The Cannabis Youth Treatment (CYT) Study: Main findings from two randomized trials

Article (PDF Available)inJournal of Substance Abuse Treatment 27(3):197-213 · November 2004with109 Reads
DOI: 10.1016/j.jsat.2003.09.005 · Source: PubMed
Abstract
This article presents the main outcome findings from two inter-related randomized trials conducted at four sites to evaluate the effectiveness and cost-effectiveness of five short-term outpatient interventions for adolescents with cannabis use disorders. Trial 1 compared five sessions of Motivational Enhancement Therapy plus Cognitive Behavioral Therapy (MET/CBT) with a 12-session regimen of MET and CBT (MET/CBT12) and another that included family education and therapy components (Family Support Network [FSN]). Trial II compared the five-session MET/CBT with the Adolescent Community Reinforcement Approach (ACRA) and Multidimensional Family Therapy (MDFT). The 600 cannabis users were predominately white males, aged 15-16. All five CYT interventions demonstrated significant pre-post treatment during the 12 months after random assignment to a treatment intervention in the two main outcomes: days of abstinence and the percent of adolescents in recovery (no use or abuse/dependence problems and living in the community). Overall, the clinical outcomes were very similar across sites and conditions; however, after controlling for initial severity, the most cost-effective interventions were MET/CBT5 and MET/CBT12 in Trial 1 and ACRA and MET/CBT5 in Trial 2. It is possible that the similar results occurred because outcomes were driven more by general factors beyond the treatment approaches tested in this study; or because of shared, general helping factors across therapies that help these teens attend to and decrease their connection to cannabis and alcohol.
Regular article
The Cannabis Youth Treatment (CYT) Study:
Main findings from two randomized trials
Michael Dennis, (Ph.D.)
a,
*
, Susan H. Godley, (Rh.D.)
a
, Guy Diamond, (Ph.D.)
b,c
,
Frank M. Tims, (Ph.D.)
e,f
, Thomas Babor, (Ph.D.)
g
, Jean Donaldson, (M.A.)
d
,
Howard Liddle, (Ed.D.)
h
, Janet C. Titus, (Ph.D.)
a
, Yifrah Kaminer, (M.D.)
g
,
Charles Webb, (Ph.D.)
g
, Nancy Hamilton, (M.P.A.)
e
, Rod Funk, (B.S.)
a
a
Chestnut Health Systems, Bloomington, IL, USA
b
Center for Family Intervention Science, ChildrenT s Hospital of Philadelphia, Philadelphia, PA, USA
c
University of Pennsylvania, Philadelphia, PA, USA
d
Center for Substance Abuse Treatment of the Substance Abuse and Mental Health Services Administration,
Rockville, MD, USA
e
Operation PAR, Inc., St. Petersburg, FL, USA
f
University of South Florida, Tampa, FL, USA
g
University of Connecticut Health Center, Farmington, CT, USA
h
Center for Treatment Research on Adolescent Drug Abuse, University of Miami School of Medicine, Miami, FL, USA
Received 11 December 2002; received in revised form 17 April 2003; accepted 13 September 2003
Abstract
This article presents the main outcome findings from two inter-related randomized trials conducted at four sites to evaluate the
effectiveness and cost-effectiveness of five short-term outpatient interventions for adolescents with cannabis use disorders. Trial 1 compared
five sessions of Motivational Enhancement Therapy plus Cognitive Behavioral Therapy (MET/CBT) with a 12-session regimen of MET and
CBT (MET/CBT12) and another that included family education and therapy components (Family Support Network [FSN]). Trial II compared
the five-session MET/CBT with the Adolescent Community Reinforcement Approach (ACRA) and Multidimensional Family Therapy
(MDFT). The 600 cannabis users were predominately white males, aged 15–16. All five CYT interventions demonstrated significant pre-post
treatment improvements during the 12 months after random assignment to a treatment intervention in the two main outcomes: days of
abstinence and the percent of adolescents in recovery (no use or abuse/dependence problems and living in the community). Overall, the
clinical outcomes were very similar across sites and conditions; however, after controlling for initial severity, the most cost-effective
interventions were MET/CBT5 and MET/CBT12 in Trial 1 and ACRA and MET/CBT5 in Trial 2. It is possible that the similar results
occurred because outcomes were driven more by general factors beyond the treatment approaches tested in this study; or because of shared,
general helping factors across therapies that helped these teens attend to and decrease their connection to cannabis and alcohol. D 2004
Elsevier Inc. All rights reserved.
Keywords: Marijuana; Manual-guided therapy; Adolescents; Treatment effectiveness; Cost-effectiveness
1. Introduction
Cannabis (including hashish, marijuana, blunts, and
other forms of tetrahydrocannabinol) is the most prevalent
psychoactive substance used by adolescents in the U.S.
(Office of Applied Studies [OAS], 2000b). In 1998, 6.8%
of U.S. 18-year-olds met criteria for past-year cannabis
0740-5472/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jsat.2003.09.005
* Corresponding author. Chestnut Health Systems, 720 West Chestnut,
Bloomington, IL 61701; Tel.: +1 309 820 3543, ext. 8 3409; fax: +1 309
829 4661.
E-mail address: mdennis@chestnut.org (M. Dennis).
Journal of Substance Abuse Treatment 27 (2004) 197 213
dependence (OAS, 2000b). Cannabis is now the leading
substance found during drug screens after adolescent
arrests (primarily for property or violent offenses—not
simple possession), emergency room admissions and
autopsies. It is also the leading substance mentioned during
adolescent treatment admissions (Bureau of Justice Sta-
tistics, 2000; Dennis, Babor, Roebuck, & Donaldson, 2002;
OAS, 2000a, 2000c). There has been increasing interest in
the development of effective outpatient treatments for
adolescents with cannabis use disorders, partially due to
the growing volume of lit erature on the social, medical,
and psychological consequences of cannabis use (Ameri-
can Academy of Pediatrics, 1999; Fergusson, Lynskey, &
Horwood, 1996).
Evaluations of existing outpatient treatment programs for
U.S. adolescent cannabis users have produced mixed results.
Some studies reported increases of 3 to 13% in cannabis use
following outpatient drug abuse treatment (Hubbard, Cava-
naugh, Craddock, & Rachel, 1985; OAS, 1995; Sells &
Simpson, 1979), whereas more recent studies (Hser et al.,
2001) have reported decreasing cannabis use by 12 to 25%.
Among the 445 adolescents followed up after a median of
42 days of outpatient treatment in the Drug Abuse Treat-
ment Outcome Study-Adolescents (Grella, Hser, Joshi, &
Rounds-Bryant, 2001; Hser et al., 2001) in the mid to late
1990s, there was a 21 to 25% reduction in cannabis use
between the year before and the year after treatment.
In addition to the large studies of existing practice,
numerous randomized and quasi-experimental field studies
have evaluated a wide variety of outpatient treatment
approaches for adolescents with alcohol or other drug use
problems during the past 10 years. These have included
cognitive behavior therapy alone and in combination with a
motivational interviewing approach (Kaminer, Burleson, &
Goldberger, 2001, 2002; Waldron, Slesnick, Brody, Turner,
& Peterson, 2001); family education and therapy approaches
(Azrin et al., 2001; Henggeler et al., 1991; Henggeler,
Clingempeel, Brondino, & Pickrel, 2002; Joanning et al.,
1992; Lewis, Piercy, Sprenkle, & Trepper, 1990; Liddle
et al., 2001; Szapocznik, Kurtines, Foote, Perez-Vidal, &
Hervis, 1983; Waldron et al., 2001); group psychoeduca-
tional approaches (Kaminer et al., 2002; Liddle et al., 2001;
Waldron et al., 2001); individual behavior therapy ap-
proaches (Azrin et al., 1994, 2001; Godley, Godley, Dennis,
Funk, & Passetti, 2002); engagement approaches (Szapo-
cznik et al., 1988); and 12-step based or Minnesota Model
therapy (Winters, Stinchfield, Opland, Weller, & Latimer,
2000). Although literature reviews (Ozechow ski & Liddle,
2000; Williams & Chang, 2000) and a meta-analysis
(Stanton & Shadish, 1997) based on pre-1997 treatment
studies suggested there might be a clear advantage to family
therapy approaches, a more recent study suggests other
interventions can be just as efficacious ( Waldron et al.,
2001). Taken as a whole, these studies suggest there may be
many effective types of treatment for this population. Yet to
date, no published treatment studies have specifically
targeted adolescent cannabis/marijuana users or provided
readily disseminable manuals for replicating them.
Studies evaluating clinical effectiveness, cost, c ost-
effectiveness, and benefit-cost of different interventions
are all critical to improving practice (Dennis, Perl, Huebner,
& McLellan, 2000; Gold, Siegel, Russell, & Weinstein,
1996). Cost-effectiveness analysis combines information on
the bsocietalQ cost of the interventions (i.e., market value of
goods and services) with the clinical outcomes. Treatments
that are more b efficientQ in ach ieving outcomes are
considered better (whether they cost more or not). As in
the clinical research literature, we know of no prior
economic evaluations of treatments designed specifically
for treating adolescent cannabis users.
In 1997, the Center fo r Substance Abuse Treatment
(CSAT) cr eated the Cannab is Youth Treatment (CYT)
cooperative agreement in response to the expanding pop-
ulation of adolescent cannabis users and the lack of short-
term (less than 3 months) outpatient treatment models
targeting adoles cents with cannabis-related problems (Clark,
Horton, Dennis, & Babor, 2002). The objectives of CYT
were to (a) develop several promising models of short-term
outpatient treatment that could be readily disseminated to the
field, and (b) conduct a field trial to estimate the cost,
effectiveness and c ost-effectiveness of these interventions.
CYT was designed to be an initial effectiveness study to allow
comparison of treatment alternatives that could be readily
used by outpatient treatment p roviders. Though a no-
treatment control might have provided more evidence about
generic efficacy, it was considered unethical in light of
previous research showing lack of improvement in untreated
or minimally treated samples (Henggeler, Pickrel, Brondino,
& Crouch, 1996; Henggler, Pickrel, & Brondino, 1999;
Kaminer, Burleson, Blitz, Sussman, & Rounsaville, 1998;
Kaminer & Burleson, 1999; Kaminer et al., 2002; Latimer,
Winters, D’Zurilla, & Nichols. 2003; Szapocznik & Kurtines,
1993; Winters et al., 2000).
