Short-Term Psychodynamic Psychotherapy for Depression:
An Examination of Statistical, Clinically Significant, and
MARK J. HILSENROTH, PH.D.,1STEVEN J. ACKERMAN, PH.D.,2MATTHEW D. BLAGYS, PH.D.,3
MATTHEW R. BAITY, M.A.,3and MEGAN A. MOONEY, M.A.4
This study investigates the effectiveness of short-term psychodynamic psychother-
apy (STPP) for depression in a naturalistic setting utilizing a hybrid effectiveness/
efficacy treatment research model. Twenty-one patients were assessed pre- and post-
treatment through clinician ratings and patient self-report on scales representing
specific DSM-IV depressive, global symptomatology, relational, social, and occupational
functioning. Treatment credibility, fidelity, and satisfaction were examined, all of which
were found to be high. All areas of functioning assessed exhibited significant and
positive changes. These adaptive changes in functioning demonstrated large statistical
effects. Likewise, changes in depressive symptoms evaluated at the patient level
utilizing clinical significance methodology were found to be high. A significant direct
process/outcome link between STPP therapist techniques and changes in depressive
symptoms was observed. Alternative treatment interventions within STPP were evalu-
ated in relation to subsequent improvements in depression and were found to be
nonsignificant. The present results demonstrate that robust statistical and clinically
significant improvement can occur in a naturalistic/hybrid model of outpatient STPP
—J Nerv Ment Dis 191:349–357, 2003
A number of studies have demonstrated the effec-
tiveness and efficacy of psychodynamic psychother-
apy for depression (Anderson and Lambert, 1995;
Barber et al., 1996; Crits-Christoph, 1992; Gaston et
al., 1998; Lueger et al., 2000; Shapiro et al., 1995). In
addition, two of these studies have made particu-
larly important contributions to psychodynamic psy-
chotherapy for depression by demonstrating direct
treatment intervention-outcome relationships. The
first study found a significant relationship between
the competent delivery of psychodynamic-expres-
sive techniques early in treatment with subsequent
improvements in depression (Barber et al., 1996),
while a second study found a significant relationship
between psychodynamically derived exploratory in-
terventions made during the middle of treatment
and less depressive symptomatology at termination
(Gaston et al., 1998).
The present study seeks to replicate and extend
these earlier findings regarding the effectiveness of
1Derner Institute of Advanced Psychological Studies, Adelphi
University, 158 Cambridge Ave., 220 Weinberg Building, Garden
City, New York 11530. Send reprint requests to Dr. Hilsenroth.
2Erikson Institute for Education and Research of the Austen
Riggs Center, Stockbridge, Massachusetts, and Harvard Medical
School, Boston, Massachusetts.
3Massachusetts General Hospital and Harvard Medical
School, Boston, Massachusetts.
4Department of Psychiatry and Behavioral Sciences, Baylor
College of Medicine, Houston, Texas.
Earlier versions of this study were presented at the annual
meetings of the Society for Psychotherapy Research, Chicago,
June 2000, and American Psychoanalytic Association, New York,
The authors thank Becky D. Baumann, Kelley L. Callahan, Erin
M. Eudell, Tracy L. Heindselman, Daniel J. Holdwick, Jr., Mollie
K. Mount, Jennifer L. Price, Candy L. Smith, and Steven R. Smith
for their participation and efforts on the Psychodynamic Psycho-
therapy Treatment Team.
THE JOURNAL OF
VOL. 191, NO. 6
THE JOURNAL OF NERVOUS AND MENTAL DISEASE
Copyright © 2003 by Lippincott Williams & Wilkins
Printed in U.S.A.
short-term psychodynamic psychotherapy (STPP)
for the treatment of depression. Posttreatment
changes from initial assessment across different do-
mains of functioning (i.e., depression, global dis-
tress, and interpersonal, social, and occupational
functioning) were expected to be moderate (d ? .5)
to large (d ?.8) in effect (Cohen, 1977). In addition,
this study is the first to empirically investigate the
two new DSM-IV (American Psychiatric Association,
1994) experimental scales on Axis V for relational
functioning (Global Assessment of Relational Func-
tioning [GARF]) and social and occupational function-
ing (Social and Occupational Functioning Assessment
Scale [SOFAS]) in regard to treatment outcome. This
study seeks to extend previous research by examining
changes in depressive symptoms at the individual pa-
tient level utilizing clinical significance methodology
(Jacobson and Truax, 1991; Jacobson et al., 1999).
Also, this study will examine the relationship between
therapist techniques with subsequent improvements in
depression from alternative models of treatment (i.e.,
cognitive-behavioral) in addition to psychodynamic in-
terventions. Finally, we will examine treatment credi-
bility, fidelity, and satisfaction.
Participants were 27 patients who received a
DSM-IV Axis I diagnosis representative of a depres-
sive spectrum disorder (major depressive disorder,
depressive disorder NOS, dysthymia, or adjustment
disorder with depressed mood; American Psychiat-
ric Association, 1994) consecutively admitted for
individual psychotherapy to a psychodynamic psy-
chotherapy treatment team (PPTT) at a community
outpatient psychological clinic. Patients were ac-
cepted into treatment regardless of disorder or co-
morbidity. Four patients prematurely terminated
their treatments (15%; after sessions 4, 4, 5, and 8)
against the advice of clinic staff. For all patients
starting psychotherapy on this treatment team, re-
gardless of diagnosis, the premature termination
rate was 18%. Two patients entered treatment utiliz-
ing antidepressant medication and were excluded
from these analyses, bringing the final sample used
in the analyses to 21.
