Drug Abuse: Research Shows Treatment Is Effective, but Benefits May Be
Overstated (Letter Report, 03/27/1998, GAO/HEHS-98-72).
Billions of dollars are spent each year to support treatment for drug
abuse and related research. A large number of large, multisite,
longitudinal studies provide evidence that drug abuse treatment is
beneficial, but reliance on self-reported data may overstate treatment
effectiveness. A large number of clients report reductions in drug use
and criminal activity following treatment. The research evidence to
support the relative effectiveness of specific treatment approaches or
settings for particular groups of drug abusers is more varied. Methadone
maintenance has been shown to be the most effective approach to treating
heroin abusers. Research on the best treatment approach or setting for
other groups of drug users, however, is less definitive. For cocaine
abusers, no pharmacological treatment has been found, but several
cognitive-behavioral treatments appear promising. A growing body of
research examining treatment interventions for adolescents suggests that
family-based therapy has potential.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: HEHS-98-72
TITLE: Drug Abuse: Research Shows Treatment Is Effective, but
Benefits May Be Overstated
DATE: 03/27/1998
SUBJECT: Drug abuse
Drug treatment
Block grants
Pharmacological research
Statistical data
Alcohol or drug abuse problems
Public health research
Methadone maintenance
Rehabilitation programs
Narcotics
IDENTIFIER: National Drug and Alcoholism Treatment Unit Survey
DOJ Drug Intervention Program
HHS Knowledge Development and Application Program
Substance Abuse Performance Partnership Grant
NIDA Drug Abuse Treatment Outcome Study
NIDA Treatment Outcome Prospective Study
SAMHSA National Treatment Improvement Evaluation Study
NIH Drug and Underage Alcohol Research Initiative
Medicaid Program
Medicare Program
National Drug Control Strategy
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
GAO/HEHS-98-72
Cover
================================================================ COVER
Report to Congressional Requesters
March 1998
DRUG ABUSE - RESEARCH SHOWS
TREATMENT IS EFFECTIVE, BUT
BENEFITS MAY BE OVERSTATED
GAO/HEHS-98-72
Drug Abuse Treatment
(108345)
Abbreviations
=============================================================== ABBREV
DARP - Drug Abuse Reporting Program
DATOS - Drug Abuse Treatment Outcome Study
HHS - Department of Health and Human Services
HIV - human immunodeficiency virus
LAAM - levo-alpha-acetylmethadol
NIDA - National Institute on Drug Abuse
NIH - National Institutes of Health
NTIES - National Treatment Improvement Evaluation Study
ONDCP - Office of National Drug Control Policy
SAMHSA - Substance Abuse and Mental Health Services Administration
TOPS - Treatment Outcome Prospective Study
VA - Department of Veterans Affairs
Letter
=============================================================== LETTER
B-279062
March 27, 1998
The Honorable Newt Gingrich
House of Representatives
The Honorable J. Dennis Hastert
House of Representatives
The Honorable Rob Portman
House of Representatives
The Honorable Charles B. Rangel
House of Representatives
Each year, the federal government, states, and private entities
invest billions of dollars in programs that combat the use of illicit
drugs. Nevertheless, illicit drug use in the United States remains a
serious and costly problem: It is estimated that in 1996, 13 million
Americans were currently using illicit drugs. The costs of drug
abuse to society--which include costs for health care, drug addiction
prevention and treatment, preventing and fighting drug-related crime,
and lost resources resulting from reduced worker productivity or
death--are estimated at $67 billion annually.
Given the serious health, economic, and social consequences of drug
abuse for the nation, you asked us to report on (1) the level of
federal support for drug abuse treatment activities; (2) the
treatment approaches and settings most commonly used and what is
known about an alternative approach--faith-based treatment; (3)
research issues affecting drug abuse treatment evaluations; and (4)
research findings on the effectiveness of drug treatment overall as
well as what is known about the effectiveness of treatment for
heroin, cocaine, and adolescent drug addiction.
To conduct our work, we reviewed and synthesized the findings on drug
abuse treatment from the largest, most comprehensive studies, which
the National Academy of Science's Institute of Medicine and the drug
treatment research community consider to be the major evaluations of
drug treatment effectiveness. It was beyond the scope of this review
to comprehensively analyze the extensive literature on drug treatment
research methodologies and study results for each specific group of
drug abusers, such as drug-abusing prisoners and women. We did not
independently evaluate the effectiveness of drug treatment programs,
nor did we verify the results reported in the studies we reviewed.
With these limitations, we did our work from October 1997 to February
1998 in accordance with generally accepted government auditing
standards.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
Billions of dollars are spent annually to support treatment for drug
abuse and related research. In 1998, 20 percent of the federal drug
control budget, $3.2 billion, supported drug abuse treatment. Over
half of federal drug treatment funds were allocated to the Department
of Health and Human Services (HHS) to support block grants to the
states, drug treatment services, and related research. An additional
one-third of treatment dollars went to the Department of Veterans
Affairs (VA) to support drug treatment services to veterans and their
inpatient and outpatient medical care. To meet the requirements of
the Government Performance and Results Act, agencies are beginning to
set goals and performance measures to monitor and assess the
effectiveness of federally funded drug treatment efforts.
Treatment services and research aim to reduce the number of current
drug abusers. Experts recognize that not all drug users require
treatment because some do not progress to abuse or dependence. Even
among those who progress to the stage of abuse, some can stop drug
use without treatment. Those who do need treatment can receive
services in a variety of settings and via two major approaches:
pharmacotherapy and behavioral therapy, with many programs combining
elements of both. Other treatment approaches, such as faith-based
strategies, are sometimes used but have not been sufficiently
evaluated to determine their effectiveness.
Measuring the effectiveness of drug abuse treatment is a complex
undertaking. The most comprehensive studies have used an
observational or quasiexperimental design, assessing effectiveness by
measuring drug use before and after treatment. Few studies have used
the most rigorous approach--random assignment to treatment and
control groups--to isolate the particular effects of treatment on
drug abuse. In most studies, the conclusions researchers can draw
are limited by factors such as reliance on self-reported data and the
time frame planned for client follow-up. Furthermore, comparisons of
study results are complicated by differences in how outcomes are
defined and measured and differences in program operations and client
factors.
A number of large, multisite, longitudinal studies provide evidence
that drug abuse treatment is beneficial, but reliance on
self-reported data may overstate treatment effectiveness.
Substantial numbers of clients report reductions in drug use and
criminal activity following treatment. For example, a study of
11,750 people entering drug treatment from 1979 to 1981 found that 40
to 50 percent of regular (weekly or more frequent) heroin and cocaine
users who spent at least 3 months in treatment reported near
abstinence during the year after treatment, and an additional 30
percent reported reduced frequency of use. This study and others
also found that clients who stay in treatment for longer periods
report better outcomes. Research on treatment effectiveness relies
heavily on client reports of drug use. When examining recent drug
use, objective tests, such as urinalysis, consistently identify more
drug users than self-reports do.
The research evidence to support the relative effectiveness of
specific treatment approaches or settings for particular groups of
drug abusers is more varied. Methadone maintenance--the approach
that has been evaluated using the most rigorous studies, randomized
clinical trials--has been shown to be the most effective approach to
treating heroin abusers. Research on the best treatment approach or
setting for other groups of drug abusers, however, is less
definitive. For cocaine abusers, no pharmacological treatment has
been found, but studies have shown that several cognitive-behavioral
treatment approaches have promise. A growing body of research
examining treatment interventions for adolescents indicates that
family-based therapy has potential.
BACKGROUND
------------------------------------------------------------ Letter :2
Recent estimates indicate that illicit drug use in the United States
remains a major problem.\1 In 1996, an estimated 13 million people
were current drug users--that is, they had used illicit drugs in the
past month--which was down from a peak of 25 million in 1979. The
number of current illicit drug users has remained relatively static
since 1992. Marijuana is the most commonly used illicit drug, with
about 10.1 million users in 1996. About half (54 percent) of the
1996 illicit drug users used marijuana only, while another 23 percent
used marijuana and one or more other drugs. The remaining 23 percent
of illicit drug users used only a drug other than marijuana. The
number of current cocaine users declined from 5.7 million people in
1985 to 1.75 million in 1996. The estimated number of crack cocaine
users in 1996 was about 668,000 and has remained steady at about this
level since 1988. However, the use of heroin has been increasing
recently, rising from 68,000 current users in 1993 to 216,000 current
users in 1996.\2
Among 12- to 17-year-old adolescents, current drug use rose from 5.3
percent in 1992 to 10.9 percent in 1995 but declined in 1996 to 9.0
percent. This decline is attributable to reductions in use among
youth aged 12 to 15; for those aged 16 and 17, there was no change in
current use from 1995 to 1996. The rate of marijuana use among
adolescents more than doubled from 1992 to 1995. By 1996, 7.1
percent of adolescents had used marijuana in the past month. The
same year, 0.6 percent of adolescents were current cocaine users, and
0.2 percent were current heroin users. Previous month use of
hallucinogens nearly doubled from 1994 to 1996, from 1.1 percent to 2
percent.
--------------------
\1 The statistical information in this section is from HHS, Substance
Abuse and Mental Health Services Administration, Office of Applied
Studies, Preliminary Results from the 1996 National Household Survey
on Drug Abuse (Washington, D.C.: HHS, July 1997). This is an annual
survey examining drug use patterns and trends within a national
sample of households, civilians living on military bases, and
residents of noninstitutional group quarters (such as shelters,
rooming houses, and dormitories). Survey results are limited by the
exclusion of groups at high risk for drug use and reliance on
self-reported data.