A national compe tition was held with awar ds going to
four site grantees (University of Connecticut Health Cen ter,
Operation PAR, Inc., Chestnut Health Systems, and Chil-
drenTs Hospital of Philadelphia) and a coordinating center
at the research division of Chestnut Health Systems. A
35-member steering committee composed of clinicians and
researchers from each grant, other research collaborators,
CSAT staff, and an independent advisory board reviewed
the 22 interventions originally proposed by the grantees and
chose five short-term interventions to implement and
compare. These interventions were selected because they
(a) could be readily manualized and f ield tested within the
4-year time frame of the study, (b) were promising based on
previously published studies or studies in progress, (c) were
demonstrated to be effective with related populations, and
(d) were recommended by expe rt panels as best practices for
adolescent treatment. The steering committee then charged
five teams with developing treatment manuals and designed
a detailed research plan.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213198
The CYT collaborative has published articles on the
studyTs design, implementation, and the reliability and
validity of the measures (Dennis, Titus , et al., 2002), the
characteristics and needs of the adolescents (Diamond,
Leckrone, & Dennis, in press; Petry & Tawfik, 2001; Tims
et al., 2002; Webb, Scudder, Kaminer, Kadden, & Tawfik,
2002), the five treatment manuals (Godley, Meyers et al.,
2001; Godley, White, Diamond, Passetti, & Titus, 2001;
Hamilton, Brantley, Tims, Angelovich, & McDougall,
2001; Liddle, 2002; Sampl & Kadden, 2001; Webb et al.,
2002), a comparison of their rationale/components (Dia-
mond et al., 2002), the reactions of clinical staff to using
them (Godley, White, et al., 2001), the relative cost of the
interventions (French et al., 2002), and the impact of the
interventions on costs to society (French et al., 2003). This
paper summarizes earlier results on the clinical character-
istics and costs, then presents the main findings in terms of
the relative clinical- and cost-effectiveness of the five CYT
interventions in the two multi-site field trials.
2. Materials and methods
2.1. Overview of research design and questions
During a 2-year period, 600 adolescents and their
families were recruited and randomized from sequential
admissions to four treatment sites: University of Connect-
icut Health Center (UCHC), Operation PAR, Chestnut
Health Systems (CHS), and ChildrenTs Hospital of
Philadelphia (CHOP). It was not logistically feasible to
implement all five conditions in any one site because of
the limits of case flow and resources. Therefore, adoles-
cents were randomly assigned within each site to one of
three treatment conditions, and the interventions were
evaluated in two trials. In Trial 1 (at UCHC and PAR),
adolescents were randomly a ssigned to one of three
treatment conditions: Motivational Enhancement Treat-
ment/Cognitive Behavior Therapy 5 Sessions (MET/
CBT5), Motivational Enhancement Treatment/Cognitive
Behavior Therapy 12 Sessions (MET/CBT12), or Family
Support Network (FSN). In Trial 2 (at CHS and CHOP),
adolescents were randomly assigned to MET/CBT5, Ado-
lescent Community Reinforce ment Approach (ACRA),
or Multidimensional Family Therapy (MDFT). Adoles-
cents were interviewed with the Global Appraisal of
Individual Needs (GAIN; Denn is, 1999) and a battery of
other measures at intake and then 3, 6, 9, and 12 months
later (see Dennis, Titus, et al., 2002 for a detailed de-
scription). Data was obtained from collateral informants,
records, and urine tests to validate the self-report data
collected at intake, 3, and 6 months. Of the eligible ado-
lescents, 85% agreed to participate. Data for the analyses
presented here were available from one or more follow-
up interviews for 99% of the adolescents (94 to 98%
per observation).
The analyses presented in this paper focus on answering
the following two core questions addressed by CYT:
1. Are there significant differences in the relative clinical
outcome effectiveness of these treatment approaches in
terms of (a) the days of abstinence over the followup
period and (b) the percent of adolescents in recovery (no
use or abuse/ dependence problems while living in the
community) at the end of the followup study?
2. Are there significant differences in the relative cost-
effectiveness of these treatment approaches in terms of
their (a) cost per day of abstinence over the follow up
period, and (b) cost per person in recovery at the end of
the study?
Analyses were conducted by site and across sites for each
of the two trials with baseline measures as covariates to
allow for individual differences, nesting conditions within
site to control for site differences, and using restricted
maximum likelihood estimation to use all of the available
data without biasing condition estimates.
2.2. Sites and participants
2.2.1. Sites
The UCHC site, located in Farmington, CT, is a major
academic medical center that has been involved in multiple
substance abuse treatment trials. The Operation PAR site,
located in St. Petersburg, is FloridaTs most comprehensive
adolescent treatment provider and also offers behavioral
healthcare services in three additional Florida counties. The
CHS site in Madison County is IllinoisT largest community-
based adolescent treatment provider and also operates other
programs in the state. CHOP, located in Philadelphia, PA, is
a leading pediatric research center and has been involved in
numerous substance abuse treatment and family therapy
trials. The sites differed in their catchment areas, with CHS
serving a rural and small urban population at three facilities
located 30 miles apart, CHOP prim arily targeting the inner
city of Philadelphia, and UCHC and PAR serving suburban
areas. When an intervention included a group component,
only study participants were included in the group sessions.
2.2.2. Target population and inclusion/exclusion criteria
The target population for this study was adolescents with
cannabis related disorders who would be appropriate for and
typically present to publicly funded outpatient treatment.
Participants were eligible for CYT if they were aged 12 to 18,
self-reported one or more DSM-IV (American Psychiatric
Association, 1994) criteria for cannabis abuse or depen-
dence, had used cannabis in the past 90 days or 90 days prior
to being sent to a controlled environment, and were
appropriate for outpatient or intensive outpatient treatment
(American Society of Addiction Medicine, 1996). Because
the goal of the study was to generalize to adolescents who
present for publicly funded outpatient treatment in the
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 199
United States, we included adolescents with alcohol and other
drug diagnoses and co-occurring psychiatric disorders (as
long as they could be managed at the outpatient level), as well
as those with only cannabis abuse diagnoses, and/or less than
weekly substance use. Adolescents were excluded if they
were inappropriate for short-term outpatient treatment or
would be unable to participate in the study. The exclusion
criteria were: (a) reported use of alcohol 45 or more of the
90 days prior to intake; (b) reported use of other drugs 13 or
more of the 90 days prior to intake; (c) reported an acute
medical or psychological problem that was likely to prohibit
full participation in treatment; (d) had insufficient mental
capacity to understand and provi de infor med consent or
participate in treatment; (e) live d outside of the programTs
catchment area; (f) had a hist ory of repeated, violent behavior
or severe conduct disorder that might put other participants at
risk; or (g) lacked sufficient ability to use English to
participate in the consent process, treatment, or research
interviews. Assuming that they met the inclusion and
exclusion criteria, both the adolescent and the parent or other
collateral were asked to participate in the study. The study
was conducted with the informed consent of the participants,
under a federal certificate of confidentiality.
2.2.3. Participant characteristics
As shown in Table 1 , participants were primarily male
(83%), white (61%) or African American (30%), enrolled in
school (87%), and currently involved in the juvenile justice
system (62%). Half were from single parent families. Most
of the adolescents began using alcohol or other drugs before
the age of 15 (85%) and used cannabis weekly or daily
(71%). Many reported engagement in risky behaviors such
as multiple sexual partners (39%), sex without barrier
protection (23%), and substance use in hazardous situations
(54%). Only about a quarter of these adolescents had
participated in substance abuse or mental health treatment
previously. There were no differences between treatment
conditions within site or across sites within the same trial on
any of these variables. Adolescen ts in the Trial 2 sites were
more likely than those in the Trial 1 sites to be African
American, female, and to be sexually active, as well as less
likely to be employed. All of the analyses presented here
were computed by trial, without any pooling of data across
the two trials.
As shown in Table 2, most participants had serious
substance-related disorders and a wide range of co-occurring
disorders. Based on self -report only, 86% met criteria for a
cannabis related disorder, and this percentage increased to
100% after considerin g additional information from parents,
juvenile justice, treatment records or other objective sources.
Over 95% self-reported one or more other problems (84%
had three or more) such as alcohol use disorders (37%), other
substance use disorders (12%), internalizing disorders (18%
major depression, 23% generalized anxiety, 9% suicidal
thoughts or actions, 14% traumatic stress disorders), and
externalizing behavioral disorders (53% conduct disorder,
38% attention def icit-hyperactivity disorder, including 30%
with both). Over half (60%) reported a history of physical,
sexual, or emotional victimizatio n (including 37% with
extensive patterns of victimization). A high percentage of
participants also reported illegal activity other than just drug
possession or use (83%) and 66% reported engaging in acts of
physical violence such as assault. There were no differences
between treatment conditions within either trial on any of
these variables. There were, however, a few significant
differences by condition within site (Operation PAR on
cannabis disorders; CHS on major depression, any illegal
activity, property and drug crime; and CHOP on weekly
alcohol use in the home by others and having sex without
protection). These differences were found for less than 3%
of the 224 tests (based on 56 variables
4 sites) or less
than expecte d by chance if using the .05 probability level
for alpha.
2.2.4. Generalizability
Comparisons with national admission data for adoles-
cents presenting for public outpatient treatment of cannabis
use problems suggest that the CYT adolescents were more
likely to be male (83% vs. 73%), non-white (39% vs. 30%),
15 years or older (85% vs. 68%), and similar in terms of the
percent breferredQ by the criminal justice system (52% vs.
51%). Clinically, they were more likely to have first used
under the age of 15 (85% vs. 78%), be using weekly at intake
(71% vs. 47%), and similar in their rates of being dependent
(46% vs. 45%) and having prior treatment episodes (26% vs.
27%). See Tims et al. (2002) for more detail.
2.3. Interventions
Below is a description of the five interventions and the
rationale underlying their selection. More detailed compari-
sons of them can be found elsewhere (Diamond et al., 2002).
2.3.1. Motivational Enhancement Treatment/Cognitive
Behavior Therapy, 5 Sessions (MET/CBT5)
MET/CBT5 (Sampl & Kadden, 2001) consisted of tw o
individual MET sessions and three group CBT sessions, with
the total duration of treatment lasting 6 to 7 weeks. Evaluated
in both trials, MET/CBT5 was an inexpensive first tier
intervention specif ically designed for the 6-week median
length of stay that occurs in much of the U.S. treatment sys-
tem. The MET component proceeds from the assumptions
that adolescents need to: (a) resolve their ambivalence about
whether they have a problem with cannabis and other
substances, and (b) increase their motivation to stop using
cannabis. Therapists using MET seek to help adolescents see
the relationship between cannabis use and its consequences
so they will conclude that the costs of cannabis use outweigh
its benef its. Several studies with adult cannabis users had
demonstrated that one or two sessions of MET were more
effective in reducing substance use than a no-treatment
control group (Copeland, Swif t, Roffman, & Stephens, 2001;
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213200
Babor et al., in press; Stephens, Roffman, & Curtin, 2000)
and that MET plus CBT was better than a control group or
MET alone (Copeland et al., 2001; Babor et al., in press).
CBT sessions were added out of concern that problem
recognition and motivation alone would not be sufficient
for adolescents who had yet to develop the necessary
coping skills to initiate and sustain change. Waldron et al.
(2001) tested a similar combination of MET and CBT
with adolescents. CBT strategies have been found to be
effective for cannabis problems in adults (Stephens,
Roffman, & Simpson, 1993). CBT also has a history of
being well received by community practitioners (Morgen-
stern & McCrady, 1992). The CBT component of this
particular intervention teaches basic skills for (a) refusing
offers of cannabis, (b) establishing a social network sup-
portive of recovery, (c) developing a plan for pleasant ac-
tivities to replace cannabis-related activities, and (d) coping
with unanticipated high-risk situations, problem solving,
and recovering from relapse, should one occur.