Eleven patients were men, and 10 were women.
Seven patients were single, seven were married, and
seven were divorced. The mean age for the current
sample was 34.43 years with a standard deviation
(SD) of 12.7. The range of DSM-IV axis I depressive
spectrum disorders in the patient sample included
major depressive disorder (N ? 10), depressive dis-
order NOS (N ? 4), dysthymia (N ? 5), and adjust-
ment disorder with depressed mood (N ? 2). Nine
patients were also diagnosed with a DSM-IV person-
ality disorder, and five others had (subclinical) per-
sonality disorder features or traits. Each participant
provided written informed consent to be included in
program evaluation research.
Treatment consisted of once- or twice-weekly ses-
sions of STPP. Treatment was organized, aided,
and informed (but not prescribed) by the technical
guidelines delineated in four treatment manuals
(Book, 1998; Luborsky, 1984; Strupp and Binder, 1984;
Wachtel, 1993). Additional technical material specific
to STPP management of depression (Luborsky et al.,
1995; Malan, 1979) was actively integrated into the
treatment of these patients with depressive symptoms.
Key features of the STPP model include (Blagys and
Hilsenroth, 2000) a) focus on affect and the expression
of emotion; b) identification of patterns in actions,
thoughts, feelings, experiences, and relationships
(these patterns were explored and formulated using
the core conflictual relationship theme [CCRT] format;
Luborsky and Crits-Christoph, 1997); c) emphasis on
past experiences; d) focus on interpersonal experi-
ences; e) emphasis on the therapeutic relationship/
alliance; f) exploration of wishes, dreams, or fanta-
sies; and g) exploration of attempts to avoid topics
or engage in activities that may hinder the progress
of therapy. In addition to these areas of treatment
focus, case presentations and symptoms are concep-
tualized in the context of interpersonal/intrapsychic
conflict (Luborsky and Crits-Christoph, 1997). Also,
when a termination date is set in the treatment, this
becomes a frequent area of intervention. Issues re-
lated to the termination are often linked to key
interpersonal, affective, and thought patterns prom-
inent in that patient’s treatment.
Treatment was not of a fixed duration, but it was
determined by clinician judgment, patient decision,
progress toward goals, and life changes. Treatment
goals were first explored during the assessment pe-
riod, and a formal treatment plan was reviewed with
each patient in the third psychotherapy session. This
treatment plan was subsequently reviewed in the
10th, 24th, 40th, 60th, and 80th sessions for changes,
additions, or deletions. Patients and therapists com-
pleted reassessment of patient functioning on a stan-
dard battery of outcome measures and process rat-
ings immediately after selected sessions prior to
these review points. At the end of treatment, all
patients receiving services from the PPTT com-
pleted an exit evaluation. All patients included in the
HILSENROTH et al.
analyses had attended a minimum of nine sessions
and had completed, at least, a ninth session reas-
sessment battery. Mean number of sessions at-
tended by these 21 patients was 30 sessions during
an average 7-month period. However, the median
number of sessions and length of treatment were
somewhat shorter at 21 sessions and 5 months, re-
spectively. Also, all sessions in this training clinic
were videotaped, not just the sessions of this study’s
Ten advanced graduate students (five men and
five women) enrolled in an American Psychological
Association-approved clinical psychology Ph.D. pro-
gram were trained in the use of STPP using the texts
described earlier. The study supervisor, a Ph.D. li-
censed psychologist with extensive training in STPP,
also treated one patient in this investigation and
utilized this treatment in a continuing case confer-
ence to augment therapist training. Each therapist
received a minimum of 3.5 hours of supervision per
week (i.e., 1.5 hours individually and 2 hours in a
group treatment team meeting) on the therapeutic
model, conceptualization, process, interpretation,
and clinical interventions. Individual and group su-
pervision focused heavily on the review of video-
taped case material and technical interventions.
The assessment process was designed to assess
depressive symptomatology, global distress, and in-
terpersonal, social, and occupational domains using
a semistructured clinical interview and standardized
measures. Clinicians (i.e., therapist and external
rater) and patient self-report assessed these various
domains of functioning. Patients and clinicians com-
pleted measures assessing symptomatic distress and
interpersonal, social, and occupational functioning
during pretreatment evaluation and again at post-
treatment (or when 90% of the treatment was com-
pleted). A more thorough description of the assess-
ment procedures (semistructured clinical interview
and assessment measures) and process utilized with
this sample are provided in greater detail elsewhere
(Ackerman et al., 2000; Hilsenroth, 2002; Hilsenroth
et al., 2000; Hilsenroth et al., 20035).
Checklist-90-Revised (SCL-90-R; Derogatis, 1994) is
a 90-item self-report inventory that assesses symp-
tom distress in a number of different domains and
problem areas using a Likert scale of 0 (not at all) to
4 (extremely). This measure contains specific sub-
scales of depression (DEP) and interpersonal sensi-
tivity (INT) and a summary score, the Global Sever-
ity Index (GSI). The mean DEP for a normal
population (N ? 974 nonpatients) is .36 (SD ? .44),
and test-retest reliability during a 1-week period uti-
lizing an outpatient sample was .82.
