ArticlePDF AvailableLiterature Review

The Equivalence of Psychodynamic Therapy and Cognitive Behavioural Therapy for Depressive Disorders in Adults: A Meta-Analytic Review.

Authors:

Abstract and Figures

Background Meta‐analyses on the relative efficacy of psychodynamic psychotherapy (PDT) and cognitive behavioral therapy (CBT) for depressive disorders are limited by heterogeneity in diagnostic samples and comparators and a lack of equivalence testing. Objective We addressed this through a meta‐analytic test of the equivalence of manualized PDT and CBT in treating adults with depressive disorders as determined by diagnostic interviews. Sensitivity analyses evaluated the impact of pretreatment differences, mixed diagnostic samples, author allegiance, study quality, year of publication and outliers on findings. Method A comprehensive literature search across multiple databases using reliable screening methods identified nine randomized controlled trials directly comparing manualized PDT and CBT for diagnosed depressive disorders in adults. Following pre‐registration, we employed random effect models for our meta‐analyses and two one‐sided test procedures for equivalence testing. Results Independent raters determined that all studies were of adequate quality. Immediately posttreatment, depressive symptoms were statistically equivalent across PDT and CBT ( k = 9; g = −0.11, 90% confidence interval [90% CI]: −0.24 to 0.02, p equivalence = .048, p NHST = .212, I ² = 32.7). At follow‐up, the longest time point within a year, depressive symptoms were neither statistically equivalent nor statistically different ( k = 6; g = −0.16, 90% CI: −0.31 to −0.02, p equivalence = .184, p NHST = .126, I ² = 0.00). Conclusion The efficacy of manualized PDT is equal to manualized CBT immediately at posttreatment for depressive disorders in the adult general population. Nevertheless, insufficient data exists to reach a conclusion regarding equivalence at follow‐up.
Content may be subject to copyright.
Received: 2 October 2023
|
Accepted: 14 January 2024
DOI: 10.1002/jclp.23649
REVIEW ARTICLE
The equivalence of psychodynamic therapy and
cognitive behavioral therapy for depressive
disorders in adults: A metaanalytic review
Martin M. Smith |Paul L. Hewitt
Department of Psychology, University of
British Columbia, Vancouver, BC, Canada
Correspondence
Martin M. Smith, Department of Psychology,
University of British Columbia, Vancouver,
BC, Canada.
Email: martin.smith@psych.ubc.ca
Abstract
Background: Metaanalyses on the relative efficacy of
psychodynamic psychotherapy (PDT) and cognitive behav-
ioral therapy (CBT) for depressive disorders are limited by
heterogeneity in diagnostic samples and comparators and a
lack of equivalence testing.
Objective: We addressed this through a metaanalytic test of
the equivalence of manualized PDT and CBT in treating adults
with depressive disorders as determined by diagnostic inter-
views. Sensitivity analyses evaluated the impact of pretreat-
ment differences, mixed diagnostic samples, author allegiance,
study quality, year of publication and outliers on findings.
Method: A comprehensive literature search across multiple
databases using reliable screening methods identified nine
randomized controlled trials directly comparing manualized
PDT and CBT for diagnosed depressive disorders in adults.
Following preregistration, we employed random effect
models for our metaanalyses and two onesided test
procedures for equivalence testing.
Results: Independent raters determined that all studies
were of adequate quality. Immediately posttreatment,
depressive symptoms were statistically equivalent across
PDT and CBT (k=9; g=0.11, 90% confidence interval
[90% CI]: 0.24 to 0.02, p
equivalence
= .048, p
NHST
= .212,
J Clin Psychol. 2024;123. wileyonlinelibrary.com/journal/jclp
|
1
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited.
© 2024 The Authors. Journal of Clinical Psychology published by Wiley Periodicals LLC.
I
2
= 32.7). At followup, the longest time point within a year,
depressive symptoms were neither statistically equivalent
nor statistically different (k=6;g=0.16, 90% CI: 0.31 to
0.02, p
equivalence
= .184, p
NHST
= .126, I
2
= 0.00).
Conclusion: The efficacy of manualized PDT is equal to
manualized CBT immediately at posttreatment for depres-
sive disorders in the adult general population. Never-
theless, insufficient data exists to reach a conclusion
regarding equivalence at followup.
