Multisystemic Treatment of Serious Juvenile Offenders: Long-Term Prevention of Criminality and Violence

Abstract

This article examined the long-term effects of multisystemic therapy (MST) vs. individual therapy (IT) on the prevention of criminal behavior and violent offending among 176 juvenile offenders at high risk for committing additional serious crimes. Results from multiagent, multimethod assessment batteries conducted before and after treatment showed that MST was more effective than IT in improving key family correlates of antisocial behavior and in ameliorating adjustment problems in individual family members. Moreover, results from a 4-year follow-up of rearrest data showed that MST was more effective than IT in preventing future criminal behavior, including violent offending. The implications of such findings for the design of violence prevention programs are discussed.

Full text

Journal
of
Consulting
and
Clinical Psychology
1995, Vol.
63. No. 4,
569-578
Copyright
1995
by the
American Psychological
Association,
Inc.
0022-006X/95/S3.00
Multisystemic Treatment
of
Serious Juvenile Offenders:
Long-Term
Prevention
of
Criminality
and
Violence
Charles
M.
Borduin
University
of
Missouri—Columbia
Barton
J.
Mann
University
of
North Carolina
at
Chapel Hill
Lynn
T.
Cpne
University
of
Missouri—Columbia
Scott
W.
Henggeler
Medical University
of
South Carolina
Bethany
R.
Fucci, David
M.
Blaske,
and
Robert
A.
Williams
University
of
Missouri—Columbia
This article examined
the
long-term
effects
of
multisystemic therapy
(MST)
vs.
individual therapy
(IT)
on the
prevention
of
criminal behavior
and
violent offending among
176
juvenile offenders
at
high
risk
for
committing additional serious crimes. Results
from
multiagent, multimethod
assess-
ment
batteries
conducted before
and
after
treatment showed that
MST was
more
effective
than
IT in
improving
key
family
correlates
of
antisocial behavior
and in
ameliorating adjustment
problems
in
individual
family
members. Moreover, results
from
a
4-year follow-up
of
rearrest data showed that
MST was
more
effective
than
IT in
preventing
future
criminal behavior, including violent
offending.
The
implications
of
such
findings for the
design
of
violence prevention
programs
are
discussed.
The
prevention
of
violent criminal acts
and
other serious
crimes
perpetrated
by
youths
has
become
a
pressing issue
on
the
national health care agenda,
as the
staggering
fiscal and so-
cial
costs
of
such crimes become evident (Children's Defense
Fund,
1992).
To
address this issue, mental health professionals
and
policy makers have
justifiably
argued
for the
promotion
of
childhood
programs that
may
prevent
the
development
of
vio-
lent
behavior (e.g., Zigler, Taussig,
&
Black,
1992).
Although
primary
prevention programs targeted
at
young children
are
certainly
needed
and
promising programs
are
currently being
examined
(e.g.,
Tolan,
Guerra,
Van
Acker,
Huesmann,
&
Eron,
1994),
the
development
of
effective
interventions with youths
who
are the
most
likely
to
perpetrate serious crimes
has
been
relatively
neglected. Clearly,
as
longitudinal studies graphically
Charles
M.
Borduin,
Lynn
T.
Cone,
Bethany
R.
Fucci, David
M.
Blaske,
and
Robert
A.
Williams, Department
of
Psychology, University
of
Missouri—Columbia;
Barton
J.
Mann,
Department
of
Psychology,
University
of
North Carolina
at
Chapel
Hill;
Scott
W.
Henggeler,
De-
partment
of
Psychiatry
and
Behavioral
Sciences,
Medical University
of
South Carolina.
This research
was
supported
by
grants
from
the
Missouri
Depart-
ment
of
Social Services
and the
University
of
Missouri—Columbia
Re-
search Council.
Sincere thanks
go to
Robert
Perry,
Gene Hamilton,
and
Alan Sirinek
of the
Missouri
13th
Judicial Circuit Juvenile Court
for
their
support
and
cooperation
and to the
therapists
and
community professionals,
too
numerous
to
mention
individually,
who
worked
on
this project.
We
also
appreciate
the
efforts
of the
many
research
assistants,
and we
extend
special thanks
to
Janette
Concepcion,
Peter Ambrose,
and
John
Draper.
Correspondence
concerning this article should
be
addressed
to
Charles
M.
Borduin, Department
of
Psychology,
21
McAlester Hall,
University
of
Missouri, Columbia, Missouri
65211.
demonstrate (e.g., Lewis, Lovely,
Yeager,
&
Famina,
1989;
Weisz,
Martin, Walter,
&
Fernandez,
1991),
serious juvenile
offenders
are,
by
far,
at the
greatest risk
for
committing addi-
tional
serious crimes. Preventing
or
attenuating
further
crimi-
nal
activity
in
such youths would
favorably
affect
their lives,
families,
and
communities.
Unfortunately,
however, interventions with serious juvenile
offenders
historically have
had
little success. Several reviews
of
the
delinquency treatment literature
in the
1970s concluded
that "nothing works" (Henggeler,
1989,
p.
84).
More recently,
Kazdin
(1987,
p.
189)
has
described several empirically driven
treatments
as
"promising"
(e.g., behavioral parent training,
cognitive-behavioral
therapy),
and
Lipsey
(1992)
has
argued
that such structured, skill-oriented treatments have demon-
strated
the
largest
effects
on
juvenile
offenders
in
general. How-
ever,
in
clinical trials with serious juvenile
offenders,
such treat-
ments
have
failed
to
produce
favorable
long-term
effects
(Bank,
Marlowe,
Reid, Patterson,
&
Weinrott,
1991;
Guerra
&
Slaby,
1990;
Weisz, Walter, Weiss, Fernandez,
&
Mikow,
1990).
More-
over,
even
favorable
outcomes
of
these treatments with less
se-
vere
antisocial behavior
are
mitigated
by findings
that univer-
sity-based successes rarely extend
to
community settings
(Weisz,
Weiss,
&
Donenberg,
1992).
In
consideration
of the
continued
difficulty
of
even well-con-
ceived
treatments
to
produce lasting change
in
youth antisocial
behavior,
investigators
have
argued that
the
major limitation
of
such
treatments
is
their relatively narrow
focus
and
failure
to
account
for the
multidetermined
nature
of
antisocial behavior
(e.g.,
Borduin,
1994;
Mulvey,
Arthur,
&
Reppucci,
1990;
Zigler
et
al.,
1992).
Overwhelming evidence supports
a
social-ecolog-
ical
view
(Bronfenbrenner,
1979)
in
which antisocial behavior
in
youths
is
conceptualized
as
multidetermined (e.g., Heng-
569

570
BORDUIN
ET AL.
geler,
1989;
Loeber
&
Dishion,
1983).
For
example, several
so-
phisticated causal modeling studies (e.g., Elliott, Huizinga,
&
Ageton,
1985; Patterson
&
Dishion, 1985;
Simcha-Fagan
&
Schwartz,
1986)
have shown that delinquency
is
linked directly
or
indirectly
with
key
characteristics
of
youths
and the
family,
peer,
school,
and
neighborhood systems
in
which youths
are
embedded.
In
light
of the
multidetermined nature
of
serious
antisocial
behavior, expecting even well-conceived
office-based
treatments
to be
effective
is
unrealistic.
Findings
from
recent primary
and
secondary prevention
studies
and a
clinical trial with serious juvenile
offenders
sup-
port
the
contention that
effective
interventions must address
the
multiple
causes
of
antisocial behavior
and be
delivered with eco-
logical
validity.
Zigler
et
al.
(1992)
concluded
in
their review
that delinquency
can be
prevented
by
early childhood interven-
tion
programs that promote children's competence across
multiple
systems
in
which they
are
embedded. Similarly,
Ol-
weus
(1992)
demonstrated that
a
large-scale, systemically ori-
ented
secondary prevention program (i.e., addressing individ-
ual,
family,
classroom, school,
and
community factors) pro-
duced
significant
reductions
in
bullying. Likewise, Henggeler
and
his
colleagues (Henggeler, Melton,
&
Smith, 1992; Heng-
geler,
Melton, Smith,
Schoenwald,
&
Hanley,
1993)
showed that
multisystemic
therapy using
the
family
preservation model
of
service
delivery
doubled
the
survival rate (i.e., percentage
of
youths
not
rearrested)
2
years
after
treatment
in a
sample
of
violent
and
chronic juvenile
offenders
at
imminent risk
of in-
carceration.
