Bone Marrow Transplants: Despite Recruitment Successes, National 
Program May Be Underutilized (18-OCT-02, GAO-03-182).		 
                                                                 
More than 30,000 people are diagnosed annually with leukemia or  
other blood, metabolic, or immune system disorders, many of whom 
may die without stem cell transplants, using stem cells from bone
marrow or another source. When a patient needs a transplant of	 
donated stem cells and no genetically compatible related donor is
available, the National Bone Marrow Donor Registry may help the  
patient search for compatible stem cells from unrelated donors.  
The National Bone Marrow Registry Reauthorization Act of 1998	 
required, among other things, that the Registry carry out a donor
recruitment program giving priority to minority and		 
underrepresented donor populations, ensure efficiency of	 
operations, and verify compliance with standards by organizations
that participate in the Registry. From 1998, when the National	 
Bone Marrow Registry Reauthorization Act was enacted, through	 
2001, the number of stem cell donors on the Registry increased	 
for all racial and ethnic groups. Although the exact number of	 
patients in need of transplants is not known, estimates suggest  
that about one-third of them use the Registry to search for	 
donors. The organizations that are involved in transplantation	 
and participate in the National Marrow Donor Program (NMDP)	 
network generally adhere to NMDP's standards and procedures. In  
2001, NMDP required 24 centers to take corrective actions because
they did not meet its standards.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-03-182 					        
    ACCNO:   A05282						        
  TITLE:     Bone Marrow Transplants: Despite Recruitment Successes,  
National Program May Be Underutilized				 
     DATE:   10/18/2002 
  SUBJECT:   Diseases						 
	     Minorities 					 
	     Performance measures				 
	     Health care programs				 
	     Health care services				 
	     Surgery						 
	     National Marrow Donor Program			 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.                                         **
**                                                              **
******************************************************************
GAO-03-182

Report to Congressional Committees

United States General Accounting Office

GAO

October 2002 BONE MARROW TRANSPLANTS

Despite Recruitment Successes, National Program May Be Underutilized

GAO- 03- 182

Page i GAO- 03- 182 Bone Marrow Transplants Letter 1 Results in Brief 3
Background 5 NMDP Has Succeeded in Increasing Recruitment of Donors,
Including Minorities, but May Not Be Able to Equalize Access to

Matches 11 National Registry May Be Underutilized 16 Organizations in the
NMDP Network Generally Comply with Its Standards and Procedures 21
Concluding Observations 25 Agency Comments 25 Appendix I Methods of
Assessing Registry Utilization 28 Method Based on Number of HLA- Identical
Sibling Transplants 28 Method Based on Number of Preliminary Searches 31
Method Based on Incidence of Disease 32 Appendix II How NMDP Achieves
Network Compliance

with Selected Standards and Procedures 34 Informed Consent of Donors and
Patients 34 Criteria for Donor Selection 35 Methods to Protect
Confidentiality 35 Marrow Collection and Transport 36 Laboratory Standards
36 Donor File Maintenance and Updates 37 Appendix III Comments from the
Health Resources and Services Administration 38 Appendix IV GAO Contact
and Staff Acknowledgments 42 Tables

Table 1: Percentage Increase in Registry Donors, by Racial and Ethnic
Group, 1998 to 2000, and Current Proportion of Groups on the Registry and
in the Population 12 Contents

Page ii GAO- 03- 182 Bone Marrow Transplants

Table 2: U. S. Patients* Utilization of the Registry, by Race and
Ethnicity, 1997 to 2000 18 Table 3: Reasons for Preliminary Search
Cancellation, January 2000 through September 2001 19 Table 4: Reasons for
Formal Search Cancellation, 1997 to 2000 20 Table 5: Alternate Approach to
the Analysis of U. S. Patients* Utilization of the Registry, by Race and
Ethnicity, 1997 to 2000 31 Table 6: Average Annual Unrelated Donor NMDP-
Facilitated Transplants and Estimated Number of Potential Recipients

for U. S. Patients with Selected Diseases Who Might Benefit from Unrelated
Stem Cell Transplants, by Race/ Ethnicity, 1997 through 2000 33 Figure

Figure 1: Theoretical Probability of Patient*s Finding at Least One
Matched Donor, by Racial and Ethnic Group, 1988 to 2001 14 Abbreviations

ABMTR Autologous Blood and Marrow Transplant Registry CPI Continuous
Process Improvement HHS Department of Health and Human Services HLA human
leukocyte antigens HRSA Health Resources and Services Administration IBMTR
International Bone Marrow Transplant Registry NMDP National Marrow Donor
Program OIG Office of Inspector General OPA Office of Patient Advocacy
PBSC peripheral blood stem cells

Page 1 GAO- 03- 182 Bone Marrow Transplants

October 18, 2002 Congressional Committees More than 30,000 people are
diagnosed annually with leukemia or other blood, metabolic, or immune
system disorders, many of whom may die without stem cell transplants,
using stem cells from bone marrow or another source. 1 When a patient
needs a transplant of donated stem cells and no genetically compatible
related (family) donor is available, the National Bone Marrow Donor
Registry (Registry) may help the patient search for compatible stem cells
from unrelated donors. Founded in 1986,

the Registry is the largest and most diverse list of potential donors in
the world. This list currently includes more than 4 million donors. 2 The
Registry is operated by the nonprofit National Marrow Donor Program (NMDP)
under contract to the Department of Health and Human Services* (HHS)
Health Resources and Services Administration (HRSA), with additional
support from the U. S. Navy. 3 NMDP coordinates stem cell transplants
through its network of more than 400 participating organizations, domestic
and foreign, involved in transplantation, including donor centers, which
recruit and manage donors; laboratories; blood

sample repositories; bone marrow collection centers; and transplant
centers. NMDP has facilitated more than 14,000 transplants since 1987.

Concerns about the Registry have been raised by the HHS Office of the
Inspector General (OIG) and in our own work. These include the extent to
which the Registry provides equality of opportunity for patients of all

racial and ethnic groups to find compatible (matched) unrelated donors,
the extent to which it is utilized by those in need of stem cell
transplantation, and the effectiveness of the management of the donor

1 The first source of stem cells for transplant was bone marrow, but now
stem cells from the bloodstream or from umbilical cord blood can also be
used. We use the term stem cell transplant to include both bone marrow
transplants and transplants involving one of these newer sources of stem
cells.

2 We use the term donors throughout this report to refer to potential
donors on the Registry, most of whom have not donated stem cells, only
expressed their willingness to do so. 3 The Navy was instrumental in the
founding of the Registry and has maintained its interest in stem cell
donation over the years. HRSA and the Navy each contributed a little less
than 20 percent of the Registry*s fiscal year 2002 funding (about $21
million and $20.5 million, respectively), with program revenue and private
sources providing the rest of the total of about $108 million.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 03- 182 Bone Marrow Transplants

centers. We reported in 1992 that the proportions of African American and
Hispanic donors on the Registry were less than their proportions in the U.
S. population. 4 This imbalance results in a decreased likelihood of an
individual from a minority group finding a match and eventually receiving
a transplant because matches are more likely to be found from among donors
of one*s own group. In an effort to address these concerns, a 1996 OIG
report recommended that HRSA and NMDP reexamine the method used to finance
the donor centers that recruit volunteers to join the Registry. 5 It
recommended a performance- based method to pay donor centers for specific
activities including monetary incentives tied to performance indicators
and emphasizing recruitment and retention of donors, especially those from
racial and ethnic minority groups.

The National Bone Marrow Registry Reauthorization Act of 1998 6 required,
among other things, that the Registry carry out a donor recruitment
program giving priority to minority and underrepresented donor
populations, ensure efficiency of operations, and verify compliance with
standards by organizations that participate in the Registry. In addition,
the act required that we conduct a study of the Registry, including an
examination of the extent to which it has increased representation of
racial and ethnic minority groups so that a member of such a group has a
probability of finding a match comparable to that of a person who is not a
member of such a group. In conducting this study, we addressed the
following questions: (1) To what extent have the program*s recruitment
efforts increased the enrollment of donors, including those from racial
and ethnic minority groups, since the 1998 act took effect, and has the
chance of finding a suitable match increased? (2) To what extent is the
Registry utilized to search for and obtain transplants? (3) Are the donor
centers and other organizations in the NMDP network complying with its
standards

and procedures? To answer these questions, we analyzed NMDP data on racial
and ethnic representation on the Registry from 1998 through 2001 and, to
provide a broader context for examining these changes, also analyzed data
on racial

4 U. S. General Accounting Office, Bone Marrow Transplants: National
Program Has Greatly Increased Pool of Potential Donors, GAO/ HRD- 93- 11
(Washington, D. C.: Nov. 4, 1992).

5 HHS OIG, National Marrow Donor Program: Financing Donor Centers, OEI-
01- 95- 00123 (Washington, D. C.: December 1996). 6 Pub. L. No. 105- 196,
112 Stat. 631 (1998).

Page 3 GAO- 03- 182 Bone Marrow Transplants

and ethnic representation in relation to the patients who searched the
Registry from 1988 through 2001. In addition, we analyzed data provided by
the International Bone Marrow Transplant Registry (IBMTR) 7 on transplants
from related donors from 1997 through 2000, which enabled us to estimate
the demand for unrelated donor transplants in the United States and relate
this estimate to Registry utilization by patients searching for donors
during this period; analyzed NMDP data on matches, canceled searches, and
transplants obtained for patients needing donors during this period;
reviewed NMDP*s standards for participating in the Registry; and reviewed
evidence of compliance with the standards and procedures by the
organizations that participate. We also interviewed officials of NMDP;
HRSA; the Department of the Navy; the American Red Cross; and selected
donor, stem cell collection, and transplant centers. We did not

independently verify the accuracy of the data provided by NMDP. We
conducted our work from June 2001 through June 2002 in accordance with
generally accepted government auditing standards.