KEYWORDS
cognitive behavioral, depression, metaanalysis, psychodynamic,
psychotherapy
1|INTRODUCTION
Depression is a widespread and costly mental health problem involving symptoms such as feelings of worthlessness,
somatic disturbances, suicide ideation, hopelessness, anhedonia, sadness, and guilt (American Psychiatric Associa-
tion, 2022;Greenbergetal.,2021;Hasinetal.,2018). Roughly 21.1% of adults in the United States will experience a
major depressive episode at some point and ~8.4% will experience a major depressive episode in a given year (Hasin
et al., 2018). This is troubling, as major depressive disorder (MDD) can severely affect an individual's employment, well
being, workplace productivity, cognitive functioning, and quality of interpersonal relationships (Clark et al., 2016;Fabbri
et al., 2023;Zlotnicketal.,2000). Likewise, MDD carries a substantial economic cost, estimated to result in $US326
billion in lost income each year in the United States (Greenberg et al., 2021). However, fortunately, several effective
treatments for MDD exist, with one being psychotherapy. For instance, metaanalytic evidence suggests that over 60%
of adults with MDD treated with psychotherapy no longer meet Diagnostic and Statistical Manual of Mental Disorders
Fourth edition (DSMIV) criteria after 12 months (Cuijpers, Karyotaki, et al., 2014). Additionally, although most of this
research is on cognitivebehavioral therapy (CBT), other psychotherapies, such as psychodynamic psychotherapy (PDT),
are also effective (Cuijpers, Karyotaki, et al., 2020).
Indeed, the American Psychological Association's Division 12 maintains that empirical support requires evidence from
randomized controlled trials (RCTs), demonstrating the effectiveness of a manualized psychotherapy for a specific mental
health problem, as compared with either a control group receiving littletono treatment or an already established
treatment (Chambless & Hollon, 1998; Tolin et al., 2015). Requiring a manual ensures that the psychotherapy is
administered in a standardized way across studies, making it possible to compare the results of different trials and
determine the overall efficacy of the therapy. RCTs provide the most powerful means of establishing a causeandeffect
relationship between an intervention and outcome as they reduce the impact of extraneous variables and guard against
confounds (i.e., factors related to both the treatment and outcomethatwerenotcontrolled).With Division 12's guidelines
in mind, there is sufficient evidence to categorize PDT as an empirically supported treatment for depression
(Fonagy, 2015). Namely, several RCTs that used PDT treatment manuals and focused on depression have demonstrated
that, on average, adults treated with PDT experienced greater reductions in depressive symptoms than adults who
received littletono treatment (for reviews, see Driessen et al., 2010,2015; Leichsenring et al., 2023).
Yet, although PDT is an empirically supported treatment for depression, the tendency for metaanalyses to
combine different comparators (e.g., CBT, counseling, bibliotherapy), outcomes (e.g., symptoms of depression, anxiety,
and eating disorders), and/or samples (e.g., adults with depressive disorders, anxiety disorders, substanceuse
2
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
disorders) make it difficult to determine the efficacy of PDT relative to CBT for depressive disorders. Furthermore, a
broader issue with metaanalyses is that trivial differences with no substantive implications can be statistically
significant with a large enough sample size. However, a solution employed widely in medicine is to metaanalytically
test for the absence of meaningful differences between treatments via equivalence testing (Treadwell et al., 2012).
Even so, a metaanalytic test of the equivalence of PDT and CBT for depressive disorders is absent from the literature,
and hence, whether PDT and CBT are equally effective treatments for depressive disorders is unclear. In the present
study, we addressed this by conducting a comprehensive metaanalytic test of the equivalence of manualized PDT
and CBT for depression severity at posttreatment and followup in diagnostically homogenous RCTs of adults with
diagnosed depressive disorders.
1.1 |Metaanalytic research on the relative efficacy of PDT and CBT for depression
Based on 385 comparisons between psychotherapies and minimaltonotreatment controls, Cuijpers, Noma, et al.
(2020) concluded that each of the 15 psychotherapies evaluated, including PDT and CBT, have comparable efficacy
for posttreatment depressive symptoms. Nonetheless, their findings were derived from indirect comparisons of
RCTs comparing psychotherapies to minimaltonotreatment controls (i.e., weak comparison RCTs) and RCTs
directly comparing psychotherapies (i.e., active comparison RCTs) provide a more stringent and appropriate efficacy
test (see Cristea, 2019; Cuijpers & Cristea, 2016), because they control for the therapeutic factors shared across
treatments that typically explain most of the variance (Cuijpers et al., 2012). Moreover, patients' background
variables generally are more similar in treatment conditions directly compared than when indirect comparisons are
made between treatments and controls (Luborsky et al., 2002). Likewise, weak comparison RCTs essentially only
evaluate if a treatment works better than littletono treatment, whereas active comparison RCTs can test why a
treatment works. Hence, when assessing efficacy, Chambless and Hollon (1998) instructed researchers to prioritize
evidence from active comparison RCTs over evidence from weak comparison RCTs.