The
common
links
of
these diverse interventions
are
their attention
to the
multiple determinants
of
antisocial
behavior
and
their delivery
of
services
in the
youths' natural
environments.
The
present study
from
the
Missouri Delinquency Project
ex-
amines
the
long-term
effects
of
multisystemic therapy (MST;
Henggeler
&
Borduin,
1990)
on the
prevention
of
criminal
ac-
tivity
in a
sample
of
predominantly serious juvenile
offenders.
The
study builds
on the
recent
MST
trial noted earlier
(Henggeler
et
al.,
1992, 1993)
as
well
as on
previous clinical
trials
demonstrating
the
efficacy
of MST in
treating inner-city
juvenile
offenders
(Henggeler
et
al.,
1986), serious juvenile
offenders
living
in
rural areas
(Scherer
et
al.,
1994),
abusive
and
neglectful
parents (Brunk, Henggeler,
&
Whelan, 1987),
and
juvenile
sexual
offenders
(Borduin, Henggeler, Blaske,
&
Stein,
1990).
In
addition
to
providing
a
needed replication
of the re-
cent
trial
of MST
with serious juvenile
offenders
(Henggeler
et
al.,
1992,
1993),
the
present study
of
MST
contains several
im-
portant methodological improvements including
(a)
a
relatively
large
sample size
to
permit certain subgroup analyses (e.g.,
re-
cidivism
for MST
completers
vs. MST
dropouts),
(b)
a
longer
follow-up
period
for
rearrest,
(c)
observational measures
of
family
relations,
and (d) a
comparison group that received
a
roughly
equivalent number
of
treatment hours.
Method
Design
A
pretest-posttest
control group design, with random assignment
to
conditions
and a
4-year follow-up
for
arrests,
was
used
to
compare
the
effectiveness
of MST
with
that
of
individual
therapy.
Participants
Two
hundred
families
with
a
12-
to
17-year-old
adolescent
offender
were
referred
to the
project
by
juvenile court personnel
and
agreed
to
participate
in a
pretreatment assessment session;
five
other
families
were
referred
to the
project
but did not
agree
to
participate.
Referrals
to the
project were made consecutively
and
included
all
families
in
which
the
youth
(a) had at
least
two
arrests,
(b) was
currently living
with
at
least
one
parent
figure, and (c)
showed
no
evidence
of
psychosis
or
dementia.
The
arrest
histories
of the
referred youths
attest
to
their
serious criminal involvement.
The
youths averaged
4.2
previous
arrests
(SD
=
1.4),
and the
mean severity
of the
most recent arrest
was 8.8 (SD
=
1.5)
on a
17-point
seriousness scale (e.g.,
1 =
truancy,
4 =
disorderly
conduct,
8 =
assault
/battery,
11
=
grand
larceny,
13 =
unarmed rob-
bery,
17
=
murder) developed
by
Hanson, Henggeler, Haefele,
and
Rod-
ick
(1984).
Moreover,
all of the
youths
had
been
detained
previously
for
at
least
4
weeks.
The
mean
age of the
youths
was
14.8 years
(SD
=
1.5); 67.5% were male; 70.0% were White,
and
30.0% were
African
American;
and
53.3% lived
with
two
parental
figures
(biological par-
ents,
stepparents,
foster
parents,
grandparents).
The
primary caretaker
of
the
youths included biological mothers (88.0%),
step-,
foster,
or
adoptive mothers
(6.5%),
other
female
relatives (3.5%),
or
biological
fathers
(2.0%).
Families averaged
3.1
children
(SD =
1.5),
and
68.8%
of
the
families
were
of
lower socioeconomic status (Class
IV or V;
Hol-
lingshead,
1975).
Of the 200
families
who
completed pretreatment assessments,
24
(12%)
subsequently refused
to
participate
in
treatment
(hereafter
re-
ferred
to as
"refusers").
The
remaining
176
families
were randomly
assigned(usingacointoss)toMST(n
=
92)
or
individual therapy (IT;
n =
84).
Of
these,
140(79.5%)completedtreatment(hereafter
referred
to as
"completers"),
and 36
(21.5%)
dropped out, defined
as
unilater-
ally
terminating
after
the first
session
(with
the
youth
or
family)
and
before
the
seventh.
Of the 36
youths
and
their families
who
dropped
out
of
treatment
(hereafter
referred
to as
"dropouts"),
15
were
from
the
MST
condition
and
21
were
from
the IT
condition (dropout
rates
for
MST
[16.3%]
and IT
[25%]
were
not
significantly
different).
We
were
not
able
to
obtain posttreatment assessment
data
from
the 36
dropouts
or the 24
refusers; however, arrest data were obtained over
the
follow-up
period
for
these youths. Analyses
of
variance
(ANOV\s)
and
chi-square
tests showed
no
between-groups
differences
in the
criminal histories
or
demographic characteristics
of MST
completers,
MST
dropouts,
IT
completers,
IT
dropouts,
and
treatment refusers.
Posttreatment assessment batteries were completed
by
90.9%
(n
=
70) of the MST
completers
and by
88.9%
(n =
56)
of the IT
completers
(these
proportions were
not
significantly
different).
Research participa-
tion
at
postassessment
was
attenuated
by the
lack
of
incentives
for the
families
to
participate (i.e.,
funds
were
not
available
for
payment
of
participants)
and by the
out-of-home placement
of 6
youths
(2
from
MST,
4
from
IT)
in
residential
facilities
of the
Division
of
Youth
Ser-
vices
of the
Missouri Department
of
Social Services. However,
the
crim-
inal
histories
and
demographics
of the
cases
that
completed both pre-
treatment
and
posttreatment assessment
batteries
(n
=126)
are
essen-
tially
the
same
as
described
for the
larger sample. Furthermore,
the
MST and IT
cases
that completed both pretreatment
and
posttreatment
assessments
did not
differ
on any
demographic variable
or
measure
of
criminal history. Analyses
of
treatment
effects
on
psychosocial
mea-
sures
are
based
on
these cases.
Treatment
Conditions
Families
who
completed
the
pretreatment assessment
and
agreed
to
participate
in
treatment were randomly assigned
to
conditions
and to
therapists within
each
condition.
The
mean numbers
of
hours
of
treat-
ment were 23.9
(SD =
8.2; range,
5 to
49)
for the MST
completers,
and
28.6
(SD =
9.8; range,
15 to 72) for the IT
completers;
these means

SPECIAL
SECTION:
MULTISYSTEMIC
TREATMENT
571
were
significantly
different,
F(
1,
139)
=
9.67,
p <
.01.
The MST
drop-
outs
and IT
dropouts
averaged 4.07
hr
(SD =
0.70)
and
4.29
hr
(SD =
1.01),
respectively,
of
treatment; these means
did not
differ
significantly.
Multisystemic
Therapy
Therapeutic interventions were
based
on the
multisystemic approach
to the
treatment
and
prevention
of
behavior problems
in
children
and
adolescents
(Henggeler
&
Borduin,
1990).
The
treatment
and
preven-
tion emphases
of MST fit
closely with
findings
from
multidimensional
causal models
of
delinquent behavior
(for
a
review,
see
Henggeler,
1989). Using interventions that
are
present-focused
and
action-ori-
ented,
MST
directly
addresses
intrapersonal
(e.g., cognitive)
and
sys-
temic (i.e.,
family,
peer,
school)
factors that
are
known
to be
associated
with
adolescent antisocial behavior. Moreover, because
different
combi-
nations
of
these factors
are
relevant
for
different
adolescents,
MST in-
terventions
are
individualized
and
highly
flexible.
Guidelines
for
designing
and
implementing
MST
interventions with antisocial
adoles-
cents
and
their
families
are
described
in
detail elsewhere
(Borduin
&
Henggeler,
1990;
Henggeler
&
Borduin, 1990).
The
provision
of MST is
consistent with
family
preservation models
of
service
delivery
(Nelson,
1991).
To
promote
cooperation
and en-
hance
generalization,
we
usually
held sessions
in the
family's home
at a
convenient
time
and in
community locations
(e.g.,
school,
recreation
center).
In
addition, services were time limited, with
an
overriding goal
of
empowering
parents with
the
skills
and
resources needed
to
indepen-
dently
address
the
inevitable
difficulties
that arise
in
raising adolescents.
Individual
Therapy
The
therapy provided
in
this condition
was
selected
to
represent
the
usual
community treatment
for
juvenile offenders
in our
judicial
dis-
trict,
and
perhaps
in
many other judicial districts
as
well (see Henggeler,
1989).