2.3.2. Motivational Enhancement Treatment/Cognitive
Behavior Therapy, 12 Sessions (MET/CBT12)
This treatment condition supplemented MET/CBT5 with
seven additional CBT sessions (CBT7; Webb et al., 2002)
in a group format, with the combined duration lasting 12 to
14 weeks. Evaluated in Trial 1, MET/CBT12 was designed
to more closely approximate group interventions used in
many community-based treatment programs. The additional
CBT sessions were designed to teach adolescents coping
skills they could use for resolving interpersonal problems
and negative affect and for addressing triggers for cannabis
use and psychological dependence. The additional
sessions address problem-solving, anger awareness, anger
management, communication skills, resistance to craving,
Table 1
Participant characteristics by condition and overall
Trial 1 Trial 2
Total CYT
(n = 600)
MET/CBT5
(n = 102)
MET/CBT12
(n = 96)
FSNM
(n = 102)
MET/CBTS
(n = 100)
ACRA
(n = 100)
MDFT
(n = 100)
Demographics
Female 19% 14% 16% 21% 20% 15% 79%
Caucasian/White 79% 71% 70% 47% 53% 47% 61%
African American/Black 9% 14% 15% 50% 44% 47% 30%
Hispanic/Latino 5% 6% 7% 2% 1% 1% 4%
Other/Mixed 7% 9% 9% 1% 2% 5% 6%
Aged 13-14 15% 16% 23% 11% 14% 13% 15%
Aged 15-16 54% 64% 52% 57% 56% 48% 55%
Aged 17-18 31% 21% 26% 32% 30% 39% 30%
Single parent family 44% 42% 49% 53% 59% 52% 50%
In school in the past 90 days 92% 94% 86% 79% 86% 85% 87%
Employed in the past 90 days 62% 60% 47% 37% 39% 36% 47%
Ever homeless/runaway 5% 10% 8% 9% 5% 5% 7%
Current CJS Involvement 58% 60% 54% 72% 62% 67% 62%
Pattern of substance use
Age of first use under 15 81% 82% 85% 80% 89% 89% 85%
Any weekly or daily substance use 72% 76% 76% 76% 73% 80% 75%
Weekly or daily marijuana use 67% 70% 74% 75% 68% 74% 71%
Weekly or daily alcohol use 22% 17% 11% 19% 15% 18% 17%
Weekly or daily use of other drugs 2% 0% 1% 0% 0% 1% 1%
Weekly or daily tobacco use 75% 68% 72% 74% 73% 78% 73%
Weekly alcohol use by others in home 31% 26% 28% 13% 26% 16% 23%
Weekly drug use by others in home 11% 6% 13% 9% 16% 10% 11%
Other risk behaviors in 90 days before intake
Sexually active 68% 60% 64% 81% 78% 81% 72%
Multiple sexual partners 32% 24% 37% 50% 47% 42% 39%
Had sex without barrier protection 21% 12% 19% 29% 23% 37% 23%
Any needle use 1% 1% 0% 0% 3% 0% 1%
Used at work, school, or while baby sitting 47% 35% 41% 27% 36% 29% 36%
Used in hazardous situation
1
52% 53% 49% 53% 60% 55% 54%
Lifetime history of behavioral interventions
Substance abuse treatment 30% 23% 28% 30% 19% 25% 26%
Mental health treatment 30% 25% 28% 23% 16% 15% 23%
Juvenile justice system involvement 86% 83% 84% 86% 80% 80% 83%
Note. Within each Trial, the differences between condition were not statistically significant.
1
Such as when driving car, using a machine, playing sports, or where the adolescent might have been forced into sex.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 201
depression management, and management of thoughts about
cannabis. In addition to producing a treatment stay and
intensity close r to existing adolescent treatment practice (at
least as intended), this dosage of MET/CBT was closer to the
combined dosage used in earlier adult studies that had proven
to be more effective than MET alone (Copeland et al., 2001;
Babor et al., in press). A key assumption underlying the
group format for CBT delivery is that adolescent skill
deficits are typically inter-personal and m ore healthy
behaviors can be learned in a safe, social context.
2.3.3. Family Support Network
FSN (Bunch, Hamilton, Tims, Angelovich, &, McDou-
gall, 1998) used MET/CBT12 to provide adolescent-
focused substance abuse treatment and added six parent
education group meetings (to improve parent knowledge
and skills relevant to adolescent problems and family
functioning), four therapeutic home visits, referral to self-
help support groups, and case management (to promote
adolescent/parent engagem ent in the treatment process).
Evaluated in Trial 1, FSN was designed to more closely
Table 2
Clinical Characteristics by Condition and Overall
1
Trial 1 Trial 2
Total CYT
(n = 600)
MET/CBT5
(n = 102)
MET/CBT12
(n = 96)
FSNM
(n = 102)
MET/CBT5
(n = 100)
ACRA
(n = 100)
MDFT
(n = 100)
Substance use disorder
Any marijuana disorder
2
86% 85% 97% 81% 80% 85% 86%
Marijuana dependence 44% 40% 46% 47% 47% 52% 46%
Marijuana abuse 42% 46% 51% 34% 33% 33% 40%
Any alcohol disorder 32% 35% 28% 41% 41% 42% 37%
Alcohol dependence 10% 6% 7% 7% 13% 12% 9%
Alcohol abuse 22% 29% 22% 34% 28% 30% 28%
Other substance use disorders 16% 6% 14% 13% 9% 13% 12%
Other substance dependence 3% 2% 4% 1% 1% 1% 2%
Other substance abuse 13% 4% 10% 12% 8% 12% 10%
Past year psychological problems
Any internal disorder 30% 30% 36% 41% 34% 30% 33%
Major depression 19% 18% 16% 27% 16% 11% 18%
Generalized anxiety disorder 19% 16% 22% 34% 21% 25% 23%
Suicidal thoughts or actions 8% 10% 10% 10% 7% 7% 9%
Any traumatic distress disorder
3
14% 11% 16% 17% 14% 12% 14%
Any external disorder 62% 55% 59% 63% 61% 64% 61%
Conduct disorder 52% 51% 47% 56% 54% 58% 53%
Attention deficit-hyperactivity disorder 41% 32% 44% 34% 38% 38% 38%
Physical, sexual or emotional victimization
Lifetime history of victimization 66% 55% 49% 63% 66% 61% 60%
Past year 42% 33% 35% 37% 40% 32% 37%
Past 90 days 20% 19% 19% 22% 26% 18% 20%
Extensive victimization
4
34% 30% 34% 42% 46% 37% 37%
Violence and illegal activity (other than possesion/use)
Acts of physical violence
5
66% 64% 63% 66% 64% 71% 66%
Any illegal activity 87% 81% 86% 84% 74% 84% 83%
Property crimes
6
61% 52% 60% 56% 41% 59% 55%
Interpersonal crimes
7
43% 44% 41% 48% 41% 50% 45%
Drug related crimes
8
66% 59% 69% 66% 57% 70% 65%
Number of substance psychological, behavioral or legal problems
9
One 2% 5% 6% 3% 8% 3% 5%
Two 14% 14% 10% 11% 9% 8% 11%
Three to twelve 84% 81% 84% 84% 81% 87% 84%
1
Within each Trial, the differences between condition were not statistically significant, such as when driving a car, using a machine, playing sports,or
where the adolescent might have been forced into sex or hurt.
2
Of the remaining 84 included in the study, 82 do meet criteria for lifetime cannabis (54) or abuse (30) if we also considered their answers to other
questions the reports of their parents treatment records, and other objective information.
3
Post traumatic distress, acute traumatic distress or disorders of extreme stress not otherwise specif ied.
4
Reporting 4 or more of the following types of victimization, traumagenic factors (e.g., multiple people, someone they trusted, fearing for life sexual
penetration people didn’t believe them for continuing fear it will reoccur.
5
Physical assault of another person within the past year.
6
Self report of or arrests related to vandalism, forgery, bad checks, shop lifting, theft, robbery, auto theft.
7
Self report of or arrests related to assault, aggravated assault with a weapon, rape, murder, and arson.
8
Self report of or arrests related to driving under the inf luence, manufacture or distribution, prostitution, gang involvement.
9
Counting each individual drug diagnosis and psychological problem, whether they have been victimized in the past year, whether they are physically
violent and whether they have had any illegal activity (other than just possession or use) during the past year.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213202
approximate the kind of bcomprehensive treatmentQ recom-
mended by adolescent treatment experts (CSAT 1992a,
1992b, 1993) for adolescents. The parent education groups
provided information on (a) adolescent development and
parentsT role, (b) substance abuse/dependence, (c) recovery
process and relapse signs, (d) family development and
functioning, (e) family organization and communication,
and (f) family systems and roles . The home visits focused
on (a) initial assessment and motivation-building, (b) family
roles and routines, and (c) assessing progress and building
commitment to change. Case management was used to
facilitate treatment attendance (reminders, transportation,
childcare), assess family needs, and make referrals to other
community services. The family component of FSN was
based on evidence that treatment outcomes were improved
when parent education was provided to at-risk adolescents,
family support interventions were added to treatment
(Brown, Myers, Mott, & Vik, 1994), and families were
actively engaged in treatment (Henggeler et al., 1991, 2002;
Liddle et al., 2001). The key assumption underlying the
FSN manual-guided approach is that in addition to problem
recognition, motivation, and improved coping skills, family
education and case management are needed to help the
adolescent achieve recovery.
2.3.4. The Adolesc ent Community Reinforcement Approach
ACRA (Godley, Meyers, et al., 2001) is composed of
10 individual sessions with the adolescent, four sessions
with caregivers (two of which are with the whole family)
and a limited amount of case management provided by the
therapist over a period of 12 to 14 weeks. Evaluated in Trial
2, ACRA incorporates elements of operant conditioning,
skills training, and a social systems approach. Three core
procedures used in ACRA sessions are (a) functional
analyses to identify the antecedents and consequences of
substance use and pro-social behaviors; (b) identifying
and reviewing clear, simple and obtainable bgoals of
counseling,Q and (c) using a rating scale to track the ado-
lescentTs satisfaction in multiple life areas to inform further
goal planning. Other procedures include identifying and
reinforcing pro-social behaviors that compete with sub-
stance use and skills training related to relapse prevention
and problem solving. Four parent sessions are devoted to (a)
an overview of the ACR A approach; (b) a review of
important parenting practices for helping adolescents stay
alcohol and drug free; (c) increasing positive communica-
tion in the family, and (d) problem-solving. Two of the four
sessions bring the parents and adolescents together for
practice of communications and problem-solving. Com-
munity Reinforcement Approaches have been effective with
adult substance users (Meyers, Miller, Hill, & Tonigan,
1999; Meyers & Godley, 2001), have been recommended as
one of the most promising approaches to treatment by
several expert panels from the Institute of Medicine (1989,
1998) and have been successfully combined with other
approaches including contingency contracting (Azrin et al.,
1994; Budney & Higgins, 1998) and family therapy
(Henggeler et al., 2002; Randall, Henggeler, Cunningham,
Rowland, & Swenson, 2001; Sisson & Azrin, 1993). The
ACRA approa ch also drew on othe r work relat ed to
effective parenting practices (Ary, Duncan, Duncan, &
Hops, 1999; Bry, Catalano, Kumpfer, Lochman, & Sza-
pocznik, 1998).