Social Adjustment Scale. The Social Adjustment
Scale (SAS; Weissman and Bothwell, 1976) is a 42-
item self-report measure that assesses social adjust-
ment in major areas of social and occupational func-
tioning. This measure contains a summary score, the
Global Adjustment Score (SASG), which is consid-
ered an overall adjustment measure of social and
DSM-IV Rating Scales. Additional details regard-
ing the reliability data of the DSM-IV scales and
aspects of related research design procedures
are reported elsewhere (Hilsenroth et al., 2000;
Hilsenroth et al., 20035). After the semistructured
clinical interview and a feedback session, each pa-
tient was rated according to DSM-IV for the pres-
ence or absence of DSM-IV Axis II psychopathology,
total number of DSM-IV major depressive episode
(MDE) symptoms (A1–A9, p. 327; American Psychi-
atric Association, 1994), and the three Axis V global
rating scales: Global Assessment of Functioning
scale (GAF; p. 32; American Psychiatric Association,
1994), GARF (p. 758; American Psychiatric Associa-
tion, 1994), and SOFAS (p. 761; American Psychiat-
ric Association, 1994). DSM-IV Axis V therapist rat-
ings (i.e., on a scale of 0 to 100) were based on the
level of functioning of patients at assessment prior
to beginning treatment. At the different treatment
review points, therapists made routine ratings of
these DSM-IV rating scales (MDE, GAF, GARF, and
SOFAS) based on the patients’ level of functioning at
that time. An independent rater scored all rating
scales (MDE, GAF, GARF, and SOFAS) used in this
study for each participant after viewing a videotape
of the clinical interview, feedback sessions, and
those sessions or treatment review representative of
when 90% of the psychotherapy had been com-
pleted. For all cases, scoring of the scales by the
second rater was completed without knowledge of
patient self-report data and the assessing clinician’s
ratings for the MDE, GAF, GARF, and SOFAS.
The Spearman-Brown correction for a one-way
random effects model intraclass correlation coeffi-
5Hilsenroth M, Baity M, Mooney M, Meyer G (2003) DSM-IV
major depressive episode criteria: An evaluation of reliability and
validity across three different rating methods. Submitted.
STPP FOR DEPRESSION
cient (ICC [1,2]; Shrout and Fleiss, 1979) was calcu-
lated to examine the reliability of the mean score for
each DSM-IV Axis V scale. These Spearman-Brown
corrected interrater reliability scores (ICC [1,2], N ?
21) for the assessment and end of treatment MDE,
GAF, GARF, and SOFAS were in the “excellent”
range (Shrout and Fleiss, 1979) at ?.74. In all anal-
yses of the MDE, GAF, GARF, and SOFAS, these
Spearman-Brown corrected interrater reliability
scores (ICC [1,2]) representing the mean of the cli-
nician and independent rater were utilized.
Treatment Fidelity: Comparative Psychotherapy
Process Scale. A more thorough description of the
development, procedures, reliability, and validity of
the Comparative Psychotherapy Process Scale
(CPPS) are reported elsewhere (Blagys et al., 20036).
The CPPS is a measure of psychotherapy process
designed to assess therapist activity, process vari-
ables, and psychotherapy techniques used and oc-
curring during the therapeutic hour. While the CPPS
is intended to primarily be a descriptive measure
(i.e., what is being done) rather than an evaluative
measure (i.e., how well it is being done), as detailed
in the scoring manual of this measure, higher scores
on the CPPS may reflect a greater competence in the
technique or intervention being employed. Devel-
oped from an extensive empirical review of the com-
parative psychotherapy process literature (Blagys
and Hilsenroth, 2000, 2002), the scale consists of 20
items to be rated on a 7-point Likert Scale, ranging
from 0 (“not at all characteristic”), 2 (“somewhat
characteristic”), 4 (“characteristic”), to 6 (“extreme-
ly characteristic”). The patient, therapist, or an ex-
ternal rater may complete the CPPS. One unique
feature of the items on the CPPS is that they were
derived from empirical studies comparing and con-
trasting psychodynamic-interpersonal– and cogni-
tive-behavioral–oriented approaches to treatment.
This measure consists of two subscales: a psychody-
namic-interpersonal subscale (PI; 10 items) and a
cognitive-behavioral subscale (CB; 10 items). The PI
subscale measures the seven domains of therapist
activity previously described as key features of the
STPP treatment model (Blagys and Hilsenroth,
2000). The CB subscale consists of items that are
significantly more characteristic of cognitive-behav-
iorally oriented therapy (Blagys and Hilsenroth,
2002). Items include a) emphasis on cognitive or
logical/illogical thought patterns and belief systems;
b) emphasis on teaching skills to patients; c) assign-
ing homework to patients; d) providing information
regarding treatment, disorder, or symptoms; e) di-
rection of session activity; and f) emphasis on future
Videotapes of sessions 3, 9, 15, 21, 27, 36, and 57
(when available) for each patient were arranged in
random order, and entire sessions were watched
and rated by the two judges independently. Immedi-
ately after viewing a videotaped session, judges in-
dependently completed the CPPS. Also, each sub-
scale (PI and CB) was coded in random order.