Relatedly, the term minimaltonotreatment control encompasses several types of controls, such as waiting
lists and notreatment. And research suggests the specific comparator used influences the magnitude of the
treatment effect found (Barth et al., 2013; Mohr et al., 2009). For instance, Furukawa et al. (2014) metaanalyzed
weak comparison RCTs on depression and found that waiting lists might be noceboconditions that perform worse
than no treatment. Accordingly, metaanalyses that treat distinct minimaltonotreatment controls as equivalent
introduce a potential source of heterogeneity that can limit generalizability. Therefore, we focus on metaanalytic
findings from active comparison RCTs of PDT and CBT for the remainder of our paper.
Steinert et al. (2017a) combined active comparison RCTs on depressive disorders, anxiety disorders, eating
disorders, personality disorders, substance abuse disorders, and posttraumatic stress disorders and tested whether
PDT was equivalent to CBT for target symptoms, general psychiatric symptoms, and psychosocial functioning.
Using an exceptionally conservative minimally important difference (MID) of g= 0.25, they found that PDT was
equivalent to CBT for posttreatment and followup target and general psychiatric symptoms. However, though
Steinert et al. (2017a) had several strengths, including controlling for researcher allegiance and conducting
equivalence testing, some viewed the clinical relevance of their results, especially as they pertained to target
symptoms, as limited (e.g., Cristea et al., 2017).
Indeed, as noted by Cristea et al. (2017), what Steinert et al. (2017a) considered target symptoms included
divergent measures of depression, substance use, suicidality, eating disorders, and body mass index, thereby
limiting the clinical relevance of their findings and potentially diluting meaningful differences. Similarly, the
diagnostically heterogeneous samples that Steinert et al. (2017a) combined make it difficult to know if their findings
generalize to a specific clinical setting (Butler et al., 2006). Lastly, some maintain, because CBT is a disorderspecific
treatment, that it only makes sense to test the equivalence of CBT relative to other psychotherapies in the context
of specific disorders (see Barlow, 2010).
SMITH and HEWITT
|
3
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Turning to narrower metaanalytic reviews, based on 10 diagnostically heterogeneous active comparison RCTs
on depression and anxiety, Tolin (2010) reported a small significant difference (g= 0.28) favoring CBT over PDT for
posttreatment depression symptoms. Similarly, based on five clinically homogenous active comparison RCTs on
depression, Cuijpers et al. (2013) found a small difference (g= 0.25) favoring CBT over PDT; however, this effect
was nonsignificant. Furthermore, Driessen et al. (2015) compared PDT with other psychotherapies(i.e., CBT,
cognitive therapy, behavior therapy, and supportive therapy) for depression and, based on 15 active comparison
RCTs, found a small significant difference favoring other psychotherapies immediately at posttreatment (g= 0.25)
and a nonsignificant difference favoring other psychotherapies(g= 0.08) at followup.
More recently, Cuijpers et al. (2023) metaanalyzed RCTs directly comparing CBT to minimaltonotreatment
controls, pharmacotherapy, and other psychotherapies. Based on seven active comparison RCTs, they reported
CBT had a small (g= 0.21) nonsignificant advantage over PDT for adults with depression as determined by
diagnostic interviews and scores above a cutoff on selfreport measures. Nonetheless, two of the seven studies
appeared to share the same data set (i.e., Dos Santos et al., 2020; Soares et al., 2018) and one was retracted (see
Soares et al., 2019). Further, the effect Cuijpers et al. (2023) included from Dos Santos et al. (2020) for
posttreatment analyses was for responders assessed 3 years after posttreatment. Moreover, relevant RCTs directly
comparing PDT to CBT for adults with diagnosed depressive disorders are absent from Cuijpers et al. (2023), such
as Gibbons et al. (2016). Finally, except for Steinert et al. (2017a), none of the aforementioned metaanalyses used
equivalence testing.
1.2 |The need for equivalence testing
1
Within the traditional frequentist null hypothesis significant testing (NHST) framework developed by Fisher
(1925) and refined by Neyman and Pearson (1933), it is mathematically impossible to show that the actual
difference between treatments, such as PDT and CBT, is zero (Lakens et al., 2018). Namely, we can never
entirely rule out the possibility that a larger sample would have detected a difference. Or, using more precise
language, NHST tests whether there is sufficient evidence to reject the null hypothesis but cannot test whether
there is enough evidence to accept the null hypothesis. Hence, using NHST to attempt to show that PDT and
CBT are equally effective treatments for depression is futile. What is possible, however, is to use equivalence
testing to determine whether the difference between PDT and CBT is small enough to be considered practically
equivalent (Linde et al., 2023).
