over 25 miles from VA medical facilities, but veterans who
lived within 5 miles of such facilities made greater use of them. 

VA hospitals and outpatient clinics that provide standard benefits
are generally less geographically accessible to veterans than private
medical facilities because they are outnumbered by private facilities
and are located throughout the country.  VA has 173 hospitals and
about 14,000 physicians, compared with the nation's over 6,000 public
and private hospitals and 550,000 private physicians.  Consequently,
veterans often travel longer distances or endure longer travel times
to obtain care in VA medical facilities than they would if they used
private providers.  For example, while dozens of community hospitals
and hundreds of private physicians are located in the 52-county area
served by VA's Dublin, Georgia, medical center, some veterans in the
area travel up to 150 miles to obtain services from the VA facility. 

VA has recently begun taking steps to improve veterans' access to VA
health care.  In improving accessibility, VA faces two basic
decisions:  where to locate new facilities and how to deliver care. 
We found that in deciding where to locate facilities, VA could target
areas that would improve convenience for existing users, improve
access for all veterans, or improve access for specific veteran
groups or eligibility categories.  For example, VA's Dublin, Georgia,
medical center has targeted two relatively densely populated areas as
possible locations for new outpatient clinics so it can improve
access to new and current users.  In deciding how to deliver care, VA
could compare the costs and other factors involved in providing care
in VA-operated facilities with those involved in contracting with
private providers.  For example, while VA-operated facilities
typically are more expensive in the short run, such facilities would
give VA more control over resources. 


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

In fiscal year 1994, VA medical care cost over $15.6 billion.  VA
facilities served about 2.4 million veterans nationwide.  VA services
included 925,000 inpatient hospital stays and about 25 million
outpatient visits. 

The VA health care system was established primarily to treat
war-related injuries and help rehabilitate veterans with disabilities
incurred or aggravated as a result of military service; such
service-connected disabilities include blindness, paralysis, and loss
of limb.  Subsequently, the Congress expanded the system so that
hospital services could be provided, to the extent space and
resources were available, to veterans who did not have
service-connected disabilities and who lacked the resources to pay. 
Today, all veterans are eligible for treatment at VA medical
facilities, but few are entitled to the full range of services under
the existing complex eligibility requirements.  For example, veterans
with service-connected disabilities are eligible for cost-free
hospital care, while high-income veterans without such disabilities
are eligible for such care but may receive it only if space and
resources are available.  Such high-income veterans are also subject
to co-payments and, if insured, their insurers must be billed for
services provided by VA facilities. 

When the VA health care system was established, there were no health
insurance programs to help veterans pay for needed care.  Private
insurance began to emerge in the 1930s and expanded rapidly in the
1950s.  In the 1960s, the Congress established the Medicare and
Medicaid programs--public insurance programs that help the elderly
and selected low-income individuals pay for health care.  By 1990, 9
out of 10 veterans had one or more alternatives to VA health care for
standard benefits (excluding such special services as psychiatric
care and prescription drugs):  about 81 percent had private health
insurance, almost 26 percent were eligible for Medicare, and 1.6
percent had Medicaid coverage. 

The VA health care system is fundamentally different from private or
public health insurance programs.  VA generally delivers health care
to patients directly using salaried physicians, nurses, and other
professionals in VA facilities.  Insurance programs, on the other
hand, provide services on a fee-for-service basis or through
contracts with private providers.  VA does not charge veterans for
services provided to treat service-connected disabilities, although
veterans may be required to make co-payments for nonservice-connected
conditions if their income exceeds prescribed thresholds.  Insurance
programs typically charge premiums to, impose deductibles on, and
require co-payments from all enrollees.  In addition, VA provides
some services that insurance programs typically do not provide.  For
example, VA covers outpatient prescription drugs and dental care that
are not covered by Medicare.  Similarly, while Medicare and most
private insurance programs provide short-term nursing home care
following hospitalization, VA may, in some instances, offer more
extensive, longer-term nursing home and domiciliary care.\1

If cost and service differences between VA and insurance programs
diminish, the importance of accessibility as a factor in veterans'
decisions on where to obtain health care could increase.  VA surveyed
the status of veterans in 1987 and found that distance to VA
facilities was one of the reasons most frequently cited by veterans
for not using VA facilities.\2 VA's survey of the status of veterans
in 1992 also showed that about one-third of the responding veterans
who received inpatient care did not choose a VA hospital because they
were too far from the VA location.\3 Similarly, over one-fourth of
the veteran respondents to VA's health care reform customer
satisfaction survey said that, given a choice of health programs and
assuming no difference in cost, they would choose private providers,
in part, because the private providers were more accessible.\4

In October 1995, VA began a major reorganization that will replace
its four regional offices with 22 integrated service networks.  Each
network will include from 5 to 11 medical centers.  The service
networks will be the basic budgetary and planning units for
delivering veterans' health care.  VA officials envision greater
emphasis on integrated delivery systems of care and on outpatient and
primary care services as a result of the reorganization. 


--------------------
\1 According to VA officials, eligibility requirements for long-term
care are very complex. 

\2 Department of Veterans Affairs, 1987 Survey of Veterans
(Washington, D.C.:  Department of Veterans Affairs, July 1989). 

\3 National Center for Veterans Analysis and Statistics, Department
of Veterans Affairs, National Survey of Veterans (Washington, D.C.: 
Department of Veterans Affairs, Apr.  1995). 

\4 Department of Veterans Affairs, Veterans Health Administration
Health Care Reform Customer Satisfaction Survey (Washington, D.C.: 
Department of Veterans Affairs:  1994). 