Regular reliability meetings were held during the
coding process to prevent rater drift. The interrater
reliabilities of the CPPS-PI and CPPS-CB subscales
were evaluated using one-way random effects model
ICC (Shrout and Fleiss, 1979) for 80 psychotherapy
sessions that were rated by both judges. Interrater
reliability scores [ICC (1)] for these 80 sessions
were in the “excellent range” (Shrout and Fleiss,
1979; ?.75) for the mean CPPS-PI and CPPS-CB
scores (both .82). Coefficient alphas from a larger
sample of (42 patients) 124 psychodynamic, cogni-
tive-behavioral, and eclectic psychotherapy sessions
were found to be high for both subscales: CPPS-PI ?
.92 and CPPS-CB ? .94.6
Patients answered two questions regarding their
confidence in the treatment they were to receive
after a socialization interview (Luborsky, 1984) be-
fore starting treatment and again at the end of the
third session after reviewing the formal treatment
plan. Patients rated their confidence in the treat-
ment on a 7-point Likert scale ranging from 1 (never)
to 7 (always). The mean score for “I feel that the
things I do in therapy will help me to accomplish the
changes I want,” was 5.3 (SD ? 1.4). The mean score
for “How confident do you feel that through your
own efforts and those of your therapist you will gain
relief from your problems,” was 5.7 (SD ? 1.3).
Patients rated their confidence in their treatment
again at the end of the third session with the mean
scores for questions 1 and 2 of 5.4 (SD ? 1.2) and 5.7
(SD ? 1.2), respectively. These results indicate, with
a score of 5 being labeled as “often” and a score of
6 labeled as “very often,” that patients were confi-
dent that this treatment would be helpful.
Ratings of therapist activity were made on the
CPPS-PI and CPPS-CB subscales for 78 of the ses-
6Blagys M, Ackerman S, Bonge D, Hilsenroth M (2003) Mea-
suring psychodynamic-interpersonal and cognitive-behavioral
therapist activity: Development of the comparative psychother-
apy process scale. Submitted.
HILSENROTH et al.
sions in this study. The mean CPPS-PI score across
these sessions was 3.56 (SD ? .79), while the mean
CPPS-CB score across the sessions was 1.21 (SD ?
.27). This difference in the two models of therapeu-
tic focus and activity was found to be significant (t ?
11.54, p ? .0001) and demonstrated a very large
effect (d ? 3.98).
Evaluation of Treatment Changes
Paired t-tests (two-tailed, p ? .05) were used to
examine all pre- and post-treatment changes. The
results are given in Table 1 for the eight outcome
scales organized in the conceptual categories of de-
pression symptoms, global distress, and interper-
sonal and social/occupational functioning. Results
revealed statistically significant change in all four of
the conceptual outcome categories. Treatment
change in depressive symptoms, as assessed by cli-
nician ratings and patient self-report, was shown to
significantly decrease in this group of treated pa-
tients (p ? .0001), and these changes were consid-
ered to be very robust in effect (d ? 1.0). Both
measures of global symptomatic distress, GAF (ther-
apist and external rater) and GSI (patient self-re-
port), were shown to significantly decrease during
the course of treatment (p ? .0001), and these
changes were also considered to be very large in
effect (d ? 1.0). Likewise, measures of interpersonal
distress, GARF (therapist and external rater) and
INT (patient self-report), were shown to signifi-
cantly decrease during the course of treatment (p ?
.001), and these changes were also considered to be
large in effect (d ? .80). The two social and occu-
pational functioning scales, SOFAS (therapist and
external rater) and SASG (patient self-report), also
showed significant changes (p ? .001), with clini-
cian rating and patient self-report demonstrating
large effects (d ? 1.30 and .85, respectively).
Clinically Significant Change in Depression
Both measures of depression were examined at
the individual patient level for clinical significance.
Prior to the calculation of clinical significance infor-
Comparison of pre- to post-treatment changes for the eight outcome scales
MDE (N ? 21)b
DEP (N ? 20)c
M S.D.M S.D.t
GAF (N ? 21)d
GSI (N ? 20)e
Global Symptom Distress
M S.D.M S.D.t
GARF (N ? 21)f
INT (N ? 20)g
SOFAS (N ? 21)h
SASG (N ? 18)i
aCohen’s d, utilizing pooled standard deviations from pre and post treatment (Cohen, 1977).
bDSM-IV Major Depressive Episode Symptoms (0–9).
cDepression Subscale of the Symptom Checklist-90-Revised.
dGlobal Assessment of Functioning scale.
eGlobal Severity Index of the Symptom Checklist-90-Revised.
fGlobal Assessment of Relational Functioning scale.
gInterpersonal Sensitivity Subscale of the Symptom Checklist-90-Revised.
hSocial and Occupational Functioning Assessment Scale.
iGlobal Adjustment Score of the Social Adjustment Scale.