From 1998, when the National Bone Marrow Registry Reauthorization Act was
enacted, through 2001, the number of stem cell donors on the Registry
increased for all racial and ethnic groups. NMDP recruitment efforts
focused on minority groups appear to have been effective in increasing the
number of donors from these populations. Since 1998 the number of donors
on the Registry has increased by 36 percent, and increases for minority
groups ranged from 30 percent to 53 percent. The total of more than 1
million minority donors listed in 2001 contrasts with the approximately
80,000 we reported in 1992. The proportional distribution of racial and
ethnic groups on the Registry was much closer to their proportional
distribution in the U. S. population at the end of 2001 than it was in our
1992 review. However, when viewed as a percentage of each group*s
proportion of the U. S. population, African Americans and

Hispanics are underrepresented by 17 and 15 percent, respectively. The
underrepresentation of minorities is somewhat mitigated by the Registry*s
efforts to have complete genetic information needed for typing on a higher
proportion of minority donors, which facilitates more rapid matching. For
all racial and ethnic groups, the theoretical probability of finding a
match has grown as the Registry size has increased, but equal access to a
match may not be attainable. Differences among racial and ethnic groups in
the rarity and variability of the genes responsible for compatibility in

7 IBMTR is not a donor registry; it records data about transplants
performed. Results in Brief

Page 4 GAO- 03- 182 Bone Marrow Transplants

transplants may mean that the Registry cannot achieve equal probability
for all groups. Further, devoting many resources in pursuit of a small
number of rare genetic types may divert resources from other efforts, such
as recruiting Caucasians and other groups with more common genetic types,
which might more readily increase the number of matches.

Although the exact number of patients in need of transplants is not known,
estimates suggest that about one- third of them utilize the Registry to
search for donors. The number of transplants facilitated by NMDP
represents about one- tenth of those we estimate to be in need of
unrelated donor transplants. These figures suggest that the Registry may
be underutilized for both searching and facilitating transplants. From
1997 through 2000, an estimated 44,740 U. S. patients were in need of
unrelated donor transplants. During this period, physicians for
approximately 15,000 U. S. patients conducted preliminary searches for
donors on the Registry, and about 4,000 of these patients obtained
unrelated donor transplants facilitated by NMDP. About 25 percent of
formal searches were not completed because stem cells were obtained from
donors or organizations without the involvement of NMDP.

The organizations that are involved in transplantation and participate in
the NMDP network generally adhere to NMDP*s standards and procedures. NMDP
monitors the compliance and performance of these organizations with its
standards by using several systems of feedback and incentives, including
site visits. Centers that deviate from NMDP*s standards may be placed on
probation or suspended or their participation in the network may be
terminated. In 2001, NMDP required 24 centers to take corrective actions
because they did not meet its standards. Further, NMDP reimburses donor
centers for services based on their performance by offering financial
incentives to centers that consistently meet donor recruitment goals and
financially penalizing centers that do not.

In its written comments on a draft of this report, HRSA stated that the
report provides an accurate and helpful overview of the status of the
National Bone Marrow Donor Registry. HRSA agreed that other efforts are
needed in addition to minority recruitment efforts in order to improve
minority access to unrelated donor transplants, but pointed out that the
Registry has complete genetic information needed for matching on higher
proportions of minority donors than it has for Caucasian donors. We have
clarified this information in the report. HRSA agreed that many patients
who could benefit from transplants do not utilize the Registry but
suggested a slightly modified method of determining the number of patients
in need of transplants. We accepted this suggestion, but note that

Page 5 GAO- 03- 182 Bone Marrow Transplants

both approaches produce virtually identical estimates of overall
underutilization. (See app. I.) HRSA also noted that many factors affect
the time required to complete a search of the Registry and that NMDP has
completed medically urgent searches in less than a month. We have included
this clarification in the report. In addition, HRSA provided technical
comments, which we have incorporated as appropriate.

Most of the diseases treated by stem cell transplantation involve
abnormalities of the blood, metabolic, or immune systems. These diseases
include several forms of cancer as well as certain nonmalignant diseases.
8 They strike all races, although one racial group or another may have a

higher incidence rate for a particular disease. 9 Not all patients with
diseases that may be cured by stem cell transplants necessarily pursue
them. Depending on a number of donor and patient characteristics, from
about 10 to 50 percent of patients are alive 5 years after transplants.
The patients who do not survive may succumb either to their diseases or to
the consequences of transplantation. Because of these low survival rates,
some patients and physicians may be reluctant to select this stressful
treatment under most or all circumstances. For most of the diseases
involved, other therapies are available that may be less invasive, carry
lower risk, or be the medically preferred initial treatment. Nevertheless,
some of these diseases are best treated by stem cell transplantation,
either initially or after other treatments have failed.

Prior to stem cell transplantation, the patient*s bone marrow and,
consequently, immune system are destroyed with radiation or chemotherapy.
The patient*s bloodstream is then infused with healthy

stem cells from a donor. Healthy stem cells can be therapeutic because
they can develop into all the components of blood, including those needed
to replace the patient*s immune system. In an *autologous* transplant,

8 Almost 90 percent of the transplants coordinated by NMDP are for types
of cancer, including, in descending order of frequency, chronic
myelogenous leukemia, acute myelogenous leukemia, acute lymphocytic
leukemia, myelodysplastic disorders, and nonHodgkin*s lymphoma. The
nonmalignant diseases most commonly treated by stem cells obtained through
NMDP are aplastic anemia and several varieties of inherited disorders of

the metabolic, immune, and blood systems. 9 For example, the age- adjusted
incidence rate per 100, 000 patients with acute myelogeneous leukemia over
the period from 1995 through 1999 is higher for Caucasians

(3.7) than for African Americans (2.9). In contrast, for patients with
myeloma, the rate is higher for African Americans (11.5) than for
Caucasians (5.2). Background

Page 6 GAO- 03- 182 Bone Marrow Transplants

these cells come from the patient*s own marrow. In a *syngeneic*
transplant, the cells come from an identical twin. For many diseases, the
most common type of transplant is an *allogeneic* transplant, which
consists of stem cells from a genetically compatible donor.

Although bone marrow was initially the only source of stem cells for
transplantation, in recent years two other sources of stem cells,
umbilical cord blood and peripheral blood stem cells (PBSC), have also
been used. In 2001, 1,215 of the transplants facilitated by NMDP (70
percent) involved marrow, 42 (2 percent) involved cord blood, and 491 (28
percent) involved

PBSC. Umbilical cord blood is collected from the placenta and umbilical
cord of a newborn and then preserved in a cord blood bank until needed by
a matched patient. The number of stem cells typically obtained from cord
blood is relatively small but is often adequate for pediatric patients.

For transplantation from cord blood, the blood is volunteered when the
blood is banked, not when it is used. The Registry began an umbilical cord
blood stem cell program in 1998. Stem cells from peripheral blood may be
obtained in numbers sufficient for transplantation when the donor is
treated with a drug that causes the cells to leave the marrow and enter
the bloodstream where they can be extracted using a process where the
stems cells are removed and the remaining components of the blood are
returned to the donor. A donor, matched to a patient, may be asked to
donate either bone marrow or PBSC, depending on the preference of the
patient*s physician. The Registry has offered PBSC to patients since 1999.

In addition to its dependence on such common determinants of treatment
success as patient age and disease severity, the outcome of a transplant
depends on the degree of match between donor and patient with respect to
particular blood cell proteins* the human leukocyte antigens (HLA)* that
are part of a person*s genetic makeup. 10 Each person has three primary
pairs (one set of three from each parent) of these antigens that play a
major role in the compatibility of a transplant. A matched donor is
defined as one for whom each of these six antigens has the same kind of
HLA. If a matched donor cannot be found, then a donor with certain types
of mismatch may be used, depending on the transplant center*s

10 An antigen is a protein found on the outside of most cells in the body
that induces the formation of antibodies. There are a number of antigens
in the human body, and HLA are a set of these. Bone Marrow and Other

Sources of Stem Cells for Transplantation

Matching Donor and Patient

Page 7 GAO- 03- 182 Bone Marrow Transplants

preferences, although usually with poorer results. In general, the more
closely related two people are, the more likely it is that their HLA will
match. At one extreme, identical twins always match, and, in fact, match
on all antigens, not just the six ordinarily focused upon. At the other
extreme, members of separate racial groups are relatively unlikely to
match one another. Full siblings can provide a six out of six match,
resulting in what is called an *HLA- identical sibling transplant,* but
only about 30 to 40 percent of patients can be expected to have a matched
sibling donor. As a result, unrelated donors with matched HLA are sought
from the registries in which their HLA type has been recorded.

The definition of a match has been refined over time as scientific
understanding of HLA increases. HLA are being typed more precisely, so
more types of HLA can now be distinguished. Thus, some of today*s matches
may be judged as mismatches in the future because better matches are
possible. This increasing refinement does not mean, however, that finding
a suitable match for transplantation is inevitably becoming more
difficult. Some kinds of mismatch may be less dangerous than others. As a
result, as research continues, there may be fewer matches by today*s
standards, but relatively harmless mismatches will be recognized as such
and used. Further, there is evidence that cord blood may not require as
exact an HLA match as is usually sought.