\2 Data from medical examiners, emergency departments, and drug
treatment facilities suggest that methamphetamine abuse may be
growing, although this has not been confirmed in national surveys.
BILLIONS OF FEDERAL DOLLARS
SUPPORT DRUG ABUSE TREATMENT
------------------------------------------------------------ Letter :3
As part of its overall drug control effort, the federal government
provides significant support for activities related to drug abuse
treatment, including grants to states, direct services, and
research.\3 Fiscal year 1998 federal funding for treatment of drug
abuse is approximately $3.2 billion, or about one-fifth of the total
drug control budget. The Congress has authorized HHS and VA to spend
the vast majority of federal drug abuse treatment funds.
--------------------
\3 The data in this section were reported in Executive Office of the
President, Office of National Drug Control Policy (ONDCP), The
National Drug Control Strategy, 1998: Budget Summary (Washington,
D.C.: ONDCP, Feb. 1998). Expenditures include some support for
activities targeting underage alcohol abuse.
ONE-FIFTH OF FEDERAL
SPENDING ON DRUG CONTROL
SUPPORTS TREATMENT
ACTIVITIES
---------------------------------------------------------- Letter :3.1
Federal spending on drug control recognizes four general areas of
emphasis: demand reduction (which includes prevention, treatment,
and related research), domestic law enforcement, interdiction, and
international cooperation. For fiscal year 1998, the federal
government budgeted a total of about $16 billion for drug control
activities.\4 The largest share of this budget--53 percent--supported
domestic law enforcement activities. Drug abuse treatment accounted
for 20 percent and prevention, for 14 percent; the remainder was
allocated to interdiction and international efforts. (See fig. 1.)
The proportion of drug control spending to reduce the demand for
drugs has remained fairly constant since the mid-1980s at about
one-third of the total.
Figure 1: Distribution of
Federal Drug Control Spending,
Fiscal Year 1998
(See figure in printed
edition.)
Source: ONDCP, The National Drug Control Strategy, 1988: Budget
Summary.
Since the early 1990s, federal spending for drug control has grown
steadily. Total federal drug control funding rose by 64 percent,
from about $9.8 billion in 1990 to about $16 billion in 1998. (See
fig. 2.) During this period, the drug treatment budget increased
slightly faster, 78 percent, growing from about $1.8 billion in
fiscal year 1990 to $3.2 billion in fiscal year 1998. An additional
$237 million above the 1998 level was requested for fiscal year 1999
treatment funding.
Figure 2: Total Federal Drug
Control Funding and Drug Abuse
Treatment Funding, Fiscal Years
1990-98
(See figure in printed
edition.)
Source: ONDCP, The National Drug Control Strategy, 1988: Budget
Summary.
--------------------
\4 State, county, and local governments and the private sector also
contribute to annual spending for drug abuse treatment. State,
county, and local governments spent about $1.4 billion in 1994 for
substance abuse treatment. Private funding sources provided about $1
billion in 1993 (the most recent year for which data are available).
The National Drug and Alcoholism Treatment Unit Survey indicated that
over half of the private funding for drug abuse treatment services
consisted of third-party payments by health insurers and health
maintenance organizations, and about 40 percent came from client
fees.
HHS AND VA RECEIVE MOST
FEDERAL FUNDS FOR DRUG ABUSE
TREATMENT ACTIVITIES
---------------------------------------------------------- Letter :3.2
Although a number of federal entities--including the Department of
Justice, the Department of Education, and the Judiciary--receive
treatment-related funding, HHS and VA receive the bulk of federal
drug abuse treatment dollars (see table 1). For fiscal year 1998,
HHS has been authorized to spend about $1.7 billion on drug abuse
treatment--54 percent of all federal treatment dollars. For the same
year, VA has received about $1.1 billion for drug abuse treatment and
related costs, which is 34 percent of the federal treatment budget.
Of the total growth in federal expenditures for drug abuse treatment
between 1994 and 1998--about $557 million--increased funding to VA
accounted for about 44 percent and to HHS, 33 percent.
Table 1
Federal Budget Authority for Drug
Treatment Activities, by Agency, Fiscal
Years 1994-99
(Dollars in Millions)
Fiscal year
----------------------------------------------------------
1999 1994-98
1994 1995 1996 1997 1998 requeste percent
Agency actual actual actual actual enacted d change
------------------- -------- -------- -------- -------- -------- -------- --------
Health and Human $1,532.9 $1,559.5 $1,471.1 $1,660.2 $1,717.7 $1,832.4 12.1
Services
Veterans Affairs\a 853.8 966.1 1,080.9 1,056.4 1,096.9 1,138.7 28.5
Defense 6.2 6.2 5.8 6.4 6.2 6.0 0.0
Education 108.2 115.4 119.2 119.8 125.8 129.9 16.3
Housing and Urban 25.0 5.4 0.0 3.0 3.2 3.2 -87.2
Development
The Judiciary 63.0 67.8 59.7 69.6 74.7 83.8 18.6
Justice 61.3 68.3 94.8 126.6 160.9 219.7 162.5
ONDCP 1.9 15.6 5.1 26.9 23.8 32.3 1,152.6
=========================================================================================
Total\b $2,652.3 $2,953.2 $2,836.6 $3,068.9 $3,209.1 $3,446.0 21
\c
-----------------------------------------------------------------------------------------
\a Includes 100 percent of medical costs provided to veterans with a
diagnosis of drug abuse when treatment is provided in a specialized
drug or substance abuse treatment program. For veterans with a
secondary or associated diagnosis of drug abuse who receive care in
other settings, only a proportion of medical costs are included.
\b Expenditures have been rounded, affecting percentages and totals.
\c Includes $148.9 million for the Social Security Administration.
Source: Compiled from data in ONDCP, The National Drug Control
Strategy, 1998: Budget Summary.
Of HHS' $1.7 billion drug treatment budget for 1998, more than half
($944 million) was dedicated to the Substance Abuse and Mental Health
Services Administration (SAMHSA) to support the treatment components
of its Substance Abuse Performance Partnership Grants to states and
the Knowledge Development and Application Program.\5 Approximately 80
percent of SAMHSA's total budget is distributed to the states through
block grants and formula programs. SAMHSA has requested an increase
of $143 million in fiscal year 1999 Substance Abuse Performance
Partnership Grants funding to make treatment available to more of
those who need it.
The Health Care Financing Administration received $360 million in
fiscal year 1998 to pay for drug abuse treatment services for
Medicaid and Medicare beneficiaries. Eighty percent of this amount
finances Medicaid treatment expenses, including all covered hospital
and nonhospital services required. The remaining 20 percent covers
Medicare hospital insurance treatment costs.
The National Institutes of Health (NIH) received about one-sixth of
HHS's drug treatment funds to conduct research in the areas of drug
abuse and underage alcohol use. For 1999, NIH has requested funding
($51 million) for its Drug and Underage Alcohol Research initiative
to expand research on underage alcohol and drug addiction among
children and adolescents, as well as chronic drug users, and to
support increased dissemination of research findings. Recognizing
the need to improve research on the infrastructure that delivers
treatment, the Congress mandated in 1992 that the National Institute
on Drug Abuse (NIDA) obligate at least 15 percent of its funding to
support research on the impact of the organization, financing, and
management of health services on issues such as access and quality of
services.
In 1998, VA was appropriated about $1.1 billion for inpatient and
outpatient medical care provided to veterans with a diagnosis of drug
abuse, as well as for drug abuse treatment services. Special
substance abuse treatment services are available at 126 medical
facilities. Additional monies support treatment research in
coordination with NIDA.
Other federal agencies that received drug treatment funds for fiscal
year 1998 include the Departments of Education and Justice (each
received more than $100 million), the federal Judiciary (about $75
million), and ONDCP (about $24 million). From 1994 to 1998,
Justice's funding rose 163 percent; moreover, its 1999 funding
request would increase its funding by another third. The Department
of Justice has requested about $83 million for fiscal year 1999 to
support its Drug Intervention Program, a new program that would
support drug testing, treatment, and graduated sanctions for drug
offenders, in an effort to break the cycle of drug abuse and
violence.
The Government Performance and Results Act was enacted in 1993 in
part as a means to improve performance measurement by federal
agencies. It requires agencies to set goals, measure performance,
and report on their accomplishments and thus should provide a useful
framework for assessing the effectiveness of federally funded drug
treatment efforts. However, demonstrating the efficient and
effective use of federal drug abuse treatment funds is particularly
challenging because most of these funds support services provided by
state and local grantees, which are given broad discretion in how
best to use them.\6 Regardless, federal agencies are now required by
the Results Act to hold states accountable for achieving federal
goals for effective treatment outcomes.
--------------------
\5 HHS expenditures cited here do not include activities principally
targeting treatment for alcohol abuse. However, for the National
Institutes of Health, underage alcohol treatment activities are
included in total spending.
\6 Substance Abuse and Mental Health: Reauthorization Issues Facing
the Substance Abuse and Mental Health Services Administration
(GAO/T-HEHS-97-135, May 22, 1997).