Although there are several forms of equivalence testing, the two onesided tests (i.e., TOST) procedure is the
most widely used and validated (Lakens, 2017a). The TOST procedure essentially involves two steps. First, based on
theory, research or expert consensus, a lower and upper equivalency bound is specified based on the MID, also
called the smallest effect size of interest. Jaeschke et al. (1989) defined the MID as the smallest difference between
treatments in the domain of interest that patients perceive as beneficial that would necessitate a change in
treatment, provided there are no troublesome side effects or excessive costs. Next, after determining the size that a
specific effect would have to exceed to be clinically relevant, two null hypotheses are tested to determine whether
the difference is smaller than the lower equivalency boundary or larger than the upper equivalency boundary. But
what should we consider the MID for depression?
Chambless and Hollon (1998) and Leichsenring et al. (2015) argued that anything smaller than a moderate
standardized difference (i.e., g= 0.50) in patientrated outcomes is inconsequential. Likewise, the National Institute
of Health and Clinical Excellence (NICE, 2004) initially maintained a difference 3 scale points on the Beck
Depression Inventorysecond edition (BDIII; Beck et al., 1996) was clinically relevant. Alternatively, based on the
opinion of 10 experienced clinical depression researchers, Driessen et al. (2013) proposed g= 0.30 as the MID
between psychotherapies for depression. Finally, Cuijpers, Turner, et al. (2014) concluded that g= 0.24 was the
smallest MID in measures of depression that patients can reliably detect.
4
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Based on this literature, we prespecified Cuijpers, Turner, et al.'s (2014) MID (g= 0.24), simply because it
provided us with the most conservative equivalence test possible. Indeed, g= 0.24 is the smallest MID ever
proposed in psychology and most medical research (Steinert et al., 2017a) and is likely an underestimate. To
illustrate, Lange and Freitag (2005) found that the median MID in 332 medical trials was g= 0.50. Also, Button et al.
(2015) studied three large clinical samples and found that the mean BDIII improvement for patients who reported
feeling better was a fivepoint decrease in BDIII scores. Given the BDIII's SD in Button et al. (2015) was 8.5, the
fivepoint change on the BDIII reported translates to g= 0.58an MID twice as large as the MID proposed by
Cuijpers, Turner, et al. (2014).
1.3 |Objectives and hypotheses
Although several metaanalyses have evaluated the efficacy of PDT relative to other psychotherapies for
depression, their findings are out of date (Cuijpers et al., 2013), based on weak comparison RCTs (e.g., Cuijpers,
Karyotaki, et al., 2020), heterogeneous patient samples, outcomes, and/or comparators (e.g., Driessen et al., 2015;
Steinert et al., 2017a; Tolin, 2010) or miss relevant studies (i.e., Cuijpers et al., 2023, see above). As well, a meta
analytic test of the equivalence of PDT and CBT for adult depressive disorders is absent from the literature, and
whether PDT and CBT are equally effective treatments for depression is thus unclear. Our objective was to address
this by metaanalytically testing whether weighted pooled standardized mean differences (SMDs) for depression
severity at posttreatment and followup (the longest followup within a year) are statistically equivalent across
manualized PDT and CBT in diagnostically homogenous samples of adults with depressive disorders using a
prespecified MID of g= 0.24 (Cuijpers et al., 2014). Based on extant findings (e.g., Cuijpers et al., 2023; Driessen
et al., 2015; Steinert et al., 2017a), we hypothesized that depressive symptoms would be equivalent between PDT
and CBT at posttreatment and followup. To increase confidence in our findings, we conducted sensitivity analyses
involving (a) removing outliers, (b) testing the impact of author allegiance, study quality, the year of publication, and
analysis type (i.e., intentiontotreat vs. completersonly), and (c) including mixed samples with and without
depressive disorders. As noted by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins
et al., 2023), when sensitivity analyses demonstrate that the overall findings are not meaningfully different based on
the different decisions made during the review process, the findings can be interpreted with greater certainty.
2|METHOD
2.1 |Protocol registration
We registered our protocol on PROSPERO (CRD42022345322) before commencing with data analysis.