In support of the Registry, NMDP manages a worldwide network consisting of
more than 400 donor centers, recruitment groups, 11 contract laboratories
where tissue is typed, apheresis centers, 12 cord blood banks, collection
centers where marrow is harvested, blood sample repositories, and
transplant centers. More than half of these organizations are donor (91)
or transplant centers (149). The relationship of these network components
to NMDP varies. Some, such as the recruitment groups, were designed to be
parts of the network and work with NMDP, whereas others, such as the
transplant centers, exist separately from the network and function
independently of NMDP except where specified by contract.

11 Recruitment groups actively seek donors, sometimes ones of a particular
ethnic or racial heritage. 12 Apheresis is a technique for separating
blood into its components, using a machine that draws blood from a vein in
a donor*s arm; filters out the desired product, such as PBSC; and returns
the remaining blood to the donor. The NMDP Network

Page 8 GAO- 03- 182 Bone Marrow Transplants

The NMDP network includes donor centers and other organizations in foreign
countries. 13 The foreign donor centers merge their files with the
Registry, contributing more than one million donors. These centers are
required to comply with NMDP policies, program standards, and other
criteria, although fees for recruiting donors and other financial
incentives and payments that go to U. S. centers are not paid to foreign
centers. 14 NMDP has also signed cooperative agreements with national
registries in

13 foreign countries. 15 Although certain data on donors recruited into
these registries are not entered into the Registry*s computer system,
these foreign registries will search their donor files on behalf of a U.
S. patient searching the Registry. In addition, 6 foreign apheresis
centers, 18 foreign bone marrow collection centers, and 36 foreign
transplant centers are affiliated with the Registry. NMDP*s affiliations
with foreign donor and transplant centers result in its facilitation of
both foreign- to- U. S. and U. S. to- foreign donations.

The existence of these international affiliations with the Registry does
not prevent U. S. transplant centers from obtaining stem cells through
foreign registries directly, that is, without going through Registry
channels. Even

domestically, the Registry is not a monopoly; other U. S. registries also
maintain lists of donors, conduct searches for stem cells, or perform both
of these functions. 16 These other registries, however, are relatively
small; often specialize in donors from particular racial or ethnic groups;
and are private, with no national requirements.

13 There are seven such donor centers. Three of these are in Germany, and
the others are located in the Netherlands, Israel, Sweden, and Norway. 14
HRSA consults with the Department of State on proposed membership of
foreign organizations. 15 These countries are Australia, Austria, Canada,
the Czech Republic, England, France, Ireland, Italy, Japan, Singapore,
Spain, Switzerland, and Taiwan.

16 For example, the American Bone Marrow Donor Registry of Mandeville,
Louisiana, is composed of a Patient Advocacy Office that coordinates and
processes search requests and a Donor Services Division that educates,
recruits, and maintains the records of donors. Moreover, it has regional
components, also called registries. Other U. S. registries include the
Caitlin Raymond International Registry of Worcester, Massachusetts, and
the Gift of Life Foundation of Boynton Beach, Florida. In addition, there
are a number of U. S. cord blood banks including ones in New York, New
Jersey, Missouri, and Massachusetts (part of the Caitlin Raymond
International Registry).

Page 9 GAO- 03- 182 Bone Marrow Transplants

The Registry serves two groups of people, donors and patients. The
Registry*s donor centers and recruitment groups recruit donors, who are
then managed by the donor centers. The Registry pays these centers and
groups for signing up donors. In view of the past underrepresentation of
minorities in the Registry, NMDP has initiated several recruitment efforts
to increase its racial and ethnic diversity. For example, it provides free
or low- cost minority- specific educational materials to donor centers and
recruitment groups. Probably the most important aspects of managing donors
are to maintain their commitment to donation so that they are locatable
and willing to donate when their stem cells are requested, to keep records
of how to contact them, and to drop from the list any individuals who are
too old 17 or no longer able or willing to donate.

A patient*s first contact with the Registry occurs when his or her
physician or a transplant center conducts a free, preliminary search of
the Registry for stem cell donors and cord blood units. The preliminary
search, which takes about 24 hours, produces a list of donors and cord
blood units that are potentially suitable for that patient. However, many
patients for whom such searches are conducted are not necessarily good
candidates for stem

cell transplants. For example, some searches may be conducted for patients
who are too sick for transplantation or who are good candidates for less
invasive therapies.

If the physician and patient decide to continue a search for an unrelated
donor (or unrelated cord blood) on the Registry, then more information
about the matching stem cells is required and a formal search is begun.
Only a physician affiliated with a transplant center in the NMDP network
may conduct a formal search of the Registry. The Registry bills the
transplant center a one- time activation fee of $600. It also bills the
center for the cost of the four or five testing components of the search
process,

each of which costs more than $100. Since several donors may have to be
tested before one is selected for the patient, these component charges may
be made repeatedly, resulting in a search costing thousands of dollars to
the transplant center, and more to the patient when the center adds its
markups. Relatively few insurance plans pay for searches; however, plans
often pay for the actual transplantation including the procurement of stem

cells. The details of the formal search and the subsequent steps in the
process possibly leading to transplantation depend on the additional
information needed; the results of laboratory tests; and the kind of stem

17 Donors are considered too old to donate at age 61. Operation of the
NMDP

Registry

Page 10 GAO- 03- 182 Bone Marrow Transplants

cells sought, whether stored blood from an umbilical cord or blood or
marrow from a living donor.

If a suitable donor or suitable cord blood unit is found, and if other
requirements in the process toward transplantation are fulfilled, 18 then
either (1) the marrow is harvested from the donor at a collection center,
(2) PBSC are collected from the donor at an apheresis center, or (3) the
cord blood is shipped from a cord blood bank. The stem cells are
transported to the transplant center, often by courier. The final step is
the infusing of the patient*s bloodstream with the selected marrow, PBSC,
or cord blood. The entire process* from the initiation of the formal
search to the transplant (infusion)* typically requires many months and
sometimes more than 1 year. However, some patients cannot wait this long
for transplants because their medical conditions are deteriorating.

During the search process, NMDP offers patient advocacy services through
two channels. Its Office of Patient Advocacy (OPA) provides several
services, including education, support, case management intervention,
financial assistance, and special advocacy projects. For example, OPA
publishes the Transplant Center Access Directory, a patient guide listing
all transplant centers in the NMDP network. The directory describes each
center*s HLA matching criteria and lists the diseases each typically
treats with unrelated donor marrow transplants. The directory also
provides information on comparable search charges and risk- adjusted
patient survival data. In addition to the services provided through OPA,
NMDP requires that each transplant center have a patient advocate on
staff. The patient advocate must be familiar with the center*s transplant
program and with issues of unrelated donor stem cell transplantation and
must not be a member of the transplant team.

A 1996 OIG review raised concerns about donor center costs and
performance. Before the review, NMDP used two methods to finance donor
centers. NMDP paid for services at some donor centers through cost- based
contracts for direct expenses, such as labor and fringe benefits and donor
expenses. Other donor centers received payments from NMDP for specified
activities, such as donor recruitment and donor search activities. The OIG
recommended that HRSA and NMDP develop a

18 These include confirmation of donor availability and willingness,
satisfactory results of laboratory tests done on the donor, and the
patient*s desire to continue the search. OIG Review

Page 11 GAO- 03- 182 Bone Marrow Transplants

payment approach for all donor centers that more directly linked funding
to performance and emphasize recruitment and retention of donors,
particularly donors from racial and ethnic minority groups. Further, the
OIG recommended that HRSA and NMDP develop procedures to monitor the
performance of donor centers and other organizations in the NMDP network.

The program*s recruitment efforts have apparently increased the number of
donors on the Registry since 1998 for all racial and ethnic groups, and
the theoretical probability of finding a match has increased steadily over
the life of the Registry. By 2001, the number of donors from each minority
group on the Registry had grown by at least 30 percent and was either
greater than or no more than 2 percentage points below its representation
in the general population. However, when viewed as a percentage of each

group*s population, African Americans and Hispanics are still
substantially underrepresented. For all racial and ethnic groups, the
theoretical probability of finding a match has grown as the Registry size
has increased, but equal access to a match may not be attainable.
Differences among racial and ethnic groups in the rarity and variability
of the genes responsible for compatibility in transplants may mean that
the Registry cannot achieve equal probability for all groups. Further, the
goal of equal access to a match conflicts to some extent with attempts to
maximize the

overall numbers of matches and transplants for the Registry. The size of
the Registry has increased since 1998 by 36 percent, and no minority group
increased by less than 30 percent. NMDP*s efforts to recruit minorities
may have substantially increased the number of donors from these
populations. Percentage increases for minorities ranged from 30 percent
for Native Americans to 53 percent for Hispanics. Caucasian donors
increased 28 percent. (See table 1.) The multiple race category had the
largest increase, 123 percent, but this may result in part from an
increase in the use of that category by those to whom it applies, rather
than solely from an increase in the availability of donors of that group.
NMDP Has Succeeded

in Increasing Recruitment of Donors, Including Minorities, but May Not Be
Able to Equalize Access to Matches

Number of Donors on Registry Has Increased for All Groups

Page 12 GAO- 03- 182 Bone Marrow Transplants

Table 1: Percentage Increase in Registry Donors, by Racial and Ethnic
Group, 1998 to 2000, and Current Proportion of Groups on the Registry and
in the Population

Race/ ethnicity Number on Registry, September 30, 1998 Number on Registry,

December 31, 2001 Percentage change

Percentage of donors on the Registry with

known race a Percentage of U. S. population b

African American 264,868 363,246 37 10 12 Asian/ Pacific Islander 194,118
287,129 48 8 4 Caucasian 1,926,675 2,460,725 28 7 69 Hispanic 252,569
386,059 53 11 13 Multiple race 34,443 76,937 123 2 2 Native American
45,478 59,112 30 2 1 Other 13,089 14,142 8 0 c 0 c Declined to specify
4,629 6,498 40 N/ A N/ A Unknown d 623,659 902,802 45 N/ A N/ A

Total 3,359,528 4,556,650 36 100 100

Note: N/ A = not applicable. a As of December 31, 2001.

b Based on 2000 U. S. Census. c Rounds to zero. d Some foreign registries
that are part of the NMDP network do not collect information on race or

ethnicity. Sources: NMDP and U. S. Bureau of the Census.