DRUG TREATMENT SERVICES ARE
PROVIDED IN A VARIETY OF
SETTINGS
------------------------------------------------------------ Letter :4
Drug addiction is a complicated disorder that includes physiological,
behavioral, and psychological aspects. For example, the
environmental cues that have been associated with drug use can
trigger craving and precipitate relapse, even after long periods of
abstinence. Despite the potential for relapse to drug use, not all
drug users require treatment services to discontinue use. For those
who do require treatment, services may be provided in either
outpatient or inpatient settings, and via two major approaches:
pharmacotherapy and behavioral therapy, with many programs combining
elements of both. Other treatment approaches, such as faith-based
strategies, have yet to be rigorously examined by the research
community.
NATURE OF DRUG ABUSE
---------------------------------------------------------- Letter :4.1
In general, drug abuse is defined by the level and pattern of drug
consumption and the severity and persistence of resulting functional
problems. A diagnosis of drug abuse is generally made when drug use
has led to social, legal, or interpersonal problems. A clinical
diagnosis of drug dependence--or addiction--is based on a group of
criteria including physiological, behavioral, and cognitive
factors.\7 In particular, drug addiction is characterized by
compulsive drug-seeking behavior. People who are dependent on drugs
often use multiple drugs and usually have substantial impairment of
health and social functioning. Furthermore, addiction is generally
accompanied by withdrawal symptoms and drug tolerance, resulting in
the need to increase the amount of drugs consumed.
Moreover, severe dependence is often associated with health
conditions or impairments in social functioning, including mental
health disorders that generally are serious and difficult to treat.
Drug abusers are more likely than nonabusers to sustain injuries; be
involved in violence and illegal activities; have chronic health
problems, including a higher risk of contracting HIV (human
immunodeficiency virus); and have difficulty holding a job.
Most scientists agree that addiction is the result of chemical and
physical changes in the brain caused by drug use. However, they
recognize that addiction extends beyond physiological components to
include significant behavioral and psychological aspects. For
example, specific environmental cues that a drug abuser associates
with drug use can trigger craving and precipitate relapse, even after
long periods of abstinence. Therefore, people receiving treatment
for drug abuse often enter treatment a number of times--sometimes in
different approaches or settings, and sometimes in the same approach
or even the same treatment facility. Often, the substance abuser
reduces his or her drug use incrementally with each treatment
episode.
Experts recognize that not all drug users require treatment to forgo
drug use because some drug users do not progress to abuse or
dependence. Even among those who progress to the stage of abuse,
some can stop drug use without treatment. This issue was addressed
in a study of Vietnam veterans' rapid recovery from heroin addiction.
Forty-five percent of enlisted Army men had tried narcotics in
Vietnam, and 20 percent reported the development of an addiction to
narcotics. However, in the first year after their return home, only
5 percent of those addicted in Vietnam remained addicted in the
United States. The author concluded that most addictions are
relatively brief, and that most drug abusers are capable of
discontinuing drug use without treatment.\8 This view is
controversial; others contend that the Vietnam veterans' experience
is an anomaly resulting from the drastic change in environment when
they returned home.
--------------------
\7 For diagnostic criteria, physicians use the Diagnostic and
Statistical Manual of Mental Disorders, DSM-IV (American Psychiatric
Association, 1994), and the International Classification of Diseases,
ICD-10 (World Health Organization, 1992). Both of these diagnostic
manuals recognize drug abuse and dependence as medical disorders.
\8 Lee N. Robins, "Vietnam Veterans' Rapid Recovery From Heroin
Addiction: A Fluke or Normal Expectation?" Addiction, Vol. 88
(1993), pp. 1041-54.
DRUG ABUSE TREATMENT
APPROACHES AND SETTINGS
---------------------------------------------------------- Letter :4.2
Data from 1992-93 on use of drug treatment in the United States (the
most current available) show that about 1.4 million people received
drug treatment during the previous year.\9 According to SAMHSA, the
individuals in drug treatment were those with the most extreme
patterns of drug use: the highest frequency of drug use, use of the
least typical drug types, and early initiation of use. Most of the
group in treatment had received treatment in multiple settings, most
commonly in drug treatment facilities and self-help groups. Only
about one-fourth of those who needed drug treatment in the previous
year reported having received it during that year.\10 Adolescents
(aged 12 to 17) were even less likely to receive needed treatment,
with 18 percent of those needing treatment receiving it.
The treatment of drug addiction can be classified under two major
approaches: pharmacotherapy and behavioral therapy. Pharmacotherapy
relies on medications to block the euphoric effects or manage the
withdrawal symptoms and cravings experienced with illicit drug use.
One such widely used medication is methadone, a narcotic analgesic
that blocks the euphoria of heroin, morphine, codeine, and other
opiate drugs and suppresses withdrawal symptoms and craving between
treatment doses. Methadone maintenance generally requires daily
clinic visits to receive the methadone dose; over time, some clients
are given take-home doses. Methadone maintenance can continue for as
long as several years, and in some cases, maintenance may last a
lifetime.
A number of other drugs have also been shown to be safe and
efficacious in the treatment of opiate addiction.
Levo-alpha-acetylmethadol (LAAM) suppresses withdrawal symptoms for
72 to 96 hours and thus can reduce clients' clinic visits to 3 days
per week. Naltrexone, like LAAM, is long-acting and can be
administered in small daily doses or in larger doses 3 times a week.
Naltrexone is believed to be most effective for highly motivated
clients, especially those with strong social supports. Buprenorphine
has been effective in clinical trials in retaining patients in
treatment and facilitating abstinence. In addition, buprenorphine
has been shown to produce less physical dependence than methadone and
LAAM.
Behavioral therapy includes various forms of psychotherapy,
contingency-based therapy, cognitive therapy, and other types of
therapies. It may include skills training and a variety of
counseling approaches, from highly structured individual or family
counseling to more informal group counseling. Some programs combine
elements of both pharmacotherapy and behavioral therapy. For
example, many methadone maintenance programs are designed to also
provide counseling services, which may include psychotherapy or
individualized social assistance. Participation in counseling
facilitates regular monitoring of client behavior, appearance, and
drug use. Some outpatient nonmethadone programs also use
pharmacological treatment, such as medications for initial
detoxification, medications to control craving, or drugs that address
psychiatric disorders such as depression or schizophrenia. Drug
abusers receiving pharmacotherapy, behavioral therapy, or both may
also participate in self-help groups, such as Alcoholics Anonymous,
Narcotics Anonymous, or Rational Recovery and are generally
encouraged to continue participation in these groups after leaving
formal treatment to help maintain abstinence and a healthy lifestyle.
A number of other, less commonly used approaches to drug treatment
offer alternatives to these established approaches. One such example
is the use of spirituality as a component of treatment. Some
researchers have acknowledged that people with a strong spiritual\11
or religious involvement seem to be at lower risk for substance
abuse, yet research in this area remains extremely limited. Experts
have yet to agree on how to define faith-based drug treatment. Some
define faith-based programs as those that are based on religious
beliefs and practices, such as Teen Challenge,\12 while others
consider any treatment approach that recognizes spirituality, such as
Narcotics Anonymous or Cocaine Anonymous, to be faith-based.
Regardless of how faith-based treatment is defined, there has not
been sufficient research to determine the results of this type of
treatment. For example, a recent research conference assessed the
evidence on spiritual treatment for alcohol and drug abuse.\13 The
panel found strong evidence for a few limited assertions: that
better treatment outcomes correlate with Alcoholics Anonymous
involvement after outpatient treatment and that meditation-based
interventions are associated with reduced levels of alcohol and drug
use. The panel concluded that the issues for future research in this
area include the definition and measurement of spiritual variables
and the possible spiritual factors that could play a role in recovery
from substance abuse.
Regardless of the approach used, drug treatment services are provided
in both inpatient and outpatient settings. Most people are served by
outpatient programs, where treatment can vary from psychotherapy at
comprehensive health centers to informal group discussions at drop-in
centers.\14 People who enter outpatient drug-free treatment generally
(though not always) have a less severe level of addiction and
associated problems than those who receive treatment in inpatient
settings. Although weekly counseling is the predominate treatment
approach available at outpatient settings, some programs also offer
pharmacological treatment and some give assistance with social needs,
including education, job training, housing, and health care.\15
Inpatient settings include hospitals as well as residential
facilities, such as therapeutic communities. Hospital-based drug
treatment is used for detoxification from drugs and to provide other
services for individuals having severe medical or psychiatric
complications. Data from 1992-93 show that, of the group reporting
drug treatment during the past year, 28 percent received treatment in
an inpatient hospital setting. Chemical dependency programs, one
type of inpatient treatment program, recognize drug problems as
having multiple causes, including physiological, psychological, and
sociocultural aspects. Treatment may last up to several weeks and
may include pharmacological intervention, education about drug
addiction, counseling, participation in self-help groups, and medical
or psychiatric services.
Long-term residential treatment programs are designed for people with
more severe drug problems--those with dependence on one or more drugs
who have failed previous treatment efforts. For example, therapeutic
communities provide treatment that is generally planned for 6 to 12
months in a residential setting. Clients are generally chronic drug
abusers who have failed at other forms of drug abuse treatment, while
staff are largely previous drug abusers. Strict behavioral
expectations and responsibilities are enforced to emphasize
appropriate social and vocational norms.
--------------------
\9 Data were collected as part of the National Household Survey on
Drug Abuse in 1992 and 1993. See Dean R. Gerstein, Mary L. Foote,
and Rashna Ghadialy, National Opinion Research Center, The Prevalence
and Correlates of Treatment for Drug Problems (Washington, D.C.:
HHS, SAMHSA, Office of Applied Studies, Apr. 1997.)