2
2.2 |Search strategy
We conducted a literature search across five databases (PsychINFO, ProQuest Dissertations and Theses, PubMed,
Cochrane Central Register of Controlled Trials, and Embase) using the following search terms and Boolean
operators: ((((((psychodynamic* or dynamic* or psychoanalytic* or interpretive or expressive or STPP or LTPP)) AND
(therapy or therapies or psychotherapy* or treatment* or counseling)) AND (study or studies or trial*)) AND
(outcome* or result or results or effect* or change*)) AND (psychiat* or mental* or psychol*)) AND (RCT* or
compar* or randomi*). We also searched for relevant articles using the published and regularly updated list of
previously identified RCTs on PDT, the Lilliengren List (see Lilliengren, 2023 for description), and previous
SMITH and HEWITT
|
5
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
systematic and metaanalytic reviews (see Supporting Information S1: Material A). On October 18, 2021, we
terminated search strategies and started data reduction. Following removing duplicates, this search yielded 13,059
studies (see Figure 1). To ensure our search was current, on September 1, 2023, we reexamined the Lilliengren List
and searched all databases for studies published between October 18, 2021 and September 1, 2023, using our key
terms. No additional relevant studies were identified.
Three trained research assistants screened the title and abstract of each study identified. To reduce the
likelihood of excluding a relevant study, screeners were instructed to consider any study that compared PDT to
CBT using an RCT design as relevant during initial screening. Specifically, screeners independently voted in pairs
whether the study was relevant (yes), potentially relevant (maybe), or irrelevant (no). Studies that received two yes
and/or maybevotes were considered eligible for fulltext review, whereas studies with two novotes were
considered irrelevant and excluded from fulltext review. Any study with one yesor maybevote and one novote
was categorized as a conflict to be resolved through team discussion. Next, the first author and a trained research
assistant screened the full text of all articles identified using the following inclusion and exclusion criteria. Studies
were included that: (a) involved the randomized assignment of individuals with diagnosed depressive disorders
established by a valid clinicianrated measure to PDT and CBT, (b) had a minimum of 10 patients per treatment
condition, (c) evaluated bona fide forms of PDT and CBT (Wampold et al., 1997),
3
(d) was in English, (e) involved at
least six sessions of manualguided PDT and CBT, and (f) contained enough information to a calculate relevant
effect sizes. Requiring a minimum of 10 patients per treatment condition is common practice in metaanalytic
reviews on psychotherapy RCTs (e.g., Driessen et al., 2020), because using randomization with small samples can
introduce error (Hsu, 1989) and inflate publication bias (Fonagy, 2015). Studies were excluded that (g) involved a
combined sample of patients with and without depressive disorders, (h) involved the unbalanced administration of
pharmacotherapy across conditions, (i) was retracted or has an erratum, (j) reported exclusively on a costbenefit
analysis, or (k) reports on couples therapy, family therapy, or any other therapy that focuses on families, couples, or
FIGURE 1 PRISMA diagram.
6
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
organizations, as opposed to the individual. Using the operational definitions of PDT and CBT provided by Cuijpers
et al. (2021), an intervention was categorized as PDT or CBT based on a close inspection of the treatment manual or
manuallike guide used.
Full texts that the first author and trained research assistant independently concluded satisfied the inclusion
criteria were included. Similarly, full texts that the first author and research assistant independently concluded did
not satisfy the inclusion criteria were excluded. Adequate agreement regarding the inclusion and exclusion of full
texts was obtained (Cohen's κ= 0.71), and discrepancies were resolved by consulting the full text and through
discussion. A total of 142 studies were excluded (see Supporting Information S1: Table B1 for justification and
Figure 1for PRISMA diagram). We excluded Jakobsen et al. (2014) as we could not access the manuals created for
the trial due to the links provided in the article no longer working and the primary author indicating they no longer
had the manuals used (Janus Jakobsen, personal communication, December 28, 2023).
2.3 |Metaanalytic procedure
Random effects metaanalyses were conducted using Comprehensive MetaAnalysis (Version 3; Borenstein
et al., 2005). To calculate weighted pooled SMDs (Hedges g) for PDT versus CBT, we used the means and SDs of
outcomes at posttreatment and followup. When this data was unavailable, we calculated Hedge's gusing raw
differences in means and standard error, independent tvalues and pvalues, or the standardized differences
reported. If only dichotomous data were reported (e.g., remission), we used the procedure described by Borenstein
et al. (2021) to calculate SMDs. If none of this data was available, we contacted the primary author and requested
the missing data.