The total of more than 1,000,000 minority donors listed in 2001 contrasts
with the approximately 80,000 we reported in 1992. As can be seen in table
1, by 2001, the proportions of both African Americans and Hispanics on the
Registry were within 2 percentage points of their proportions in the 2000
U. S. population. The proportions of other minorities on the Registry were
either approximately equal to or exceeded their proportions in the
population. While the differences between Registry and population levels
of representation for African Americans and Hispanics reflect improved
representation of these groups, the 2- percentage point differences still
indicate a substantial underrepresentation in comparison with their
proportions in the U. S. population. Specifically, in 1992, the
proportions of African Americans and Hispanics, both at 4 percent of the
Registry, were 8 and 5 percentage points lower, respectively, than their
proportions in the U. S. population (which were 12 and 9 percent,
respectively). This translated to a 67 percent underrepresentation for
African Americans and a 56 percent underrepresentation for Hispanics. The
current 2- percentage point differences on the Registry for these groups
translate to a 17 percent

Page 13 GAO- 03- 182 Bone Marrow Transplants

underrepresentation for African Americans and a 15 percent
underrepresentation for Hispanics. 19 For all racial and ethnic groups the
theoretical probability of a patient*s

finding at least one matched donor has increased every year since 1988 but
has leveled off somewhat since 1998. 20 The increase in theoretical
probability represents significant progress in raising the likelihood of a
match. It reflects inclusion in the Registry of the most common genetic
types over the period when the Registry was small and new, and

recruitment efforts were beginning. The leveling off likely reflects the
fact that for all groups, after years of recruitment activity, improvement
now occurs mainly when rare types are added. (See fig. 1.)

19 For example, the African American difference of 2 percentage points is
17 percent of that group*s 12 percent share of the population. 20 This
probability was computed in 2001 by considering all patients who had
searched the

Registry by that time and, using NMDP*s matching criteria, asking what
proportion of these would have found a match during each year of the
Registry*s existence, given the donors on the Registry during that year.
This theoretical probability has advantages over the observed proportion
of matches as a measure of access by patients. One is that a large and
representative number of searching patients are repeatedly applied to the
Registry over its history so that any fluctuations cannot be a result of
fluctuations in the numbers or kinds of patients searching from year to
year. Another advantage is that today*s definition of a

match has been applied throughout the years covered so that any
fluctuations cannot be a result of changes in that definition over the
years. Theoretical Probability of Finding a Match Has

Increased over Life of Registry

Page 14 GAO- 03- 182 Bone Marrow Transplants

Figure 1: Theoretical Probability of Patient*s Finding at Least One
Matched Donor, by Racial and Ethnic Group, 1988 to 2001

Source: NMDP.

Nevertheless, the theoretical probability of finding a match varies by
race, ranging in 2001 from under 60 percent for African Americans to over
80 percent for Caucasians. This probability has always been higher for
Caucasian patients than for patients in any minority group, in part,
perhaps, because of Caucasians* greater numbers and level of
representation on the Registry. The theoretical probability of finding a
matched donor has been lowest for African American patients. This is
because, in addition to their smaller numbers and lower level of
representation on the Registry, their rarer and more varied HLA
combinations make matching harder.

Page 15 GAO- 03- 182 Bone Marrow Transplants

Because of genetic differences among racial and ethnic groups, there is
reason to believe that patients from some minority groups, notably African
Americans, may never have the same probability of finding matches, and
therefore of access to transplants, as Caucasian patients, regardless of
the

efforts made to recruit them. Any patient is more likely to find a match
in his or her own racial and ethnic group than in another group, so
patient matching rates depend, to some extent, on the number of people in
the

patient*s group on the Registry. All minorities are at a disadvantage for
this reason. Further, some minority groups, such as African Americans, are
known to have more rare and more varied HLA combinations than do
Caucasians. The likelihood of finding a match from among a group of
racially or ethnically defined donors declines with the rarity and number
of possible genetic types found among the members of that group.

In addition to these factors related to finding a match, there are other
factors that may contribute to differences in access to a transplant. Some
of these depend on the characteristics of those who volunteer for the
Registry. For example, donors from different groups may differ in their
tendency to be available (locatable, willing, and physically able) when
called upon to actually donate. Other possible factors involve the
attitudes, health, medical care, resources, and preferences of the
patients. Patients of different groups may differ in their tendency to
engage the health care system at all, to seek help early enough in their
illnesses, or to search the Registry as opposed to pursuing other options.
It may be possible to effect changes in these factors, thereby moving
closer to the goal of equal opportunity for all racial and ethnic groups.

However, not only is the goal of equal access to transplants for all
groups difficult to attain, but it also may conflict with the statutory
goal of maximizing the number of patients who find a match and thereby
maximizing the number of transplants facilitated. Recruiting donors with
the rare HLA combinations that may be needed for minorities is difficult.
Large numbers of donors must be recruited and retained in the Registry in
order to identify and add each rare genetic type to the donor pool, so the

cost of recruiting such donors the incremental cost of adding these rare
genetic types to the donor pool is large. Thus, devoting many resources in
pursuit of a small number of rare genetic types may divert resources from
other efforts, such as recruiting Caucasians and other groups with more
common genetic types, which might more readily increase the number of
matches.

Because of the difficulty encountered in finding matches for minority
patients, NMDP engages in a number of initiatives to increase the Equal
Access for All

Groups May Be Unattainable

Page 16 GAO- 03- 182 Bone Marrow Transplants

Registry*s diversity. It conducts outreach, recruitment, and educational
efforts directed towards minorities. In addition, NMDP has initiated a
program to pay the full costs of HLA tissue typing for minority donors.

Although the difficulty in finding matches for minority patients may be
unavoidable, it may be mitigated somewhat by the efforts of the Registry
to increase the number of donors on whom it has complete HLA typing. The
vast majority of actual donations are obtained from by donors whose HLA is
fully typed. 21 When only these donors are considered, each minority
constitutes a larger portion of the Registry than its representation in
the population. Therefore, because access to a match depends upon, for the
most part, the fully typed donors on the Registry, access for minorities
may be somewhat better than might be assumed by looking at the Registry as
a whole.

Although the exact number of patients in need of transplants from
unrelated donors is not known, the number of patients utilizing the
Registry to search for matches is about one- third of the estimated number
of patients in need of unrelated donor transplants. About one- tenth of
the number of patients estimated to be in need of unrelated donor
transplants obtain transplants facilitated by NMDP. These figures suggest
that the Registry may be underutilized for both searching and facilitating
transplants. 22 Physicians for approximately 15,000 U. S. patients
requested preliminary searches of the Registry from 1997 through 2000.
This number represents 34 percent of the 44,740 U. S. patients estimated
to be in need of stem cell transplants from unrelated donors in that 4-
year period. About

4,000, or 27 percent, of the patients whose physicians searched the
Registry eventually received transplants facilitated by NMDP. However, a
significant proportion of searches were not completed because stem cells
were obtained from donors or organizations without the involvement of
NMDP.

21 A fully typed donor is one for whom all crucial antigens are determined
at the time the donor volunteers. 22 See app. I for an explanation of how
we estimated the number of patients in need. National Registry May

Be Underutilized

Page 17 GAO- 03- 182 Bone Marrow Transplants

From 1997 through 2000, physicians carried out preliminary searches for 34
percent of the number of U. S. patients estimated to be in need of
transplantation from unrelated donors at any time during that period. The
number of transplants facilitated by NMDP for all U. S. patients was 9
percent of the number estimated to be in need. The precise number of
patients in need of unrelated donor transplants is not known. However,
there is a greater than 10 to 1 ratio between the number of such patients
estimated to be in need and the number of transplants facilitated by

NMDP. This suggests that the Registry may be underutilized, as many more
U. S. patients may need unrelated donor transplants than obtain them
through the Registry. 23 The ratio of the number of preliminary searches
to the number of patients in need varied by race and ethnicity. Among
specific racial and ethnic groups, the percentage of preliminary searches
was highest for Caucasian patients (35 percent), and was lowest for

Hispanic patients (24 percent) and Native American patients (24 percent).
(See table 2.) We do not know why these apparent disparities in search
rates exist.