   PROFILE OF VA FACILITY USERS
------------------------------------------------------------ Letter :3

While many veterans received medical care during the early 1990s,
relatively few obtained their care from VA facilities.  Veterans who
used VA facilities generally had lower incomes and were less likely
to have private insurance than veterans who used non-VA facilities. 
Also, almost half the veterans who used VA facilities had
service-connected disabilities.  Most veterans lived over 25 miles
from VA facilities, but veterans who lived within 5 miles of such
facilities made greater use of the facilities. 


      MOST VETERANS USED NON-VA
      FACILITIES
---------------------------------------------------------- Letter :3.1

VA's survey of the status of veterans in 1992 showed that 54.8
percent of the 27 million veterans nationwide received medical care
in 1992 but few received care in VA facilities.  About 90 percent of
the veterans who received either inpatient or outpatient care got it
from non-VA sources, as did over 80 percent of the veterans who
received both types of care.  The survey also showed that of the 10
percent of veterans who received care in VA facilities, 5.6 percent
received VA care exclusively and 4.4 percent received both VA and
non-VA care.  VA officials said that, while only 10 percent of all
veterans used VA facilities in 1992, a larger percentage of veterans
used VA facilities over a 3-year period; the reason for this is that
many veterans do not require medical services every year.  The
officials also said that some veterans who received both VA and
non-VA care needed resource-intensive care and had been transferred
for that purpose to VA facilities. 


      VA FACILITY USERS GENERALLY
      HAD LOWER INCOMES THAN USERS
      OF NON-VA FACILITIES
---------------------------------------------------------- Letter :3.2

Using VA patient treatment records and Internal Revenue Service
information, we examined the incomes of the 2.2 million veterans who
used VA medical facilities in 1991 and found that two-thirds had
incomes under $20,000, and about one-third of these had incomes under
$5,000.\5 However, VA's survey of veterans in 1992 showed that only
about 32 percent of all veterans had gross family incomes below
$20,000.  This suggests that most VA facility users had lower incomes
than non-VA facility users. 

VA's survey also showed that veterans' use of VA facilities decreased
as incomes increased.  For example, over 44 percent of veterans with
incomes under $10,000 received inpatient care from VA facilities,
compared with 2.5 percent of those with incomes over $50,000; and 32
percent of veterans with incomes under $10,000 received outpatient
care in VA facilities, compared with 1.7 percent with incomes over
$50,000.  Moreover, cost was the reason most frequently cited by
veterans who participated in VA's survey for choosing a VA hospital
for inpatient care:  over 19 percent cited cost as a reason for
choosing a VA hospital, while less than 1 percent cited cost as a
reason for choosing private or public hospitals. 


--------------------
\5 VA Health Care:  A Profile of Veterans Using VA Medical Centers in
1991 (GAO/HEHS-94-113FS, Mar.  29, 1994). 


      VA FACILITY USERS WERE LESS
      LIKELY TO HAVE INSURANCE
---------------------------------------------------------- Letter :3.3

Veterans who used VA facilities were also less likely than users of
non-VA facilities to have insurance.  VA's survey of veterans in 1992
showed that about 49 percent of all veterans had private insurance
alone, 12 percent had public insurance alone,\6 29 percent had both,
and 9 percent had no insurance.\7 The survey also showed that the
availability of insurance increased as income increased.  For
example, about 27 percent of veterans with income under $10,000 had
no insurance, compared with 1.3 percent with income over $50,000. 
Conversely, about 14 percent of veterans with income under $10,000
had private insurance, compared with over 76 percent with income over
$50,000. 


--------------------
\6 Public insurance includes Medicare, Medicaid, and the Civilian
Health and Medical Program of the Uniformed Services (CHAMPUS). 

\7 Type of insurance was unknown for about 1 percent of the veterans. 


      VETERANS WITH
      SERVICE-CONNECTED
      DISABILITIES WERE MORE
      LIKELY TO USE VA FACILITIES
---------------------------------------------------------- Letter :3.4

Many veterans who used VA facilities had service-connected
disabilities.  In 1994, we reported that 1 million of the 2.2 million
veterans who received care in VA facilities in 1991 had
service-connected disabilities.\8 In addition, VA's survey of
veterans in 1992 showed that veterans with service-connected
disabilities were more likely than those without them to use VA
facilities.  Almost 72 percent of the veterans who reported they had
a service-connected disability used medical facilities, compared with
53 percent who did not have such disabilities.  About 24 percent of
veterans with service-connected disabilities received inpatient care
exclusively in VA facilities, compared with less than 7 percent who
did not have such disabilities.  Similarly, about 16 percent of
veterans with service-connected disabilities received outpatient care
exclusively in VA, compared with slightly over 3 percent who did not
have such disabilities. 


--------------------
\8 GAO/HEHS-94-113FS, Mar.  29, 1994. 


      VA FACILITY USERS OFTEN
      TRAVELED LONG DISTANCES
---------------------------------------------------------- Letter :3.5

Veterans' use of VA facilities was influenced by the distances they
had to travel to the facilities.  Our analysis of 1990 census data
showed that about 50 percent of all veterans lived over 25 miles from
a VA hospital, including 6 percent who lived over 100 miles away, and
34 percent lived over 25 miles from a VA clinic.\9 Many of those
veterans traveled long distances to use VA facilities.  For example,
44 percent of VA facility users lived over 25 miles from VA hospitals
providing acute medical and surgical care, and 32 percent lived over
25 miles from outpatient clinics that provided such services. 