2.3.5. Multidimensional Family Therapy
As imp lemented in CYT, MDFT (Liddle, 2002) is com-
posed of 12 to 15 sessions (typically six with the adoles-
cent, three with parents, and six with the whole family) and
case management provided over a period of 12 to 14 weeks.
Evaluated in Trial 2, MDFT proceeds in three phases: (a)
setting the stage (engaging adolescents, engaging parents,
building alliances with all members of the system, iden-
tifying goals of treatment), (b) working the themes for
adolescents (trust/mistrust, abandonment and rejection,
disillusionment and past hurts, motivation and self-agency,
hope or lack of hope for the future, credibility) and families
(preparing for adolescent-parent communications, managing
conversation, shifting from high conf lict to affective issues,
developing positive experiences/interactions with each
other, tying conversation and themes to drug use), and (c)
sealing the changes (preparing for term ination, reviewing
treatment work, preparing for future challenges). MDFT is
based on research linking reductions in adolescentsT drug
and problem behavior to changes in parenting practices
(Schmidt, Liddle, & Dakof, 1996), therapist-adolescent
alliance (Diamond & Liddle, 1996), and the use of culturally
specific themes to engage African American males (Jackson-
Gilfort, Liddle, Tejeda, & Dakof, 2001) and females (Dakof,
2000). Unlike FSN, where family therapy was added to an
adolescent-focused treatment component, MDFT integrates
treatment for substance use into family therapy. A key
assumption of MDFT is that adolescents are involved in
multiple systems (e.g., family, peers, school, welfare, legal)
that produce multiple risk factors that can best be addressed
in a family-based, developmental-ecological, multiple sys-
tems approach. MDFT had promising data from studies
conducted prior to and concurrent with CYT and has been
demonstrated to be more effective than multi-family educa-
tion, adolescent group therapy, or CBT only (Liddle, 2001;
Liddle et al., 2001; Liddle, Rowe, Henderson, Dakof, &
Ungaro, in press). It has also been identified as a Best Prac-
tice by the U.S. Department of Health and Human Services
(http://phs.os.dhhs.gov/ophs/BestPractice/mdft_miami.htm),
a SAMHSA Model Program (http://modelprograms.samhsa.
gov), and an effective drug abuse treatment approach by
NIDATs Behavioral Therapies Development Program (http://
www.nida.nih.gov/BTDP/Effective).
2.3.6. Staff characteristics
Therapists employed in this study had a range of
educational backgrounds (20% doctorates, 50% masters,
30% bachelors) and averaged 7 years of clinical experience.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 203
Though most of the therapists were experienced clinicians,
this was the first time most had used a manual-guided
therapy (Godley, White, et al., 2001).
2.3.7. Staff training and supervision of clinical staff
After training for the intervention they were to deliver,
clinical staff taped their sessions for review. Each treat-
ment clinical coordinator reviewed audio or videotapes of
all sessions provided by each therapist until he or she was
certified as proficient in that intervention. Weekly super-
vision continued throughout the study and included
review of at least two therapy tapes per month to prevent
therapist drift. During tape reviews, the clinical coordina-
tors completed treatment-specific rating forms to monitor
adherence and provide feedback to therapists. Since sites
and the clinical coordinators were geographically dis-
tanced from each other, the group supervision meetings
occurred during frequent conference calls. These meetings
provided clinical coordinators the means of en suring
consistency in the delivery of an intervention across
disparate locations.
2.3.8. Treatment received by condition
Table 3 shows the type and quantity of services received
by conditions based on the daily service logs completed by
each therapist. These data demonstrate that the conditions
differed significantly and as planned with regard to modal-
ity of service s (e.g., participant only, multiple participant
groups, multi-family groups, parent/collateral only, family
counseling, case management), hours of contact, length of
stay, days of contact, and treatment completion. In Trial 1,
the conditions were designed to compare a range of dosages
and modalities. When compared to MET/CBT5 partici-
pants, MET/CBT12 participants were provided more group
treatment, days of contact, and had longer lengths of stay.
FSN was the only intervention to provide multi-family,
collateral, family therapy, and case management services
and provided the most total hours of service. In Trial 2,
each condition focused on different modalities. MET/CBT5
was the briefest individual and group approach without any
family involvement, while ACRA was primarily provided
on an individual basis to the adolescent with some parent
only and family sessions, and case management. MDFT
provided the most hours of family therapy. Of the
adolescents assigned to one of the four 12- to 14-week
treatment interve ntions , 52% had lengths of stay that
reached 90 days and 86% stayed 6 or more weeks. For
the 6- to 7-week MET/CBT5 condition, 62% stayed in
treatment for 6 or more weeks. The steering committee
agreed that treatment completion woul d be defined as
having completed 75% of the planned dosage for a given
treatment. This was defined as at least 200 min of therapy
for MET/CBT5, at least 400 min of therapy for MET/
CBT12, ACRA and MDFT, and at least 800 min for FSN.
Based on this definition, 71% of the adolescents completed
treatment, 22% received a partial dosage (i.e., less than the
standard above), and 5% were randomized but n ever
participated in any hours of their assigned treatment.
The bottom of Table 3 also repeats the average societal
cost per episode of care as estimated by French and
colleagues (2002). This estimate considers the market value
of all direct program resources (e.g., personnel, supplies and
materials, contracted services, buildings and facilities,
equipment, and miscellaneous items) after partialling out
the research costs and dividing the remaining costs among
the three interventions within a given site. There were
significant differences in the costs of treatment by con-
ditions across and within each site in Trial 1. In Trial 2, the
differences were significant across and within sites, but in a
different order by site. For example, MET/CBT5 cost the
least at Site 3, but ACRA cost the least at Site 4. It is also
interesting to note that the average cost across conditions
was much higher at Site 4 than Site 3 ($2,118 vs. $1,194).
Below in the data source section is a summary of how costs
were estimated; see French et al., 2002 for a detailed
discussion of the method as well as other geograp hical and
management factors that may also explain these differences.
It is important to acknowledge that the cost estimates do
not reflect potential changes in other costs to society (e.g.,
subsequent treatment, missing school, dentition) discussed
elsewhere (French et al., 2003). We have previously
demonstrated that the direct costs of the CYT interventions
(reported in Table 3) were largely offset by changes in other
costs to society in 12 to 30 months (see Dennis, 2003;
French et al., 2003). However, these savings came from
important but tertiary outcomes (e.g., reductions in service
utilization, arrests, days in detention), occur in only a
fraction of the cases, and were often related to multiple other
co-occurring problems. No value could be placed on the
core clinical outcomes (e.g., days of abstinence or percent in
recovery) that are the focus of this paper and an value could
not be econometrically modeled because the changes in
costs society (from the tertiary outcomes) were not
significantly correlated (r b.10, n.s.d.) with the changes in
clinical outcomes. In other words, while knowledge about
benefit costs is useful in its own right, it is not sufficient to
evaluate the clinical effectiveness or cost-effectiveness of
the interventions.
2.4. Data and measures
2.4.1. Data source
Data were collected from several sources including
participant interviews, collateral interviews, urine tests,
service logs, and other process measures (see Dennis,
Titus, et al., 2002 for a complete list). The participant
characteristics, diagnoses, and primary outcomes were
measured with the GAIN (Dennis, 1999; www.chestnut.
org/li/gain). The GAIN is a standardized semi-structured
interview with 8 main sections (background, substance use,
physical health, risk behaviors, mental health, environ-
ment, legal, vocational) that is designed to support diagnosis,
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213204
placement, outcome monitoring, and economic analysis.
Self-reports of cannabis use were consistent at intake and
various followup waves (kappa of .7 to .9) with family/
collateral reports, on-site urine tests, and gas chromatog-
raphy/mass spectrometry tests for delta-9-tetrahydrocanna-
binol (D
9
-THC; Buchan, Dennis, Tims, & Diamond, 2002).
A test-retest reliability study of key GAIN measures with
210 adolescent outpatients revealed consistent reports of
days of cannabis use (rho = .7), days of alcohol use (rho = .7),
lifetime abuse/dependence symptoms (rho = 2.7), and life-
time dependence diagnosis (Kappa = .6; Dennis, Titus, et al.,
2002). The GAIN has also been found to accurately predict
diagnoses of co-occurring psychiatric disorders that were
made by independent staff blind to GAIN findings (kappas
ranging from .69, adjustment disorder, to 1.0, ADHD;
Shane, Jasi ukaitis, & Green, 2003). The GAINTs dimen-
sional measures of psychological problems using parent
report data correlated .56 to .68 with their counterparts in
the Child Behavior Checklist; Achenbach & Edelbrock,
1983; Diamond et al., in press). Self-reported treatment
utilization data from the GAIN have also been found to be
largely consistent with staff/agency records (r = .78; Godley
et al., 2002).
Cost estimates are based on data collected with the Drug
Abuse Treatment Cost An alysis Pro gram (DATCAP;
French, 2001; www.DATCAP.com). Designed to measure
both the accounting and opportunity costs of a substance-
abuse treatment program based on standard economic
principles, DATCAP has been widely used with adult
treatment, but this was its first use with adolescent programs.
It is used to colle ct data on pro gram resources (e.g.,
personnel, supplies and materials, contracted services, build-
ings and facilities, equipment, and miscellaneous items),
program revenues and client case flow. In CYT, the DATCAP
was supplemented with service contact logs completed by
therapists and case managers (see Dennis, Titus, et al., 2002).
A more detailed discussion of the use of DATCAP in CYT
has been reported elsewhere (see French et al., 2002, 2003).