STPP FOR DEPRESSION
mation, to address concerns of pretreatment score
regression to the mean, each of the pretreatment
scores utilized in this stage of data analysis were
adjusted according to standard psychometric proce-
dures (Speer, 1992). In this formula, evaluation
scores were “true score adjusted” to attenuate any
regression effects. Reliable Change Index (RCI;
Jacobson and Truax, 1991) scores then were calcu-
lated for each variable (using the adjusted pretest
scores). An RCI score exceeding 1.96 suggests that
the test score change was psychometrically reliable,
reflected real change, and was not the product of
random error (p ? .05, two tailed). Each posttreat-
ment test score then was examined to determine
whether it fell below the cutoff score for a func-
tional distribution, within 2 SD of the normative
mean. Patients who met both of these criteria (i.e.,
reliable change and moved within 2 SD of the nor-
mative mean) were considered to have achieved
clinically significant change. RCIs were examined to
determine whether any patients reliably deterio-
rated during treatment.
All calculations of clinical significance for the DEP
utilized the normative mean, standard deviation, and
test-retest reliability data reported in the method. Cal-
culation of RCI for the MDE required a slight modifi-
cation in the computation of the standard error of
difference (Sdiff) used in the denominator of this
formula. A number of authors (Jacobson et al., 1999;
Kadera et al., 1996; Tingey et al., 1996) have recom-
mended alternative reliability estimates when calcu-
lating RCI for measures that do not have test-retest
reliability estimates available for nonclinical popu-
lations. Since we were unaware of any test-retest
reliability data from nonclinical subjects for the
DSM-IV MDE symptoms, we instead utilized a
pooled mean interrater reliability coefficient (ICC
[1,2], .86 and .97, respectively) from the evaluation
and final session rating in the computation of Sdiff.
Similar to the modification of the RCI, there is
currently a lack of normative mean and standard
deviation data available for the DSM-IV MDE symp-
toms necessary to establish a functional distribution
criteria point. Regarding this selection of a func-
tional distribution criteria, Jacobson et al. (1999)
have recently noted that criterion calculation may
be irrelevant for any clinical problem in which ex-
ceeding a predetermined cutoff point on a scale
automatically guarantees change to a normal band
of functioning. When applying this rationale specif-
ically to the MDE symptoms, each of the depressive
spectrum disorders utilized in this study does have
criteria points at which functioning is considered
“nonclinical,” all of which would require the pres-
ence of less than two MDE criteria (0–1). Table 2
reports the frequency and percentage of patient re-
liable change, movement into a functional distribu-
tion, clinical significance, and deterioration through
the course of psychotherapy for each depression
Approximately three fourths (71% or greater) of
the patients who completed treatment showed reli-
able change and movement into a functional distri-
bution in either clinician-rated or self-reported lev-
els of depressive symptomatology. A very high level
of clinical significance was also shown for clinician-
rated depressive symptoms (71%), while almost two
thirds (65%) of patients achieved clinically signifi-
cant change based on self-report. None of the pa-
tients showed any deterioration during psychother-
apy in either of the assessment modalities. When we
apply a more conservative standard to calculate clin-
ical significance rates, as recently recommended by
Westen and Morrison (2001), by adding the four
prematurely terminating patients to the total num-
ber of patients in the denominator (i.e., intent-to-
treat sample rather than only treatment completers),
we still obtain a high level of clinically significant
change on clinician- (60%) and patient- (54%) rated
Treatment Process and Outcome Relationship
The next analyses in this study examined the re-
lationship between therapist activity and technique
with changes in depressive symptomatology. Clini-
cian-rated and patient self-report pretreatment
scores of depression symptoms were adjusted for
regression to the mean prior to these analyses as
part of the previously described RCI methodology. A
mean score was tabulated across all viewed sessions
for each treatment case on the CPPS-PI and
CPPS-CB subscales, and these average amounts of
therapist technique were examined in relation to
Clinically significant change in depression symptoms
CriterionMDE (N ? 21)a
RCI ? 1.96c
DEP (N ? 20)b
aDSM-IV Major Depressive Episode Symptoms (0–9).
bDepression Subscale of the Symptom Checklist-90-Revised.
cNumber of individuals who reliably improved after adjusting
pretest scores for regression to the mean.
dNumber of individuals who fell within 2 standard deviations of
the general population mean.
eNumber of individuals who reliably improved and fell within 2
standard deviations of the general population mean.
fNumber of individuals who reliably deteriorated during treat-
HILSENROTH et al.
that patient’s reliable degree of change in depressive
symptomatology (MDE-RCI and DEP-RCI).
As reported in Table 3, results demonstrated that
higher mean levels of PI techniques across the treat-
ment were significant and positively related to
amount of reliable change in clinician-rated and pa-
tient self-reported changes in depression symptoms
(r ? .57, p ? .006 and r ? .49, p ? .03, respectively).
These findings also revealed that mean levels of CB
interventions across the course of treatment were
nonsignificant and negatively related to amount of
reliable change in clinician-rated and patient self-
reported changes in depression symptoms (both r ?
?.33, p ? .15).
To better understand the specific aspects of PI
technique that were most related to subsequent
changes in depression symptoms, we undertook two
post hoc, exploratory analyses to examine this issue.
The results of the first stepwise regression analysis
revealed that the CPPS-PI item “The therapist en-
courages the patient to experience and express feel-
ings in the session” to be significantly related to the
criterion variable MDE-RCI (R ? .62, R2 ? .39, F ?