Living closer to a VA facility significantly increases the likelihood
that a veteran will use VA health care and VA officials told us that
some veterans move to be closer to VA medical facilities.  For
example, about 11 percent of all veterans lived within 5 miles of a
VA hospital, and they accounted for 22 percent of the facility users. 
Similarly, 17 percent of veterans lived within 5 miles of a VA
outpatient clinic and accounted for 26 percent of the clinic users. 
The likelihood and frequency of VA use decline significantly among
veterans living more than 5 miles away.  Veterans with
service-connected disabilities and low-income veterans, however, are
less sensitive to distance. 


--------------------
\9 VA Health Care:  How Distance From VA Facilities Affects Veterans'
Use of VA Services (GAO/HEHS-96-31, Dec.  20, 1995). 


   VA MEDICAL FACILITIES ARE LESS
   ACCESSIBLE THAN PRIVATE
   FACILITIES
------------------------------------------------------------ Letter :4

VA medical facilities are located throughout the country and are
outnumbered by private-sector facilities.  Consequently,
private-sector health care is usually more convenient to veterans
than VA health care.  In some urban areas, for example, VA operates
one or two hospitals, while the private sector may have a dozen or
more hospitals and hundreds of locations that provide primary and
specialty care.  While private-sector hospital and outpatient
providers are less plentiful in rural areas than in urban areas, they
still generally outnumber VA providers. 


      VA HEALTH CARE FACILITIES
      ARE SCATTERED
---------------------------------------------------------- Letter :4.1

VA hospital and outpatient clinics are often located hundreds of
miles from each other, as shown in figure 1. 

   Figure 1:  Locations of VA
   Medical Centers and
   Freestanding Outpatient Clinics

   (See figure in printed
   edition.)



   (See figure in printed
   edition.)

Some areas of the United States have more extensive VA medical
coverage than others.  Of VA's 158 medical centers,\10 124, or 78
percent, are located in urban areas and 33 are in rural areas.\11

Similarly, of VA's 187 freestanding outpatient clinics,\12 123, or
almost 66 percent, are located in urban areas and 58 are in rural
areas.\13 VA has no medical facilities in 147, or about 46 percent,
of the country's 323 urban areas. 

Private-sector medical facilities also significantly outnumber VA
facilities.  For example, in contrast to VA's 173 hospitals, there
are over 6,000 public and private hospitals nationwide. 

Moreover, private-sector health care programs generally target
specific geographic markets and place a priority on developing
extensive networks of providers.  They typically strive to provide
primary care within a few miles or minutes of their enrollees in
urban markets and within a reasonable distance or time, albeit
somewhat longer and farther, in rural areas.\14

Further, while they realize that enrollees are willing to travel
longer distances for specialty care, they also realize that such
services must be "convenient" to attract enrollees and remain
competitive with other private programs. 


--------------------
\10 VA medical centers typically include one or more hospitals and a
hospital-based outpatient clinic. 

\11 One VA medical center is located in Puerto Rico. 

\12 Freestanding outpatient clinics are facilities that are
physically separate from a VA medical center; that is, they are not
located on the medical center property. 

\13 VA also operates six freestanding outpatient clinics outside the
United States:  two in Puerto Rico, two in the Virgin Islands, one in
Guam, and one in the Philippine Islands. 

\14 Primary care is a point of entry into the health care system for
nonemergency care. 


      COMPARISON OF VETERANS'
      ACCESS TO VA AND PRIVATE
      FACILITIES IN SELECTED
      MARKETS
---------------------------------------------------------- Letter :4.2

Veterans generally have less access to VA medical facilities than to
private facilities, as the following market examples show. 
Nevertheless, VA officials said that veterans often have better
access through VA facilities than through private facilities to such
specialty services as mental health care, rehabilitation for the
blind, and post-traumatic stress disorder care. 


         AN URBAN AREA WITH A VA
         MEDICAL CENTER
-------------------------------------------------------- Letter :4.2.1

Veterans living in urban areas with one or more VA medical centers,
such as Chicago, Illinois, may not have to travel long distances for
care in VA facilities but may have to endure long travel times. 
Moreover, private facilities are generally more accessible. 

VA operates four medical centers in the Chicago primary metropolitan
statistical area, which encompasses nine counties with almost 700,000
veterans.\15 Together, the four medical centers provide services
ranging from primary care to highly specialized services, such as
psychiatric inpatient care and treatment for spinal cord injuries. 
The centers provide 2,600 acute-, psychiatric-, and extended-care
beds.  In 1993, almost 17,000 veterans in the Chicago metropolitan
area received VA inpatient services, and almost 70,000 received VA
outpatient services.  While over 77 percent of the veterans in the
Chicago metropolitan area live within 20 miles of a medical center,
traffic congestion makes these centers difficult and time-consuming
to reach.  When public transportation is used, travel times may
exceed 1-1/2 hours. 

Private facilities are often more convenient.  Over 100 hospitals and
thousands of primary care and specialist physicians are located in
the Chicago primary metropolitan statistical area.  In addition,
insurance companies and managed care programs have formed networks of
hospitals and physicians to offer enrollees a wide range of services
located near their homes.  Patients enrolled in one of Aetna's
managed care plans, for example, have a choice of 45 hospitals and
more than 450 sites where they can obtain primary care within the
Chicago metropolitan area.  (See fig.  2.) An executive from another
major managed care plan serving the same area told us that the plan's
goal is to provide two primary care physicians within 5 miles of
every enrollee. 

   Figure 2:  Locations of Aetna
   Hospitals and VA Facilities in
   the Chicago Nine-County Primary
   Metropolitan Statistical Area

   (See figure in printed
   edition.)


--------------------
\15 The Office of Management and Budget designates areas of the
country with a large population nucleus and surrounding communities
that have a high degree of economic and social integration as
metropolitan statistical areas. 