All of the
offenders
in
this condition received individual therapy
that focused
on
personal,
family,
and
academic issues.
The
therapists
offered
support, feedback,
and
encouragement
for
behavior change.
Their theoretical orientations were
an
eclectic blend
of
psychodynamic
(e.g., promoting insight
and
expression
of
feelings),
client-centered
(e.g.,
building
a
close relationship, providing empathy
and
warmth),
and
behavioral (e.g., providing social approval
for
school attendance
and
other positive
behaviors)
approaches.
Although there were some
variations
in the
treatment strategies used
by the
therapists (e.g., some
therapists provided less empathy
or
were more directive than other
therapists),
the
common thread
of
their approaches
was
that
the
inter-
ventions
focused
on the
individual
adolescent rather than
on the
systems
in
which
the
adolescent
was
embedded.
Therapists
MST
was
provided
by
three
female
and
three
male graduate students
(ages ranged
from
23 to
31
years;
M
=
26)
in
clinical psychology.
One
of
the
therapists
was
Native American,
and the
others were White. Each
had
approximately
1.5
years
of
direct clinical experience with children
or
adolescents before
the
study.
The six
therapists served
in the
study
for
an
average
of
16
months
(range,
12
to 24
months).
Therapist supervi-
sion
was
provided
by
Charles
M.
Borduin
in a
3-hr weekly group meet-
ing
and
continued throughout
the
course
of the
investigation. During
these meetings,
the
therapists
and
supervisor reviewed
the
goals
and
progress
of
each
case,
observed
and
discussed
selected
videotaped
or
audiotaped therapy sessions,
and
made decisions about
how
best
to fa-
cilitate
the
family's
progress.
Interventions
in the IT
group were provided
by
three female
and
three
male
therapists
(ages
ranged
from
25 to 33
years;
M =
28)
at
local men-
tal
health outpatient agencies, including
the
treatment services branch
of
the
juvenile
court.
One of the
therapists
was
African
American,
and
the
others were White. Each therapist
had a
master's
degree
(or
equiva-
lent training)
in
either counseling psychology, social work,
or
another
mental health-related
field, and had
approximately
4
years
of
direct
clinical
experience
with
adolescents.
The six
therapists
voluntarily
served
in the
study
for an
average
of 17
months (range,
11
to 28
months).
These
therapists
attended
2.5-hr
weekly
case reviews with
the
treatment
coordinator
from
the
juvenile court
to
discuss
the
goals
and
progress
of
each
case.
Treatment
Integrity
To
sustain
the
integrity
of
MST, therapists documented each thera-
peutic contact
by
summarizing what transpired
and how
much
progress
had
been made
in
meeting
the
goals
of
treatment; ongoing clinical
su-
pervision
and
feedback were provided throughout
the
investigation.
To
monitor
the
integrity
of
IT,
therapists
were required
to
provide monthly
reports
summarizing
the
nature
of
therapeutic contacts,
who was
pres-
ent
at the
contacts,
and
adolescent progress
in
meeting
the
goals
of
treat-
ment;
and the
project
director
(Charles
M.
Borduin)
met
periodically
with
the
therapists
to
review selected videotapes
of
sessions
and to en-
sure that
the
therapists adhered
to
their stated treatment plans. Adher-
ence
to
these treatment plans
was
also promoted
by the
juvenile court
treatment
coordinator,
who met
weekly
with
the
therapists
in the IT
condition.
Although
it was not
possible
to
include
an
independent assessment
of
the
integrity
of
either
MST or IT, the
therapists
in
both conditions
completed
a
checklist
for
each
of
their cases
to
indicate
the
systems
directly
addressed
during
the
course
of
treatment
(i.e.,
individual, mar-
ital,
family,
peer,
school)
and the
general issues addressed
in
each iden-
tified
system. These checklists revealed that, among
the MST
complet-
ers
(n =
77),
none
of the
cases
had
received interventions
in
only
one
system,
26
(33.8%)
had
received interventions
in two
different
systems
(most
often
family
and
school),
and 51
(66.2%)
had
received interven-
tions
in
three (most often
family,
school,
and
peer)
or
more systems.
In
contrast,
among
the IT
completers
(n
=
63),
57
(90.5%)
of
the
cases
had
received direct interventions
in
only
one
system
(always
the
individual
adolescent),
and the
other
6
cases
(9.5%)
had
received interventions
in
two
systems (always
the
adolescent
and the
family).
The
checklists
revealed
a
virtually
identical pattern
of
interventions
for the
cases
that
had
dropped
out of
each treatment condition. Notwithstanding
the
lim-
itations
of
using therapist
reports
to
assess
treatment integrity,
the
pat-
tern
of
interventions reported
by
therapists
in
each condition suggests
that each treatment
was at
least aimed
at the
intended systems.
Research
Procedures
All
families
who
were referred
to the
project were initially
contacted
by
phone
or a
home visit. Families were informed that 1.5-hr research
assessments
would
be
conducted shortly before treatment began
and
shortly
after
treatment
had
ended.
It was
emphasized that
the
family's
participation
in the
research
was
voluntary
and
that
refusal
to
partici-
pate
(or
exercising
the
right
to
discontinue participation
at any
time)
would
not
jeopardize
the
receipt
of
treatment services
or
result
in any
sanctions
from
the
court.
The
adolescents remained under
the
jurisdic-
tion
of the
court regardless
of
their families' decisions about participat-
ing
in the
research
assessments
or in
treatment.
Research
assistants
received approximately
20 hr of
training before
their
first
family
contact
to
standardize
the
assessment
procedures
and
to
recognize
and
attenuate circumstances (e.g.,
fatigue,
reading
problems)
that threatened
the
validity
of
the
assessments.
The
pretreat-
ment
assessment
session
was
scheduled
at the
family's
convenience
ei-
ther
in
their home
or in a
youth
center
in
their neighborhood;
the
vast
majority
(91%)
of the
families
in the MST and IT
groups completed
the

572
BORDUIN
ET AL.
assessment
in
their
homes.
At the
outset
of the
session,
a
research assis-
tant
explained
the
general procedure
and
purpose
of
the
assessment
and
obtained
written consent
or
assent
from
the
family
members. During
each
assessment,
a
series
of
self-report
instruments
and
behavior rating
inventories
were administered
in a
random order
to the
parent
(or
parents)
and
adolescent.
In
addition,
the
parent
(or
parents)
and
ado-
lescent were videorecorded
as
they
discussed
and
jointly completed
an
unrevealed
differences
task.
The
posttreatment
assessment
was
con-
ducted
at the
same location
and
with
the
same measures
as the
pretreat-
ment
assessment
within
1
week
of the
completion
of
treatment. Follow-
up
assessments
using
police
and
court records
of
adolescent criminal
activity
were
conducted approximately
4
years
after
treatment
had
been
completed.
One
of the
adolescent's teachers also completed
a
paper-and-pencil
instrument
before
and
after
treatment.
The
teacher
was
randomly
se-
lected
from
a
list
of the
adolescent's current teachers.
The
teacher
was
told that
the
adolescent
was a
participant
in a
study
of
adolescent
socialization.
Outcome Measures
A
multiagent,
multimethod assessment battery
was
used
to
obtain
outcome
measures related
to the
instrumental
and
ultimate goals
(Rosen
&
Proctor,
1981)
of
MST. Instrumental goals, which
are
theory
driven,
included improved individual adjustment
of the
adolescent
and
parent
(or
parents),
improved
family
relations,
and
improved relations
between
the
adolescent
and his or her
peers. Ultimate goals, which
are
common
to all
treatments
of
juvenile
offenders,
included
decreases
in
the
rate
and
seriousness
of
adolescent criminal activity.
Individual
Adjustment
Psychiatric
symptomatology.
Symptomatology
in
mothers,
fathers
(when
present),
and
adolescents
was
assessed
through self-reports
on
the
Symptom
Checklist—90—Revised
(SCL-90-R;
Derogatis,
1983).
The
Global Severity Index, which represents
the
best single indicator
of
the
respondent's psychiatric
functioning,
was
used
to
provide
an
overall
symptom
score
for
each
family
member.
Adolescent
behavior
problems. Behavior problems
in
adolescents
were
assessed through mothers' reports
(total
score)
on the
89-item
Re-
vised
Behavior
Problem Checklist
(RBPC;
Quay
&
Peterson,
1987).
The
measure discriminates between violent
and
nonviolent delinquents
(Blaske,
Borduin, Henggeler,
&
Mann, 1989)
and
predicts serious
offense
history
in
delinquents (Hanson
et
al.,
1984).