12.09, p ? .003). The results of a second stepwise
regression analysis revealed that the CPPS-PI item
“The therapist addresses the patient’s avoidance of
important topics and shifts in mood” to be signifi-
cantly related to the criterion variable DEP-RCI
(R ? .51, R2 ? .26, F ? 6.35, p ? .02).
At completion of treatment, each patient was
asked to rate his or her level of satisfaction with the
psychotherapy on a ?4 to ?4 Likert scale for three
questions: a) “How unhelpful or helpful has therapy
been for you?” b) “Overall, how satisfied or dissat-
isfied have you been with therapy?” and c) “In gen-
eral, how productive do you feel the sessions have
been with your therapist?” The mean scores on
these treatment satisfaction questions were very
positive (3.5, 3.6, and 3.6), with small standard devi-
ations (.68, .59, and .51, respectively), and most pa-
tients responded with one of the two highest possi-
ble ratings, either a 3 or 4.
This is one of the first studies to examine treat-
ment credibility, fidelity, and satisfaction within a
naturalistic/effectiveness model of STPP for depres-
sion, all of which were found to be high. Also, when
evaluating psychotherapy outcomes, treatment attri-
tion should be considered. In this study, the treat-
ment termination rate for these depressed patients
(15%) is low in relation to general practice (Garfield,
1994; Olfson and Pincus, 1994; Owen and Kohutek,
1981) and in outcome research (DeRubeis et al.,
1999; Elkin et al., 1989; Westen and Morrison, 2001).
Changes in the four domains of depressive symp-
toms, global symptomatic distress, and interper-
sonal and social/occupational functioning showed
substantial improvements and large statistical ef-
fects (p ? .001 and d ? .80). In addition, the findings
of this study support the clinical utility of the two
DSM-IV Axis V experimental scales, GARF and
SOFAS, as outcome variables in treatment studies.
Almost all (86% and 80%) of those patients who
completed treatment demonstrated reliable change
or scored within a functional distribution in regard
to depression symptoms. Likewise, the percentage
of patients exhibiting clinically significant change in
depression (71% and 65%) through the course of
psychotherapy was high. This was still the case
when a more restrictive definition of clinical signif-
icance was calculated (60% and 54%) using a conser-
vative estimate (i.e., intent-to-treat sample rather
than only treatment completers) that is uncommon
in contemporary psychotherapy outcome research
(Westen and Morrison, 2001). The statistical and
clinical significance results of this study were con-
sistent with prior research on the efficacy and effec-
tiveness of psychodynamic psychotherapy for the
management of depression (Anderson and Lambert,
1995; Barber et al., 1996; Crits-Christoph, 1992;
Gaston et al., 1998; Lueger et al., 2000; Shapiro et al.,
1995). Indeed, these posttreatment changes and lev-
els of clinical significance compare favorably
with other modalities of treatment for depression
(DeRubeis et al., 1999; Elkin et al., 1989; Ogles et al.,
1995; Westen and Morrison, 2001).
In addition, these changes in depressive symp-
toms were significantly related to therapist tech-
niques. This is the third study, from three indepen-
dent research groups, that has found a direct link
between psychodynamic interventions and subse-
Treatment techniques in relation to subsequent changes in
MDE-RCI (N ? 21)a
r ? .57, p ? .006
r ? ?.33, p ? .15
aReliable Change in number of Major Depressive Episode Symp-
toms (0–9) after adjusting pretest scores for regression to the mean.
bReliable Change in the SCL-90-R Depression Subscale after ad-
justing pretest scores for regression to the mean.
cMean Comparative Psychotherapy Process Scale: Psychodynam-
ic-Interpersonal Process Subscale across psychotherapy sessions.
dMean Comparative Psychotherapy Process Scale: Cognitive-Be-
havioral Process Subscale across psychotherapy sessions.
DEP-RCI (N ? 20)b
r ? .49, p ? .03
r ? ?.33, p ? .15
STPP FOR DEPRESSION
quent changes in depressive symptoms (Barber et
al., 1996; Gaston et al., 1998). Further, it appears in
this study that specific therapist techniques directed
toward achieving and maintaining session focus on
the exploration and expression of affect were most
related to positive changes in depressive symptom-
atology. These interventions provide important in-
formation concerning applied clinical practice and
are consistent with a psychodynamic model of
change (Blagys and Hilsenroth, 2000; Book, 1998;
Fosha, 2002; Luborsky, 1984; Luborsky and Crits-
Christoph, 1997; Luborsky et al., 1990, 1995; Malan,
1979; Strupp and Binder, 1984; Wachtel, 1993),
whereby a supportive environment and relationship
are developed with the therapist that may allow the
patient to better tolerate the expression and explo-
ration of painful affect. When this painful affect is
engaged or avoided, then interventions are focused
“in the moment” (including issues related to the
therapeutic relationship) for further expression and
Assessment of CB interventions was conducted to
evaluate one potential competing hypothesis for the
basis of patient change. It is important to note that
the negative, nonsignificant relationship between
CB interventions and change in depressive symp-
toms that was observed needs to be understood
within the context of this specific study. These find-
ings indicate that within a study of STPP for outpa-
tient depression, the very limited amount of CB
interventions utilized within this larger psycho-
dynamic treatment did not contribute to outcome.