23 NMDP has used a similar method of estimation and draws a similar
conclusion about possible underutilization. Estimates Suggest about

One- Third of Patients in Need Search the Registry and about One- Tenth
Receive Transplants

Page 18 GAO- 03- 182 Bone Marrow Transplants

Table 2: U. S. Patients* Utilization of the Registry, by Race and
Ethnicity, 1997 to 2000 Race/ ethnicity

Estimated number of patients without matched sibling

donor a (patients in need) Actual

number of preliminary

searches Ratio of number

of preliminary searches to

number of patients in need

Actual number (percentage) of

preliminary searches resulting in

formal searches

Actual number (percentage) of

preliminary searches resulting in

NMDPfacilitated transplants

Ratio of number of

NMDPfacilitated transplants to

number of patients in

need

African American 5,397 1,694 0.31 958 (57) 256 (15) 0.05 Asian/ Pacific
Islander 1,666 439 0.26 270 (62) 96 (22) 0.06 Caucasian 30,929 10,844 0.35
7,079 (65) 3,321 (31) 0.11 Hispanic 5,613 1,366 0.24 840 (61) 317 (23)
0.06 Native American 329 80 0.24 56 (70) 20 (25) 0.06 Other 806 365 0.45
213 (58) 39 (11) 0.05

Total 44,740 15,231 b 0.34 9,623 c (63) 4,056 d (27) 0.09

a For Caucasians, the number of HLA- identical sibling transplants
multiplied by the number of patients expected to be without matched
sibling donors for each such transplant was derived from data obtained
from the Statistical Center of the IBMTR and Autologous Blood and Marrow
Transplant Registry (ABMTR). (The analysis has not been reviewed or
approved by the Advisory Committees of the IBMTR and ABMTR.) See appendix
I for a description of this method of estimation. For each of the other
groups, the number was derived by assuming that the group*s need is the
same as it is for

Caucasians and in proportion to the group*s representation in the U. S.
population. b Includes 443 preliminary searches, not included elsewhere in
the column, from patients of unknown

race/ ethnicity. c Includes 207 formal searches, not included elsewhere in
the column, from patients of unknown

race/ ethnicity. d Includes 7 transplants, not included elsewhere in the
column, from patients of unknown

race/ ethnicity. Source: GAO analysis of data from the Statistical Center
of the IBMTR and ABMTR and NMDP.

About one- fifth of the number of patients estimated to be in need
formally searched the Registry (9, 623 out of 44, 740). Less than one-
tenth of those estimated to be in need ultimately received NMDP-
facilitated transplants. The numbers and percentages of preliminary
searches that progressed to formal searches from 1997 through 2000 are
presented by racial and ethnic group in table 2. The overall rate of
progression from preliminary to formal search is 63 percent. Further,
4,056 of the 15,231 U. S. patients (27 percent) for whom preliminary
searches were conducted from 1997 through 2000 eventually received NMDP-
facilitated transplants. This number corresponds to 9 percent of the
number of patients estimated to be in need of unrelated transplants during
that period.

Page 19 GAO- 03- 182 Bone Marrow Transplants

Reasons for cancellation of preliminary searches or formal searches vary.
Although clinical reasons, such as a change in medical condition, are the
most commonly cited explanations for cancellation of both preliminary and
formal searches, 24 another relatively frequent reason is that stem cells
are obtained from a provider other than NMDP, such as a related donor or
another registry. (See tables 3 and 4.) We do not know the proportion of
these cases that used a related donor, and some cases may not have been
able to find a potential match at NMDP. However, it is likely that in at
least some of these cases, NMDP might have facilitated a transplant if the
patient*s transplant center had not selected another registry to provide
the stem cells, thus representing another kind of possible
underutilization of NMDP. Lack of donor availability* not finding any
potential matches* and financial reasons are not commonly cited as reasons
for cancellation of either kind of search, although it is possible that
patients with limited

financial resources or insurance may not be encouraged to make preliminary
searches.

Table 3: Reasons for Preliminary Search Cancellation, January 2000 through
September 2001

Reason for cancellation Number of preliminary search cancellations
Percentage of preliminary

search cancellations

No donor available 105 7 Another provider 317 20 Patient stable 383 25
Financial reasons 114 7 Personal reasons a 187 12 Deterioration/ death 160
10 Other 284 18

Total 1,550 100

Note: These data are based on a survey, conducted by OPA, of individuals
making preliminary searches. a Personal reasons for preliminary search
cancellations include decisions made by physicians and

patients. Source: NMDP.

24 Cancellation of a preliminary search means that 45 days have occurred
since the search without a formal search having been initiated.
Cancellation of a formal search means that the transplant center has
submitted a particular form indicating a desire to terminate the search
and included a reason for doing so. Reasons for Cancellation

Vary and Include Obtaining Stem Cells from a Provider Other than NMDP

Page 20 GAO- 03- 182 Bone Marrow Transplants

Table 4: Reasons for Formal Search Cancellation, 1997 to 2000 Reason for
cancellation Number of formal search

cancellations Percentage of formal search cancellations

No donor available 131 3 Another provider 1,200 25 Financial reasons 22 0
Personal reasons a 733 15 Deterioration/ death 2,096 44 Alternative
therapy 357 7 Other 262 5

Total 4,801 100

a Personal reasons for formal search cancellations include decisions made
by physicians, patients and patients* families. Source: NMDP.

Several factors may influence a decision to obtain stem cells from a
provider outside the NMDP network, including the source of stem cells
preferred by the physician, the costs involved, and the timeliness of the
response. Outside providers may need to be used when the physician sees
cord blood as a viable alternative source to bone marrow or PBSC because
some cord blood banks do not list their cord blood units with NMDP. 25
Search and procurement costs can also be a factor. Administrators of

transplant centers that have done non- NMDP- affiliated transplants told
us that other registries charge less for searches than NMDP does. For
example, we were told that only a few other registries worldwide charge a
search activation fee in addition to their charges for the specific
medical procedures needed to confirm that a particular donor is healthy
and matched to the patient. In addition, the cost of stem cell procurement
at NMDP tends to be higher. One transplant center director told us that
the center pays about $13,000 for stem cells obtained directly from
overseas registries and about $21,000 for NMDP stem cells. However, even
when NMDP is not paid for a formal search or for stem cells, it may still
have been utilized. An official at NMDP informed us that it is possible
for a transplant center to determine the NMDP- affiliated registry at
which a foreign (but not domestic) potential match is registered on the
basis of a

25 Although several blood banks list their cord blood units with NMDP,
others, including the largest, the New York Blood Center, do not, and thus
NMDP cannot facilitate transplants from those banks.

Page 21 GAO- 03- 182 Bone Marrow Transplants

preliminary search and to contact the foreign registry directly to obtain
the stem cells. Moreover, that official stated that some transplant
centers may do this regularly. Thus, although NMDP may not be recorded as
having facilitated the transplants that result, its role in helping to
locate donors in such cases means that its utilization is somewhat greater
than the record

suggests. Timeliness can be another factor. A few center administrators
mentioned that NMDP takes longer to provide stem cells than do other
registries. For example, one administrator told us that the time it takes
to obtain a donor sample for testing at the transplant center* an
important component of the overall search process* can be a week longer
for NMDP than for a foreign registry, depending on whether NMDP judges the
search to be urgent. Waiting this additional week can be frustrating for
those at the transplant center who are anxious to determine whether they
have a confirmed match or will have to continue searching. Another
director told us that stem cells from non- NMDP providers are more likely
to be received by the date the transplant center requests them than are
stem cells from NMDP. NMDP has attempted to shorten its time from formal
search

initiation to transplant and reports that its median time has decreased
from 4.8 months from 1992 through 1993 to 3.7 months in 2000. The optimal
time frames for patients vary. Some may not be urgent, but NMDP has shown
that it is possible to complete urgent searches in less than a month and
reports that it expects to begin offering urgent searches as an option to
transplant centers.

Organizations that participate in the NMDP network generally comply with
the standards and procedures it has established. In order to encourage
adherence, NMDP uses various mechanisms to monitor compliance and
performance. These include site visits, the Continuous Process Improvement
(CPI) program, and incident reports, as well as a financial incentive
system designed to improve the performance of donor centers. The results
of the selected site visits, analysis of CPI measures, and incident report
summaries we reviewed show that the organizations in the NMDP network
generally adhere to NMDP*s standards and procedures. In general, NMDP
ensures compliance by taking action against noncompliant organizations.
(See app. II for examples of how NMDP uses these systems to achieve
compliance with respect to selected activities.) In 2001, NMDP required 24
donor and transplant centers to take corrective actions because they did
not meet its standards. The incentive system encourages compliance by
linking donor center reimbursement to performance. Organizations in the

NMDP Network Generally Comply with Its Standards and Procedures

Page 22 GAO- 03- 182 Bone Marrow Transplants

NMDP uses several mechanisms to encourage the compliance and performance
of the participating organizations in its network. NMDP staff members
conduct site visits to donor centers to monitor the centers* compliance
with NMDP*s standards and procedures and to provide feedback about the
results. It also employs the CPI program to assess and provide feedback at
donor, transplant, and bone marrow collection centers. Further, NMDP
monitors incident reports from donor, transplant, and collection centers
and may take corrective action including, in serious cases, suspension or
termination.

According to NMDP officials, NMDP staff members conduct site visits at
donor centers approximately every 2 years to assess donor center
compliance with program standards and procedures. NMDP staff members
review the organization of the program (such as its support and staffing
structure), recruitment activities (such as performance against goals and
donor drive compliance), donor management activities (such as management
of patient- related donor search requests, confidentiality procedures, and
records management), and billing and reimbursement to determine adherence
to NMDP*s standards and procedures. They also compare performance against
goals for various recruitment activities. Upon completion of these visits,
NMDP staff members discuss the results with the center staff and provide a
summary report. Centers that are noncompliant are advised of the problems
and are required to submit corrective action plans to NMDP that address
the problems. Our review of donor center site visit reports indicates that
the reports identified problems and the corrective actions required of the
centers to meet NMDP criteria.

Since 1998, NMDP has conducted additional site visits at transplant
centers to verify the accuracy of the data that the transplant centers
submit electronically to NMDP. NMDP staff members compare the data from
the centers* records with the data from NMDP*s computer system.

During these visits, NMDP staff members may also review other activities,
such as the signing of patient consent forms. The site visits are
scheduled for each transplant center every 4 years. NMDP plans to issue
its first

annual report on the results of the first cycle of site visits in
September 2002.