\10 A person who met at least one of the following criteria was
classified as needing treatment: dependence, needle use, having
received treatment, and heavy use.
\11 SAMHSA defines spirituality as involvement in socially desirable
activities or processes that are beyond the details of daily life and
personal self-interest. According to SAMHSA, ethical behavior,
consideration for the interests of others, community involvement,
helping others, and participating in organized religion are all ways
in which spirituality can be expressed. See "Matching Treatment to
Patient Needs in Opiate Substitution Therapy," Treatment Improvement
Protocol Series, No. 20 (Washington, D.C.: HHS, SAMHSA, Center for
Substance Abuse Treatment, 1995).
\12 Teen Challenge is a drug-free residential treatment program for
drug abusers. Founded in 1961, Teen Challenge bases its drug
treatment approach on the belief that only devotion to Jesus Christ
can form the basis for the development of a healthy personality and
satisfying life. Through centers across the country, adolescents and
adults receive drug detoxification, followed by rehabilitation and
training. The program employs a strict system of rules and
activities to build self-discipline and personal responsibility and
Bible study to encourage spiritual growth and development.
\13 "Scientific Progress in Spiritual Research," a three-part series
from July 1996 to July 1997, sponsored by the National Institute for
Healthcare Research.
\14 Typically called outpatient drug-free or outpatient nonmethadone
treatment, this type of approach encompass all outpatient treatment
except methadone maintenance. Although called drug-free, outpatient
programs can include pharmacological treatment, such as desipramine
to moderate cocaine craving and clonidine to treat withdrawal from
narcotics.
\15 The Institute of Medicine reported in 1995 that there are
opiate-dependent patients in outpatient drug-free treatment settings.
Research from the early 1990s found that 10 percent of clients in
outpatient drug-free settings reported having used opiates within the
past 30 days. See Institute of Medicine, The Development of
Medications for the Treatment of Opiate and Cocaine Addictions
(Washington, D.C.: Institute of Medicine, 1995), p. 99.
RESEARCH ISSUES MAKE ASSESSMENT
OF TREATMENT EFFECTIVENESS
DIFFICULT
------------------------------------------------------------ Letter :5
The study of drug treatment programs is complicated by a number of
challenging methodological and implementation issues. Evaluations of
treatment effectiveness can use one of several methodologies,
depending on the specific questions to be addressed. Thus, the
appropriateness of the study design and how well the evaluation is
conducted determine the confidence to be placed in the research
findings. In particular, studies of the validity of self-reported
data demonstrate that information on treatment outcomes collected by
self-report should be interpreted with some caution. The ability to
compare the results of effectiveness studies is also influenced, and
often limited, by differences in how outcomes are measured, how
programs are operated, and client variables.
QUALITY OF EVIDENCE VARIES
BY STUDY DESIGN
---------------------------------------------------------- Letter :5.1
Drug treatment effectiveness research conducted over the past 2
decades has used a variety of designs, including randomized clinical
trials, simple or controlled observation, and quasiexperimental
designs.\16 Selection of the study design depends on a number of
factors, including the questions being addressed and the resources
available to fund the study. Methodologists agree that randomized
clinical trials are the most rigorous study designs and therefore
offer the strongest support for their findings. Studies that rely on
a simple observational design produce less definitive findings but
can provide a good indication of the operation of drug treatment
programs as well as information on treatment outcomes. A
quasiexperimental design, the most frequently used in field settings,
falls somewhere in between.
Randomized clinical studies are designed to isolate the effects of a
treatment by randomly assigning individuals to either a control
group--receiving no treatment or an alternative treatment--or to a
group that receives the treatment being studied. This study design
has been used in the assessment of methadone maintenance for treating
heroin addiction. Randomized trials are often used to study the
efficacy of a treatment, asking the question, "Can it work?" Although
such studies provide the most definitive information about whether
particular treatments are effective, they are not widely used in drug
treatment research. According to an analysis by the Lewin Group,
among the reasons cited for the limited use of randomized trials are
the difficulties in obtaining informed consent from drug abusers and
the perceived ethical issue of randomly assigning people who are
seeking drug treatment to a control group in which no treatment or a
treatment regimen not of the client's choice is provided.\17
Simple and controlled observation designs typically employ a
repeated-measures methodology, whereby the researchers collect
information on drug use patterns and other criteria from clients
before, during, and after treatment. Generally, controlled
observation studies examine multiple treatment groups, and simple
observation studies follow a single treatment group without a
nontreatment comparison group. Observational studies provide
information about the effectiveness of treatments when implemented in
uncontrolled, or real-world, conditions. Observational design has
been used to assess treatment provided in all four of the major
treatment settings: residential therapeutic communities and
outpatient methadone maintenance, outpatient drug-free, and inpatient
chemical dependency programs.
Quasiexperimental study designs generally have a comparison group, a
key feature of strong research design, but an investigator does not
randomly assign individuals to treatment and comparison groups.
Instead, comparisons are made between possibly nonequivalent client
groups or by using statistical techniques that adjust for known
differences in client characteristics. Even in a quasiexperimental
design, a repeated-measures methodology might be used in comparing
the behaviors of the same group of drug abusers before, during, and
after treatment. A quasiexperimental design is often applied in
evaluations of naturally occurring events, such as introducing a new
treatment approach or closing a treatment program. Such a design
allows greater confidence (than observation alone) that any
differences detected are due to treatment but not as much confidence
as random assignment of clients to treatment and comparison groups.
Quasiexperimental study designs have been used to assess the
effectiveness of both methadone maintenance programs and therapeutic
communities as well as outpatient drug-free programs.
--------------------
\16 One distinction among these methodologies is the analytic
techniques that often are used. Quasiexperimental designs and
controlled observations usually use multivariate analysis, whereas
simple observations usually use only univariate or bivariate
analysis. See Institute of Medicine, Treating Drug Problems
(Washington, D.C.: Institute of Medicine, 1990), p. 186.
\17 Henrick J. Harwood, Sharon L. Carothers, and Christine Lee,
Random Assignment Studies of Drug Abuse Treatment: Progress in the
Application to Psychosocial Components of Care, Report to the
National Institute on Drug Abuse (Fairfax, Va.: The Lewin Group,
Inc., 1994).
TREATMENT EVALUATIONS DEFINE
AND MEASURE OUTCOMES
DIFFERENTLY
---------------------------------------------------------- Letter :5.2
Treatment program goals generally include a wide range of issues,
such as reducing drug use, reducing criminal behavior, and improving
employment status. Most researchers have agreed that reducing drug
use from the level it would have been without treatment (harm
reduction) is a valid goal of drug treatment and an indication of
program success. In addressing this issue, researchers acknowledge
that abstinence from illicit drug consumption is the central goal of
all drug treatment, but they contend it is not the only acceptable
goal of treatment, since total abstinence from drug use may be
unrealistic for many users. According to the Institute of Medicine,
"an extended abstinence, even if punctuated by slips and short
relapses, is beneficial in itself and may serve as a critical
intermediate step toward lifetime abstinence and recovery."\18
Even with harm reduction as the common objective, treatment outcome
measures vary among--and sometimes within--treatment programs.
Operationalizing the outcome measures is also done differently across
programs, which makes it difficult to compare treatment outcomes of
different programs. For example, one program may measure reduction
in drug use by examining the frequency of drug use, while another may
choose to focus on reduced relapse time. Major drug treatment
studies use other outcomes as well to measure treatment
effectiveness, ranging from reductions in criminal activity to
increased productivity. Indicators for these outcome measures also
vary by study. (See table 2.)
Table 2
Types of Outcome Measures Used to Assess
Effectiveness of Drug Treatment
Indicator
Outcome measure -----------------------------------------------------
Drug use Reduced frequency of drug use
Reduced amount of drug used
Reduced relapse time
Abstinence
Criminal activity Fewer arrests
Fewer convictions
Parole or probation status
Health and safety Improved medical status and general improvement in
health (for example, fewer hospitalizations and
doctor and emergency room visits)
Improved mental health status (for example, improved
mood, cognition, and personality traits and fewer
psychotic states)
Improved behavior associated with risk of HIV
infection
Improved public safety (for example, incidence of
drug-related fires, car crashes, accidents, and
trauma)
Social and interpersonal Positive changes in social values and networks
skills
Improved relationships with family, friends, and
employers
Productivity Increased days of employment
Enrollment in training program or school
Increased school attendance
Improved grades and overall performance
--------------------------------------------------------------------------------
Another issue related to measuring treatment outcomes is concern
about the time frame for client follow-up. Since drug addiction is
commonly viewed as a life-long disease, many argue that long-term
follow-up is needed to fully assess treatment outcomes. However,
many of those who complete treatment programs are lost in the
follow-up assessment period. Treatment assessment periods vary
considerably, ranging from a 1-year follow-up for most studies to a
12-year follow-up for a subset of clients in one of the major studies
we reviewed. The research literature indicates difficulties in
tracking drug abusers even for 1-year follow-up periods. For
example, of the group selected for follow-up interviews in the Drug
Abuse Treatment Outcome Study (DATOS), only 70 percent actually
completed the interview.
--------------------
\18 Institute of Medicine, Treating Drug Problems, p. 129.