This finding should not be generalized beyond this
limited context. The positive relationship between
CB interventions and outcome has been demon-
strated in previous research (DeRubeis and Feeley,
1990; Feeley et al., 1999; Tang and DeRubeis, 1999).
However, we thought it was important to assess the
impact of these alternative treatment interventions
within psychodynamic psychotherapy. We believed
this was an important methodological issue to ad-
dress because prior research has shown a few tech-
niques historically understood to be psychodynamic
in nature (either interventions distinctive to or em-
phasized significantly more in PI therapy than CB
treatments; Blagys and Hilsenroth, 2000) employed
within a CB treatment have been significantly re-
lated to patient improvements (Ablon and Jones,
1998; Castonguay et al., 1996; Gaston et al., 1998;
Hayes and Strauss, 1998; Jones and Pulos, 1993).
One limitation of this study was that the patient
sample primarily suffered from mild to moderate
levels of distress and impairments in functioning.
Further research is necessary using inpatient sam-
ples exhibiting severe levels of distress and func-
tional impairment to extend the implications of the
present findings. In addition, the small sample size
and open-ended psychodynamic treatment provided
by advanced graduate trainees will necessitate fu-
ture research to ascertain whether these changes
are generalizable to other treatment settings, with
therapists possessing greater levels of experience or
varying treatment modalities. Finally, the lack of an
experimental design does not allow us to conclu-
sively rule out the potential impact of common fac-
tors on our observed treatment-related effects.
These limitations notwithstanding, this treatment
study is one of the first to integrate the assessment,
technique, and training aspects of an efficacy model
within a naturalistic setting (Seligman, 1996). The
incorporation of these efficacy features in this oth-
erwise naturalistic treatment delivery setting pro-
vides important information regarding the nature of
the treatment that is not often evaluated in general
psychotherapy effectiveness studies. As such, this
study represents a more naturalistic examination of
STPP for depression as delivered in an outpatient
community clinic. The present results demonstrate
that robust statistical and clinically significant
improvement can occur in STPP for depression. In
addition, a significant positive relationship between
psychodynamic techniques and subsequent changes in
depressive symptomatology was observed in a manner
consistent with a psychodynamic model of change
(Blagys and Hilsenroth, 2000; Book, 1998; Fosha, 2002;
Luborsky, 1984; Luborsky and Crits-Christoph, 1997;
Luborsky et al., 1990, 1995; Malan, 1979; Strupp and
Binder, 1984; Wachtel, 1993).
Ablon J, Jones E (1998) How expert clinician’s prototypes of an
ideal treatment correlate with outcome in psychodynamic and
cognitive-behavior therapy. Psychother Res 8:71–83.
Ackerman S, Hilsenroth M, Baity M, Blagys M (2000) Interaction
of therapeutic process and alliance during psychological as-
sessment. J Pers Assess 75:82–109.
American Psychiatric Association (1994) Diagnostic and statis-
tical manual of mental disorders (4th ed). Washington, DC:
American Psychiatric Association.
Anderson E, Lambert M (1995) Short-term dynamically oriented
psychotherapy: A review and meta-analysis. Clin Psychol Rev
Barber J, Crits-Christoph P, Luborsky L (1996) Effects of thera-
pist adherence and competence on patient outcome in brief
dynamic therapy. J Consult Clin Psychol 64:619–622.
Blagys M, Hilsenroth M (2000) Distinctive features of short-term
psychodynamic-interpersonal psychotherapy: A review of the
comparative psychotherapy process literature. Clin Psychol
Sci Pract 7:167–188.
Blagys M, Hilsenroth M (2002) Distinctive features of short-term
cognitive-behavioral psychotherapy: An empirical review of
the comparative psychotherapy process literature. Clin Psy-
chol Rev 22:671–706.
Book H (1998) How to practice brief psychodynamic psycho-
therapy: The core conflictual relationship theme method.
Washington, DC: American Psychological Association.
HILSENROTH et al.
Castonguay L, Goldfried M, Wiser S, Raue P, Hayes A (1996) Download full-text
Predicting the effect of cognitive therapy for depression: A
study of unique and common factors. J Consult Clin Psychol
Cohen J (1977) Statistical power analysis for the behavioral
sciences (2nd ed). New York: Academic Press.
Crits-Christoph P (1992) The efficacy of brief dynamic psycho-
therapy: A meta-analysis. Am J Psychiatry 149:151–158.
Derogatis L (1994) Symptom checklist-90-revised: Administra-
tion, scoring, and procedures manual (3rd ed). Minneapolis,
MN: National Computer Systems.
DeRubeis R, Feeley M (1990) Determinants of change in cogni-
tive therapy for depression. Cogn Ther Res 14:469–482.
DeRubeis R, Gelfand L, Tang T, Simons A (1999) Medications
versus cognitive behavior therapy for severely depressed out-
patients: Mega-analysis of four randomized comparisons. Am J
Elkin I, Shea T, Watkins J, Imber S, Sotsky S, Collins J, Glass D,
Pilkonis P, Leber W, Docherty J, Fiester S, Parloff M (1989)
National Institute of Mental Health Treatment of Depression
Collaborative Research Program: General effectiveness of
treatments. Arch Rev Psychiatry 46:971–982.
Feeley M, DeRubeis R, Gelfand L (1999) The temporal relation of
adherence and alliance to symptom change in cognitive ther-
apy for depression. J Consult Clin Psychol 67:578–582.