2.4.2. Clinical outcomes
The two clinical outcomes used were (a) days of
abstinence between the randomization date and the
12-month followup interview and (b) whether the adoles cent
was in recovery at the end of the study. Days of abstinence
(from cannabis, alcohol and other drugs) were summed
across all fou r quarterly followup waves, using t he
adolescentTs average days abstinent to fill in any missing
data. Being in recovery at the end of the study was defined
as living in the community (vs. incarceration, inpatient
treatment, or other controlle d environment) and reporting no
past month substance use, abuse or dependence problems at
the 12-month interview. For the 6% of the adolescents who
Table 3
Treatment dosage by condition within each trial
Trial 1 Trial 2
Total
MET/CBT5
(n = 102)
MET/CBT12
(n = 96)
FSNM
(n = 102)
MET/CBT5
(n = 100)
ACRA
(n = 100)
MDFT
(n = 100)
Mean Hours
Participant only
b
1.8 1.9 1.8 1.8 5.2 2.8 2.5
Multiple participant group
a,b
3.1 7.3 7.8 2.1 0.0 0.0 3.4
Multi-family group
a
0.0 0.0 7.8 0.0 0.0 0.0 1.3
Parent/collateral only
a,b
0.1 0.0 0.4 0.0 1.4 1.3 0.5
Family counseling
a,b
0.0 0.0 3.8 0.0 1.4 5.4 1.8
Total therapy sessions hours
a,b
5.0 9.2 21.6 3.8 7.9 9.5 9.5
Case management
a,b
0.5 0.5 1.8 1.0 2.8 4.6 1.9
Total of any service hours
a,b
5.5 9.7 23.4 4.8 10.7 14.2 11.4
Length of stay (mean days)
a,b,d
44.4 78.4 89.7 41.1 73.4 77.5 67.4
Percent with 6+ weeks
a,b
65% 83% 89% 59% 85% 85% 78%
Percent with 90+ days
a,b
2% 55% 73% 1% 39% 40% 35%
Days of contact
Days of 1+ Therapy Sessions
a,b
4.2 7.7 11.6 3.4 7.9 9.7 7.4
Days of Any Contact
a,b
5.5 9.3 18.9 7.8 13.6 24.2 13.2
Treatment completion
a,c
Completed (column %) 87% 67% 79% 60% 61% 70% 71%
Partial (column %) 11% 31% 15% 31% 35% 26% 25%
None (column %) 2% 2% 6% 9% 4% 4% 5%
Treatment Episode cost
a,b,e,f
$1,113 $1,185 $3,246 $1,558 $1,408 $2,002 $1,758
a
Significant differences ( p b .05) by condition in Trial 1.
b
Significant differences ( p b .05) by condition in Trial 2.
c
Based on 0 minutes for none, and the percent below (partial) and equal or above (Full) the following amount of therapy treatment: 200 minutes for MET/
CBT5, 400 for MET/CBT12, ACRA and MDFT, and 800 for FSN.
d
Based on days from randomization to last therapy session of treatment.
e
Significant difference ( p b .05) by site and condition within site 0 but in different directions (see Appendix at http://www.chestnut.org/li/downloads/
Dennis
_
et
_
al
_
in
_
press
_
CYT
_
MF
_
Appendix.pdf ).
f
Economic cost from a societal perpective as reported in French et al. 2002.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 205
did not complete their 12-month interview, data from their
previous followup interview was used to determine their
recovery status.
2.4.3. Cost effectiveness
The two economic outcomes used were the (a) cost per
day of abstinence (CPDA) over the 12-month followup
period and (b) cost per person in recovery at the end of the
study. The economic costs of each episode of care was
estimated in 1999 dollars from a societal perspective (i.e.,
market value of goods and services used) and previously
reported (French et al., 2002). Overall estimates were made
for each of the three conditions within each of the four sites
(12 total). Within site and condition, these costs were
prorated based on the hours of therapy sessions received by
a given adolescent, divided by the average hours for all
individuals in the same condition and site. CPDA was then
calculated by dividing the adolescentTs prorated costs (PC)
by his or her total days abstinent (TDA; i.e., for individual i,
CPDA
i
=PC
i
/ TDA
i
). Since it is impossible to divide by a
0/1 measure, CPPR was calculated as the individualTs
prorated costs divided by the percent of individuals for the
same therapy condition (within site) that were in recovery
(PR); that is, for individual i, in therapy condition t and
site s, CPPR
i
=PC
i
/PR
ts
.
2.5. Procedures
2.5.1. Staff training and supervision of research staff
All field staff were centrally trained to use the GAIN
interview by its developer and supervised by both a cross-
site and a local research coordinator. The cross-site research
coordinator reviewed and certified each of the on-site
research coordinators using taped interviews; then the local
coordinator reviewed and certified the siteTs research staff
through tape review or direct observation. Questions related
to data collection and research procedures were addressed by
local site coordinators and the cross-site coordinator con-
ducted an annual review of all research materials to maximize
adherence to the studyTs research methods. Data analysts
regularly reviewed raw data to identify any implementation
problems that could be addressed with additional training
(Dennis, Titus, et al., 2002 ). Economic data was collected
by an independent team under the direction of Dr. French
(a health economist and author of the DATCAP).
2.5.2. Recruitment
Participants were recruited from the existing case flow of
the sites and through outreach to the juvenile justice system,
schools, doctors and public service announcements from
1998 to 2000. Of the 1244 adolescents screened, 44% were
ineligible based on the inclusion and exclusion criteria (20%
being too severe for outpatient treatment, 24% not being
severe enough). Of the 702 who were eligible, 600 (85%)
agreed to participate. See Dennis, Titus, et al., (2002) for
more details.
2.5.3. Randomization
Within each site, eligible adolescents were assigned to
one of the three local conditions using a randomly ordered
list that was generated by independent research staff at the
coordinating center using Microsoft Excel. To prevent any
bias in the assignmen t process, research staff were only able
to assign an adolescent after he or she was determined
eligible and had completed the intake assessments. Since
clinical staff needed to be trained in the specific interven-
tion they were providing, they could not be blind to a
participantTs assignment. Assignment logs were kept in a
locked file cabinet and were never access ible to clinical
staff. To prevent bias at followup, tracking and followup
logs were maintained separately from assignment logs.
Unique identification numbers were assigned to every
adolescent screened and used by the coordinating center to
audit the randomization process.
2.5.4. Data collection and processing
Intake and 3-, 6-, 9- and 12-month followup interviews
were conducted by research staff. While parti cipation in the
study was voluntary, participants were compensated ($25 to
$50 depending on site) for each follow up assessment, and
to maximize followup completion rates, an extensive
tracking protocol was used (see Dennis, Titus, et al.,
2002; Scott, in press). Of the 600 adolescents randomized,
one or more followup interviews were completed on 99%
(n = 597), including 98% at 3 months, 97% at 6 months,
96% at 9 months, and 94% at 12 months.
2.5.5. Analytic procedures
Following the Consolidated Standards of Reporting
Trials (Moher, Schultz, & Altman, 2001), all analyses were
conducted with an bintent-to-treatQ approach. Thus, all
adolescents were included in analyses as assigned, including
approximately 5% who did not actually receive any treat-
ment. Analyses were conducted with SPSS (2001) by site
and across sites for each of the two trials. The baseline
clinical measures (days of abstinence or % in recovery) were
included as covariates to allow for individual differences.
Within each trial, site differences were modeled with a
dummy variable. Reflecting the randomized block design,
conditions were modeled as ne sted within site, which
produces a statistic for the significance of site effects,
conditions across site effects, and conditions within site
effects. Logistic regression was used to analyze differences
for the percent in recovery at 12 months, as this is a
dichotomous outcome. Where there were significant differ-
ences by conditi on, Tukey multiple range tests were
conducted to verify which condition or conditions were
different in pair-wise comparisons. Statements about the
size of an effect or trends are based on Cohen’s (1988) effect
size f (for multiple groups), with 0.10 being considered
small, 0.20 moderate, and 0.40 or more large.
In CYT, about one in five adolescents report no use in
the month immediately preceding intake. Rev iew of these
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213206
cases suggests that this most likely was due to a
combination of recent arrest/urine monitoring, or because
an adolescent entered treatment from a controlled environ-
ment (e.g., following residential treatment; detention,
hospital disch arge). Several alternative analysis were con-
ducted with past 90 day scales as covariates to verify that
these findings held with alternative approaches and mea-
sures. They did not change the overall pattern of results and
are available from JSAT in an on-line appendix at http://
www.chestnut.org/LI/downloads/Dennis
_
et
_
al
_
CYT
_
MF
_
Appendix.pdf ).
Since there was no control or treatment as usual
condition, cost and cost-effectiveness measures were com-
pared to the average for a given site or trial as recommended
by Gold and colleagues (1996). In both of the cross-site
analyses and three of the four site analyses, the results show
that one condition economically bdominatesQ the others (i.e.,
it is both the most effective and least expensive), so cost-
effectiveness ratios are neither necessary nor appropriate
(Gold et al., 1996). For the remaining site, we have reported
a cost-effectiveness ratio (CER) in the text to evaluate the
return of increased costs in terms of increased effectiveness.
This ratio is calculated as the difference between the
conditionTs cost and the site average (DC) over the differ-
ence in the conditionT s outcomes and the siteTs average
outcome (DO; i.e., for treatment t in site s, CER = (C
t
C
s
)/
(O
t
O
s
)=DC/DO). Conditions with lower ratios are
considered to be the more bcost-effective,Q which means that
further investment in them has a higher than expected rate of
return in terms of the changes in outcomes.
3. Results
3.1. Clinical outcome analysis
Fig. 1 shows the general pattern of clinical outcomes
across sites and conditions by quarter. The days of abs-
tinence per quarter (the solid line on top) increased from 52
(of 90) in the quarter before intake to an average of 65 days
per quarter (+24%) across the four followup periods. The
overall change occurred during active treatment (from
intake to month 3) and was stable across followup, though
individuals did vary (intraclass correlation coefficient
[ICC] = .47). The percent of adolescents in recovery at
each interview increased from 3% at intake to an average of
24% across the four followup periods. Again, across
conditions and sites, change occurred during active treat-
ment, was stable across followup waves, and individual
adolescents continued to move in and out of recovery
(ICC = .33). Fig. 2, summarized below, shows the difference
by condition for each trial in the total days of abstinence.
Abstinence is summed across the four followup waves
and the percent in recovery at the end of the study in month
12. (Site level data is available from JSAT in an on-line
appendix at http://www.ch estnut.org/LI/downloads/Dennis
_
et
_
al
_
CYT
_
MF
_
Appendix.pdf ).
In Trial 1, the total days of abstinence (summed across
the four followup waves) was not significantly different by
site or condition (within or across sites). The percent in
recovery at the end of the study was significantly different
by condition overall (CohenTs f = 0.12, p b .05) with MET/
CBT5 (27%) having the highest percent in recovery,
followed by FSN (22%) and MET/CBT12 (17%). However,
the pair-wise differences were not large enough to reach
significance using a Tukey multiple range test. These
f indings held both across and within sites.
In Trial 2 , the total days of abstinence were not
significantly different by site or condition (within or across
sites). The percent in recovery was not significantly differ-
ent by condition across sites, though there was a small trend
(CohenTs f = 0.16) for ACRA (34%) to have a slightly
higher percent of participants in recovery than MET/CBT5
(23%) and MDFT (19%). This finding was driven by site 3
(CHS), where within site there was a moderate sized
significant difference by condition (CohenTs f = .20, p b .05)
0
10
20
30
40
50
60
70
80
90
Intake 3 6 9 12
Days Abstinent Per Quarter.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
% in Recovery at the End of the Quarter.
Days Abstinent
Percent in Recovery
Fig. 1. Illustration of treatment and stability effects across trials, sites and conditions (n = 597).
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 207
Fig. 2. Clinical outcomes by condition.
Fig. 3. Economic efficiency of CYT Therapies across site.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213208
with ACRA (40%) having a higher percent in recovery
than MDFT (22%) and MET/CBT5 (18%). However, the
pair-wise differences were not large enough to r each
significance using a Tukey multiple range test.