         AN URBAN AREA WITH A
         FREESTANDING VA
         OUTPATIENT CLINIC
-------------------------------------------------------- Letter :4.2.2

Veterans living in urban areas with only a freestanding VA outpatient
clinic, such as Rockford, Illinois, often travel long distances to
other VA facilities to obtain needed inpatient and specialty care. 
Such care is often available from nearby private providers. 

VA operates one outpatient clinic in the Rockford metropolitan
statistical area, which encompasses three counties with about 40,000
veterans.  The VA clinic provides a broad range of primary care and
some specialty services.  In 1993, about 400 veterans in the Rockford
metropolitan area received VA inpatient services, and about 2,200
received VA outpatient services.  While almost 74 percent of these
veterans lived within 10 miles of the clinic, they had to travel
about 70 miles to obtain inpatient care or outpatient specialty
services not available at the Rockford clinic.  The nearest VA
medical centers are in Madison, Wisconsin, and North Chicago,
Illinois. 

According to local health care officials, private-sector care is
generally available within minutes of most of the Rockford
population.  Rockford has three general medical and surgical
hospitals that, together, have almost 1,000 hospital beds.  Each has
developed a network of providers.  For example, the SwedishAmerican
Health Alliance System includes 3 hospitals and 22 primary care sites
in the Rockford metropolitan area and operates or is affiliated with
many retail pharmacies, acute-care centers, and home-care and
long-term care facilities in the Rockford area.  (See fig.  3.)

   Figure 3:  Locations of
   SwedishAmerican Health Alliance
   Hospitals and Primary Care
   Sites and VA Facilities in the
   Three-County Rockford,
   Illinois, Metropolitan
   Statistical Area

   (See figure in printed
   edition.)


         AN URBAN AREA WITHOUT A
         VA FACILITY
-------------------------------------------------------- Letter :4.2.3

Veterans living in urban areas with no VA medical facilities, such as
Salem, Oregon, often travel long distances to obtain VA health care. 
Private providers are often nearby. 

VA operates no hospital or outpatient clinic in the Salem primary
metropolitan statistical area, which encompasses two counties with
about 46,000 veterans.  The Salem metropolitan area, situated in VA's
Portland, Oregon, service area, covers 27 counties in northwest
Oregon and southwest Washington.  In 1993, about 600 veterans from
the Salem metropolitan area received VA inpatient services, and about
2,700 received VA outpatient services.  These veterans had to travel
from 20 to over 70 miles to get to the Portland facility. 

Many private medical facilities are located in the Salem metropolitan
area, typically, according to officials of these facilities, within
20 minutes of Salem residents.  One of Oregon's largest hospitals--a
454-bed facility that serves as a referral point for smaller
hospitals in surrounding communities--is located within the Salem
city limits, and there are three smaller hospitals in the
metropolitan area as well.  Physicians are widely distributed and
within easy reach of most residents.  In addition, the area has
several private-sector managed care plans that have developed
networks of providers to make care convenient for their enrollees. 
ODS Health Plans, for example, offers its enrollees a choice of 4
hospitals and 56 primary care sites in Salem.  (See fig.  4.)

   Figure 4:  Locations of ODS
   Health Plan Hospitals and
   Primary Care Sites in the
   Salem, Oregon, Primary
   Metropolitan Statistical Area
   and the Nearest VA Facility

   (See figure in printed
   edition.)


         A RURAL AREA WITH A VA
         MEDICAL CENTER
-------------------------------------------------------- Letter :4.2.4

Veterans living in rural areas served by a single VA medical center,
such as central Georgia, often travel long distances to obtain care
in VA facilities because such facilities often serve large geographic
areas.  Private facilities scattered throughout the rural areas may
be more convenient. 

VA operates a medical center in Dublin, Georgia, a rural community
about 135 miles southeast of Atlanta.  The medical center is the only
VA medical facility in a 52-county area with about 130,000 veterans. 
The facility provides primary and secondary medical, surgical, and
psychiatric care, as well as extended-care services.  In 1993, about
3,700 veterans in the Dublin service area received VA inpatient
services, and about 14,000 received VA outpatient services.  The
veterans came from as far away as 70 miles to the north, 150 miles to
the east, 100 miles to the south, and 140 miles to the west. 
Moreover, veterans needing highly specialized care are referred to
the VA medical center in Augusta, Georgia, about a 2-1/2-hour drive
from Dublin. 

Private medical facilities are also scattered throughout the Dublin
VA medical center service area and are more accessible to veterans,
with the exception of veterans living in Dublin, than is VA's medical
center.  In 1993, there were more than 40 medical and surgical
hospitals and almost 1,400 physicians in the area.  In addition, Blue
Cross/Blue Shield of Georgia, which operates a preferred provider
network throughout the state, offers its enrollees a choice of 10
hospitals and 79 primary care sites in the 52-county area.  (See fig. 
5.)

   Figure 5:  Locations of Blue
   Cross/Blue Shield Hospitals and
   Primary Care Sites and the VA
   Facility in the 52-County
   Dublin, Georgia, VA Medical
   Center Service Area

   (See figure in printed
   edition.)


   OPTIONS VA MIGHT EXPLORE FOR
   IMPROVING ACCESSIBILITY OF VA
   HEALTH CARE
------------------------------------------------------------ Letter :5

In February 1995, VA issued an interim policy that encouraged its
field offices to employ all means at their disposal, consistent with
funding availability and federal law, to improve veterans' access to
VA health care.  VA also authorized its offices to establish
VA-operated clinics as well as VA-funded or VA-reimbursed private
clinics, group practices, or individual practitioners.  The directors
of VA's newly created 22 networks are exploring ways to better
integrate service delivery. 