Fosha D (2002) The activation of affective change processes in
AEDP (accelerated experiential-dynamic psychotherapy). In J
Magnavita (Ed), Comprehensive handbook of psychotherapy.
Psychodynamic and object relations psychotherapies (Vol 1).
New York: John Wiley & Sons.
Garfield S (1994) Research on client variables in psychotherapy.
In SL Garfield, AE Bergin (Eds), Handbook of psychotherapy
and behavior change. New York: Wiley.
Gaston L, Thompson L, Gallagher D, Cournoyer L, Gagnon R
(1998) Alliance, technique, and their interactions in predicting
outcome of behavioral, cognitive, and brief dynamic therapy.
Psychother Res 8:190–209.
Hayes A, Strauss J (1998) Dynamic systems theory as a paradigm
for the study of cognitive change in psychotherapy: An appli-
cation of cognitive therapy for depression. J Consult Clin
Hilsenroth M (2002) Adelphi University: Psychodynamic Psycho-
therapy Process and Outcome Research Team. In P Fonagy, J
Clarkin, A Gerber, H Kachele, R Krause, E Jones, R Perron, E
Allison (Eds), An open door review of outcome studies in
psychoanalysis (2nd ed, pp 241–247). London: International
Hilsenroth M, Ackerman S, Blagys M, Baumann B, Baity M, Smith
S, Price J, Smith C, Heindselman T, Mount M, Holdwick D
(2000) Reliability and validity of DSM-IV axis V. Am J Psychi-
Jacobson N, Roberts L, Berns S, McGlinchey J (1999) Methods
for defining and determining the clinical significance of treat-
ment effects: Description, application, and alternatives. J Con-
sult Clin Psychol 67:300–307.
Jacobson N, Traux P (1991) Clinical significance: A statistical
approach to defining meaningful change in psychotherapy re-
search. J Consult Clin Psychol 59:12–19.
Jones E, Pulos S (1993) Comparing the process in psycho-
dynamic and cognitive-behavioral therapies. J Consult Clin
Kadera S, Lambert M, Andrews A (1996) How much therapy is
really enough? A session-by-session analysis of the psycho-
therapy dose-effect relationship. J Psychother Pract Res
Luborsky L (1984) Principles of psychoanalytic psychotherapy:
Manual for supportive/expressive treatment. New York: Basic
Luborsky L, Barber J, Crits-Christoph P (1990) Theory-based
research for understanding the process of dynamic psycho-
therapy. J Consult Clin Psychol 58:281–287.
Luborsky L, Crits-Christoph P (1997) Understanding transfer-
ence: The core conflictual relational theme method (2nd ed).
Washington, DC: APA.
Luborsky L, Mark D, Hole A, Popp C, Goldsmith B, Cacciola J
(1995) Supportive-expressive dynamic psychotherapy of de-
pression: A time-limited version. In J Barber, P Crits-Christoph
(Eds), Dynamic therapies for the psychiatric disorders (axis
I). New York: Basic.
Lueger R, Lutz W, Howard K (2000) The predicted and observed
course of psychotherapy for anxiety and mood disorders.
J Nerv Ment Dis 188:127–134.
Malan D (1979) Individual psychotherapy and the science of
psychodynamics. London: Butterworths.
Ogles B, Lambert M, Sawyer J (1995) Clinical significance of the
National Institute of Mental Health Treatment of Depression
Collaborative Research Program data. J Consult Clin Psychol
Olfson M, Pincus H (1994) Outpatient psychotherapy in the
United States: 2. Patterns of utilization. Am J Psychiatry
Owen P, Kohutek K (1981) The rural mental health dropout. J
Rural Com Psychol 2:38–41.
Seligman M (1996) Science as an ally of practice. Am Psychol
Shapiro D, Rees A, Barkham M, Hardy G, Reynolds S, Startup M
(1995) Effects of treatment duration and severity of depres-
sion on the maintenance of gains after cognitive-behavioral
and psychodynamic-interpersonal psychotherapy. J Consult
Clin Psychol 63:378–387.
Shrout P, Fleiss J (1979) Intraclass correlations: Uses in assess-
ing rater reliability. Psychol Bull 86:420–428.
Speer D (1992) Clinically significant change: Jacobson and Truax
(1991) revisited. J Consult Clin Psychol 60:402–408.
Strupp H, Binder J (1984) Psychotherapy in a new key. New
York: Basic Books.
Tang T, DeRubeis R (1999) Sudden gains and critical session in
cognitive-behavioral therapy for depression. J Consult Clin
Tingey R, Lambert M, Burlingame G, Hansen N (1996) Assessing
clinical significance: Proposed extensions to method. Psycho-
ther Res 6:109–123.
Wachtel P (1993) Therapeutic communication: Principles and
effective practice. New York: Guilford.
Weissman M, Bothwell S (1979) Assessment of social adjustment
by patient self-report. Arch Rev Psychiatry 33:1111–1115.
Westen D, Morrison K (2001) A multidimensional meta-analysis
of treatments for depression, panic, and generalized anxiety
disorder: An empirical examination of the status of empirically
supported therapies. J Consult Clin Psychol 69:875–899.
STPP FOR DEPRESSION