3.2. Cost-effectiveness analysis
Given the differences by condition in cost presented
earlier (see Table 3 ) and the similarity of clinical out-
comes ( Fig. 2), the next logical step was to consider
the economic efficiency with which the conditions achieve
their clinical outcomes. Across trials and conditions, the
average cost of CYT interventions per day of abstinence
achieved over the next 12 months was $8.72 per day and
the average cost per person in recovery at the end of the
study was $8,231. Fig. 3 shows the difference by condition
for each trial in these measures and is summarized below.
(Site level data and detailed calculations are available from
JSAT in an on-line appendix at http://www.chestnut.org/LI/
downloads/Dennis
_
et
_
al
_
CYT
_
MF
_
Appendix.pdf ).
In Trial 1, the average cost per day of abstinence over the
12 months post intake was $8.79 and varied significantly by
condition (CohenTs f = 0.48, p b .05). Based on Tukey range
tests, the primary difference was that MET/CBT5 ($4.91)
and MET/CBT12 ($6.15) had significantly lower cost per
day of abstinence than FSN ($15.13). This pattern held at
both Site 1 ($5.75 and $7.67 vs. $17.04; f = 0.40, p b .05)
and Site 2 ($4.17 and $5.00 vs. $13.80; f = 0.63, p b .05).
The average cost per person in recovery at the end of the
study was $8,846 and varied significantly by condition
(CohenTs f =0.72, p b .05), with MET/CBT5 ($3,958)
costing significantly less per person in recovery than
MET/CBT12 ($7,377) and both of the MET/CBT models
costing significantly less per person in recovery than FSN
($15,116). This pattern held at Site 1 ($3,495 vs. $9,257 vs.
$18,284; f = 0.67, p b .05) and Site 2 ($4,369 vs. $5,914 vs.
$12,899; f = 0.81, p b .05). Given that MET/CBT5 econom-
ically dominated the other conditions (i.e., it was both the
most effective and least expensive), cost-effectiveness ratios
were neither necessary nor appropriate.
In Trial 2, the average cost per day of abstinence was
$8.65 and varied significantly by condition overall (CohenTs
f = 0.22, p b .05); while there was a trend for ACRA ($6.62)
to have a lower cost per day of abstinence than MET/
CBT5 ($9.00) or MDFT ($10.38), the pair-wise compari-
sons were not significant using the more conservative
criterion in Tukey range tests. Part of the probl em was that
across conditions there were large site differences (Site 3 =
$5.15 vs. Site 4 = $12.23) and differences in the pattern by
site. The Site 4 (CHOP) results parallel the cross-site
findings, with ACRA having a lower average cost per day of
abstinence than MET/CBT5 or MDFT ($8.09 vs. $15.83 vs.
$12.79; f = 0.23, p b .05). The differences between the
ACRA and MDFT were significant in pair-wise Tukey
range testing, but MET/CBT5 was between them in terms of
cost effectiveness and was not significantly different than
either. In Site 3, MET/CBT5 was less expensive than ACRA
or MDFT in pair-wise comparisons ($839 vs. $1,237 vs.
$1,428), but also had a (non-significant) trend to be less
effective (257 vs. 281 vs. 271 days abstinent). While it had a
lower average cost per day of abstinence ($3.86 vs. $5.36
vs. $5.94; f = 0.21, p b .05), when we controlled for the
average performance for the site, MET/CBT5 was actually
less cost-effective (i.e., [condition cost-average cost] /
[condition effect-average effect]) than ACRA ($26.34 vs.
$4.10 per additional day of abstinence over average). The
average cost per person in recovery at the end of the study
was $7,615 and varied significantly by condition (CohenTs
f = 0.78, p b .06), with ACRA ($4,460) being lower than
MET/CBT5 ($6,611) and both being lower than MDFT
($11,775) in Tukey range tests. While there were still major
site differences in magnitude, the above order and signif i-
cance findings were replicated at Site 3 ($3,123 vs. $4,673
vs. $6,490; f = 0.61) and Site 4 ($6,029 vs. $8,016 vs.
$17,979; f = 0.83, p b .05). Both across and within sites,
ACRA economically dominated MET/CBT5 and MDFT
and MET/CBT5 economically dominated MDFT.
4. Discussion
4.1. Summary
This study examined the relative clinical effectiveness
and cost-effectiveness of five short-term (90 days or less)
outpatient treatments for adolescents with cannabis use
disorders in two randomized trials with 600 adolescents
from four sites. All five CYT interventions demonstrated
significant pre-post treatment effects that were stable in
terms of increasing days of abstinence during the 12 months
after they were randomized to a treatment intervention and
the percent of adolescents in recovery at the end of the
study. Overall , the clinical outcomes were very similar
across sites and conditions. The effect sizes were generally
small (CohenTs f = .1) and varied by measure and site. Such
findings are also consistent with earlier studies with adults
that compared multiple approaches to substance abus e
treatment (sharing many comm on components/approaches;
e.g., Babor & Del Boca, 2003; Crits-Christoph et al., 2001;
Cooney, Babor, DiClemente, & Del Boca, 2003). Alter-
native analyses using mixed models, repeated measures, and
scales (instead of individual items) to reduce measurement
error and increase statistical power were consistent with
these findings. Though the findings from these analyses
were similar in direction and magnitude of effect sizes, the
increased statistical power of these analyses did identify
some additional statistically significant differences. (See
appendix noted above for more detail.)
Given the similarity in clinical outcomes and large dif-
ferences in the costs of the interventions, cost-effectiveness
was also examined. There were significant cost differences
by condition in each of the 4 sites. When treatment costs
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 209
were combined with clinical outcomes to estimate the cost
per day of abstinence over the 12-month followup period
and cost per person in recovery at the last followup interval,
moderate ( CohenT s f = .2) to large ( f z .4) significant
differences by condition were revealed. In Trial 1, MET/
CBT5 and to a lesser extent MET/CBT12, were more cost-
effective than FSN. In Trial 2, ACRA and to a lesser extent
MET/CBT5, were more cost-effective than MDFT.
4.2. Strengths and limitations
This study has both strengths and limitations. It offers
initial evidence that: (a) the five treatmen ts can be delivered
in the manner intended and can be differentiated quantita-
tively and qualitatively; (b) the costs of these treatments
differ in predictable ways associated with their intensity but
all are roughly within the bounds now commonly spent on
adolescent outpatient treatment (see French et al., 2002); (c)
the treatments are reasonably acceptable to the adoles-
cents and their families as evidenced by participation and
retention rates reported above; (d) many (though certainly
not all) of these adolescent cannabis users show significant
improvement in substance use and in other me asures
bduringQ treatment (months 0–3) and these improvements
are sustained for significant periods following treatment
completion; (e) the amount and duration of the clinical
improvements were very similar between sites and across
treatments; and (f) the cost-effectiveness differences are
moderate to large.
The primary methodological limitations of this study are
its reliance on participant self-report, the generalizability of
the cost estimates to non-experimental settings, and the lack
of a no-treatment control group. While self-reports are
generally valid and typically more sensitive to change than
biometric measures, collateral reports, or records (Del Boca
& Noll, 2000; Dennis, Titus, et al., 2002; Rouse, Kozel, &
Richards, 1985), the study would have been strengthened by
having multiple types of measures available at all followup
waves. While the economists tried to control for research
and start up costs, there were large differences by site and
ideally these cost estimates need to be replicated as part of
more natural/on-going efforts to implement these interven-
tions. CYT was also one of the first economic evaluations of
adolescent treatment. Further work is need to better under-
stand the differences in cost-effectiveness (based on direct
costs and clinical outcomes) as reported here and benefit-
cost effectiveness (based on a range of other tertiary
outcomes—but not the core clinical outcomes). Both
clinical and economic outcomes might vary with other
populations and settings. From a methodological perspec-
tive, it would have been preferable to have a no-treatment
control group; however, this was considered unethical at the
time of CYT and it is unlikely that any Institutional Review
Board or Principal Investigator would go back to compare
the CYT interventions to a no-treatment control. However,
comparisons with other interventions based on research and/
or practice are clearly warranted. Replications are already
under way to evaluate each of the CYT interventions. These
include a 22-site study of MET/CBT5 in multiple popula-
tions and settings, MET/CBT5 and MDFT in early
intervention programs, MET/CBT12, FSN, ACRA, and
MDFT in several day and juvenile drug court programs, and
ACRA in several continuing care studies.
4.3. Implications and suggestion s for future study
Contrary to expectations based on reviews of studies that
had minimal treatment comparison groups (Kaminer, 2001;
Kaminer & Burleson, 1999; Kaminer et al., 1998), we found
only limited evidence that simply increasing the dosage of
treatment had a differential effect on substance use and
associated problems.
Despite concerns in the literature that group therapy
might produce iatrogenic effects (e.g., Dishion, McCord, &
Poulin, 1999), all three group therapy conditions were
associated with reduced substance use and probl ems during
the three mont h treatment phase and these changes were
stable during the 12 month followup. No evidence was
found that there were iatrogenic effects from the group
therapy relative to individual and family therapy.
Although it was expected that the longer and resource
intensive family treatment approaches w ould be more
effective, family treatments did not prove to be consistently
superior to the other inte rventions. In Trial 1, MET/CBT5
was as effective as FSN and in Trial 2; ACRAwas as effective
as the more family systems focused MDFT. We believe these
findings can be explained, at least in part, because all the
interventions were designed to be developmentally appro-
priate and implemented with a high level of quality assurance
as was common in the early family therapy studies.
Each treatment team expected that their respective
intervention would be significantly more effective than the
others, but the clinical outcomes over time were relatively
similar. It is possible that these similarities occurred because
treatment outcomes are driven more by general factors (e.g.,
Frank & Frank, 1991) rather than the kinds of differences
found in the treatment approaches tested in this study.
Alternatively, the similar findings may be due to shared
bother factorsQ (e.g., days in subsequent treatment, incar-
ceration, juvenile justice monitoring, systematic and struc-
tured ways of helping teens to attend to and decrease their
connection to cannabis and alcohol).
When relapse patterns were further examined over the
followup period, it was found that half of the ad olescents
went in and out of periods of recovery and relapse one or
more times after discharge. Two thirds were still reporting
substance use or related problems at the 12-month
followup interview. Thus, while the CYT interventions
were relatively effective as initial interventions, they were
not enough to interrupt all futu re s ubstance use and
problems for many adolescents. For significant subgroups
of clinically referred teens, the conceptualization of their
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213210
drug problems as a chronic condition (Kazdin, 1987)
suggests the need to focus more on monitoring and re-
intervention or continuing care.