Consistent with the Cochrane Handbook for Metaanalytic Reviews, when the SD was not reported and when
it was not possible to estimate the SD from the statistics reported, we imputed the SD by using the weighted
pooled mean SD from other included studies that used the same measure (Higgins et al., 2023). We operationalized
followup as the longest period within a year. For pooled effects involving three or more RCTs, we used the Knapp
Hartung adjustment (Knapp & Hartung, 2003) to correct the standard error. Simulation studies (e.g., IntHout
et al., 2014; Langan et al., 2019) imply KnappHartung adjustments reduce the probability of false positives,
particularly when the number of included studies is small (Borenstein et al., 2021). Further, the KnappHartung
adjustment leads to wider confidence intervals (CIs), which, when conducting equivalence testing, makes it more
difficult to reject the null hypothesis of nonequivalence. For studies that reported standard errors but not SDs, we
calculated SDs by multiplying the standard error by the square root of the sample size (Higgins et al., 2023).
For RCTs assessing depressive symptoms using more than one measure, we generated a sequence in which
more commonly used and reliable measures were given priority over rarely used and less reliable measures
(Cuijpers, 2016). This ensured that only one measure of depression was extracted for each study and guarded
against violations of the assumption of independent observations (Card, 2012). We preferentially preferred
assessorrated measures of depression rated by independent observers blinded to the patient's treatment condition
over selfreport measures. When assessorrated measures were not rated by independent observers blinded to the
patient's treatment condition, we preferentially preferred the Beck Depression Inventory (Beck, 1961) over the
Hamilton Rating Scale for Depression (HRSD; Hamilton et al., 1960), given concerns over the HRSD's validity
(see Bagby et al., 2004). The measures included in the overall analyses are indicated by subscripts in Table 1and
effects specific to blinded assessor ratings, assessor ratings, selfreports, the BDI, and the HRSD are presented
separately in Table 2. The first author and a trained research assistant independently rated each included study's
level of allegiance and quality using the MultiLevel Allegiance Rating Scale (MARS) provided by Steinert et al.
(2017a) and the RCT Quality Rating Scale (RCTQR; Kocsis et al., 2010).
To test whether SMDs at posttreatment and followup were equivalent, we used the TOSTER software
package for R (Lakens, 2017b). Cuijpers, Turner, et al.'s (2014) proposed MID of g= 0.24 was used to calculate
SMITH and HEWITT
|
7
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TABLE 1 Characteristics of RCTs in metaanalysis.
Sample Measures
N
PDT
N
CBT
Diagnoses
Mean
age
Female
%
Number of
sessions
Quality
rating
Allegiance
score PDT manual CBT manual Outcome
Cooper
et al.
(2003)
45 42 Postpartum MDD
(DSMIII),
EPDS 12
28.0 100.0 10 33.5 0 Psychodynamic therapy
(Cramer et al., 1990;
Stern, 1995)
Cognitive behavioral
therapy (Hawton
et al., 1989)
EPDS
a,b,c
SCIDD
d
Driessen
et al.
(2013)
177 164 MDD (DSMIV),
HAMD score 14
38.9 70.1 16 37.5 0 Short psychodynamic
supportive
psychotherapy (de
Jonghe, 2005)
Cognitive behavioral
therapy (Molenaar
et al., 2009)
HRSD
a,b,d
IDSSR
c
Fonagy
et al.
(2020)
73 20 MDD (DSMIV),
HAMD > 14,
PHQ9>10
38.3 67.7 16
c
33 0 Dynamic interpersonal
therapy (Lemma et al.,
2017; Luyten &
Blatt, 2012)
Cognitive behavioral
therapy (Beck &
Haigh, 2014)
HRSD
d
Gallagher
Thompson
and
Steffen
(1994)
21 31 Depressed family
caregivers, major,
minor, or
intermittent
depressive
disorder (RDC),
BDI 10
62.0 92.0 1620 26.5 1 Timelimited
psychotherapy (Mann,
1973; Rose &
DelMaestro, 1990)
Cognitive behavioral
therapy (Beck et al.,
1987; Lewisohn et al.,
1985; Gallagher &
Thompson, 1982)
BDI
a,c
HRSD
d
SADSC
b
Gallagher and
Thompson
(1982)
10 10 Depressed elders,
MDD (RDC),
HAMD score 14,
BDI score 17
68.3 70.0 16 27.5 1 Brief relational and insight
therapy (Bellak &
Small, 1965;
Wolberg, 1965)
Cognitive therapy
(Beck et al., 1979;
Emery, 1981)
BDI
a,b,c
HRSD
d
Gibbons
et al.
(2016)
118 119 MDD (DSMIV),
QIDSC score 11
36.2 75.1 16 38.5 1 Supportive expressive
PDT (Book, 1998;
Luborsky, 1984)
Cognitive behavioral
therapy (Beck, 1995)
HRSD
a,d
8
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TABLE 1 (Continued)
Sample Measures
N
PDT
N
CBT
Diagnoses
Mean
age
Female
%
Number of
sessions
Quality
rating
Allegiance
score PDT manual CBT manual Outcome
Leuzinger
Bohleber
et al.
