ArticlePDF Available

The Role of the Therapeutic Alliance in Psychotherapy and Pharmacotherapy Outcome: Findings in the National Institute of Mental Health Treatment of Depression Collaborative Research Program

Authors:

Abstract

The relationship between therapeutic alliance and treatment outcome was examined for depressed outpatients who received interpersonal psychotherapy, cognitive-behavior therapy, imipramine with clinical management, or placebo with clinical management. Clinical raters scored videotapes of early, middle, and late therapy sessions for 225 cases (619 sessions). Outcome was assessed from patients' and clinical evaluators' perspectives and from depressive symptomatology. Therapeutic alliance was found to have a significant effect on clinical outcome for both psychotherapies and for active and placebo pharmacotherapy. Ratings of patient contribution to the alliance were significantly related to treatment outcome; ratings of therapist contribution to the alliance and outcome were not significantly linked. These results indicate that the therapeutic alliance is a common factor with significant influence on outcome.
Journal
of
Consulting
and
Clinical Psychology
1996, Vol.
64, No. 3,
532-539
Copyright
1996
by the
American
Psychological
Association, Inc.
0022-006X/96/S3.00
The
Role
of the
Therapeutic
Alliance
in
Psychotherapy
and
Pharmacotherapy
Outcome:
Findings
in the
National
Institute
of
Mental
Health
Treatment
of
Depression
Collaborative
Research
Program
Janice
L.
Krupnick,
Stuart
M.
Sotsky,
Sam
Simmens,
and
Janet
Moyer
George Washington University Medical Center
Irene
Elkin
University
of
Chicago
John
Watkins
Atlanta
Center
for
Cognitive Therapy
Paul
A.
Pilkonis
University
of
Pittsburgh School
of
Medicine
The
relationship between therapeutic alliance
and
treatment outcome
was
examined
for
depressed
outpatients
who
received interpersonal psychotherapy, cognitive-behavior therapy, imipramine with
clinical management,
or
placebo
with clinical management. Clinical raters scored videotapes
of
early,
middle,
and
late therapy sessions
for 225
cases
(619 sessions). Outcome
was
assessed
from
patients'
and
clinical evaluators' perspectives
and
from depressive symptomatology. Therapeutic
alliance
was
found
to
have
a
significant
effect
on
clinical outcome
for
both psychotherapies
and for
active
and
placebo pharmacotherapy. Ratings
of
patient contribution
to the
alliance were
signifi-
cantly related
to
treatment outcome; ratings
of
therapist contribution
to the
alliance
and
outcome
were
not
significantly
linked.
These
results indicate that
the
therapeutic alliance
is a
common factor
with
significant
influence
on
outcome.
The
therapeutic
alliance,
denned
broadly
as the
collaborative
bond between therapist
and
patient,
is
widely
considered
to be
an
essential ingredient
in the
effectiveness
of
psychotherapy.
In
the
ongoing debate about
the
relative importance
of
modality-
specific
techniques versus common factors
for
therapeutic suc-
cess
in all
psychotherapies,
the
therapeutic alliance
has
been
de-
scribed
as
"the most promising
of the
common elements
for
future
investigation"
(Bordin,
1976)
and as
"the quintessential
integrative
variable"
(Wolfe
&
Goldfried,
1988)
in
that
it is be-
lieved
to
influence
outcome across
a
range
of
psychotherapies,
despite their therapeutic
and
technical
differences.
In
a
meta-analysis
of
therapeutic alliance studies, Horvath
and
Symonds
(1990)
concluded that
the
therapeutic alliance
had
been
significantly
associated
with
outcome
not
only across
a
number
of
investigations
but
also across
different
types
of
psy-
chotherapy.
Many
of the
studies they reviewed, however,
in-
cluded
therapists
who
described their orientations
as
"eclectic"
or who
combined treatments with
different
orientations. Thus,
Janice
L.
Krupnick, Stuart
M.
Sotsky,
Sam
Simmens,
and
Janet
Moyer,
Department
of
Psychiatry
and
Behavioral Sciences
George
Washington University Medical Center; Irene Elkin, School
of
Social
Service Administration, University
of
Chicago;
John Watkins, Atlanta
Center
for
Cognitive Therapy; Paul
A.
Pilkonis,
School
of
Medicine,
University
of
Pittsburgh.
An
earlier version
of
this article
was
presented
at the
annual meeting
of the
Society
for
Psychotherapy Research, Berkeley, California, June
1992.
This work
was
supported
by
Grant MH44296
from
the
National
Institute
of
Mental Health
(NIMH).
We
acknowledge
the
NIMH Treatment
of
Depression Collaborative
Research Program, which
was a
multisite program initiated
and
spon-
sored
by the
Psychosocial Treatments Research Branch, Division
of Ex-
tramural Research Programs (now
part
of the
Mood,
Anxiety,
and
Per-
sonality
Disorders Research Branch, Division
of
Clinical Research),
NIMH.
The
program
was
funded
by
cooperative agreements with
six
participating sites:
George
Washington University,
MH
33762; Univer-
sity
of
Pittsburgh,
MH
33753;
University
of
Oklahoma,
MH
33760;
Yale
University,
MH
33827; Clarke Institute
of
Psychiatry,
MH
38231;
and
Rush
Presbyterian-St.
Luke's Medical Center,
MH
35017.
The