Family
Relations
Perceived
family
functioning.
Parental
and
adolescent perceptions
of
family
relations were evaluated with
the
30-item
Family
Adaptability
and
Cohesion Evaluation
Scales—II
(FACES-II;
Olson,
Portner,
&
Bell,
1982),
which
assesses
the
constructs
of
cohesion
and
adaptability.
Fol-
lowing
the
recommendations
of
Henggeler, Burr-Harris, Borduin,
and
McCallum
(1991),
we
treated
adaptability
and
cohesion
as
linear scales
in
subsequent statistical analyses. Family composite ratings
of
adapt-
ability
and
cohesion were created
by
averaging together
the
scores
of the
individual
family
members
on
each scale.
Observed
family
interactions. Observational measures were based
on
the
family
members' videorecorded discussion
on the
nine-item
Un-
revealed
Differences
Questionnaire—Revised
using
procedures
de-
scribed
in
previous publications (Blaske
et
al.,
1989?
Mann, Borduin,
Henggeler,
&
Blaske,
1990).
Three reliable dimensions
of
family
in-
teraction derived
from
factor
analysis were
assessed.
Supportiveness
represents
the
observed encouragement
and
respect
between members
of a
family
dyad.
Verbal
activity
reflects
the
amount
of
verbal activity
between
members
of a
dyad.
Conflict-hostility
reflects emotional nega-
tivity
resulting
from
a
clash
of
opposing interests
and
ideas within
a
dyad.
For
subsequent analyses, composite
(factor)
scores were derived
from
the
variables constituting each
factor.
Peer
Relations
Maternal
and
teacher perceptions
of the
adolescent's
peer relations
were
evaluated
with
the
13-item
Missouri Peer Relations Inventory
(MPRI;
Borduin, Blaske, Cone, Mann,
&
Hazelrigg,
1989).
The
MPRI
measures three
factor
analytically derived dimensions
of
peer relations:
emotional
bonding,
aggression,
and
social maturity;
the
construct
va-
lidity
of
these dimensions
has
been supported
in
studies
of
serious juve-
nile
offenders
(e.g.,
Blaske
et
al.,
1989;
Henggeler
et
al.,
1992).
For
sub-
sequent
analyses,
factor
scores were derived
from
the
mothers'
and
teachers'
ratings
on
each dimension
of
peer relations
and
were averaged
across
the two
groups
of
respondents
(the
mean
r
between mothers'
and
teachers'
ratings
was
.39).
Criminal Activity
Juvenile
court, local police,
and
Department
of
Public
Safety
(state
police) records, collected
an
average
of
3.95 years
(SD =
1.03; range,
2.04
to
5.41)
from
the
time
of the
adolescent's
release
from
juvenile
court supervision (i.e., probation), were used
to
obtain data
on
post-
probation arrests. Arrest
data
for
each
offender
were anchored
by the
point
of
release
from
probation (i.e., within
2
weeks
of
treatment ter-
mination
for 96% of
completers
and an
average
of 6
months
from
the
time
of
referral
for
dropouts
and
refusers)
to
provide
a
distinct begin-
ning
for the
follow-up
period
for
treatment completers, dropouts,
and
refusers.
Adolescents
with
longer
follow-up
time periods were among
the
earlier participants
in the
project, whereas youths
with
shorter fol-
low-ups
were generally among
the
later referrals
to the
project.
Although
we
were able
to
track
189
(94.5%)
of the
adolescents
to the
end of the
follow-up
period,
11
adolescents were lost
to
follow-up
after
2
to 3
years
of
tracking. Given that
we had
obtained
follow-up
data
on
each
of the
11
adolescents
(6
recidivists,
5
nonrecidivists)
for
approxi-
mately
2.5
years,
we
decided
not to
drop these adolescents
from
our
follow-up
sample.
Results
Preliminary Analyses
Pretreatment
Comparability
of
Treatment Groups
Analyses
were
completed
to
examine
whether
participants
assigned
to MST
(n
= 92) and
those
assigned
to IT (n = 84)
differed
at
pretreatment
on
criminal
history
and
demographic
characteristics
as
well
as on
measures
of
individual
adjustment,
family
relations,
and
peer relations.
(These
analyses
collapsed
across
the
treatment completers
and
dropouts
in
each
condi-
tion.)
ANOVAs
and
chi-square
tests
revealed
no
significant
differences
between
participants
in the two
treatment
conditions.
Attrition
ANOVAs
were
used
to
examine
whether
the MST
completers
(n
=
77),
MST
dropouts
(n
=
15),
IT
completers
(«
=
63),
IT
dropouts
(n =
21),
and
treatment
refusers
(n = 24)
differed
on
any of the
pretreatment assessment measures
of
individual
adjustment,
family
relations,
or
peer relations
(as
noted pre-
viously,
these groups
did not
differ
in
their
criminal
histories
or

SPECIAL
SECTION:
MULTISYSTEMIC
TREATMENT
573
demographic
characteristics).
Across
a
large number
of
tests,
no
differences
emerged.
Treatment
Outcomes
Instrumental
Outcomes
Repeated measures multivariate analyses
of
variance
(MANOVAs)
and
ANOVAs
were used
to
evaluate whether sig-
nificant
changes pre-
to
postassessment
were experienced
by the
70 MST
youths
and
families
or 56 IT
youths
and
families
who
completed pretreatment
and
posttreatment
assessments. Table
1
presents
the
means
and
standard deviations
for the
measures
of
instrumental outcomes (i.e., individual adjustment,
family
relations, peer
relations)
at
pre-
and
posttreatment assessments.
Significant
MANOVAs
for the
effect
of
time were found
on the
measures
of
observed
mother-adolescent
relations,
F(3,
123)
=
4.84,
p <
.003; observed father-adolescent relations,
F(
3,62)
=
7.08,
p <
.001;
and
observed mother-father relations,
F(3,
60) =
7.84,
p <
.001.
MANOVAs
for the
interaction between
treatment
group
and
time were significant
for the
FACES-II,
F(2,
124)
=
3.04,
p <
.05; observed
mother-adolescent
re-
lations,
F(
3,123)
=
4.99,
p <
.003; observed father-adolescent
relations,
F(3,
62) =
3.42,
p <
.023;
and
observed
mother-
father
relations, F(3,
60) =
2.98,
p <
.038.
The
results
of the
ANOVAs
for the
effect
of
time
and for the
Treatment Group
X
Time
interaction
are
shown
in
Table
1.
The
following
discussion
addresses
the
results
of the
ANOVAs
that showed
a
significant
interaction
effect,
with
within-group
t
tests used
to
evaluate
change
over time
for
each group.
Individual
adjustment.
As
shown
in
Table
1,
significant
in-
teraction
effects
were
found
for
mothers'
and
fathers' reports
of
psychiatric
symptomatology
(SCL-
90-R);
mothers
and
fathers
in
the MST
group showed decreases
in
their symptoms
from
pre-
to
posttreatment, whereas their counterparts
in the IT
group showed either
an
increase (mothers)
or no
change
(fathers)
in
their symptoms.
In
addition,
a
significant interac-
tion
effect
emerged
for
mothers' reports
of
adolescent behavior
problems; mothers
in the MST
group reported
a
decrease
in
adolescent behavior problems
from
pre-
to
posttreatment,
whereas
mothers
of
youths receiving
IT
reported
an
increase
in
behavior
problems.
Family
relations.
Significant
interaction
effects
were
ob-
served
for
both measures
of
perceived
family
functioning
(FACES-II).
Families receiving
MST
reported increases
in
fam-
ily
cohesion
and
adaptability
at
posttreatment, whereas
re-
ported
family
cohesion
and
adaptability decreased
in the IT
condition.
On the
observational measures,
the
analyses generally indi-
cated that families
in the MST
group evidenced many more
positive
changes
in
their dyadic interactions than
did
families
in
the
IT
group.
Specifically,
in the MST
group,
mother-adoles-
cent dyads, father-adolescent dyads,
and
mother-father dyads
showed
increased supportiveness
and
decreased conflict-hostil-
ity
from
pre-
to
posttreatment.
In
contrast, dyadic relations
for
families
in the IT
group either deteriorated (decrease
in
mother-adolescent supportiveness, increase
in
father-adoles-
cent conflict-hostility)
or
showed
no
change
(on
measures
of
supportiveness
and
conflict-hostility).
Peer
relations.
The
composite measures
of
adolescent
peer
relations
(MPRI)
did not
show
any
significant
interaction
effects.