RELIANCE ON SELF-REPORTED
DATA HAS LIMITATIONS
---------------------------------------------------------- Letter :5.3
With all types of study designs, data collection issues can hamper
assessments of treatment effectiveness. The central debate regarding
data collection on the use of illicit drugs surrounds the common use
of self-reported data. A recent NIDA review of current research on
the validity of self-reported drug use highlights the limitations of
data collected in this manner.\19 According to this review, recent
studies conducted with criminal justice clients (such as people on
parole, on probation, or awaiting trial) and former treatment clients
suggest that 50 percent or fewer current users accurately report
their drug use in confidential interviews. In general, self-reports
are less valid for the more stigmatized drugs, such as cocaine; for
more recent rather than past use; and for those involved with the
criminal justice system.
The largest studies of treatment effectiveness, which have evaluated
the progress of thousands of people in drug treatment programs, have
all relied on self-reported data. That is, the drug abuser is
surveyed when entering treatment, and then again at a specified
follow-up interval. In general, individuals are asked, orally or in
writing, to report their drug use patterns during the previous year.
Self-reports of drug use may be subject to bias both prior to and
following treatment and can be either over- or understated. Drug
abusers may inflate their current level of drug use when presenting
for treatment if they believe that higher levels of use will increase
the likelihood of acceptance into treatment. Drug use may also be
underreported at treatment intake or follow-up. Motivations cited
for underreporting include the client's desire to reflect a positive
outcome from treatment and the perception of a strong societal stigma
associated with the use of particular drugs.
As questions have developed about the accuracy of self-reported
data,\20
researchers have begun using objective means to validate the data
collected in this manner, although these methods also have
limitations. Generally, a subgroup of the individuals surveyed after
treatment is asked to provide either a urine sample or a hair sample,
which is then screened for evidence of drug use. The results from
the urinalysis or hair analysis are then compared against
self-reports of drug use. Some researchers believe that it may be
possible to systematically adjust self-reported data to correct for
the biases exposed by urinalysis or hair analysis, although this
technique is not currently in use.
Recent major studies of drug treatment effectiveness have used
urinalysis to validate self-reported data. For example, the National
Treatment Improvement Evaluation Study (NTIES) found that
self-reports of recent drug use (in the past 30 days) for opiates and
cocaine were lower than current drug use as revealed by urinalysis.
However, the self-reports of substance use over the entire follow-up
period (that is, use on at least five occasions) yielded an
equivalent or higher rate of use than the results of analyzing urine
specimens collected at the follow-up interview. (See table 3.) Other
studies found similar underreporting of drug use. The Treatment
Outcome Prospective Study (TOPS), which followed people entering
treatment in the early 1980s, reported that 40 percent of the
individuals testing positive for cocaine 24 months after treatment
had reported using the drug in the previous 3 days.
Table 3
Comparison of Drug-Positive Urine Tests
With Self-Reported Drug Use, NTIES Study
Cocaine Opiates
Data collection method at (including (including
follow-up crack) heroin)
------------------------------ ------------ ------------
Urine test 28.7% 16.2%
Self-report of use in last 30 20.4 11.3
days
Self-report of use during 33.5 16.5
period since treatment
----------------------------------------------------------
Source: National Opinion Research Center at the University of
Chicago and the Research Triangle Institute, NTIES Brief Report
(Washington, D.C.: HHS, SAMHSA, Center for Substance Abuse
Treatment, Feb. 1997).
Despite the discrepancies observed, each of the data collection
methods used to measure treatment effectiveness has particular
weaknesses. As shown above, validation studies indicate that
self-reports of current drug use underreport drug use. At the same
time, researchers emphasize that client reporting on use of illicit
drugs during the previous year (the outcome measure used in most
effectiveness evaluations) has been shown to be more accurate than
reporting on current drug use. In comparison, urine tests can
accurately detect illicit drugs for about 48 hours following drug
use. However, urinalysis does not provide any information about drug
use during the previous year. In addition, individual differences in
metabolism rates can affect the outcomes of urinalysis tests. Hair
analysis has received attention because it can detect drug use over a
longer time--up to several months. However, unresolved issues in
hair testing include variability across drugs in the accuracy of
detection, the potential for passive contamination, and the relative
effect of different hair color or type on cocaine accumulation in the
hair.\21
To examine the validity of self-reported data on other outcome
measures, NTIES researchers compared self-reports on arrests to
official arrest records and found 80 percent agreement, with
underreporting of arrest histories most frequent among individuals
interviewed in prison or jail and among men under 25 years of age.
Researchers also compared self-reports of treatment completion,
primary drug use, and demographic data with program records and found
high levels of concordance between records and individual
self-reports; for example, 92 percent agreed on whether a client
completed the prescribed treatment.
--------------------
\19 HHS, NIH, The Validity of Self-Reported Drug Use: Improving the
Accuracy of Survey Estimates, National Institute on Drug Abuse
Research Monograph Series 167 (Washington, D.C.: HHS, 1997).
\20 The research literature prior to the mid-1980s showed drug use
self-reports to be generally valid, while studies conducted since
then have raised concerns about validity. The apparent change in
validity may be due in part to improved urinalysis testing that now
detects drug use more accurately. It is also possible that
individuals were more willing to admit use of illicit drugs in the
past, when societal reaction toward drug use was not as strong as it
is today. Even today, researchers are not in agreement on the
limitations of self-reported data. For example, the researchers for
DATOS, the most recently completed study of drug treatment,
acknowledged limitations to self-reported data but asserted that most
data so obtained are reasonably reliable and valid.
\21 We have reported on the limitations of using self-reported data
in estimating the prevalence of drug use. We concluded that hair
testing merited further evaluation as a confirmatory measure. See
Drug Use Measurement: Strengths, Limitations, and Recommendations
for Improvement (GAO/PEMD-93-18, June 25, 1993).
VARIATION IN PROGRAM
OPERATIONS AND CLIENT
FACTORS MAKES COMPARISONS
DIFFICULT
---------------------------------------------------------- Letter :5.4
Research results often do not account for the tremendous variation in
program operations, such as differences in standards of treatment,
staff levels and expertise, and level of coordination with other
services. For example, surveys of the dosages used in methadone
maintenance programs have shown that a large proportion of programs
use suboptimal or even subthreshold dosages, which would likely
result in poorer treatment outcomes than those of programs that
provide optimal dosage levels to their clients. Similarly,
outpatient drug-free programs operate with different numbers and
quality of staff and have varying levels of coordination with local
agencies that offer related services that are generally needed to
support recovering abusers. An outpatient drug-free program that has
close ties with local services, such as health clinics and job
training programs, is likely to have better treatment outcomes than a
program without such ties.
Assessing treatment effectiveness is also complicated by differences
in client factors. Researchers recognize that client motivation and
readiness for treatment, as well as psychiatric status, can
significantly affect the patient's performance in treatment. For
example, unmotivated clients are less likely than motivated ones to
adhere to program protocols and to continue treatment. In studies of
pharmacotherapy for opiate addiction, researchers have found that
patients with high motivation to remain drug-free--such as health
professionals, parolees, and work-release participants--have better
treatment outcomes.\22
--------------------
\22 Much research has focused on the issue of matching patients to
treatment with the goal of providing the most appropriate and highest
quality care, while maximizing cost-efficiency by providing the least
expensive effective treatment. However, research suggests that
patient variables such as sociodemographics and drug use history have
not been predictive of success in one treatment versus another. More
recent research is focusing on program factors, including the
services provided and the ability of programs to retain clients in
treatment, since longer treatment episodes have consistently been
associated with better outcomes.
STUDIES INDICATE BENEFITS FROM
TREATMENT, BUT EVIDENCE VARIES
ON BEST APPROACHES FOR SPECIFIC
GROUPS
------------------------------------------------------------ Letter :6
Major studies have shown that drug treatment is beneficial, although
concerns about the validity of self-reported data suggest that the
degree of success may be overstated. In large-scale evaluations
conducted over the past 20 years, researchers have concluded that
treatment reduces the number of regular drug users as well as
criminal activity. In addition, these studies demonstrate that
longer treatment episodes are more effective than shorter ones.
Research also indicates that the amount and strength of evidence
available to support particular treatment approaches for specific
groups of drug abusers vary.
CONSISTENT EVIDENCE SHOWS
DRUG TREATMENT IS
BENEFICIAL, BUT OUTCOMES MAY
BE OVERSTATED
---------------------------------------------------------- Letter :6.1
Numerous large-scale studies that examined the outcomes of treatment
provided in a variety of settings have found drug treatment to be
beneficial. Clients receiving treatment report reductions in drug
use and criminal activity, with better treatment outcomes associated
with longer treatment duration. However, studies examining the
validity of self-reported data suggest that a large proportion of
individuals do not report the full extent of drug use following
treatment. Therefore, the findings from these major studies of
treatment effectiveness--all of which relied on self-reported data as
the primary data collection method--may be somewhat inflated.
MAJOR STUDIES REPORT
REDUCTIONS IN DRUG USE
AND CRIME FOLLOWING
TREATMENT
-------------------------------------------------------- Letter :6.1.1
Comprehensive analyses of the effectiveness of drug treatment have
been conducted by several major studies over a period of nearly 30
years: DATOS, NTIES, TOPS, and the Drug Abuse Reporting Program
(DARP) (see table 4). These large, multisite studies were designed
to assess drug abusers on several measures before, during, and after
treatment. These studies are generally considered by the Institute
of Medicine and the drug treatment research community to be the major
evaluations of drug treatment effectiveness, and much of what is
known about typical drug abuse treatment outcomes comes from these
studies.