prin-
cipal NIMH
collaborators
were Irene Elkin,
Coordinator
(now
at
Univer-
sity
of
Chicago);
M.
Tracie Shea, Associate Coordinator
(now
at
Brown
University);
John
P.
Docherty (now
at New
York
Hospital—Cornell
Medical Center);
and
Morris
B.
Parloff
(now
at
American University).
The
principal investigators
and
project
coordinators
at the
three par-
ticipating research sites were
as
follows:
Stuart
M.
Sotsky
and
David
Glass, George Washington University; Stanley
D.
Imber
and
Paul
A.
Pilkonis, University
of
Pittsburgh;
and
John
T.
Watkins
and
William
Leber, University
of
Oklahoma.
The
principal investigators
and
project
coordinators
at the
three research sites responsible
for
training thera-
pists were
as
follows:
Myrna Weissman
(now
at
Columbia
University),
Eve
Chevron,
and
Bruce
J.
Rounsaville,
Yale
University; Brian
F.
Shaw
and T.
Michael
Vallis,
Clarke Institute
of
Psychiatry;
and Jan A.
Fawcett
and
Philip Epstein, Rush
Presbyterian-St.
Luke's Medical Center. Col-
laborators
in the
data
management
and
data
analysis
aspects
of the
pro-
gram were
C.
James
Klett,
Joseph
F.
Collins,
and
Roderic
Gillis
of the
Veterans Administration Cooperative Studies Program, Perry Point,
Maryland.
Correspondence concerning this article should
be
addressed
to
Janice
L.
Krupnick,
who is now at the
Department
of
Psychiatry,
Georgetown
University
Medical Center,
3800
Reservoir
Road,
NW,
Washington,
DC
20007.
532
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THERAPEUTIC ALLIANCE
533
the
extent
to
which
the
therapeutic alliance
affects
outcome
for
carefully
denned,
specific
treatments
has
remained unclear.
A
substantial empirical literature
attests
to the
fundamental role
played
by the
alliance
in
psychodynamic psychotherapy
(Luborsky
&
Auerbach,
1985),
but far
fewer
studies have
ex-
plored
the
relationship between alliance
and
outcome
in
inter-
personal, cognitive,
or
behavioral treatments. Furthermore,
there have been
no
empirical investigations examining
the po-
tential relationship between alliance
and
outcome
in
pharma-
cotherapy.
Recent studies comparing
the
strength
of the
alliance
in
different
treatment modalities have produced mixed results.
Salvio,
Beutler,
Wood,
and
Engle
(1992)
found
no
significant
differences
in
alliance levels
in a
comparison
of
cognitive versus
experiential group therapies
for
depression, whereas
in
single
"significant"
sessions
of
cognitive
behavior
and
psychodynamic
interpersonal therapies, Raue, Castonguay,
and
Goldfried
(1993)
found
significantly
higher
total
alliance scores
for
cogni-
tive
behavior sessions than
for
psychodynamic interpersonal
sessions.
In
studies
of the
relationship between therapeutic alli-
ance
and
outcome, there
is
support
from
some,
but not
all, stud-
ies
of
cognitive therapy (Castonguay
et
al.,
1993; DeRubeis
&
Feeley,
1989;
Safran
&
Wallner,
1991).
In
pharmacotherapy,
a
positive
doctor-patient
relationship
has
been viewed recently
as
a
factor
in
compliance,
but
active medication
is
considered
the
primary
agent
of
change (Docherty
&
Feister,
1985).
Earlier,
however,
Downing
and
Rickels
(1978)
suggested that factors
quite
distinct
from
pharmacological properties, such
as the
doc-
tor-patient
bond, might
affect
the
response
to
both active med-
ication
and
placebo.
Only
two
studies have compared
the
relationship
of
alliance
to
outcome among
different
standardized treatments.
In the Na-
tional Institute
of
Mental Health (NIMH) Treatment
of De-
pression Collaborative Research Program
(TDCRP)
of
inter-
personal psychotherapy, cognitive behavior therapy,
and
active
and
placebo pharmacotherapy,
Krupnick
et al.
(1994)
con-
ducted
a
pilot study using
a
subset
of the
cases used
for the
pres-
ent
investigation
in an
extreme contrast group design. There
was
a
significant
relationship between alliance
and
outcome
when
ratings were pooled across treatments,
but
within treat-
ments
this relationship
was
significant
only
for
interpersonal
therapy.
Mean alliance ratings were
significantly
higher
for
psy-
chotherapy than
for
pharmacotherapy.
In a
study
of
cognitive,
behavioral,
and
brief dynamic psychotherapies
for
elderly
de-
pressed patients, Marmar, Gaston, Gallagher,
and
Thompson
(1989)
found
a
significant
relationship between alliance
and
outcome when ratings were pooled across treatments,
but
within
treatments
the
relationship
was
significant
only
in
cog-
nitive
therapy.
In
the
current study,
we
sought
to
elucidate
further
the
role
played
by the
therapeutic alliance
in the
outcome
of the
four
treatments
in the
NIMH TDCRP.
We
anticipated that
findings
from
this study might
differ
from
our
pilot results because
of
the
different
study design
and
much larger sample, that
is,
all
patients
who
completed
at
least
two
sessions.
We
were inter-
ested
in
learning whether
the
specific
effect
of the
therapeutic
alliance
in
interpersonal therapy would
be
sustained
and
whether
the
alliance
effect
might also
be
demonstrable
in
cog-
nitive
behavior therapy
and
pharmacotherapy with this larger
sample,
and
including
the
entire range
of
therapy outcomes.
To
our