Ultimate
Outcomes
Measures
of
ultimate outcome were
based
on
arrest
data
that
were
collected during
follow-up.
Survival
functions. Survival analysis
(based
on the
LIFE-
TEST procedure;
SAS
Institute,
1991)
was
used
to
obtain
the
cumulative
survival functions
(or
survival curves)
for
partici-
pants
who
were randomly assigned
to the MST (n =
92)
or IT
(n
= 84)
groups, whose average follow-up periods were 1447.4
days
and
1425.2 days, respectively. This analysis collapsed
across
treatment
completers
and
dropouts
in
each group
to
pro-
vide
a
conservative
test
of
treatment
effects.
The
cumulative sur-
vival
function
represents
the
proportion
of
participants surviv-
ing
arrest
(i.e.,
not
arrested)
in
each group
by the
length
of
time
(in
days)
from
release
from
probation.
A
log-rank
test
revealed
that
the
survival functions
for the two
groups were
significantly
different,
x2(
1, N =
176)
=
46.39,
p <
.0001.
As
depicted
in
Figure
1,
youths
in the MST
group were
at
lower risk
of
arrest
(i.e., more
likely
to
"survive")
during
follow-up
than were
youths
in the IT
group.
By the end of 4
years
(1,460
days),
71.4%
of the
youths
in the IT
group
had
been arrested
at
least
once,
compared
with
26.1%
of the
youths
in the MST
group.
We
conducted another survival analysis
to
compare
MST
completers
and IT
completers
to
each other,
as
well
as to MST
dropouts,
IT
dropouts,
and
treatment refusers. Thus,
in
this
analysis,
we
included each
of the
adolescents
(N
=
200)
who
had
participated
in the
pretreatment assessment.
A
log-rank
test revealed that
the
overall
set of
differences
between
the
sur-
vival
functions
for the five
groups
was
highly significant,
x2(4,
N
=
200)
=
58.89,
p <
.0001.
Pairwise
comparisons
of the
sur-
vival
curves shown
in
Figure
2
revealed that
the MST
complet-
ers
were
at
lower risk
of
arrest during
follow-up
than were
IT
completers,
x2
(1,
^
=
140)
=
49.95,
p <
.0001;
MST
dropouts,
X2(
1,
TV
= 92) =
9.66,
p <
.002;
IT
dropouts,
x2(
1, N = 98) =
40.98,
p <
.0001;
or
treatment refusers,
\2(l,
N =
101)
=
61.20,
p <
.0001.
MST
dropouts also were
at
lower risk
of
arrest
during
follow-up than were treatment refusers,
\2(l,
N = 39)
=
4.38,
p <
.04;
or IT
dropouts,
x2(
1,
N = 36) =
2.80,
p <
.09,
although
the
latter comparison only approached significance.
IT
completers were
not
significantly
different
from
IT
dropouts
(p
=
.34),
MST
dropouts
(p=
.15),
or
refusers
(p
=
.28).
At
4
years
of
follow-up,
the
overall recidivism
rate
for MST
completers
(22.1%)
was
less than one-third
the
overall
rate
for
IT
completers
(71.4%),
IT
dropouts
(71.4%),
or
treatment
re-
fusers
(87.5%)
and
approximately one-half
the
overall
rate
for
MST
dropouts
(46.6%).
Number
and
seriousness
of
arrests.
Additional analyses
ex-
amined
the
number
of
arrests
and the
seriousness
(based
on
the
17-point
seriousness scale noted
previously)
of
those arrests
among recidivists
in the MST and IT
groups
(completers
and
dropouts were combined
in
each
group).
The first set of
analy-
ses
revealed that recidivists
in the MST
group
had
been arrested
less
often
(M
=
1.71,
SD =
1.04) during
follow-up
than their
counterparts
in the IT
group
(M =
5.43,
SD =
3.62),
F(
1, 82)
=
10.36,
p <
.002.
In
addition,
a
similar pattern emerged when

574
BORDUIN
ET AL.
Table
1
Group
Means, Standard Deviations,
and F
Values
for
Treatment Completers
on
Measures
of
Instrumental Outcomes
Multisystemic
therapy
completers
Measure
(n
Pre
=
70)
Post
Individual
therapy completers
(«
=
Pre
'56)
Post
Repeated
ANOVA
F
Time
Group
X
Time
Individual
adjustment
SCL-
90-R
(z
scores)
Mother
M
SD
Father
M
SD
Adolescent
M
SD
RBPC
(z
scores)
M
SD
0.12
1.02
0.06
0.90
0.03
0.94
0.17
0.74
-0.15"
0.97
-0.07"
0.77
-0.15
0.79
-0.54"
0.81
0.04
1.17
0.06
1.05
-0.05
0.98
-0.15
0.80
0.20b
1.26
0.19
1.09
-0.07
1.03
0.64b
0.85
0.31
0.05
6.02*
0.33
4.16*
4.44*
0.36
4.97*
Family relations
FACES-II
(mean
of
mother,
father,
and
adolescent
z
scores)
Cohesion
M
SD
Adaptability
M
SD
Observational measures (factor
scores):
mother-adolescent
Support!
veness
M
SD
Verbal
activity
M
SD
Conflict-hostility
M
SD
Observational measures (factor
scores):
father-adolescent
Supportiveness
M
SD
Verbal
activity
M
SD
Conflict-hostility
M
SD
Observational measures (factor scores):
Mother-father
Supportiveness
M
SD
Verbal
activity
M
SD
Conflict-hostility
M
SD
-0.11
0.86
-0.03
0.72
-0.09
1.01
-0.05
1.03
0.09
1.01
0.06
1.03
0.10
1.07
0.15
0.93
-0.01
1.00
0.12
0.94
0.27
0.96
0.14"
0.88
0.13"
0.86
0.23"
0.90
-0.16
1.05
-0.54"
0.76
1.06"
1.27
0.70
1.04
-0.63"
0.82
0.79"
1.03
0.10
0.86
-0.54*
0.57
0.11
0.82
0.04
0.85
0.10
0.99
0.05
0.97
-0.11
0.98
-0.07
0.97
-0.12
0.90
-0.18
1.06
-0.09
1.06
-0.15
1.07
-0.04
1.11
-0.08"
0.74
-0.16"
0.71
-0.14"
0.93
0.26
0.94
-0.22
0.85
0.23
0.90
0.26
0.94
0.27"
0.79
0.31
0.82
-0.22
1.03
-0.03
1.03
0.42
0.58
1.78
2.65
14.13***
19.91***
13.01***
0.68
17.49***
0.85
2.24
3.83*
5.49*
6.42**
3.07
5.30*
9.18**
1.31
6.66**
6.33**
1.65
4.34*
(table
continues)

SPECIAL
SECTION: MULTISYSTEMIC TREATMENT
575
Table
1
(continued)
Measure
Multisystemic
Individual
therapy
completers
therapy
completers
(n
= 70) (n = 56)
Repeated
ANOVA
F
Pre
Post
Pre
Post Time Group
X
Time
Peer relations
MPRI
(mean
of
mother
and
teacher
factor
scores)
Peer bonding
M
SD
Peer aggression
M
SD
Peer maturity
M
SD
0.24
1.52
-0.15
2.02
0.04
1.87
0.34
1.07
-0.11
1.88
0.17
1.81
0.41
1.24
0.10
1.74
-0.07
2.19
0.44
1.13
0.32
1.65
0.24
1.88
0.34
0.08
0.92
0.76
1.56
0.14
Note.
The
univariate
dfe
for
each measure
are as
follows: mother
SCL-90-R
and
observed
mother-adolescent
relations
(1,
123); father
SCL-90-
R and
observed father-adolescent relations
(1,
64); adolescent
SCL-90-R,
RBPC,
FACES-II,
and
MPRI
(1,
125);
observed mother-father relations
(1,
62). SCL-90-R
=
Symptom
Checklist—90—Revised;
RBPC
=
Revised Behavior Problem Checklist;
FACES-II
=
Family Adaptability
and
Cohesion
Evaluation
Scales-II;
MPRI
=
Missouri
Peer Relations Inventory.
*
Significant
decrease
from
pretreatment
to
posttreatment.
b
Significant
increase
from
pretreatment
to
posttreatment.
*p<.05.