Table 4
Characteristics of Major Drug Treatment
Studies
Follow-
up
interval
Number Number after Treatment Research
Study of of treatmen approach/ organization(s Research
Study period\a programs clients t setting\b ) sponsor
------- -------- -------- -------- -------- -------------- -------------- --------
DATOS 1991-93 99 10,010 1 year\c R(LT), MM, Research NIDA
ODF, R(ST) Triangle
Institute
NTIES 1993-95 78\d 4,411 1 year MM, C, R(LT), National SAMHSA
R(ST), ODF Opinion
Research
Center at the
University of
Chicago;
Research
Triangle
Institute
TOPS 1979-81 41 11,750 1 year, R(LT), MM, ODF Research NIDA
2 years, Triangle
3-5 Institute
years
DARP 1969-73 52 44,000 1 year, TC, MM, ODF, D Institute of National
then 3- Behavioral Institut
12 years Research at e of
Texas Mental
Christian Health\e
University
-----------------------------------------------------------------------------------------
\a Clients were accepted into the study program during these years.
\b Key: C = correctional, D = detoxification, MM = methadone
maintenance, ODF = outpatient drug-free, R(LT) = long-term
residential, R(ST) = short-term residential, RSM = residential
"social model," and TC = therapeutic community.
\c Additional follow-up to determine long-term outcomes is planned
through the DATOS cooperative, a set of agreements between NIDA and
three collaborating research sites.
\d Includes evaluations of facilities supported by the Center for
Substance Abuse Treatment within SAMHSA.
\e DARP was transferred to NIDA when that agency was created in 1974.
Source: See appendix for bibliographic references.
These federally funded studies were conducted by research
organizations independent of the groups operating the treatment
programs being assessed. Although the characteristics of the studies
vary somewhat, all are based on observational or quasiexperimental
designs. The most recently completed study, DATOS, is a longitudinal
study that used a prospective design and a repeated-measures
methodology to study the complex interactions of client
characteristics and treatment elements as they occur in typical
community-based programs.\23 NTIES, completed in March 1997, was a
congressionally mandated, 5-year study that examined the
effectiveness of treatment provided in public programs supported by
SAMHSA.
All of these studies relied on self-report as the primary data
collection method. That is, drug abusers were interviewed prior to
entering treatment and again following treatment, and asked to report
on their use of illicit drugs, their involvement in criminal
activity, and other drug-related behaviors. As described previously
in this report, studies examining the validity of self-reported data
suggest that many individuals do not report the full extent of drug
use following treatment.\24 Since results from the major studies of
treatment effectiveness were not adjusted for the likelihood of
underreported drug use (as revealed by urinalysis substudies), the
study results that follow may overstate reductions in drug use
achieved by drug abusers. Researchers contend that the bias in
self-reports on current drug use is greater than the bias in
self-reports on past year use and that therefore the overall findings
of treatment benefits are still valid.
Each of these major studies attributed benefits to drug treatment
when outcomes were assessed 1 year after treatment. They found that
reported drug use declined when clients received treatment from any
of three drug treatment approaches--residential long-term, outpatient
drug-free, or outpatient methadone maintenance--regardless of the
drug and client type.\25 As shown in table 5, DATOS, the study most
recently completed, found that the percentage of individuals
reporting weekly or more frequent drug use or criminal activity
declined following treatment.
Table 5
Percentage of DATOS Clients Reporting
Regular Drug Use and Criminal Activity
Before and After Treatment
Year
Year prior following
to treatment treatment
------------------------------ ------------ ------------
Heroin users in outpatient 89.4 27.8
methadone treatment\a
Cocaine users in long-term 66.4 22.1
residential treatment
Cocaine users in outpatient 41.7 18.3
drug-free treatment
Predatory illegal activity by 40.5 15.9
clients in long-term
residential treatment\b
----------------------------------------------------------
Note: In separate multivariate analyses, lower levels of drug use
and crime for people in treatment for 3 months or more, when compared
with those in treatment for less than 3 months, were shown to be
statistically significant (that is, not likely to have occurred by
chance alone).
\a Lower levels of heroin use among people still in treatment during
the follow-up year were statistically significant.
\b Less criminal activity among people in long-term residential
treatment for 6 months or more was statistically significant.
Source: Robert L. Hubbard and others, "Overview of 1-Year Follow-Up
Outcomes in the Drug Abuse Treatment Outcome Study (DATOS),"
Psychology of Addictive Behaviors, Vol. II, No. 4 (1997).
Previous studies found similar reductions in drug use. For example,
researchers from the TOPS study found that across all types of drug
treatment, 40 to 50 percent of regular heroin and cocaine users who
spent at least 3 months in treatment reported near abstinence during
the year after treatment, and an additional 30 percent reported
reducing their use. DARP found that in the year after treatment,
abstinence from daily opiate use was reported by 64 percent of
clients in methadone programs, 61 percent in therapeutic communities,
and 56 percent in outpatient drug-free programs. NTIES found that 50
percent of clients in treatment reported using crack cocaine five
times or more during the year prior to entering treatment, while 25
percent reported such use during the year following treatment.
The major studies also found that criminal activity declined after
treatment. DATOS found that reports of criminal activity declined by
60 percent for cocaine users in long-term residential treatment at
the 1-year follow-up. Only 17 percent of NTIES clients reported
arrests in the year following treatment--down from 48 percent during
the year before treatment entry. Additionally, the percentage of
clients who reported supporting themselves primarily through illegal
activities decreased from 17 percent before treatment to 9 percent
after treatment. DARP found reported reductions in criminal activity
for clients who stayed in treatment at least 3 months.
--------------------
\23 The DATOS researchers note that this research methodology
provides more rigorous evidence than is provided by simple
observational design. In addition to descriptive information on
clients when they enter treatment and their behaviors before, during,
and after treatment, "the prospective cohort research design also
provides strong support for evaluative and causal inferences." See
Patrick M. Flynn and others, "Methodological Overview and Research
Design for the Drug Abuse Treatment Outcome Study (DATOS),"
Psychology of Addictive Behaviors, Vol. II, No. 4 (1997), p. 233.
\24 A large percentage of the clients participating in these studies
were involved with the criminal justice system. For example, 56
percent of DATOS clients reported being on probation or parole or
awaiting trial when they entered treatment, and 31 percent were
referred into treatment by the courts. Research suggests that
self-reported data tend to be the least reliable for those involved
with the criminal justice system.
\25 In 1990, the Institute of Medicine concluded there was no
evidence to suggest that hospital-based chemical dependency programs,
a type of inpatient treatment, were either more or less effective
than chemical dependency programs not situated in hospitals.
LONGER TREATMENT EPISODES
HAVE BETTER OUTCOMES, BUT
TREATMENT DURATION IS
LIMITED BY CLIENT
DROP-OUT
-------------------------------------------------------- Letter :6.1.2
Another finding across these studies is that clients who stay in
treatment longer report better outcomes. For the DATOS clients that
reported drug use when entering treatment, fewer of those in
treatment for more than 3 months reported continuing drug use than
those in treatment for less than 3 months (see table 6). DATOS
researchers also found that the most positive outcomes for clients in
methadone maintenance were for those who remained in treatment for at
least 12 months.
Table 6
Impact of Treatment Duration on Outcomes
for DATOS Clients
Percentage
reduction in
Percentage Percentage the number of
reduction in reduction in clients
the number of the number of reporting
self-reported self-reported criminal
Treatment heroin users cocaine users activity in
setting in the year in the year the year
and following following following
duration treatment treatment treatment
---------- -------------- -------------- --------------
Outpatient methadone
----------------------------------------------------------
Less than 57.7 68.2 80.8
3 months
More than 72.1 64.6 71.9
3 months
Long-term residential
----------------------------------------------------------
Less than 54.9 54.5 69.5
3 months
More than 81.6 82.2 78.5
3 months
Outpatient drug-free
----------------------------------------------------------
Less than 56.2 57.2 70.1
3 months
More than 86.4 86.8 82.5
3 months
----------------------------------------------------------
Note: Percentage reductions are based on the number of clients
reporting drug use or predatory criminal activity in the year before
treatment (weighted).
Source: Hubbard and others, "Overview of 1-Year Follow-Up Outcomes
in DATOS," Psychology of Addictive Behaviors, p. 271.
Earlier studies reported similar results. Both DARP and TOPS found
that reports of drug use were reduced most for clients who stayed in
treatment at least 3 months, regardless of the treatment setting. In
fact, DARP found that treatment lasting 90 days or less was no more
effective than no treatment at facilitating complete abstinence from
drug use and criminal behavior during the year following treatment.
Although these studies show better results for longer treatment
episodes, they found that many clients dropped out of treatment long
before reaching the minimum length of treatment episode recommended
by those operating the treatment program. For example, a study of a
subset of DATOS clients found that all of the participating methadone
maintenance programs recommend 2 or more years of treatment, but the
median treatment episode by clients was about 1 year. Long-term
residential programs participating in DATOS generally recommended a
treatment duration of 9 months or longer, while outpatient drug-free
programs recommended at least 6 months in treatment; for both program
types, the median treatment episode was 3 months. TOPS found that in
the first 3 months of treatment, 64 percent of outpatient drug-free
program clients and 55 percent of therapeutic community clients
discontinued treatment. For clients receiving methadone maintenance
treatment, drop-out rates were somewhat lower--32 percent--in the
first 3 months.