Acknowledgments
Financial assistanc e for this study was provided by SAM -
HSAs Center for Substance Abuse Treatment (CSAT) Grant
Nos. TI11317,TI11320, TI11321, TI11323,and TI11324. The
opinions expressed herein are those of the authors and do
not reflect official positions of the government. CYT would
not have been possible without the contributions and
collaboration of staff from the federal government, research
institutes and participating programs: specifically, James
Fraser, Mark Godley, James Herrell, Ron Kadden, Kerry
Anne McGeary, Nancy Petry, M. Christopher Roebuck,
Susan Sampl, and Christy Scott, who helped with the studyTs
conception, design, and supervision; Nancy Angelovich,
Dana Becker, Laura (Bunch) Brantley, Betty Buchan, Joseph
A. Burleson, David Hodgkins, Caroline Hunter Williams,
Tracy L. Karvinen, Jodi (Johnson) Leckrone, Barbara
McDougall, Robert Meyers, Susan Panichelli-Mindel,
Melissa Siekmann, Meleney Scudder, Zeena Tawfik, James
West, Michelle White, and William White, who helped
supervise the data collection and clinical interventions; Peter
Bergstrom, Jane Ungemack, Joan Unsicker, and Kelli Varney
Wright, who helped to prepare and edit the manuscript; A.
Thomas McLellan, Robert Stephens, and Holly Waldron,
who served as an ex ternal advisory board; and Randolph
Muck, who took over as the proje ct officer during the last
year. Finally, we would like to thank the other CYT line
staff, adolescents, and families who participated in this
study. More information on the project can be found at
www.chestnut.org/li/cyt.
References
Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the child
behavior checklist and revised child behavior profile. Burlington, VT7
University of Vermont, Department of Psychiatry.
American Academy of Pediatrics. (1999). Marijuana: A continuing concern
for pediatricians. Pediatrics, 104, 982 985.
American Psychiatric Association. (1994) . Diagnostic and statistical
manual of mental disorders. (4th ed.). Washington, DC7 Author.
American Society of Addiction Medicine. (1996). Patient placement
criteria for the treatment of psychoactive substance disorders (2nd
ed.). Chevy Chase, MD7 Author.
Ary, D. V., Duncan, T. E., Duncan, S. C., & Hops, H. (1999). Adolescent
problem behavior: The inf luence of parents and peers. Behaviour
Research and Therapy, 37, 217 230.
Azrin, N. H., Donohue, B., Teichner, G. A., Crum, T., Howell, J., &
DeCato, L. A. (2001). A controlled evaluation and description of
individual-cognitive problem solving and family-behavior therapies in
dually-diagnosed conduct-disordered and substance-dependent youth.
Journal of Child and Adolescent Substance Abuse, 11, 1 43.
Azrin, N. H., McMahon, P., Donohue, B., Besalel, V., Lapinski, K., Kogan,
E., Acierno, R., & Galloway, E. (1994). Behavior therapy for drug
abuse: A controlled treatment-outcome study. Behaviour Research and
Therapy, 32, 857 866.
Babor, T. F., & Del Boca, F. K. (Eds.). (2003). Treatment matching in
alcoholism. New York7 Cambridge University Press.
Babor, T. F., Carroll, K., Christiansen, K., Kadden, R., Litt, M., McRee, B.,
Miller, M., Roffman, R., Solowij, N., Steinberg, K., Stephens, R.,
Vendetti, J., Donaldson, J., & Herrell, J. (in press). Brief treatments for
cannabis dependence: Findings from a randomized multi-site trial.
Journal of Consulting and Clinical psychology.
Brown, S. A., Myers, M. G., Mott, M. A., & Vik, P. W. (1994). Correlates
of success following treatment for adolescent substance abuse. Applied
& Preventive Psychology, 3, 61 73.
Bry, B. H., Catalano, R. F., Kumpfer, K. L., Lochman, J. E., & Szapocznik,
J. (1998). Scientific findings from family prevention intervention
research. In R. Ashery, E. B. Robertson, & K. L. Kumpfer (Eds.),
Drug abuse prevention through family interventions (NIH Publication
No. 97-4135). NIDA Research Monograph (pp. 103–129). Rockville,
MD: National Institute on Drug Abuse.
Buchan, B. J., Dennis, M. L., Tims, F. M., & Diamond, G. S. (2002).
Cannabis use: Consistency and validity of self-report, on-site urine
testing, and laboratory testing. Addiction, 97(Suppl. 1), 98 108.
Budney, A. J., & Higgins, S. T. (1998). A community reinforcement plus
vouchers approach: Treating cocaine addiction. Rockville, MD7
National Institute on Drug Abuse.
Bunch, L., Hamilton, N., Tims, F., Angelovich, N., & McDougall, B.
(1998). CYT: A multi-site study of the effectiveness of treatment for
cannabis dependant youth, Family Support Network. St. Petersburg,
FL7 Operation PAR, Inc.
Bureau of Justice Statistics. (2000). Sourcebook of criminal justice statistics
1999 (27th ed.). Washington, DC7 U.S. Department of Justice.
Center For Substance Abuse Treatment (CSAT). (1992a). Empowering
Families, Helping Adolescents: Family-Centered Treatment of Adoles-
cents With Alcohol, Drug Abuse, and Mental Health Problems. DHHS
Publication No. 92–1745, Technical Assistance Publication Series No. 6.
Rockville, MD7 Author.
Center For Substance Abuse Treatment. (1992b). Guidelines for the
Treatment of Alcohol- and Other Drug-Abusing Adolescents. DHHS
Publication no. 93–2010, Treatment Improvement Protocol Series
(TIPS) 4. Rockville, MD7 Author.
Center for Substance Abuse Treatment. (1993). Screening and assessment
of alcohol and other drug abusing adolescents. DHHS Publication No.
93–2009, Treatment Improvement Protocol Series (TIPS) 3. Rockville,
MD7 Author.
Clark, H. W., Horton Jr., A. M., Dennis, M. L., & Babor, T. F. (2002).
Moving from research to practice just in time: The treatment of
cannabis use disorders comes of age. Addiction, 97(Suppl. 1), 1 3.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (pp.
222 237). Hillsdale, NJ7 Lawrence Erlbaum.
Cooney, N. L., Babor, T. F., DiClemente, C. C., & Del Boca, F. K. (2003).
Clinical and scientific implications of Project MATCH. In T. F. Babor,
& F. K. Del Boca (Eds.), Treatment matching in alcoholism. (pp.
222 237). Cambridge, UK7 Cambridge University Press.
Copeland, J., Swift, W., Roffman, R., & Stephens, R. (2001). A randomized
controlled trial of brief cognitive-behavioral interventions for cannabis
use disorder. Journal of Substance Abuse Treatment, 21, 55 64.
Crits-Christoph, P., Siqueland, L., McCalmont, E., Gastfriend, D. R.,
Frank, A., Moras, K., Barber, J. P., Blaine, J., & Thase, M. E. (2001).
Impact of psychosocial treatments on associated problems of cocaine-
dependent patients. Journal of Consulting and Clinical Psychology, 69,
825 830.
Dakof, G. A. (2000). Understanding gender differences in adolescent drug
abuse: Issues of comorbid ity and family functioning. Journal of
Psychoactive Drugs, 32, 25 32.
Del Boca, F. K., & Noll, J. A. (2000). Truth or consequences: The validity
of self-report data in health services research on addictions. Addiction,
95(Suppl. 3), S347 S360.
Dennis, M. L. (1999). Global Appraisal of Individual Needs (GAIN):
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213 211
Administration guide for the GAIN and related measures (Version
1299). Bloomington, IL7 Chestnut Health Systems.
Dennis, M. L. (on behalf of the CYT steering committee) (2003). Short-
(3-12 month) and long- (30 month) term outcomes of the Cannabis
Youth Treatment (CYT) Experiment. Presented at the 65th Annual
Scientific Meeting of the College of Problems on Drug Dependence,
Bal Harbour, FL. Retrieved from http://www.chestnut.org/LI/Posters/
Dennis%20CYT%2030%mon%20findings.pps.
Dennis, M. L., Babor, T., Roebuck, M. C., & Donaldson, J. (2002).
Changing the focus: The case for recognizing and treating marijuana
use disorders. Addiction, 97(Suppl. 1), S4 S15.
Dennis, M. L., Perl, H. I., Huebner, R. B., & McLellan, A. T. (2000).
Twenty-five strategies for improving the design, implementation and
analysis of health services research related to alcohol and other drug
abuse treatment. Addiction, 95(Suppl. 3), S281 S308.
Dennis, M. L., Titus, J. C., Diamond, G., Donaldson, J., Godley, S. H.,
Tims, F., Webb, C., Kaminer, Y., Babor, T., Roebuck, M. C., Godley,
M. D., Hamilton, N., Liddle, H., Scott, C. K., & Steering Committee,
C. Y. T. (2002). The Cannabis Youth Treatment (CYT) experiment:
Rationale, study design, and analysis plans. Addiction, 97(Suppl. 1),
S16 S34.
Diamond, G. S., Godley, S. H., Liddle, H. A., Sampl, S., Webb, C., Tims,
F. M., & Meyers, R. (2002). Five outpatient treatment models for
adolescent marijuana use: A description of the Cannabis Youth
Treatment interventions. Addiction, 97(Suppl. 1), S70 S83.
Diamond, G. S., Leckrone, J., & Dennis, M. L. (in press). The Cannabis
Youth Treatment study: Clinical and empirical developments. In R.
Roffman & R. Stephens (Eds.), Cannabis dependence: Its nature, conse-
quences, and treatment. Cambridge, UK: Cambridge University Press.
Diamond, G. S., & Liddle, H. A. (1996). Resolving a therapeutic impasse
between parents and adolescents in multidimensional family therapy.
Journal of Consulting and Clinical Psychology, 64, 481 488.
Dishion, T., McCord, J., & Poulin, F. (1999). When interventions harm: Peer
groups and problem behavior. American Psychologist, 54, 755 764.
Fergusson, D. M., Lynskey, M. T., & Horwood, L. J. (1996). The short-term
consequences of early onset cannabis use. Journal of Abnormal Child
Psychology, 24, 499 512.
Frank, J. D., & Frank, J. B. (1991). Persuasion and healing (3rd ed.).
Baltimore7 The Johns Hopkins Press.
French, M. T. (2001). Drug Abuse Treatment Cost Analysis Program
(DATCAP): Program version users manual (7th ed.). Coral Gables, FL7
University of Miami.
French, M. T., Roebuck, M. C., Dennis, M., Diamond, G., Godley, S.,
Liddle, H. A., & Tims, F. (2003). Outpatient marijuana treatment for
adolescents: Economic evaluation of a multisite field experiment.
Evaluation Review, 27, 421 459.
French, M. T., Roebuck, M. C., Dennis, M., Diamond, G., Godley, S.,
Tims, F., Webb, C., & Herrell, J. M. (2002). The economic cost of
outpatient marijuana treatment for adolescents: Findings from a
multisite experiment. Addiction, 97(Suppl. 1), S84 S97.
Godley, M. D., Godley, S. H., Dennis, M. L., Funk, R., & Passetti, L.
(2002). Preliminary outcomes from the assertive continuing care
experiment for adolescents discharged from re sidential treatment.
Journal of Substance Abuse Treatment, 23, 21 32.