In developing plans for improving veterans' access to VA health care,
the directors face two basic decisions:  where to locate new
facilities and how to operate them.  Each decision involves the
assessment of a variety of key factors. 


      TARGET POPULATION IS KEY
      FACTOR IN DECIDING WHERE TO
      LOCATE NEW FACILITIES
---------------------------------------------------------- Letter :5.1

A key factor that could influence VA's decisions about where to
locate new medical facilities is the population to be targeted.  VA
has several options:  (1) improve convenience for current users, (2)
attract new users by improving access for all veterans, or (3)
improve access for specific groups of veterans, such as those in
selected eligibility categories or those residing in medically
underserved areas. 


         IMPROVING ACCESS FOR
         CURRENT USERS
-------------------------------------------------------- Letter :5.1.1

In improving convenience for existing users by making VA medical
facilities more accessible, VA could decide, for example, to help
veterans traveling great distances or times to receive care or to
help the largest number of veterans, without regard to travel
distances or times. 

To improve convenience for current users with long travel distances
or times, VA would have to establish criteria for determining
reasonable travel distances or times and then assess the number of
current users residing in areas beyond those distances or times. 
These criteria could vary depending on whether the area targeted is
urban or rural.  Facilities could be located so they could provide
the maximum number of existing users with improved access, as
illustrated in the following examples. 

  In a densely populated urban area like Chicago, Illinois,
     one-quarter of a medical center's current users might live 10 to
     20 miles from the facility.  However, traffic congestion and
     transportation barriers might cause these veterans to travel an
     hour or more one way to obtain care, especially if they needed
     to use public transportation.  A medical center could decide to
     make outpatient care available within 30 minutes' average travel
     time.  That is, the center could establish facilities nearer to
     large numbers of current users who now have to travel
     significantly longer than 30 minutes. 

  In a largely rural area, such as the one served by the VA medical
     center in Dublin, Georgia, many veterans must travel extensive
     distances to obtain VA care.  In fiscal year 1994, over 23
     percent of the outpatient visits at Dublin were by veterans
     living in or around Macon and Albany, Georgia, which are located
     approximately 50 and 100 miles away, or about 1 to 2 hours,
     respectively, from the medical center.  A rural VA medical
     center like Dublin could significantly reduce the time veterans
     needed to travel for outpatient care by adding access points in
     key population centers, such as Macon or Albany. 

To establish delivery sites that improve access for the largest
number of current users, VA could identify large concentrations of
veterans currently served.  These concentrations would be likely to
be within close proximity of existing medical centers.  For example,
a medical center in an urban area could elect to establish a nearby
clinic because over one-third of its current users resided in that
area.  Providing a new facility in a nearby location could help
alleviate overcrowding at the medical center. 


         ATTRACTING NEW USERS
-------------------------------------------------------- Letter :5.1.2

Medical facilities could also be located to attract new users.  VA
could identify areas that have a large number of veterans who do not
now receive VA care.  The target population could be based on the
total number of veterans in an area, as shown in the following
examples. 

  Officials from the Dublin, Georgia, VA medical center told us their
     goal was to improve the accessibility of VA health care services
     for new and current users.  Consequently, Dublin targeted
     population centers in its service area, such as Macon and
     Albany, as possible locations for new outpatient clinics. 

  Portland, Oregon, VA medical center officials told us their plans
     were designed to retain existing users and attract new veterans. 
     As a result, Portland officials planned on adding access points
     in major urban areas of their service area, such as Salem, from
     which they would be likely to draw the most veterans. 

VA also could target a population on the basis of a predetermined
percentage of nonusers, regardless of how many veterans resided in
the area.  For example, a VA medical center could choose to establish
a delivery site in a remote area where less than a prescribed
percentage of veterans, such as 5 percent, obtained care at the
medical center.  By providing an access point closer to where the
veterans lived, the medical center could encourage more veterans to
use its health care services. 


         IMPROVING ACCESS FOR
         SELECTED ELIGIBILITY OR
         OTHER GROUPS
-------------------------------------------------------- Letter :5.1.3

VA's assessment of areas in which to locate new medical facilities
could also consider differences in veterans' eligibility status and
other factors.  In general, veterans with service-connected
disabilities, low-income veterans, and certain other "mandatory care"
veterans, such as World War I veterans and veterans exposed to toxic
substances, have the highest priority for receiving VA health care. 
High-income veterans without service-connected disabilities have the
lowest priority. 

VA could improve access for one or more of the high-priority groups. 
For example, a medical center could target veterans with
service-connected disabilities.  The center could establish medical
facilities in areas with high concentrations of such veterans,
providing more convenient access.  Moreover, the size of the facility
could be based on the estimated number of such veterans residing in
the targeted area. 

VA medical centers also could choose to provide better health care
access for veterans living in areas where community providers are
unavailable.  For example, although many veterans have public or
private health insurance, those that live in underserved areas may
not be able to obtain care because of shortages of private providers. 
To improve access for such veterans, a VA medical center could
identify veterans living in areas designated Health Professional
Shortage Areas\16 and establish a medical facility in the area.\17


--------------------
\16 Health Professional Shortage Areas are geographic areas and
population groups recognized by the federal government as having an
acute shortage of health care personnel. 

\17 In VA Health Care:  Retargeting Needed to Better Meet Veterans'
Changing Needs (GAO/HEHS-95-39, Apr.  21, 1995), we discussed how
well the VA health care system is meeting the health care needs of
veterans and options available for reconfiguring the system to better
meet veterans' needs. 