Researchers note that drug abuse treatment outcomes should be
considered comparable to those of other chronic diseases; therefore,
significant dropout rates should not be unexpected. These results
are similar to the levels of compliance with treatment regimens for
people with chronic diseases such as diabetes and hypertension. A
review of over 70 outcome studies of treatment for diabetes,
hypertension, and asthma found that less than 50 percent of people
with diabetes fully comply with their insulin treatment schedule,
while less than 30 percent of patients with hypertension or asthma
comply with their medication regimens.\26
--------------------
\26 Charles O'Brien and A. Thomas McLellan, "Myths About the
Treatment of Addiction," The Lancet, Vol. 347 (1996), pp. 237-40.
RESEARCH SUGGESTS THAT
OUTPATIENT TREATMENT
REDUCES DRUG USE AS MUCH
AS RESIDENTIAL TREATMENT,
BUT COSTS VARY WIDELY
-------------------------------------------------------- Letter :6.1.3
A 1990 Institute of Medicine assessment of the treatment literature
concluded that despite the heterogeneity of the programs and their
clients, treatment outcomes are "qualitatively similar" regardless of
whether treatment is provided in a residential or outpatient
setting.\27 In 1997, an ONDCP report showed that 34 percent of
clients in outpatient treatment were no longer "heavy users"
following treatment, while 38 percent of clients in residential
settings reported the same.\28 Evidence from the recent DATOS study
confirmed that reported reductions in cocaine use were similar for
outpatient drug-free and residential settings when clients remained
in treatment for at least 3 months. Researchers point out, however,
that more severe drug abusers may receive treatment in residential
treatment settings than in outpatient settings, making such
comparisons difficult.
However, analysis of the data from DATOS showed mixed results on the
impact of treatment on drug-related criminal activity.\29 Clients in
long-term residential treatment for at least 6 months were
significantly less likely than clients who did not complete more than
13 weeks of treatment to report engaging in an illegal activity in
the year after treatment. In contrast, clients in methadone or
drug-free treatment in an outpatient setting who remained for at
least 6 months were not significantly less likely to report engaging
in illegal activity than clients who did not complete more than 13
weeks of treatment in these settings.
Although the available evidence does not show sharp differences in
outcomes, studies do show wide variation in treatment costs for
inpatient and outpatient settings. A recent NTIES study found that
costs per day were lowest in outpatient settings, where the average
treatment period is several months. In contrast, short-term (1
month) residential treatment costs were much higher, resulting in a
cost per treatment episode that was double the cost of outpatient
treatment episodes. (See table 7.)
Table 7
Costs of Substance Abuse Treatment by
Approach and Setting
Approximate
Cost for one Approximate cost for
Treatment day of treatment treatment
setting treatment duration episode
---------------- ------------ ------------ ------------
Outpatient $13 300 days $3,900
methadone
Outpatient drug- 15 120 days 1,800
free
Long-term 49 140 days 6,800
residential
Short-term 130 30 days 4,000
residential
----------------------------------------------------------
Source: National Opinion Research Center at the University of
Chicago and the Research Triangle Institute, NTIES Brief Report.
Regardless of the findings of similar outcomes and great variation in
costs, there is still reason to support residential treatment for
certain patients. In some cases, residential treatment may be
required for optimum treatment outcomes, such as for drug abusers
with severe substance-related problems, those who have failed in
outpatient treatment, or those with severe psychosocial impairments.
In contrast, patients with greater psychosocial stability and less
substance-related impairment appear to benefit most from nonhospital
and nonresidential treatment.\30
--------------------
\27 Institute of Medicine, Treating Drug Problems, p. 15.
\28 ONDCP, Reducing Drug Use in America (Washington, D.C.: ONDCP,
Oct. 1997).
\29 Multivariate analytic techniques were used to adjust for
differences in the population characteristics of the treated and
comparison groups. For these analyses, the comparison group was all
clients completing 1 to 13 weeks of treatment.
\30 Mim J. Landry, Overview of Addiction Treatment Effectiveness
(Washington, D.C.: HHS, SAMHSA, Office of Applied Studies, 1997),
pp. iv-v.
EVIDENCE VARIES ON THE BEST
TREATMENT APPROACHES FOR
SPECIFIC GROUPS OF DRUG
ABUSERS
---------------------------------------------------------- Letter :6.2
Research provides strong evidence to support methadone maintenance as
the most effective treatment for heroin addiction. However, research
on the most effective treatment interventions for other groups of
drug abusers is less definitive. Promising treatment approaches for
other groups include cognitive-behavioral therapy for treatment of
cocaine abuse and family-based therapy for adolescent drug users.
RESEARCH SUPPORTS
METHADONE MAINTENANCE AS
THE MOST EFFECTIVE
TREATMENT FOR HEROIN
ADDICTION
-------------------------------------------------------- Letter :6.2.1
A number of approaches have been used in treating heroin addiction.
Methadone maintenance, however, is the treatment most commonly used,
and numerous studies have shown that those receiving methadone
maintenance treatment have better outcomes than those who go
untreated or use other treatment approaches--including detoxification
with methadone.\31 Methadone maintenance has been shown to reduce
heroin use and criminal activity and improve social functioning. HIV
risk is also minimized, since needle usage is reduced. Proponents of
methadone maintenance also argue that reductions in the use of
illicit drugs and associated criminal behavior help recovering drug
abusers focus on their social and vocational rehabilitation and
become reintegrated into society.\32
However, outcomes among methadone programs have varied greatly, in
part because of the substantial variation in treatment practices
across the nation. Many methadone clinics have routinely provided
clients dosage levels that are lower than optimum--or even
subthreshold--and have discontinued treatment too soon. In late
1997, an NIH consensus panel concluded that people who are addicted
to heroin or other opiates should have broader access to methadone
maintenance treatment programs and recommended that federal
regulations allow additional physicians and pharmacies to prescribe
and dispense methadone.
Similarly, several studies conducted over the past decade show that
when counseling, psychotherapy, health care, and social services are
provided along with methadone maintenance, treatment outcomes improve
significantly. However, the recent findings from DATOS suggest that
the provision of these ancillary services--both the number and
variety--has eroded considerably during the past 2 decades across all
treatment settings. DATOS researchers also noted that the percentage
of clients reporting unmet needs was higher than that in previous
studies.
There are other concerns associated with methadone maintenance. For
example, methadone is often criticized for being a substitute drug
for heroin, which does not address the underlying addiction.
Additional concerns center on the extent to which take-home methadone
doses are being sold or exchanged for heroin or other drugs.
--------------------
\31 The Institute of Medicine noted that "the most convincing results
about the efficacy of methadone maintenance . . . come from a
handful of clinical experiments that are widely separated in time and
place but that consistently yield very distinctive findings. In
these studies, heroin-dependent, heavily criminally involved
populations who were randomly assigned to methadone maintenance or a
control condition (an outpatient nonmethadone modality) demonstrated
clinically important and statistically significant differences in
favor of methadone on the gauges of drug use, criminal activity, and
engagement in socially productive roles such as employment,
education, or responsible child rearing." See Institute of Medicine,
Treating Drug Problems, p. 143.
\32 Experts recognize that a large percentage of patients in
methadone programs are also cocaine users. One study found that 39
percent of patients reported having used cocaine prior to methadone
treatment; while in methadone treatment, cocaine use varied widely.
COGNITIVE-BEHAVIORAL
TREATMENTS SHOW PROMISE
FOR COCAINE ADDICTION
-------------------------------------------------------- Letter :6.2.2
Evidence of treatment effectiveness is not as strong for cocaine
addiction as it is for heroin addiction. No pharmacological agent
for treating cocaine addiction or reducing cocaine craving has been
found. However, an accumulating body of research points to
cognitive-behavioral therapies as promising treatment approaches for
cocaine addiction.
In an earlier report, we noted that treatments used for other drug
dependencies, such as methadone maintenance, have not proven useful
for treating cocaine dependency.\33 Although a number of
pharmacotherapies have been studied and some have proven successful
in one or more clinical trials, no medication has demonstrated
substantial efficacy once subjected to several rigorously controlled
trials. Nor has any medication used in combination with one or more
cognitive-behavioral therapies proven effective in enhancing cocaine
abstinence. Researchers are hopeful, however, that a pharmacological
agent for treating cocaine addiction will be developed.\34
Without a pharmacological agent, researchers have relied on
psychotherapeutic approaches to treat cocaine addiction. Studies
have shown that clients receiving three cognitive-behavioral
therapies have demonstrated prolonged periods of abstinence and high
rates of retention in treatment programs. The cognitive-behavioral
therapies, based largely on counseling and education, include (1)
relapse prevention, which focuses on teaching clients how to identify
and manage high-risk, or "trigger," situations that contribute to
drug relapse; (2) community reinforcement/contingency management,
which establishes a link between behavior and consequence by
rewarding abstinence and reprimanding drug use; and (3)
neurobehavioral therapy, which addresses a client's behavioral,
emotional, cognitive, and relational problems at each stage of
recovery.
These programs have shown promise in curbing drug use. One relapse
prevention program showed cocaine-dependent clients were able to
remain abstinent at least 70 percent of the time while in treatment.
A community reinforcement/contingency management program showed that
42 percent of the participating cocaine-dependent clients were able
to achieve nearly 4 months of continuous abstinence, while a
neurobehavioral program showed that 38 percent of the clients were
abstinent at the 6-month follow-up.