(2019)
40 34 MDD, dysthymia, or
double depression
(DSMIV),
BDI > 17,
QIDSC>9
40.4 67.9 80 PDT
e
60 CBT
e
37.5 0 Psychoanalytic
psychotherapy
(Taylor, 2010)
Cognitive behavioral
therapy
(Hautzinger, 2013)
BDIII
c
QIDSC
a,b,d
Shapiro
et al.
(1994)
58 57 MDD (DSMIII),
BDI > 16
40.5 52.1 816 33.5 0 Psychodynamic
Interpersonal Therapy
(Hobson, 1985;
Goldberg et al., 1984)
Cognitive Therapy
(Beck et al., 1979)
BDI
a,b,c
Thompson
et al.
(1987)
30 31 Depressed elders,
MDD (RDC),
HAMD score 14,
BDI score 17
67.4 67.4% 1620 24.5 1 Brief psychodynamic
therapy (Horowitz &
Kaltreider, 1979)
Cognitive therapy
(Beck et al., 1979)
BDI
a,c
HRSD
d
Note: Negative allegiance scores indicate an allegiance toward CBT; positive allegiance scores indicate an allegiance towards PDT.
Abbreviations: BDI, Beck's (1961) Beck Depression Inventory; BDIII, Beck et al.'s (1996) Beck Depression InventorySecond Edition; CBT, cognitive behavioral therapy; DSMIIIR,
American Psychiatric Association's (1987) Diagnostic and Statistical Manual of Mental Disorders (3rd edition, revised); DSMIV, American Psychiatric Association's (2000) Diagnostic and
Statistical Manual of Mental Disorders (4th edition); EPDS, Cox et al.'s (1987) Edinburgh Postnatal Depression Scale; HRSD, Hamilton et al.'s (1960) Rating Scale for Depression17 item;
IDSSR, Rush et al.'s (1986) Inventory of Depressive SymptomologySelfReport; MD, major depressive disorder; N
CBT
, the number of participants included in the analysis at posttreatment
in the cognitive behavioral therapy condition; N
PDT
, number of patients included in the analysis at posttreatment in the psychodynamic therapy condition; PDT, psychodynamic
psychotherapy; PHQ9, Kroenke et al.'s (2001) Patient Health Questionnaire; QIDSC, Rush et al.'s (2003) clinicianrated quick inventory of depressive symptoms; RCTs, randomized
controlled trials; RDC, Spitzer's (1978) research diagnostic criteria; SADSC, Spitzer and Endicott's (1977) Schedule for Affective Disorders and Schizophrenia Change interview; SCIDD,
Spitzer's (1992) Structured Clinical Interview for DSMIIIR depression section.
a
Included in overall analysis at posttreatment.
b
Included in overall analysis at followup.
c
Included in selfreported depression analysis.
d
Included in assessorrated depression analysis.
e
Patients received up to 80 sessions of PDT and up to 60 sessions of CBT in the first year.
SMITH and HEWITT
|
9
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TABLE 2 Summary of overall effects of PDT versus CBT for depressive symptoms at posttreatment and followup: Hedges' g.
a
Variable kN
PDT
N
CBT
g90% CI Q
T
I
2
k
TF
Trim and Fillestimates
[90% CI] Equivalence p
Equivalence
outcome NHST p
NHST
outcome
Posttreatment
Depressive disorders 9 572 508 0.11 [0.238; 0.023] 11.9 32.7 1 0.10 [0.231; 0.029] .048 Equivalent .212 No difference
MDD 7 511 443 0.06 [0.203; 0.075] 9.2 34.7 0 0.06 [0.203; 0.075] .019 Equivalent .476 No difference
Adults (1850 years) 6 511 436 0.08 [0.240; 0.083] 9.6 47.9 0 0.08 [0.240; 0.083] .051 Not equivalent .461 No difference
Assessorrated
depression
8 514 451 0.03 [0.184; 0.116] 11.8 40.6 1 0.00 [0.165; 0.163] .012 Equivalent .719 No difference
HRSD 6 429 375 0.04 [0.183; 0.111] 7.8 35.6 1 0.00 [0.178; 0.180] .011 Equivalent .706 No difference
Blind assessorrated
depression
3 231 173 0.08 [0.423; 0.261] 5.8 65.8 0 0.08 [0.423; 0.261] .222 Not equivalent .734 No difference
Selfreported depression 7 367 382 0.16 [0.277; 0.039] 4.0 0.0 1 0.15 [0.266; 0.030] .129 Not equivalent .073 No difference
BDI 5 157 163 0.27 [0.452; 0.088] 1.7 0.0 1 0.25 [0.429; 0.077] .606 Not equivalent .071 No difference
Followup
Depressive disorders 6 262 256 0.16 [0.306; 0.017] 0.9 0.0 2 0.14 [0.274; 0.000] .184 Not equivalent .126 No difference
MDD 4 208 200 0.17 [0.329; 0.007] 0.5 0.0 0 0.17 [0.329; 0.007] .230 Not equivalent .184 No difference
Adults (1850 years) 4 233 219 0.14 [0.290; 0.016] 0.2 0.0 1 0.12 [0.261; 0.011] .134 Not equivalent .236 No difference
Assessorrated
depression
5 203 200 0.13 [0.306; 0.040] 3.0 0.0 1 0.10 [0.272; 0.072] .155 Not equivalent .275 No difference
Selfreported depression 5 238 227 0.12 [0.270; 0.031] 1.8 0.0 2 0.08 [0.213; 0.051] .095 Not equivalent .260 No difference
Note: Negative values favor CBT and positive values favor PDT.