**p<.01.
***p<.001.
the
comparison included treatment completers only; recidivists
who
had
completed
MST had
fewer
arrests
(M =
1.57,
SD =
0.85)
during
follow-up
than
did
recidivists
who had
completed
IT
(M =
4.41,
SD =
3.89),
F(l,
60) =
10.42,
p <
.002. Other
between-groups
comparisons (e.g., treatment completers
vs.
dropouts,
dropouts
vs.
refusers)
of
recidivists revealed
no
sig-
nificant differences
on
number
of
arrests.
The
second
set of
analyses indicated that recidivists
in the
MST
group
had
been arrested
for
less serious crimes
(M
=
5.17,
SD
=
5.01)
during
follow-up
than their counterparts
in the IT
group
(Af
=
9.40,
SD =
3.37),
F(l,
82) =
20.10,
p <
.0001.
Similarly,
when completers only were compared,
the
analysis
showed
that recidivists
who had
completed
MST had
been
ar-
rested
for
less serious
offenses
(M =
6.35,
SD =
4.67)
than
re-
1-0
-T
MST
0-0
1
1
1 1 1 1 1 1
r
0 200 400 600 800
1000 1200 1400 1600 1800
DAYS
SURVIVED
Figure
1.
Survival
functions
for
multisystemic therapy
(MST)
and in-
dividual
therapy
(IT)
groups. Completers
and
dropouts
are
combined
in
each group.
MST
Completers
O
ce
o.
<
0.4
-
DC
U)
0 200 400 600 800
1000 1200 1400 1600 1800
DAYS
SURVIVED
Figure
2.
Survival functions
for
multisystemic therapy
(MST)
com-
pleters,
MST
dropouts, individual therapy
(IT)
completers,
IT
drop-
outs,
and
treatment refusers.

576
BORDUIN
ET AL.
cidivists
who
completed
IT
(M =
9.67,
SD =
3.38),
F( 1, 60) =
12.80,
p <
.001.
Thus,
MST was
more
effective
than
IT in re-
ducing
the
number
and
seriousness
of
crimes among
those
youths
who
were arrested.
A
r
rests
for
violent
crimes.
In
light
of our
considerable inter-
est in
preventing violent
offending
in
this high-risk sample,
a
hierarchical
multiple regression analysis
was
used
to
evaluate
the
effect
of
treatment
on
violent
offending.
The
number
of ar-
rests
for
violent crimes (i.e., rape, attempted
rape,
sexual
as-
sault,
aggravated assault, assault
and
battery)
during
follow-up
served
as the
dependent variable,
and the
independent variables,
in
order
of
entry, were
(a) the
number
of
arrests
for
violent
crimes
before treatment
and (b)
treatment group (coded
as a
dummy
variable,
MST or IT,
that collapsed across treatment
completers
and
dropouts
in
each
group).
This regression anal-
ysis
revealed
that,
even
after
the
effect
of
pretreatment arrests
for
violent crimes
was
controlled, treatment group
was a
highly
significant
predictor
of
arrests
for
violent crimes during
follow-
up,
F(2,
173)
=
11.74,
p <
.0008.
Youths
who
participated
in
MST
were less
likely
to be
arrested
for
violent crimes
following
treatment than were youths
who
participated
in IT.
When this
analysis
was
repeated
using
treatment
completers
only,
treat-
ment
group
was
again
a
significant
predictor,
F( 2,
137)
=
8.66,
p
<
.003, with
MST
completers
having
a
lower likelihood
of
violent
offending
during
follow-up
than
IT
completers.
Factors
associated with ultimate outcome. Hierarchical
multiple
regression analyses were used
to
evaluate
the
effects
of
potential
moderators
(age,
race,
social
class,
gender,
pretreat-
ment
arrests)
of MST
effectiveness.
The
dependent variable
was
the
number
of
posttreatment arrests.
For
each regression,
a
dummy
variable representing treatment group (collapsing
across treatment completers
and
dropouts
in
each group)
was
the
first
variable entered,
the
moderating variable
was
entered
second,
and the
cross-product term
of the
treatment group
and
the
moderating variable
was
entered
last.
The
significance
of
change
in
R2
for the
cross-product term indicated whether
MST
was
differentially
effective
with youths
and
families
from
different
backgrounds.
In no
case
did the
cross-product term
contribute
a
significant
portion
of
additional variance. Thus,
for
example,
MST was
equally
effective
with youths
of
different
gender
and
ethnic backgrounds.
A
similar pattern emerged
when
these analyses were repeated using treatment completers
only.
Discussion
The
findings
clearly
demonstrate
the
impact
of MST on key
family
correlates
of
antisocial behavior
and on
individual
ad-
justment
in
family
members.
At
posttest,
MST had
highly
fa-
vorable
effects
on
perceived
family
relations (increased cohe-
sion
and
adaptability)
and
observed
family
interactions
(increased
supportiveness
and
decreased conflict-hostility
across
family
dyads).
Moreover,
MST
resulted
in
decreased
symptomatology
in
parents (based
on
self-reports)
and de-
creased behavior problems
in the
youths (based
on
parental
reports).
Most importantly,
however,
MST
produced
long-
standing change
in
youths' criminal behaviors.
'Youths
treated
with
MST
were significantly less
likely
than
comparison
coun-
terparts
to be
rearrested within
4
years after
treatment
termina-
tion, and, when rearrested,
had
committed
significantly
less
se-
rious
offenses.
In
addition,
the
relative
efficacy
of MST (as re-
flected
in
the
number
of
posttreatment
arrests)
was not
moderated
by
measured demographic characteristics
(i.e.,
race,
gender,
age, social class, pretreatment
arrests),
suggesting that
MST was not
differentially
effective
with youths
and
families
of
divergent
backgrounds.
During
follow-up,
youths
who
dropped
out of MST
were
at
higher
risk
of
arrest than were
MST
completers (who averaged
almost
20
more hours
in
treatment)
but
were
at
lower risk
of
arrest
than were
IT
dropouts
(who
received about
the
same amount
of
treatment
as MST
dropouts)
or
treatment refusers.
It is
possible
that
a
short dose (i.e., approximately
4 hr) of
involvement
in MST
may
help
to
reduce
the
risk
of
later criminal activity
for
some
ju-
venile
offenders.
Such
a
suggestion
is
consistent with other
re-
search indicating that
48% to 58% of
psychotherapy patients show
measurable improvement
in
four
to
seven sessions, regardless
of
the
ultimate duration
of
treatment
(Howard,
Kopta,
Krause,
&
Orlinsky,
1986).
Caution must
be
expressed about this suggestion,
however,
because
it
does
not
explain
why MST
dropouts were
at
lower
risk
of
arrest than were
IT
dropouts. Perhaps some
of the
youths
and
families
in the MST
group decided
to
drop
out
of
treat-
ment because they
had
obtained needed skills
or
resources,
al-
though
the
therapist believed otherwise. Certain features
of MST
may
have also contributed
to the
relatively lower risk
of
rearrest
among
MST
dropouts. Indeed,
in
contrast
to
individual therapy,
MST
focuses
on
youths' social systems throughout treatment
and
intervenes directly
in
those
systems. Moreover,
from
the
outset
of
treatment,
MST
attempts
to
empower
family
members and,
within
a
context
of
support
and
skill
building, emphasizes
the
need
for
behavior change across
key
systems linked with antisocial
be-
havior.
Measurement
of a
broader range
of
factors (e.g.,
family
expectations
of
treatment, therapist-family
relationship)
may en-
hance
our
understanding
of
outcomes among
MST
dropouts.
The
effectiveness
of MST in
reducing criminal activity
in
high-risk
youths
has
important implications regarding
the de-
sign
of
violence prevention programs
for
such youths.
If, as
sug-
gested earlier,
a
major shortcoming
of
most interventions
for
preventing
and
treating delinquency
has
been their neglect
of
the
multiple determinants
of
antisocial behavior, then
the
suc-
cess
of MST may be
linked with
its
comprehensive nature;
that
is,
the
results
of MST may be due to its
explicit
focus
on
ame-
liorating
key
social-ecological factors associated with delin-
quency,
including behavior problems, parental disturbance,
problematic
family
relations, association with deviant peers,
and
poor school performance. This proposition
is
supported,
in
part,
by
instrumental
outcomes
demonstrating significant
effects
on
youth behavior problems, parental symptomatology,
and
family
relations.
On the
other hand, favorable changes
in
adolescents' peer
re-
lations were
not
observed, which
was
unexpected
in
light
of the
central role
of
peer relations
in
causal models
of
delinquency,
the
clear intent
of MST to
decrease youths' association with
antisocial peers,
and
earlier
findings
regarding
the
capacity
of
MST to
improve peer relations
(Henggeler
et
al.,
1986,
1992).