NMDP monitors the operations and performance of its centers through the
CPI program. The program includes nine goals to increase the efficiency of
key activities in the search and donation process and measures performance
against these goals. For example, at donor centers, NMDP measures the
timeliness of registering new donors, resolving search NMDP Monitors
Network

Compliance and Performance

Page 23 GAO- 03- 182 Bone Marrow Transplants

related requests, and processing requests for HLA blood typing. At
transplant centers, NMDP measures the time it takes to resolve and report
confirmatory testing results. NMDP also monitors post- transplant data

submission through CPI. These outcome data are used in research studies to
analyze outcomes for donors and patients. NMDP also monitors the accuracy
and timeliness with which donor and transplant centers submit donor and
patient blood samples to NMDP*s research repository. NMDP provides regular
feedback to donor and transplant centers concerning their performance on
CPI measures. For example, each center receives a monthly report
summarizing the results of its activities, along with those

of all other centers, in the previous month. The reports allow centers to
analyze how consistently they perform and to compare their results to
those of other centers in the network. NMDP also conducts a year- end
analysis to provide feedback to centers.

Through its CPI program, NMDP monitors whether organizations in its
network meet goals for timeliness and may recommend corrective actions for
centers that do not meet these goals. 26 A year- end analysis of the CPI
program shows that during 2001 almost half (44 of 91) of donor centers met
all nine CPI goals for the search process. In addition, 20 more donor
centers met eight of nine goals, and 9 others met seven of nine goals.
According to NMDP, the remaining 18 donor centers (20 percent) that met
six or fewer goals were the focus of technical assistance to improve their
performance. Our analysis shows that 5 of the 91 donor centers (5 percent)
were placed on review or probation for failing to meet CPI goals in 2001.

Our analysis also shows that NMDP placed 18 of the 129 transplant centers
(14 percent) on probation. Eight of these were placed on probation for
failure to meet CPI goals for the search process, seven for failure to
meet CPI measures concerned with timely submission of recipient follow- up
information, and three for problems related to the accuracy and timeliness
of submissions of donor and patient research blood samples.

NMDP supplements these activities with incident reports, which are written
accounts of deviations from policies and standards that are categorized by
the nature of a deviation and include, but are not limited to,

26 Under CPI, NMDP allows up to three notices of noncompliance with a
particular standard and sets interim goals to be met within a specified
review period after each notice. After a third notice, centers are placed
on probation. Failure to meet the requirements of the probationary period
may result in termination.

Page 24 GAO- 03- 182 Bone Marrow Transplants

categories such as confidentiality concerns, customer service, and product
transport. NMDP uses incident reports to track deviations from its
standards by recording the specifics of incidents. NMDP staff members
follow up and investigate incidents. In addition, an NMDP committee
reviews a summary report of incidents twice a year to identify developing
trends that may affect an individual center or the entire network. Since

NMDP reviews center participation annually, the committee may follow up on
deviations from NMDP*s standards or take action such as probation,
suspension, or termination during the reapplication process. We reviewed a
summary of incidents categorized by type of problem and the corrective
actions taken to resolve them. For example, one incident involved an
operating room staff member administering less appropriate blood, rather

than the donor*s own blood, which was available for that purpose, during a
marrow harvest. NMDP monitored an investigation at the hospital to ensure
that the problem would be addressed.

To improve the operation of its donor centers, NMDP ties their
reimbursement to their performance. In 1997, NMDP instituted a new
reimbursement system that links payment to performance on CPI goals for
all donor centers. 27 NMDP pays donor centers a fee for each activity to
recruit donors for the Registry, such as signing up donors, typing their
tissues, maintaining their files, and other activities related to
confirming that the donors identified as potential matches for a searching
patient actually match and are medically cleared for donation. NMDP pays
each

donor center a recruitment fee of $28 and $10 for every minority and
Caucasian donor, respectively, recruited up to the number specified in its
recruitment goal. NMDP establishes annual recruitment goals for each donor
center based on the demographics of the local population. When donors are
recruited, the donor centers that do not register a specific percentage of
the new donors within a certain period incur financial penalties. For
example, the CPI goal for registering new donors is to register at least
85 percent of them within 35 days of the date on which they volunteer.
NMDP would reduce the total recruitment fee it pays to donor centers that
register less than 85 percent of new donors within this time frame. NMDP
data show that in May 2001, 98 percent of all donor

27 The HHS OIG recommended that HRSA standardize contracts between NMDP
and donor centers for donor services to improve the cost efficiency of the
centers and to link payment to performance. HRSA included this requirement
in its 1997 contract with NMDP to operate the Registry. Donor Center

Reimbursement Is Linked to Performance

Page 25 GAO- 03- 182 Bone Marrow Transplants

centers met this goal. In addition, NMDP pays incentives to donor centers
for retaining donors at various points in the donation process.

In spite of progress in recruiting minority donors, racial and ethnic
disparities in the Registry remain, due in part to differences in the
genetic variability within groups. Thus, differences among racial and
ethnic groups in the probability of obtaining transplants will likely
continue. Many in need of transplants may not search the Registry; those
that do often do not obtain them, and for those that obtain them, the
transplants may not be facilitated by NMDP. Although NMDP enhances the
quality of its network

by actively monitoring the compliance and performance of the component
organizations, it has not attained the level of utilization that might be
expected.

In its written comments on a draft of this report, HRSA stated that the
report provides an accurate and helpful overview of the status of the
National Bone Marrow Donor Registry. HRSA agreed that recruitment of
donors cannot be the sole strategy for improving access to unrelated donor
transplants for minority patients or those with unusual antigens, and
cited the need for other efforts to supplement recruitment activities.
However, HRSA noted that the Registry consists of two distinct groups of
donors, those who are fully HLA typed and those who are less than fully

typed. Since the vast majority of actual donors are selected from the
fully typed portion, minority racial and ethnic groups therefore make up a
larger proportion of the Registry than their representation in the U. S.
population. We have noted in the report that, because of this, access for
minorities may be somewhat better than might be assumed by looking at the
Registry as a whole.

With regard to underutilization of the Registry, HRSA agreed that many
patients who could benefit from unrelated donor transplants never consult
the Registry or do so too late in the course of their illnesses. HRSA
suggested a slightly modified method for estimating the number of patients
in need. We modified table 2 in accordance with its suggestions, but note

that both approaches produce virtually identical estimates of overall
utilization. (See app. I.)

Finally, HRSA noted that many factors affect the time required to complete
a search of the Registry. While searches frequently take many months and
the median search time has decreased, NMDP has completed medically urgent
searches in less than a month, on a pilot basis, and reports that it
Concluding

Observations Agency Comments

Page 26 GAO- 03- 182 Bone Marrow Transplants

expects to begin offering urgent searches as an option to transplant
centers. We have revised the report to include this clarification. HRSA
also provided technical comments, which we incorporated as appropriate.
HRSA*s comments are reprinted in appendix III.

We are sending this report to the Administrator of HRSA, the NMDP Chief
Executive Officer, and other interested persons. We will also make copies
available to others upon request. In addition, the report will be
available at

no charge on the GAO Web site at http:// www. gao. gov. If you or your
staff members have any questions about this report, please call me at
(202) 512- 7119. Key contributors to this assignment are listed in
appendix IV. Janet Heinrich

Director, Health Care* Public Health Issues

Page 27 GAO- 03- 182 Bone Marrow Transplants

List of Committees The Honorable Robert C. Byrd Chairman The Honorable Ted
Stevens Ranking Minority Member Committee on Appropriations United States
Senate

The Honorable Edward M. Kennedy Chairman The Honorable Judd Gregg Ranking
Minority Member Committee on Health, Education,

Labor, and Pensions United States Senate

The Honorable C. W. Bill Young Chairman The Honorable David Obey Ranking
Minority Member Committee on Appropriations House of Representatives

The Honorable W. J. *Billy* Tauzin Chairman The Honorable John D. Dingell
Ranking Minority Member Committee on Energy and Commerce House of
Representatives

Appendix I: Methods of Assessing Registry Utilization Page 28 GAO- 03- 182
Bone Marrow Transplants

Registry utilization is the extent to which patients in need of unrelated
stem cell transplants search the Registry or obtain NMDP- facilitated
transplants. In determining utilization, it is necessary to use indirect
methods to calculate the number of patients in need because it is
impossible to determine this number directly. For example, although we may
be able to obtain data on the number of patients who have been diagnosed
with certain blood and immune system diseases, we are unable to determine
the number for whom stem cell transplants are the best treatment.

One measure of the utilization of the Registry is the extent to which the
number of patients obtaining transplants facilitated by the Registry is as
high as it could be. The maximum possible utilization of the Registry
would be indicated if the number of U. S. patients conducting preliminary
searches was approximately equal to the estimated number of patients
needing unrelated donor transplants. A second measure of utilization is
the extent to which patients search the Registry. The method we used to
assess the two aspects of utilization* searching

the Registry and obtaining an NMDP- facilitated transplant* is also used
by NMDP. It involves estimating the number of patients in need of
unrelated donor transplants by using data on the number of HLA- identical
sibling transplants obtained from IBMTR. 1 This method and two alternative
methods that are also used by NMDP to assess utilization by U. S.
patients, one based on the number of preliminary searches conducted and
the other based on the incidence of disease, are described here.