Godley, S. H., Meyers, R. J., Smith, J. E., Godley, M. D., Titus, J. M.,
Karvinen, T., Dent, G., Passetti, L., & Kelberg, P. (2001). The Adolescent
Community Reinforcement Approach (ACRA) for adolescent cannabis
users. Rockville, MD7 Center for Substance Abuse Treatment.
Godley, S. H., White, W. L., Diamond, G., Passetti, L., & Titus, J. (2001).
Therapists’ reactions to manual-guided therapies for the treatment of
adolescent marijuana users. Clinical Psychology: Science and Practice,
8, 405 417.
Gold, M. R., Siegel, J. E., Russell, L. B., & Weinstein, M. C. (Eds.). (1996).
Cost-effectiveness in health and medicine. New York7 Oxford Uni-
versity Press.
Grella, C. E., Hser, Y. I., Joshi, V., & Rounds-Bryant, J. (2001). Drug
treatment outcomes for adolesce nts with comorbid mental and
substance use disorders: Findings from the DATOS-A. Journal of
Nervous and Mental Disease, 189, 384 392.
Hamilton, N., Brantley, L., Tims, F., Angelovich, N., & McDougall, B.
(2001). Family Support Network (FSN) for adolescent cannabis users.
Rockville, MD7 Center for Substance Abuse Treatment.
Henggeler, S. W., Borduin, C. M., Melton, G. B., Mann, B. J., Smith, L. A.,
Hall, J. A., Cone, L., & Fucci, B. R. (1991). Effects of multisystemic
therapy on drug use and abuse in serious juvenile offenders: A progress
report from two outcome studies. Family Dynamics of Addiction
Quarterly, 1, 40 51.
Henggeler, S. W., Clingempeel, W. G., Brondino, M. J., & Pickrel, S. G.
(2002). Four-year follow-up of multisystemic therapy with sub-
stance-abusing and substance-dependent juvenile offenders. Jour-
nal of the American Academy of Child and Adolescent Pyschiatry, 41,
868 874.
Henggeler, S. W., Pickrel, S. G., & Brondino, M. J. (1999). Multisystemic
treatment of substance-abusing and dependent delinquents: Outcomes,
treatment, fidelity, and transportability. Mental Health Services
Research, 1, 171 184.
Henggeler, S. W., Pickrel, S. G., Brondino, M. J., & Crouch, J. L. (1996).
Eliminating (almost) treatment dropout of substance abusing or
dependent delinquents through home-based multisystemic therapy.
American Journal of Psychiatry, 153, 427 428.
Hser, Y. I., Grella, C. E., Hubbard, R. L., Hsieh, S. C., Fletcher, B. W.,
Brown, B. S., & Anglin, M. D. (2001). An evaluation of drug
treatments for adolescents in four U.S. cities. Archives of General
Psychiatry, 58, 689 695.
Hubbard, R. L., Cavanaugh, E. R., Craddock, S. G., & Rachel, J. V. (1985).
Characteristics, behaviors, and outcomes for youth in the TOPS:
Treatment services for adolescent substance abusers. In A. S. Friedman,
&G.M.Beschner(Eds.),Treatment services for adolescent
substance abusers. Rockville, MD7 National Institute on Drug Abuse.
Institute of Medicine. (1989). Prevention and treatment of alcohol problems:
Research opportunities. Washington, DC7 National Academy Press.
Institute of Medicine. (1998). Bridging the gap between practice and
research: Forging partnerships with community-based drug and alcohol
treatment. Washington, DC7 National Academy Press.
Jackson-Gilfort, A., Liddle, H. A., Tejeda, M. J., & Dakof, G. A. (2001).
Facilitating engagement of African American male adolescents in
family therapy: A cultural theme process study. Journal of Black
Psychology, 27, 321 340.
Joanning, H., Quinn, W., Thomas, F., & Mullen, R. (1992). Treating
adolescent drug abuse: A comparison of family systems therapy, group
therapy, and family drug education. Journal of Marital and Family
Therapy, 18, 345 356.
Kaminer, Y. (2001). Adolescent substance abuse treatment: Where do we
go from here? Psychiatric Services, 52, 147 149.
Kaminer, Y., & Burleson, J. A. (1999). Psychotherapies for adolescent
substance abusers: 15-month follow-up of a pilot study. American
Journal on Addictions, 8, 114 119.
Kaminer, Y., Burleson, J. A., Blitz, C., Sussman, J., & Rounsaville, B. J.
(1998). Psychotherapies for adolescent substance abusers: A pilot study.
Journal of Nervous and Mental Disease, 186, 684 690.
Kaminer, Y., Burleson, J. A., & Goldberger, R. (2001). Psychotherapies for
adolescents with alcohol and other substance abuse: Three-and nine-
month post treatment outcomes. Alcoholism: Clinical and Experimental
Research, 25S, 90A.
Kaminer, Y., Burleson, J. A., & Goldberger, R. (2002). Cognitive-
behavioral coping skills and psychoeducation therapies for adoles-
cent substance abuse. Journal of Nervous and Mental Disease, 190,
737 745.
Kazdin, A. E. (1987). Treatment of antisocial behavior in children: Current
status and future directions. Psychological Bulletin, 102, 187 202.
Latimer, W. W., Winters, K. C., D’Zurilla, T., & Nichols, M. (2003).
Integrated family and cognitive-behavioral therapy for adolescent
substance abusers: A stage-I efficacy study. Drug and Alcohol
Dependence, 71, 303 317.
M. Dennis et al. / Journal of Substance Abuse Treatment 27 (2004) 197–213212
Lewis, R. A., Piercy, F. P., Sprenkle, D. H., & Trepper, T. S. (1990).
Family-based interventions for helping drug-abusing adolesce nts.
Journal of Adolescent Research, 5, 82 95.
Liddle, H. A. (2001). Advances in family-based therapy for adolescent
substance abuse: Findings from the Multidimensional family therapy
research program. In L. S. Harris (Ed.), Problems of drug dependence
2001: Proceedings of the 63
rd
annual scientific meeting (NIDA
Research Monograph No. 182, pp. 113–115). Bethesda, MD7 National
Institute on Drug Abuse.
Liddle, H. A. (2002). Multidimensional Family Therapy (MDFT) for adoles-
cent cannabis users. Rockville, MD7 Center for Substance Abuse Treat-
ment, Substance Abuse and Mental Health Services Administration.
Liddle, H. A., Dakof, G. A., Parker, K., Diamond, G. S., Barrett, K., &
Tejeda, M. (2001). Multidimensional family therapy for adolescent
substance abuse: Results of a randomized clinical trial. American
Journal of Drug and Alcohol Abuse, 27, 651 688.
Liddle, H. A., Rowe, C. L., Ungaro, R. A., Dakof, G. A., & Henderson C.
(2004). Early intervention for adolescent substance abuse: Pretreatment
to posttreatment outcomes of a randomized controlled trial comparing
Multidimensional Family Therapy and peer group treatment. Journal of
Psychoactive Drugs, 36, 2–37.
Meyers, R. J., & Godley, M. D. (2001). The community reinforcement
approach. In R. Meyers & W. Miller (Eds.), A community reinforcement
approach to addition treatment (pp. 1 7). Cambridge, UK7 Cambridge
University Press.
Meyers, R. J., Miller, W. R., Hill, D. E., & Tonigan, J. S. (1999). Community
reinforcement and family training (CRAFT): Engaging unmotivated
drug users in treatment. Journal of Substance Abuse, 10, 291 308.
Moher, D., Schultz, K. F., & Altman, D. G. (2001). The CONSORT
statement: Revised recommendations for improving the quality of
reports of parallel group randomized trials. Journal of the American
Medical Association, 285, 1987 1991.
Morgenstern, J., & McCrady, B. S. (1992). Curative factors in alcohol and
drug treatment: Behavioral and disease model perspectives. British
Journal of Addiction, 87, 901 912.
Office of Applied Studies. (1995). National Household Survey on Drug
Abuse (NHSDA). Rockville, MD7 Substance Abuse and Mental Health
Services Administration.
Office of Applied Studies. (2000a). Mid-year 2000 Emergency Department
Data from the Drug Abuse Warning Network (DAWN). Rockville, MD7
Substance Abuse and Mental Health Services Administration.
Office of Applied Studies. (2000b). National Household Survey on Drug
Abuse: Main Findings 1998. Rockville, MD7 Substance Abuse and
Mental Health Services Administration.
Office of Applied Studies. (2000c). Year-end 1999 Medical Examiner Data
from the Drug Abuse Warning Network (DAWN). Rockville, MD7
Substance Abuse and Mental Health Services Administration.
Ozechowski, T. J., & Liddle, H. A. (2000). Family-based therapy for
adolescent drug abuse: Knowns and unknowns. Clinical Child and
Family Psychology Review, 3, 269 298.
Petry, N. M., & Tawf ik, Z. (2001). A comparison of problem gambling and
non-problem gambling youth seeking treatment for marijuana abuse.
Journal of the American Academy of Child and Adolescent Psychiatry,
40, 1324 1331.
Randall, J., Henggeler, S. W., Cunningham, P. B., Rowland, M. D., &
Swenson, C. C. (2001). Adapting multisystemic therapy to treat
adolescent substance abuse more effectively. Cognitive & Behavioral
Practice, 8, 359 366.
Rouse, B. A., Kozel, N. J., & Richards, L. G. (Eds.). (1985). Self-report
methods of estimating drug use: Meeting current challenges to validity.
Rockville, MD7 National Institute on Drug Abuse.
Sampl, S., & Kadden, R. (2001). Motivational Enhancement Therapy and
Cognitive Behavioral Therapy (MET-CBT-5) for adolescent cannabis
users. Rockville, MD7 Center for Substance Abuse Treatment,
Substance Abuse and Mental Health Services Administration.
Schmidt, S. E., Liddle, H. A., & Dakof, G. A. (1996). Changes in parenting
practices and adolescent drug abuse during multidimensional family
therapy. Journal of Family Psychology, 10, 12 27.
Scott, C. K. (in press). A replicable model for achieving 90%+ follow-up
rates with substance abusers: Its impact on internal validity and study
costs. Drug and Alcohol Dependence.
Sells, S. B., & Simpson, D. D. (1979). Evaluation of treatment outcome for
youths in the Drug Abuse Reporting Program (DARP): A follow-up
study. In G. M. Beschner & A. S. Friedman (Eds.), Youth drug
abuse: Problems, issues, and treatment (pp. 571 628). Lexington, MA7
DC Heath.
Shane, P., Jasiukaitis, P., & Green, R. S. (2003). Treatment outcomes
among adolescents with substance abuse problems: The relationship
between comorbidities and post-treatment substance involvement.
Evaluation and Program Planning, 26, 393 402.
Sisson, R. W., & Azrin, N. H. (1993). Community reinforcement training
for families: A method to get alcoholics into treatment. In T. J. OTFarrell
(Ed.), Treating alcohol problems: Marital and family interventions
(pp. 34 53). New York7 Guilford Press.
SPSS. (2001). Statistical program for the social sciences, version 10.1.
Chicago, IL7 SPSS.