Perhaps,
in the
present study, criminal behavior
was
influenced
primarily
by
strengthening
the
family
so
that
the
deleterious
effects
of
association with deviant peers were
buffered.
Alterna-
tively,
relevant peer relations constructs
may not
have been

SPECIAL
SECTION:
MULTISYSTEMIC
TREATMENT
577
tapped adequately because
of the
dearth
of
well-validated
mea-
sures
in the
literature.
At any
rate,
the
effect
of MST on
peer
relations
of
juvenile
offenders
should
be
clarified
by
several clin-
ical
trials that
are
currently
in
progress.
Similarly, current clin-
ical
trials
are
also examining whether improved school perfor-
mance
is
associated
with favorable long-term outcome
as
pos-
ited
by the MST
model. Taken together,
findings
from
these
studies should elucidate
the
linkages between changes
in
juve-
nile
offenders'
extrafamilial systems
and
changes
in
their anti-
social behavior
after
MST.
A
second implication
of the
present
findings for the
design
of
violence
prevention programs pertains
to the
accessibility
and
ecological validity
of
services. Traditionally,
as
Melton
and
Pag-
liocca
(1992)
emphasized, mental health services
for
juvenile
offenders
either have been inaccessible
(i.e.,
office
based)
or
have
provided interventions that have little bearing
on the
nat-
ural
ecology
of
youths (e.g., residential treatment centers,
incarceration).
In
contrast,
by
using
the
family
preservation
model
of
service delivery,
MST was
provided
in
natural com-
munity
contexts (e.g., home, school, recreation center).
The de-
livery
of
services
in the
natural ecology
of
youths
has
several
advantages including
the
promotion
of
family
cooperation
and
the
acquisition
of
more accurate data regarding
the
assessment
of
identified
problems
and the
results
of
treatment interven-
tions
(Henggeler
&
Borduin, 1990). Indeed,
two
independent
American
Psychological Association
(APA)
Task Force
reports
have
emphasized
the
importance
of
providing children's mental
health
services that both recognize
the
natural ecology
of the
child
and
diminish barriers
to
access
(Henggeler,
1994;
Roberts,
1994).
As
described earlier,
we are
contending that
the
favorable
re-
sults
of
this study were largely
due to two
crucial aspects
of
MST:
its
comprehensive nature
and
ecologically valid delivery.
However,
it
must
be
noted that
the
design
of
this study con-
founds
the
examination
of
this issue,
as the
comparison treat-
ment
(i.e.,
office-based
individual
therapy)
was
neither compre-
hensive
nor
delivered
in
adolescents' natural ecologies.
A
study,
for
example, that compared
the
effects
of MST
with
the
effects
of
a
less comprehensive home-based treatment (e.g., Home-
builders;
Kinney,
Haapala,
&
Booth,
1991)
would address
the
issue
of
whether
both
comprehensiveness
and
ecological validity
are
necessary conditions
of
success. Moreover, therapist moti-
vation
and
commitment
may
have been confounded
in
this
study.
Indeed,
as
described elsewhere (Henggeler
&
Borduin,
1990),
therapist characteristics such
as
motivation, social
facil-
ity,
intelligence,
and flexibility are
viewed
as
crucial
to
success-
ful
outcome
in
MST. Nevertheless,
one can
safely
assume that
these characteristics
are not
sufficient
for
favorable outcome
with
delinquents,
as
many previous studies that
did not
obtain
favorable
long-term results must have
had
motivated
and
com-
petent therapists. Thus, although
the
design
of
this study
did
not
permit
a
determination
of the
critical conditions
of
positive
therapeutic outcome,
we
suggest that successful interventions
for
serious antisocial behavior
in
youths must
be
comprehen-
sive,
ecologically valid,
and
delivered with
skill
and
persistence.
In
conclusion,
the
results
of
this study indicate that
a
com-
prehensive
intervention, addressing
the
multiple determinants
of
antisocial behavior
in
youths' naturally occurring systems,
can
successfully
reduce criminal activity
and
violent
offending
in
serious juvenile offenders. When considered along with other
findings
regarding MST, conclusions from recent
APA
Task
Force reports,
and
federal
(Center
for
Mental Health Services,
RFA
No. SM
94-01)
and
foundation (e.g., England
&
Cole,
1992)
initiatives
to
reform mental health services
for
youths,
the
present
findings
have clear implications.
The
restrictive,
narrowly
focused,
and
family-blaming
practices
that have dom-
inated children's mental health services have
not
been
effective.
For
the
optimization
of
positive outcomes
in the
treatment
and
prevention
of
serious clinical problems, current practices must
be
changed
to
emphasize child-centered, family-focused, com-
prehensive,
flexible, and
ecologically valid services.
References
Bank,
L.,
Marlowe,
J. H.,
Reid,
J. B.,
Patterson,
G. R., &
Weinrott,
M. R.
(1991).
A
comparative evaluation
of
parent-training interven-
tions
for
families
of
chronic delinquents. Journal
of
Abnormal Child
Psychology,
19,
15-33.
Blaske,
D.
M.,
Borduin,
C.
M.,
Henggeler,
S. W., &
Mann,
B. J.
(1989).
Individual,
family,
and
peer characteristics
of
adolescent
sex
offenders
and
assaultive
offenders.
Developmental
Psychology,
25,
846-855.
Borduin,
C. M.
(1994).
Innovative models
of
treatment
and
service
delivery
in the
juvenile justice system. Journal
of
Clinical
Child
Psy-
chology,
23,
19-25.
Borduin,
C.
M.,
Blaske,
D.
M.,
Cone,
L.,
Mann,
B.
J.,
&
Hazelrigg,
M. D.
(1989).
Development
and
validation
of
a
measure
of
adolescent
peer
relations:
The
Missouri
Peer
Relations
Inventory.
Unpublished
manuscript, Department
of
Psychology, University
of
Missouri—
Columbia.
Borduin,
C.
M.,
&
Henggeler,
S. W.
(1990).
A
multisystemic
approach
to the
treatment
of
serious delinquent behavior.
In R. J.
McMahon
& R. D.
Peters
(Eds.),
Behavior disorders
of
adolescence:
Research,
intervention,
and
policy
in
clinical
and
school settings (pp.
63-80).
New
\brk:
Plenum.
Borduin,
C.
M.,
Henggeler,
S.
W.,
Blaske,
D.
M.,
&
Stein,
R.
(1990).
Multisystemic
treatment
of
adolescent sexual
offenders.
Interna-
tional
Journal
of
Offender
Therapy
and
Comparative Criminology,
34,
105-113.
Bronfenbrenner,
U.
(1979).
The
ecology
of
human
development:
Exper-
iments
by
nature
and
design. Cambridge,
MA:
Harvard University
Press.
Brunk,
M.,
Henggeler,
S.
W.,
&
Whelan,
J. P.
(1987).
A
comparison
of
multisystemic
therapy
and
parent training
in the
brief treatment
of
child abuse
and
neglect. Journal
of
Consulting
and
Clinical
Psychol-
ogy,
55,311-318.
Children's Defense Fund.
(1992).
The
state
of
America's children:
1992. Washington,
DC:
Author.
Derogatis,
L.
R.(1983).
TheSCL-90-R:Administration,
scoring,
andpro-
cedures
manual-II.
Towson,
MD:
Clinical Psychometric Research.
Elliott,
D.
S.,
Huizinga,
D.,
&
Ageton,
S. S.
(1985).
Explaining delin-
quency
and
drug use. Beverly Hills,
CA:
Sage.
England,
M.
J.,
&
Cole,
R. F.
(1992).
Building
systems
of
care
for
youth
with
serious mental illness. Hospital
and
Community
Psychiatry,
43,
630-633.
Guerra,
N.
G.,
&
Slaby,
R. G.
(1990).
Cognitive mediators
of
aggres-
sion
in
adolescent
offenders:
2.
Intervention.
Developmental
Psychol-
ogy,
26,
269-277.
Hanson,
C.
L.,
Henggeler,
S.
W.,
Haefele,
W. F, &
Rodick,
J. D.
(1984).
Demographic, individual,
and
family
relationship correlates
of
seri-
ous and
repeated crime among adolescents
and
their siblings. Journal
of
Consulting
and
Clinical
Psychology,
52,
528-538.
Henggeler,
S. W.
(1989).
Delinquency
in
adolescence. Newbury Park,
CA:
Sage.

578
BORDUIN
ET AL.
Henggeler,
S. W.
(1994).