For the years from 1997 through 2000, we estimated the number of Caucasian
patients in need of unrelated donor transplants based on the average
annual number of Caucasian HLA- identical sibling transplants performed
during those years. To obtain this estimate, we multiplied the number of
HLA- identical sibling transplants, for Caucasians, by the number of
patients of that group that genetic theory predicts* on the basis of the
average number of children born to the women of that group* are

1 This registry, located in Milwaukee, registers bone marrow transplants,
not donors like the other registries discussed in this report. The data
used in our estimations were obtained from the Statistical Center of IBMTR
and Autologous Blood and Marrow Transplant Registry (ABMTR). The analysis
has not been reviewed or approved by the Advisory Committees of the IBMTR
and ABMTR. Appendix I: Methods of Assessing Registry

Utilization Method Based on Number of HLAIdentical Sibling Transplants

Appendix I: Methods of Assessing Registry Utilization Page 29 GAO- 03- 182
Bone Marrow Transplants

in need of unrelated donor transplants for every Caucasian HLA- identical
sibling transplant in the United States.

The average number of children born to Caucasian women over a lifetime
during the years from 1989 through 1995 was 1.7925. 2 Subtracting the
individual who is in need of a transplant gives n = 0.7925 as the number
of siblings available to be transplant donors. The likelihood of a match
between two siblings is 25 percent because each child inherits one- half
of each parent*s HLA genes, resulting in a one out of four chance of
having the same HLA genes as a sibling has. Therefore, the probability
that no

sibling HLA identically matches the one in need is P = (0.75) n . For a
Caucasian patient, P = (0. 75) 0.7925 = 0. 796134.

The number of patients in need of unrelated stem cell transplants is equal
to the number of sibling donor transplants multiplied by P/( 1 P).

Therefore, for every HLA- identical sibling transplant recorded for a
Caucasian patient, there will be 0.796134/( 1 0.796134) = 3.90518 patients
in need of unrelated donor transplants. Because there were 7,920 sibling
transplants performed for Caucasian patients from 1997 through 2000, we

estimate that 3.90518( 7, 920) = 30, 929 Caucasian patients were in need
of stem cell transplants during that period. The estimates for other
racial and ethnic groups are presented in table 2. Because minorities
generally have less access to health care 3 and may therefore have less
access to sibling

transplants specifically, these estimates were obtained by assuming that 2
We determined this average by taking the median Caucasian fertility rate
for the years from 1989 through 1995. Fertility rates for non- Hispanic
Caucasians were not available for earlier years. We did not include rates
for years after 1995 because we do not think many of the transplants
occurring during the years 1997 through 2000 were done for patients born
after 1995. Because fertility rates tended to be higher before about 1973,
when some of the patients seeking transplants during the period of our
analysis, 1997 through 2000, were born, the use of the 1989 through 1995
rates results in an underestimation of the average number of siblings and
a consequent overestimation of the number of patients in need of

unrelated donor transplants. The effect of this consideration of the 1989
through 1995 rates is counterbalanced to an unknown extent because (1) the
fertility rates count half siblings and dead siblings as well as living
full siblings and (2) the fertility rates count all of a woman*s live
births, including those that occur after the patient needs a transplant.
The effect of these two counterbalancing considerations is to overestimate
of the number of

siblings available to donate and underestimate the number of patients in
need. The net effect of the choice of 1989 through 1995 rates and the
considerations concerning the fertility rates on the estimation of the
number of patients in need is not known. 3 See, for example, R. M.
Mayberry, F. Mili, and E. Ofili, *Racial and Ethnic Differences in

Access to Medical Care,* Medical Care Research and Review, vol. 57,
Supplement 1 (2000), pp. 108- 145.

Appendix I: Methods of Assessing Registry Utilization Page 30 GAO- 03- 182
Bone Marrow Transplants

each minority group*s need for unrelated donor transplants is proportional
to the Caucasian group*s need. The estimates were obtained by multiplying
the number of persons in the minority group by the proportion of
Caucasians in need of unrelated donor transplants. This approach
implicitly assumes that differences across groups in fertility rates are
of negligible importance in computing the numbers of patients in need of
unrelated donor transplants.

An alternative approach assumes that minorities and Caucasians have equal
access to HLA- identical sibling transplants. Based on this assumption,
this approach derives the needs of minorities for unrelated

donor transplants directly from their observed numbers of HLA- identical
sibling transplants. In doing so, it allows for the possibility that each
group has its own disease incidence rates and that the differences among
groups in their relative levels of sibling donations reflect these rates,
not differences in access. (See table 5.) This approach, while utilizing
somewhat different assumptions from the method above, produces a virtually
identical estimate of the underutilization of the Registry (10 percent
versus 9 percent).

Appendix I: Methods of Assessing Registry Utilization Page 31 GAO- 03- 182
Bone Marrow Transplants

Table 5: Alternate Approach to the Analysis of U. S. Patients* Utilization
of the Registry, by Race and Ethnicity, 1997 to 2000 Race/ ethnicity

Estimated number of patients without matched sibling

donor a, b (patients in need) Actual

number of preliminary

searches Ratio of

number of preliminary searches to

number of patients in

need Actual number

(percentage) of preliminary

searches resulting in

formal searches

Actual number (percentage) of

preliminary searches resulting in

NMDPfacilitated transplants

Ratio of number of

NMDPfacilitated transplants to

number of patients in

need

African American 1,880 1,694 0.90 958 (57) 256 (15) 0.14 Asian/ Pacific
Islander 1,355 439 0.32 270 (62) 96 (22) 0.07 Caucasian 35,964 10,844 0.30
7,079 (65) 3,321 (31) 0.09 Hispanic 1,593 1,366 0.86 840 (61) 317 (23)
0.20 Native American 97 80 0.82 56 (70) 20 (25) 0.21 Other 290 365 1.26
213 (58) 39 (11) 0.13

Total 41,179 15,231 c 0.37 9,623 (63) 4,056 (27) 0.10

Note: This table presents an alternate approach to that given in table 2.
a Number of HLA- identical sibling transplants multiplied by the number of
patients expected to be without matched sibling donors for each such
transplant was derived from data obtained from the Statistical Center of
the IBMTR and Autologous Blood and Marrow Transplant Registry (ABMTR).

(The analysis has not been reviewed or approved by the Advisory Committees
of the IBMTR and ABMTR.) b Numbers based on the HLA- identical sibling
transplants of the designated race/ ethnicity plus a

portion of those of unknown race/ ethnicity. These unknowns submitted
record forms that did not ask about race/ ethnicity. The unknowns can be
assumed to be similar in racial/ ethnic distribution to the other
patients, and so we distributed them among the racial/ ethnic groups
according to that distribution. It can therefore be assumed that there is
no racial bias in this estimation method. c Includes 443 preliminary
searches, not included elsewhere in the column, from patients of unknown

race/ ethnicity. Source: GAO analysis of data from the Statistical Center
of the IBMTR and ABMTR and NMDP.

The second method used by NMDP to assess Registry utilization is based
simply on the annual number of patients conducting preliminary searches.
In order to use this method, one must assume that this number directly
represents those in need of unrelated donor transplants. One cannot assess
the extent to which those in need search the Registry on the basis of this
number since the number itself is the number of patients searching.
However, one can assess the extent to which those in need obtain
NMDPfacilitated

transplants by considering the annual percentage of patient searches that
result in NMDP- facilitated transplants. This method yields an estimate of
the patients searching who obtain NMDP- facilitated transplants of 27
percent. (See table 5.) Method Based on

Number of Preliminary Searches

Appendix I: Methods of Assessing Registry Utilization Page 32 GAO- 03- 182
Bone Marrow Transplants

Although this approach has been used by NMDP as a way of assessing
utilization, officials at NMDP observe that the validity of this approach
to utilization assessment is limited by the freedom with which patients
can choose whether to search. These officials point out that preliminary
searches are performed for some patients who are not good candidates for
transplant and that other patients who should submit preliminary searches
probably do not. Because of the lack of correspondence between the number
of patients in need and the number performing preliminary searches, this
estimate is not likely to be as accurate as the other two.

The third method used by NMDP is based on an estimate of the annual number
of U. S. patients newly diagnosed from 1997 through 2000 with selected
diseases that might benefit from unrelated stem cell transplants. 4 The
estimated number of potential recipients for each disease is obtained from
disease incidence estimates, with adjustments for the likelihood that

(1) the patient is young enough to benefit from transplantation, (2)
disease severity is not so great as to make transplantation futile, and
(3) an HLAidentical sibling donor is available, thereby making unrelated
donor transplant unnecessary. The ratio of the annual number of
NMDPfacilitated transplants for U. S. patients diagnosed with these
selected diseases during this period to the estimated number of new U. S.
patients with the diseases is used to assess utilization. 5 (See table 6.)
The ratio, for

all patients with the selected diseases, corresponds to an estimated
percentage of candidates obtaining transplants* 10 percent* that is very
close to the estimate obtained by the first method. The validity of this
third method is constrained by the limited number of diseases for which
data are available. 4 R. P. Gale, *Potential Utilization of a National
HLA- Typed Donor Pool for Bone Marrow

Transplantation,* Transplantation, vol. 42, no. 1 (1986), pp. 54- 58. 5 We
have not related the number of potential recipients estimated in this
third way with the numbers of preliminary searches for patients with the
selected diseases, only with the

numbers of transplants. Method Based on Incidence of Disease

Appendix I: Methods of Assessing Registry Utilization Page 33 GAO- 03- 182
Bone Marrow Transplants

Table 6: Average Annual Unrelated Donor NMDP- Facilitated Transplants and
Estimated Number of Potential Recipients for U. S. Patients with Selected
Diseases Who Might Benefit from Unrelated Stem Cell Transplants, by Race/
Ethnicity, 1997 through 2000

Race/ ethnicity Acute

lymphocytic leukemia

Acute myelogenous

leukemia Chronic

myelogenous leukemia

NonHodgkin*s lymphoma Total

African American 8 12 18 4 42

Caucasian 121 197 161 68 547

Hispanic 24 12 14 2 52

Native American 1 1 1 0 3

Asian/ Pacific Islander 6 5 4 1 16

Other 2 1 2 0 5 Total number of NMDP- facilitated transplants 162 228 200
75 665

Estimated number of new U. S. patients with selected diseases who might
benefit from unrelated stem cell transplants 1,359 662 761 4,081 6,863

Percentage of new patients who receive NMDPfacilitated transplants 12 34
26 2 10

Sources: GAO analysis of data from NMDP and the National Cancer
Institute*s Surveillance, Epidemiology, and End Results program.