      RESOURCES, MARKET
      CONDITIONS, AND PRIVATE
      PROVIDER WILLINGNESS TO
      CONTRACT WITH VA ARE KEY
      FACTORS IN DECIDING HOW TO
      DELIVER CARE
---------------------------------------------------------- Letter :5.2

VA has identified two alternative approaches for expanding access. 
First, it could establish VA-operated facilities, using either
VA-owned or VA-leased space.  Second, it could contract with non-VA
providers to provide care.\18 VA's decisions in this regard are
likely to be primarily influenced by three factors--resources, market
conditions, and the willingness of private providers to contract with
VA. 


--------------------
\18 Currently, VA has authority to contract for health care services
under prescribed conditions, for example, when specialized medical
resources or routine treatment of certain veteran groups, such as
those with service-connected conditions, is required.  VA recently
developed a legislative proposal to expand its contracting authority. 


         RESOURCES
-------------------------------------------------------- Letter :5.2.1

The availability of resources is a key factor affecting VA's
decisions whether to operate new VA medical facilities or contract
for care.  In general, VA could assess the cost implications on a
short- and long-term basis, taking into account start-up and
operating costs.  VA could also consider the potential for fraud,
waste, and abuse. 

VA-operated facilities typically require a substantial capital
investment, which increases VA's costs over the short term.  For
example, VA would have to build, purchase, or lease a facility.  It
would also have to staff and equip the facility. 

VA-operated medical facilities could be more costly to operate than
contracting for care.  Although performing valid cost comparisons is
complicated, one VA medical center concluded that certain types of
health care could be more expensive when delivered in a VA medical
center.  The medical center contracted with a community provider to
furnish certain health care services to veterans at a capitation rate
of $178.  A medical center official estimated that the cost to
provide the same services in the VA facility would be substantially
more. 

Similarly, in their Fiscal Year 1995 Independent Budget for the
Department of Veterans Affairs,\19 veterans' service organizations
estimated that adding 132 VA-operated leased clinics to VA's direct
delivery system over a 4-year period could cost $346 million--$137
million more than the $209 million the veterans' service
organizations estimated it would cost to provide comparable services
through contracts with the private sector. 

VA-operated medical facilities would also be likely to result in less
accessibility for veterans than could be achieved through
contracting, because of high start-up and operating costs.  VA could
operate its own facility in an area or, for the same dollar
expenditures, could contract for care at multiple locations,
significantly enhancing the geographic accessibility of services to
veterans living in the area.  Managed care plans in the areas we
visited typically contracted for care with geographically dispersed
networks of physicians rather than operating their own facilities at
only a few locations.  In Salem, for example, most managed care plans
contracted with an independent practice association that represented
virtually all physicians practicing in the area.  Thus, the plans
were able to offer their enrollees numerous locations where they
could obtain care. 

On the positive side, VA-operated facilities could give VA more
control over resources, potentially lessening the risk of fraud,
waste, and abuse.  VA has had problems in administering contracts and
sharing agreements.  For example, in a 1987 audit of scarce medical
specialist contracts, VA's Inspector General reported that VA medical
centers had paid for services they had not received and had not
established controls to ensure that contractor performance and
billing complied with contract terms.\20 Our July 1992 follow-up to
the Inspector General's report found that VA lacked assurance that
these problems were identified and corrected.\21

In addition, contracting has sometimes led to less VA control over
the utilization of health care services to veterans because VA has
had difficulty controlling practice patterns of private physicians. 
Chicago VA officials told us this was a major factor in their
decision to replace a contract clinic that was reimbursed on a
fee-for-service basis with a VA-operated clinic in Rockford,
Illinois, in 1994.  According to VA officials, reviews of the
contract clinic's medical records indicated that many veterans
continued to receive treatment at the clinic after their conditions
had stabilized.  Moreover, attempts to convince contractor physicians
to discharge stable patients were unsuccessful, they said, because VA
lacked direct administrative control over the physicians.  While this
problem could be avoided if VA contracted on a capitation basis, VA
officials said that capitation contracts could lead to other
problems, such as diminished services. 


--------------------
\19 Annually, four congressionally chartered veterans' service
organizations--American Veterans of World War II, Korea, and Vietnam;
Disabled American Veterans; Paralyzed Veterans of America; and the
Veterans of Foreign Wars of the United States--prepare and submit to
the Congress an independent budget proposal for VA. 

\20 Office of the Inspector General, Department of Veterans Affairs,
Audits of Selected Aspects of VA's Program for Sharing Scarce Medical
Resources, Report No:  7AM-A99-089 (Washington, D.C.:  Department of
Veterans Affairs, July 15, 1987). 

\21 VA Health Care:  Inadequate Controls Over Scarce Medical
Specialist Contracts (GAO/HRD-92-114, July 29, 1992). 


         MARKET CONDITIONS
-------------------------------------------------------- Letter :5.2.2

The stability of market conditions and the competitive environment in
which VA operates could affect its health care delivery decisions. 
In general, VA would need to ensure that its decisions included
sufficient flexibility, since market conditions vary over time. 
Also, VA would need to ensure that its decisions considered the
availability of existing providers in a particular market. 

Operating its own facilities could limit VA's ability to expand or
relocate operations promptly in response to changing market
conditions.  Expanding or relocating services could be more time
consuming and difficult than modifying or renegotiating a contractual
arrangement with a private provider because clinics would need to be
constructed or leased, equipped, and staffed. 

Moreover, VA may be at a disadvantage in recruiting primary care
physicians to operate its facilities.  Only about 20 percent of VA's
physicians are primary care physicians, while about 60 percent of
managed care plans' physicians are typically primary care physicians. 
Private-sector officials we spoke with believed that VA would have
difficulty increasing the number of primary care physicians to staff
its facilities because of the shortage of such physicians and because
VA currently is unable to match the salaries and benefits offered by
managed care plans. 