A
consensus: Introduction
to the APA
Task
Force
report
on
innovative models
of
treatment
and
service delivery
for
children, adolescents,
and
their families. Journal
of
Clinical Child
Psychology,
23,
3-6.
Henggeler,
S.
W.,
&
Borduin,
C. M.
(1990).
Family
therapy
and
beyond:
A
multisystemic
approach
to
treating
the
behavior problems
of
chil-
dren
and
adolescents.
Pacific Grove,
CA:
Brooks/Cole.
Henggeler,
S.
W.,
Burr-Harris,
A.
W.,
Borduin,
C.
M.,
&
McCallum,
G.
(1991).
Use of the
Family Adaptability
and
Cohesion Evaluation
Scales
in
child clinical
research.
Journal
of
Abnormal Child
Psychol-
ogy,
19,
53-63.
Henggeler,
S.
W.,
Melton,
G. B., &
Smith,
L. A.
(1992).
Family preser-
vation
using multisystemic therapy:
An
effective
alternative
to
incar-
cerating
serious
juvenile
offenders.
Journal
of
Consulting
and
Clinical
Psychology,
60,
953-961.
Henggeler,
S.
W.,
Melton,
G.
B.,
Smith,
L.
A.,
Schoenwald,
S.
K.,
&
Hanley,
J. H.
(1993).
Family preservation using multisystemic treat-
ment: Long-term follow-up
to a
clinical trial with serious juvenile
offenders.
Journal
of
Child
and
Family Studies,
2,
283-293.
Henggeler,
S.
W.,
Rodick,
J.,
Borduin,
C.
M.,
Hanson,
C.,
Watson,
S.,
&
Urey,
J.
(1986).
Multisystemic treatment
of
juvenile
offenders:
Effects
on
adolescent behavior
and
family
interaction. Developmental
Psychology,
22,
132-141.
Hollingshead,
A. B.
(1975).
The
four-factor
index
of
social status.
Un-
published
manuscript,
Yale
University,
New
Haven,
CT.
Howard,
K.
I.,
Kopta,
S.
M.,
Krause,
M.
K.,
&
Orlinsky,
D. E.
(1986).
The
dose-effect relationship
in
psychotherapy. American
Psycholo-
gist,
41,
159-164.
Kazdin,
A. E.
(1987).
Treatment
of
antisocial behavior
in
children:
Current status
and
future
directions. Psychological Bulletin,
102,
187-203.
Kinney,
J.,
Haapala,
D.,
&
Booth,
C.
(1991).
Keepingfamilies
together:
The
Homebuilders
model.
New
\brk:
Aldine
De
Gruyter.
Lewis,
D.
Q,
Lovely,
R.,
Yeager,
C.,
&
Famina,
D.
(1989).
Toward
a
theory
of the
genesis
of
violence:
A
follow-up study
of
delinquents.
Journal
of
the
American Academy
of
Child
and
Adolescent
Psychiatry,
25,431-436.
Lipsey,
M. W.
(1992,
November).
The
effects
of
treatment
on
juvenile
delinquents:
Results from
meta-analysis.
Paper presented
at the Na-
tional Institute
of
Mental Health Meeting
for
Research
to
Prevent
Youth
Violence, Bethesda,
MD.
Loeber,
R.,
&
Dishion,
T.
(1983).
Early
predictors
of
male delinquency:
A
review.
Psychological
Bulletin,
94,
68-99.
Mann,
B.
J.,
Borduin,
C.
M.,
Henggeler,
S.
W.,
&
Blaske,
D. M.
(1990).
An
investigation
of
systemic conceptualizations
of
parent-child
co-
alitions
and
symptom
change.
Journal
of
Consulting
and
Clinical
Psychology,
58,
336-344.
Melton,
G.
B.,
&
Pagliocca,
P. M.
(1992).
Treatment
in the
juvenile
justice system: Directions
for
policy
and
practice.
In J. J.
Cocozza
(Ed.),
Responding
to the
mental health needs
of
youth
in the
juvenile
justice
system
(pp.
107-139).
Seattle,
WA:
National Coalition
for
the
Mentally
111
in the
Criminal Justice System.
Mulvey,
E. P.,
Arthur,
M.
A.,
&
Reppucci,
N. D.
(1990).
Review
of
programs
for the
prevention
and
treatment
of
delinquency.
Office
of
Technology Assessment, Washington,
DC:
U.S. Government Printing
Office.
Nelson,
K. E.
(1991).
Populations
and
outcomes
in five
family
preser-
vation
programs.
In K.
Wells
& D. E.
Biegel
(Eds.),
Family preserva-
tion services: Research
and
evaluation (pp.
72-91).
Newbury Park,
CA:
Sage.
Olson,
D.
H.,
Portner,
J.,
&
Bell,
R.
(1982).
FACES-II.
In D. H.
Olson,
H. I.
McCubbin,
H. L.
Barnes,
A.
Larsen,
M.
Muxen,
& M.
Wilson
(Eds.),
Family inventories (pp.
5-24).
St.
Paul: University
of
Minne-
sota,
Department
of
Family Social Science.
Olweus,
D.
(1992).
Bullying among
school
children:
Intervention
and
prevention.
In R. D.
Peters,
R. J.
McMahon,
& V. L.
Quinsey
(Eds.),
Aggression
and
violence throughout
the
life
span
(pp.
100-125).
New-
bury Park,
CA:
Sage.
Patterson,
G.
R.,
&
Dishion,
T. J.
(1985).
Contributions
of
families
and
peers
to
delinquency. Criminology,
23,
63-79.
Quay,
H.
C.,
&
Peterson,
D. R.
(1987).
Manual'for
the
RevisedBehavior
Problem Checklist. Coral Gables,
FL:
University
of
Miami.
Roberts,
M.
C.
(1994).
Models
for
service delivery
in
children's mental
health: Common characteristics. Journal
of
Clinical Child
Psychol-
ogy,
23,212-219.
Rosen,
A.,
&
Proctor,
E. K.
(1981).
Distinctions between treatment
outcomes
and
their implications
for
treatment evaluation. Journal
of
Consulting
and
Clinical
Psychology,
49,
418-425.
SAS
Institute.
(1991).
SAS
user's guide: Statistics.
Gary,
NC:
Author.
Scherer,
D.
G.,
Brondino,
M.
J.,
Henggeler,
S.
W,
Melton,
G.
B.,
&
Hanley,
J. H.
(1994).
Multisystemic
family
preservation therapy:
Preliminary
findings
from
a
study
of
rural
and
minority serious ado-
lescent
offenders.
Journal
of
Emotional
and
Behavioral Disorders,
2,
198-206.
Simcha-Fagan,
Q, &
Schwartz,
J. E.
(1986).
Neighborhood
and
delin-
quency:
An
assessment
of
contextual
effects.
Criminology,
24,
667-
703.
Tolan,
P.
H.,
Guerra,
N.,
Van
Acker,
R.,
Huesmann,
L.
R.,
&
Eron,
L. D.
(1994,
June). Proximal
effects
of
four
levels
of
intervention
on
the
prevention
of
urban
children's
aggressive behavior.
Paper
pre-
sented
at the
meeting
of the
Society
for
Research
in
Child
and
Ado-
lescent Psychopathology, London.
Weisz,
J.
R.,
Martin,
S.
L.,
Walter,
B.
R.,
&
Fernandez,
G. A.
(1991).
Differential
prediction
of
young adult arrests
for
property
and
per-
sonal crimes: Findings
of a
cohort follow-up study
of
violent boys
from
North Carolina's Willie
M
Program. Journal
of
Child
Psychol-
ogy
and
Psychiatry,
32,
783-792.
Weisz,
J.
R.,
Walter,
B.
R.,
Weiss,
B.,
Fernandez,
G.
A.,
&
Mikow,
V. A.
(1990).
Arrests among emotionally disturbed violent
and
assaultive
individuals
following minimal versus lengthy intervention through
North
Carolina's
Willie
M
Program. Journal
of
Consulting
and
Clin-
ical
Psychology,
58,
720-728.
Weisz,
J.
R.,
Weiss,
B.,
&
Donenberg,
G. R.
(1992).
The lab
versus
the
clinic.
American Psychologist,
47,
1578-1585.
Zigler,
E.,
Taussig,
C.,
&
Black,
K.
(1992).
Early childhood interven-
tion:
A
promising preventative
for
juvenile delinquency. American
Psychologist,
47,
997-1006.
Received
October
15,
1993
Revision
received
July
29,
1994
Accepted
December
14,
1994
•