Appendix II: How NMDP Achieves Network Compliance with Selected Standards
and Procedures

Page 34 GAO- 03- 182 Bone Marrow Transplants

NMDP requires that the organizations participating in its network comply
with its standards and procedures. This appendix discusses how NMDP
achieves the compliance by network organizations with standards and
procedures for obtaining the informed consent of donors and patients,
donor selection criteria, confidentiality of records, collection and
transportation of marrow, laboratory standards, and maintenance of donor
files in the Registry.

At each stage of the search process, NMDP requires donors to sign informed
consent statements for procedures performed at the donor and transplant
centers. 1 A volunteer must sign an informed consent form before being
listed as a donor on the Registry, and also before the

collection of blood for initial and follow- up testing, infectious disease
testing, and participation in research. In addition, consent must be
obtained before notifying the transplant center that a donor is willing to
proceed to marrow donation and before the administration of anesthesia.
Consent must also be obtained before collecting blood specimens for

research and before any proposed procedure for which the donor has not
previously given consent. According to NMDP officials, during each donor
center site visit, NMDP

staff members review about 35 randomly selected donor files. NMDP staff
members check that each donor has signed all appropriate consent forms for
the stages of the recruitment and search process the donor has completed.
According to an NMDP official, since NMDP began performing site visits in
1998, missing or unsigned donor consent forms occurred in only a few
cases, indicating that a high level of compliance has been achieved. The
number of missing consent forms is not readily available because
cumulative data are not permanently stored. Transplant centers are
responsible for obtaining informed consent from each transplant patient,
for collecting research blood samples that are sent to the NMDP
repository, and for submitting baseline and follow- up data to the
Registry. Some of the centers have separate consent forms specifically for
the research samples and clinical data, whereas others incorporate consent
for the research samples and clinical data into the informed consent
document the patient signs for the transplant.

1 Informed consent refers to the process of helping an individual weigh
the risks against the benefits of a procedure or treatment. By signing a
form, an individual consents to undergo a procedure after being fully
informed of the risks and benefits. Appendix II: How NMDP Achieves Network

Compliance with Selected Standards and Procedures

Informed Consent of Donors and Patients

Appendix II: How NMDP Achieves Network Compliance with Selected Standards
and Procedures

Page 35 GAO- 03- 182 Bone Marrow Transplants

NMDP is currently collecting information on how transplant centers are
handling the informed consent process for the research samples and
clinical data submitted to NMDP. This information will be analyzed, and
NMDP will evaluate whether changes in policies or procedures should be
made to the consent process for obtaining NMDP data and research blood
samples.

In order to be considered for stem cell donation, donors must be from age
18 through 60 and in good health. Individuals with serious illness or
those who are significantly overweight are disqualified. The donor must
provide a medical history and acknowledge in writing that the history is
accurate. Pertinent donor medical information is evaluated for acceptance
or

deferral according to NMDP medical eligibility standards and criteria set
by the medical director at the local donor center. NMDP monitors whether
registered donors have filled out the appropriate medical history
questionnaires, but NMDP does not store cumulative data on the number of
missing medical history questionnaires. During each donor center site
visit, NMDP staff members check a random number of

health history questionnaires. However, NMDP is limited in how it monitors
the donor selection process. Although NMDP tracks the number of donors who
are unavailable for medical reasons, it cannot determine whether an
unavailable donor*s medical condition was preexisting, and therefore
should have been caught in the health screening at the time the donor
volunteered, or whether the donor*s health changed during the period
between registration and a request for testing prior to donation.

NMDP requires that each participating donor center have a system for
safeguarding donor confidentiality. The Registry identifies donors by code
number only. Donor centers maintain donor identity and location and limit

access to this information by using locked file cabinets and locked rooms.
NMDP also requires that each participating transplant center have a system
of confidentiality in place to protect the privacy of patients. It
provides that transplant patient identification should not appear on
papers or publications, and the patient*s name and location should not be

disclosed to the donor( s). Criteria for Donor

Selection Methods to Protect Confidentiality

Appendix II: How NMDP Achieves Network Compliance with Selected Standards
and Procedures

Page 36 GAO- 03- 182 Bone Marrow Transplants

Organizations responsible for marrow collection and transport must meet
certain participation criteria in order to be affiliated with NMDP. Among
other things, participating cord blood banks must be accredited and
licensed or registered by the Food and Drug Administration for collection
of autologous blood. Marrow collection centers must provide emergency and
intensive care services and must be accredited by the Joint Commission on
Accreditation of Healthcare Organizations. In addition, each collection
center must have a licensed medical director, an experienced marrow
collection team that regularly collects bone marrow, and a designated site
for management of collection activities.

NMDP has established standards to ensure the proper collection and
transportation of marrow. These require that bone marrow collection
centers have experienced personnel to collect marrow and adequate
resources to support collection and management activities. In addition,
NMDP requires that collection centers maintain written standard operating
procedures and policies for collecting, testing, labeling, and
transporting marrow.

Laboratories responsible for HLA tissue typing must meet certain criteria
in order to be affiliated with NMDP. Participating HLA typing laboratories
must be accredited by the American Society for Histocompatibility and
Immunogenetics (ASHI) 2 or the European Foundation for Immunogenetics for
techniques required by NMDP. Laboratories must also comply with all state
and federal regulations, including the Clinical Laboratory Improvements
Amendments of 1988 (or their non- U. S. equivalent) for infectious disease
testing, blood typing, red cell antibody screening, and other tests
required by NMDP.

As part of NMDP*s quality control program, participating laboratories must
type blind samples provided by NMDP. The laboratories must maintain
monthly error rates less than or equal to 1.5 percent. If a laboratory
fails to meet quality control and quality assurance standards established
by ASHI or NMDP, NMDP may require that laboratory to submit a corrective
action plan. After the period allowed for corrective action, the
laboratory*s contract with NMDP may be terminated if it still does not
meet the standards.

2 ASHI is an accrediting body that has established standards that all
histocompatibility laboratories must meet if their services are to be
considered acceptable. Marrow Collection

and Transport Laboratory Standards

Appendix II: How NMDP Achieves Network Compliance with Selected Standards
and Procedures

Page 37 GAO- 03- 182 Bone Marrow Transplants

From February 2000 through April 2002, NMDP suspended five laboratories
responsible for HLA tissue typing. The length of suspension ranged from 1
to 9 weeks, and reasons for suspension were related to electronic
communication problems, overdue samples, and poor turnaround time.

NMDP*s central database is updated when new donors are recruited and when
information on existing donors changes or donors are deleted from the
Registry. Information about newly recruited donors includes donor

identification numbers, demographic data, and the donors* HLA types.
According to NMDP procedures, domestic donor centers submit data on donors
daily through NMDP*s central database. Donor File

Maintenance and Updates

Appendix III: Comments from the Health Resources and Services
Administration

Page 38 GAO- 03- 182 Bone Marrow Transplants

Appendix III: Comments from the Health Resources and Services
Administration

Appendix III: Comments from the Health Resources and Services
Administration

Page 39 GAO- 03- 182 Bone Marrow Transplants

Appendix III: Comments from the Health Resources and Services
Administration

Page 40 GAO- 03- 182 Bone Marrow Transplants

Appendix III: Comments from the Health Resources and Services
Administration

Page 41 GAO- 03- 182 Bone Marrow Transplants

Appendix IV: GAO Contact and Staff Acknowledgments

Page 42 GAO- 03- 182 Bone Marrow Transplants

Marcia Crosse, (202) 512- 3407 The following staff members made important
contributions to this work: Donna Bulvin, Charles Davenport, Donald
Keller, Kelly Klemstine, Behn Miller, and Roseanne Price. Appendix IV: GAO
Contact and Staff

Acknowledgments GAO Contact Acknowledgments

(290057)

The General Accounting Office, the investigative arm of Congress, exists
to support Congress in meeting its constitutional responsibilities and to
help improve the performance and accountability of the federal government
for the American people. GAO examines the use of public funds; evaluates
federal programs and policies; and provides analyses, recommendations, and
other assistance to help Congress make informed oversight, policy, and
funding decisions. GAO*s commitment to good government is reflected in its
core values of accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO*s Web site (www. gao. gov) contains abstracts
and fulltext files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as *Today*s Reports,* on its
Web site daily. The list contains links to the full- text document files.
To have GAO e- mail

this list to you every afternoon, go to www. gao. gov and select
*Subscribe to daily E- mail alert for newly released products* under the
GAO Reports heading.

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to: U. S. General Accounting Office 441 G Street NW, Room LM
Washington, D. C. 20548 To order by Phone: Voice: (202) 512- 6000

TDD: (202) 512- 2537 Fax: (202) 512- 6061

Contact: Web site: www. gao. gov/ fraudnet/ fraudnet. htm E- mail:
fraudnet@ gao. gov Automated answering system: (800) 424- 5454 or (202)
512- 7470 Jeff Nelligan, managing director, NelliganJ@ gao. gov (202) 512-
4800

U. S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.
C. 20548 GAO*s Mission Obtaining Copies of

GAO Reports and Testimony

Order by Mail or Phone To Report Fraud, Waste, and Abuse in Federal
Programs Public Affairs
*** End of document. ***