Further, VA-operated facilities might duplicate private-sector
services in areas where the private sector has sufficient capacity,
requiring VA to compete for existing patients.  Several
private-sector officials we spoke with questioned the need to open
new VA outpatient facilities, especially in urban areas where the
private sector already provides sufficient access.  In these cases,
contracting with existing providers might be a more viable option for
improving access. 

Conversely, VA-operated facilities might be a more viable option in
areas that are underserved by the private sector.  Some rural and
inner-city areas, for example, suffer from a shortage of physicians. 
In other areas, physicians may have full practices and, as a result,
be unlikely to take on additional patients.  Thus, if VA wanted to
improve access for veterans in such areas it might be able to do so
only by establishing its own facility. 


         WILLINGNESS OF PRIVATE
         PROVIDERS TO CONTRACT
         WITH VA
-------------------------------------------------------- Letter :5.2.3

Finally, even if VA chose to do so, it could have difficulty
contracting for veterans' health care.  Several private-sector
officials indicated that providers would be cautious about entering
into capitation arrangements with VA because the risk of financial
loss would be too high.  They said that the VA patient population is
perceived as being sicker and, therefore, more expensive to care for
than the general population.  At the same time, they questioned the
willingness of VA to provide an adequate level of reimbursement,
particularly under a capitation arrangement. 

A medical center in Chicago recently encountered such an obstacle. 
The VA medical center proposed establishing a VA-operated outpatient
clinic in its service area after failing to interest three local
health maintenance organizations (HMO) in providing services on a
capitation basis.  The HMOs were not interested because they believed
the medical histories of the medical center's patients represented
too high an underwriting risk.  The HMOs, according to medical center
officials, were unwilling to develop a rate for patients outside
their normal risk definition. 

In addition, officials at two managed care plans expressed concern
about contracting with VA because doing so could adversely affect
veteran patients' care.  Non-VA physicians who lacked admitting
rights in VA hospitals would be required to relinquish control of
patients needing inpatient care to VA physicians.  Officials at one
managed care plan said that their physicians would not be likely to
be receptive to such an arrangement because they would lose the
ability to ensure continuity of care for their patients.  Similarly,
officials from another plan said that contracting only for primary
care would result in veterans' seeing physicians from two different
systems--their plan and VA.  This, they said, would make tracking
their patients difficult. 

Finally, some private-sector officials believed that many providers
would be reluctant to contract with VA because of their unwillingness
to deal with excessive government contracting requirements and
regulations. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

Veterans' access to VA health care could improve significantly if
medical centers employed all means at their disposal to expand
access, as VA's February 1995 interim policy encourages centers to
do.  While medical centers have numerous options in locating new
facilities, selecting a target population, such as current or new
users, poses a difficult policy choice.  Also, medical centers'
decisions on how to operate new facilities--directly or by
contracting--require the evaluation of several key factors on a
facility-specific basis to ensure care is delivered in the most
appropriate manner.  Overall, medical centers will be likely to use a
variety of options tailored specifically to the variabilities of
local conditions. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :7

We obtained comments on a draft of this report from VA officials,
including the Deputy Secretary for Health.  The officials agreed with
the basic concepts in our report.  They cautioned, however, that
evaluating veterans' access to health care and improving access may
be more complex, in reality, than the report appears to suggest.  For
example, they said that the availability of a large number of
private-sector medical facilities does not ensure that veterans can
receive needed care from such facilities, noting that veterans often
need such special services as psychiatric care and rehabilitation for
blindness.  We revised the report to show that we focused on
veterans' access to standard health care benefits.  VA officials also
suggested some technical changes, primarily for clarification.  We
incorporated the suggestions as appropriate. 


---------------------------------------------------------- Letter :7.1

Copies of this letter are being sent to the Chairmen and Ranking
Minority Members of the House and Senate Committees on Veterans'
Affairs and the Secretary of Veterans Affairs.  Copies will be made
available to others upon request. 

Please call me at (202) 512-7101 if you have any questions or need
additional assistance.  Other GAO contacts and contributors to this
report are listed in appendix II. 

Sincerely yours,

David P.  Baine
Director, Federal Health Care
 Delivery and Quality Issues


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

To obtain information on users of VA medical facilities, we reviewed
VA's studies of the status of veterans in 1987 and 1992\22 and
previous GAO VA health care reports.\23

To compare the accessibility of VA and private-sector facilities, we
categorized VA's existing facilities into four general markets: 
urban and rural areas with medical centers and urban and rural areas
with freestanding outpatient clinics.  We also identified urban areas
where VA operates no medical facilities.  We visited the following
locations: 

  Chicago, Illinois--an urban area where VA operates a medical
     center,

  Rockford, Illinois--an urban area where VA operates a freestanding
     outpatient clinic,

  Salem, Oregon--an urban area where VA has no medical facilities,
     and

  Dublin, Georgia--a rural area where VA operates a medical center. 

For purposes of this report, we considered metropolitan statistical
areas and primary metropolitan statistical areas as urban areas.  The
Office of Management and Budget defines a metropolitan statistical
area as either (1) an area that includes at least one city with a
population of at least 50,000 or (2) a Census Bureau-defined
urbanized area of at least 50,000 inhabitants and a total
metropolitan population of 100,000 (75,000 in New England). 
Metropolitan statistical areas with populations of 1 million or more
are called consolidated metropolitan statistical areas if separate
component areas, or primary metropolitan statistical areas, can be
identified in the areas.\24 We used the primary metropolitan
statistical area rather than the consolidated metropolitan
statistical area to determine whether VA had medical facilities in
urban areas.  For example, both Portland and Salem, Oregon, are part
of the same consolidated metropolitan statistical area but are
separate primary metropolitan statistical areas.  VA has a medical
center located in Portland, and we classified the Portland primary
metropolitan statistical area as an urban area with a VA medical
center.  The Salem primary metropolitan statistical area has no VA
facility, and we classified it as an urban area with no VA medical
facilities. 