--------------------
\33 Cocaine Treatment: Early Results From Various Approaches
(GAO/HEHS-96-80, June 7, 1996).
\34 For example, in recent animal research, a new immunization
procedure has demonstrated positive effects in blocking the stimulant
effects of cocaine. When vaccinated, rats produced antibodies that
acted like biological "sponges" or blockers, diminishing by more than
70 percent the amount of cocaine reaching the brain. NIDA has
reported that the cocaine vaccine project was expected to have begun
human studies in the first quarter of 1998.
FAMILY THERAPY IS UNDER
STUDY FOR ADOLESCENT DRUG
ABUSERS
-------------------------------------------------------- Letter :6.2.3
Adolescent drug abusers are similar to adult drug abusers in that
they are likely to use more than one type of illicit drug and to have
coexisting psychiatric conditions. In other ways, they differ from
adult drug abusers. Adolescents may have a shorter history of drug
abuse and thus less severe symptoms of tolerance, craving, and
withdrawal. In addition, they usually do not show the long-term
physical effects of drug abuse. Despite a number of studies on the
topic, little is known about the best way to treat adolescent drug
abusers. Researchers believe that adolescents have special treatment
needs; however, research has not shown any one method or approach to
be consistently superior to others in achieving better treatment
outcomes for adolescents. Among the wide variety of treatment
approaches and settings used for adolescents, family-based therapies
show promise.
Historically, adolescents have been referred to residential treatment
settings, which may range from group-home living with minimal
professional involvement to a setting that provides intensive
medical, psychiatric, and psychosocial treatment 24 hours a day.
Experts now recognize that many adolescents can be successfully
treated in an outpatient treatment setting, where treatment may range
from less than 9 hours per week to regular sessions after school to
intensive day programs that provide more than 20 hours of treatment
per week. Although not thoroughly evaluated, pharmacotherapy may
also be used to treat adolescent drug abuse. Researchers believe
that self-help or peer support groups, such as Alcoholics Anonymous,
are important adjuncts to treatment for adolescents.
The relative effectiveness of alternative approaches for treating
adolescents remains uncertain.\35 An earlier study of adolescents
found that residential treatment resulted in more substantial and
consistent reductions in drug use, drug-related problems, and illegal
activity than did outpatient drug-free programs.\36 In contrast, the
American Academy of Child and Adolescent Psychiatry acknowledged in
its 1997 treatment practice parameters that research on drug
treatment for adolescents has failed to demonstrate the superiority
of one treatment approach over another.\37 Studies show that success
in treatment seems to be linked to the characteristics of program
staff, the availability of special services, and family
participation.
Many experts believe that family-based intervention shows promise as
an effective treatment for adolescent drug abusers. Family-based
intervention is based on the assumption that family behaviors
contribute to the adolescent's decision to use drugs. Many
researchers believe that family interventions are critical to the
success of any treatment approach for adolescent drug abusers, since
family-related factors--such as parental substance use, poor
parent-child relations, and poor parent supervision--have been
identified as risk factors for the development of substance abuse
among adolescents. Family relationships may be the primary target
for intervention or one of many target areas. A 1995 literature
review suggests that family intervention can engage and retain drug
abusers and their families in treatment, significantly reducing drug
use and related areas of problem behavior.\38 Further, a 1997
meta-analysis and literature review held family therapy to be
superior to other treatment modalities.\39 However, NIDA points out
in a soon-to-be published article that further research is needed to
identify the best approach to treating adolescent drug abusers.\40
--------------------
\35 As a component of the DATOS study, the Research Triangle
Institute is gathering information on treatment outcomes for 3,300
adolescents in treatment at 30 programs in six cities. Results have
not yet been reported.
\36 These findings result from a study of a subset of TOPS clients,
consisting of 375 participants aged 19 or below.
\37 Oscar Bukstein, M.D. (principal author) and the Washington Group
on Quality Issues, "Practical Parameters for the Assessment and
Treatment of Children and Adolescents With Substance Use Disorders,"
Journal of the American Academy of Child and Adolescent Psychiatry,
Vol. 36, No. 10, Supplement, Oct. 1997, pp. 1405-1565.
\38 Howard A. Liddle and Gayle A. Dakof, "Efficacy of Family
Therapy for Drug Abuse: Promising but Not Definitive," Journal of
Marital and Family Therapy, Vol. 21, No. 4 (1995), pp. 511-43.
\39 M.D. Stanton and W.R. Shadish, "Outcome, Attrition, and
Family/Couples Treatment for Drug Abuse: A Meta-Analysis and Review
of the Controlled, Comparative Studies," Psychology Bulletin, Vol.
122 (1997), pp. 170-91.
\40 Naimah Z. Weinberg, M.D., and others, "Adolescent Substance
Abuse: A Review of the Past 10 Years," Journal of the American
Academy of Child and Adolescent Psychiatry (forthcoming).
CONCLUSIONS
------------------------------------------------------------ Letter :7
With an annual expenditure of more than $3 billion, the federal
investment in drug abuse treatment is an important component of the
nation's drug control efforts, and monitoring the performance of
treatment programs can help ensure that progress toward the nation's
goals is being achieved. Research on the effectiveness of drug abuse
treatment, however, is highly problematic, given the methodological
challenges and numerous factors that influence the results of
treatment. Although studies conducted over nearly 3 decades
consistently show that treatment reduces drug use and crime, current
data collection techniques do not allow accurate measurement of the
extent to which treatment reduces the use of illicit drugs.
Furthermore, research literature has not yet yielded definitive
evidence to identify which approaches work best for specific groups
of drug abusers.
AGENCY AND OTHER COMMENTS
------------------------------------------------------------ Letter :8
NIDA, SAMHSA, VA, and a private consultant with expertise in drug
treatment issues generally acknowledged that methodological and
implementation issues make the evaluation of treatment effectiveness
difficult. SAMHSA and NIDA also provided extensive and helpful
technical comments, which we incorporated into a substantially
revised final report.
---------------------------------------------------------- Letter :8.1
As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days
from the date of this letter. At that time, we will send copies to
interested parties and make copies available upon request.
If you have any questions about this report, please call me at (202)
512-7119. Other contributors to this report include Rosamond Katz
and Jenny Grover.
Marsha Lillie-Blanton
Associate Director
Health Services Quality and
Public Health Issues
BIBLIOGRAPHIC REFERENCES FOR
SELECTED STUDIES
==================================================== Appendix Appendix
For additional information on the four major studies that we
reviewed, see the sources cited below.
DRUG ABUSE TREATMENT OUTCOME
STUDY
------------------------------------------------ Appendix Appendix:0.1
"Drug Abuse Treatment Outcome Study (DATOS)." Psychology of Addictive
Behaviors, Vol. 11, No. 4 (1997), pp. 211-323.
THE NATIONAL TREATMENT
IMPROVEMENT EVALUATION STUDY
------------------------------------------------ Appendix Appendix:0.2
National Opinion Research Center at the University of Chicago. The
National Treatment Improvement Evaluation Study--Final Report.
Prepared for the Center for Substance Abuse Treatment, SAMHSA, in
collaboration with the Research Triangle Institute, Mar. 1997.
TREATMENT OUTCOME
PROSPECTIVE STUDY
------------------------------------------------ Appendix Appendix:0.3
Hubbard, R.L. "Evaluation and Treatment Outcome." Substance Abuse:
A Comprehensive Textbook, 2nd ed. Baltimore, Md.: Williams &
Wilkins, 1992, pp. 596-611.
Hubbard, R.L., and others. Drug Abuse Treatment: A National Study
of Effectiveness. Chapel Hill, N.C.: University of North Carolina
Press, 1989.
Ginzburg, H.M. "Defensive Research--The Treatment Outcome
Prospective Study (TOPS)." Annals of the New York Academy of
Sciences, Vol. 311 (1978), pp. 265-69.
DRUG ABUSE REPORTING PROGRAM
------------------------------------------------ Appendix Appendix:0.4
Simpson, D.D. "Drug Treatment Evaluation Research in the United
States." Psychology of Addictive Behaviors, Vol. 7 (1993), pp.
120-28.
Simpson, D.D., and S.B. Sells, eds. Opioid Addiction and Treatment:
A 12-Year Follow-up. Malabar, Fla.: Robert E. Krieger, 1990.
Simpson, D.D., and S.B. Sells. "Effectiveness of Treatment for Drug
Abuse: An Overview of the DARP Research Program." Advances in
Alcohol and Substance Abuse, Vol. 2 (1992), pp. 7-29.
RELATED GAO PRODUCTS
Drug Courts: Overview of Growth, Characteristics, and Results
(GAO/GGD-97-106, July 31, 1997).
Drug Control: Observations on Elements of the Federal Drug Control
Strategy (GAO/GGD-97-42, Mar. 14, 1997).
Substance Abuse Treatment: VA Programs Serve Psychologically and
Economically Disadvantaged Veterans (GAO/HEHS-97-6, Nov. 5, 1996).
Drug and Alcohol Abuse: Billions Spent Annually for Treatment and
Prevention Activities (GAO/HEHS-97-12, Oct. 8, 1996).
Cocaine Treatment: Early Results From Various Approaches
(GAO/HEHS-96-80, June 7, 1996).
At-Risk and Delinquent Youth: Multiple Federal Programs Raise
Efficiency Questions (GAO/HEHS-96-34, Mar. 6, 1996).
*** End of document. ***