The four study locations were selected as a nonrandom judgmental
sample representing VA's typical market categories and a diversity of
geographic areas.  In Chicago and Dublin, we met with local VA
officials to discuss the accessibility of VA's health care facilities
and options for improving access.  For Rockford, we met with local
and North Chicago VA Medical Center (the parent facility) officials. 
For Salem, we met with officials from the VA medical center in
Portland, Oregon, the closest VA medical facility. 

In each location, we also met with private-sector hospital and
managed care plan officials to obtain information about the local
health care market and the accessibility of their health care
services.  In each market area, we used the provider directory of one
managed care plan to determine the number of primary care sites and
hospitals available to the plan's enrollees.  A provider directory is
a listing of the hospitals, physicians, and other health care
providers affiliated with the managed care plan.  We graphically
depicted these facilities and VA's facilities on maps of the four
locations.  The list of organizations contacted follows. 


--------------------
\22 1987 Survey of Veterans (July 1989) and National Survey of
Veterans (Apr.  1995). 

\23 We made extensive use of two reports:  GAO/HEHS-94-113FS, Mar. 
29, 1994, and GAO/HEHS-96-31, Dec.  20, 1995. 

\24 As of June 1993, there were 250 metropolitan statistical areas
and 18 consolidated metropolitan statistical areas consisting of 73
primary metropolitan statistical areas in the United States. 


   CHICAGO, ILLINOIS
--------------------------------------------------------- Appendix I:1


      VA
------------------------------------------------------- Appendix I:1.1

  Lakeside VA Medical Center

  North Chicago VA Medical Center

  Crown Point Outpatient Clinic


      PRIVATE SECTOR
------------------------------------------------------- Appendix I:1.2

  Blue Cross/Blue Shield of Illinois

  Chicago HMO

  Aetna Health Plans

  Northwestern HealthCare Network

  HealthNetwork

  Highland Park Hospital


   ROCKFORD, ILLINOIS
--------------------------------------------------------- Appendix I:2


      VA
------------------------------------------------------- Appendix I:2.1

  Rockford Outpatient Clinic


      PRIVATE SECTOR
------------------------------------------------------- Appendix I:2.2

  SwedishAmerican Health Alliance System

  Rockford Health System


   SALEM, OREGON
--------------------------------------------------------- Appendix I:3


      VA
------------------------------------------------------- Appendix I:3.1

  Portland VA Medical Center


      PRIVATE SECTOR
------------------------------------------------------- Appendix I:3.2

  HMO Oregon (Blue Cross/Blue Shield of Oregon)

  Kaiser Permanente

  ODS Health Plans

  PacifiCare Health Systems

  SelectCare Health Plans

  Salem Hospital


   DUBLIN, GEORGIA
--------------------------------------------------------- Appendix I:4


      VA
------------------------------------------------------- Appendix I:4.1

  Carl Vinson VA Medical Center


      PRIVATE SECTOR
------------------------------------------------------- Appendix I:4.2

  Fairview Park Hospital

  Blue Cross/Blue Shield of Georgia (Macon, Georgia)

To determine how far veterans in our study areas lived from a VA
medical facility, we computed the distance from zip codes where the
veterans lived to the closest VA medical facility in the area (that
is, the Chicago primary metropolitan statistical area; the Rockford
metropolitan statistical area; the Salem primary metropolitan
statistical area; and Dublin service areas, as defined by VA).  If VA
had no facilities in the area, we computed the distance to the
nearest VA facility.  For example, in the Chicago, Illinois, primary
metropolitan statistical area, we computed the distance from the zip
codes where veterans lived to the closest of VA's four medical
facilities in the area; in the Salem, Oregon, area, which had no VA
facility, we computed the distance from where the veterans lived to
the VA facility in Portland, Oregon, the VA facility closest to
Salem. 

We also used these zip codes to determine the number of veterans
living and receiving services in the study areas.  We applied 1990
U.S.  census data to these zip codes to determine the number of
veterans living in each study area.  We applied data from VA's fiscal
year 1993 Outpatient and Patient Treatment databases\25 to these zip
codes to determine the number of veterans who received VA outpatient
and inpatient services in each study area. 

To identify the factors VA considers in deciding where to locate new
medical facilities and whether to add new VA-operated facilities or
contract with private providers, we reviewed VA policy directives and
other guidance and interviewed VA central office and selected medical
center officials.  We also reviewed prior GAO reports and other
studies.  In addition, we met with VA Office of General Counsel
officials to discuss VA's contracting authority.  We also discussed
contracting and expanding VA's direct delivery system with officials
from the private-sector organizations we visited. 


--------------------
\25 VA's Outpatient File lists all veterans who received VA
outpatient services each year.  VA's Patient Treatment File lists all
veterans discharged from a VA Medical Center each year.  Both
databases track veterans' zip codes. 


GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
========================================================== Appendix II

GAO CONTACTS

Paul R.  Reynolds, Assistant Director, (202) 512-7109
Robert T.  Ferschl, Evaluator-in-Charge, (312) 220-7632
Byron S.  Galloway, (202) 512-7247

STAFF ACKNOWLEDGMENTS

In addition to those named above, the following individuals made
important contributions to this report:  Abigail Ohl provided
assistance with data gathering and analysis, and Ann McDermott,
Angela Pun, and Joan Vogel prepared the graphics. 


*** End of document. ***