Abbreviations: BDI, Beck's (1961) Beck Depression Inventory and Beck et al.'s (1996) Beck Depression InventorySecond Edition; CBT, cognitive behavioral therapy; CI, confidence
interval; HRSD, Hamilton's (1960) Rating Scale for Depression17 item; I
2
, percentage of heterogeneity; k, number of studies; k
TF
, number of imputed studies as part of trim and fill
method; MDD, major depressive disorder; N
CBT
, total number of participants in the kCBT samples; NHST p, null hypothesis significance testing pvalue calculated using the
KnappHartung adjustment; N
PDT
, total number of participants in the kPDT samples; Q
T
, measure of heterogeneity; PDT, psychodynamic therapy; Tim and Fillestimates, Hedges' g
adjusted for publication bias.
a
According to the randomeffects model.
10
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
equivalence boundaries (i.e., 0.24 to 0.24). Consistent with recommendations (Cristea et al., 2017), we adopted the
conservative strategy of preferentially preferring completersonly data over intentiontotreat data. For all
weighted pooled estimates, we evaluated heterogeneity by calculating Q
T
(the overall heterogeneity among
weighted mean effects) and I
2
(the percentage of variance across studies attributable to heterogeneity). Likewise,
we evaluated publication bias by visually inspecting funnel plots, testing Egger's regression to the intercept and
using trimandfill procedures to obtain effect size estimates after correcting for publication bias. Symmetry near
the top of the funnel plot and asymmetry near the bottom suggest publication bias, as does a significant Egger's
regression coefficient.
To evaluate the sensitivity of our findings to the various analytic decisions made, we tested the extent to which
findings changed after removing outliers, which we operationalized as any study with a 95% CI outside the bounds
of the 95% CI for the weighted pooled estimate. Additionally, we tested the degree to which results were impacted
by pretreatment differences, author allegiance, study quality, and data type (intentiontotreat vs. completers only)
via metaregression with maximum likelihood estimation. Further, we tested the impact of including mixed samples
with and without depressive disorders on findings.
2.4 |Description of studies
Studies were published between 1982 and 2020, with a median publication year of 2003. The mean age of patients
was 46.7 (SD = 15.0; range = 2868.3) years old. Most patients were female (73.5%; SD = 14.3%). Four of the nine
included studies did not report ethnicity (i.e., Cooper et al., 2003; LeuzingerBohleber et al., 2019; Shapiro
et al., 1994; Thompson et al., 1987). Of the five studies that did report ethnicity, the mean percentage of ethnic
minority (i.e., the proportion of participants who did not identify as belonging to the majority ethnicity) was 23.9%
(SD = 23.5%; range = 0.0%48.9%).
Depressive disorders were diagnosed using the Structured Clinical Interview for DSMIV Axis 1 disorders
(k=2; Firstetal., 1997),theScheduleforAffectiveDisordersandSchizophrenia(k= 3; Endicott, 1978), the
MiniInternational Neuropsychiatric Interview Plus (k= 2; Sheehan et al., 1998), the Diagnostic Interview
Schedule (k= 1; Eaton & Kessler, 1985), and the Structured Clinical Interview for American Psychiatric
Association's (1987)DiagnosticandStatisticalManualofMental Disorders (third edition [DSMIII], revised)
(k=1; Spitzer, 1992). Four studies included patients who met DSMIV criteria for MDD, two included
patients who met the research diagnostic criteria for MDD (RDC; Spitzer, 1978), one included patients who
met DSMIII criteria for MDD, one included patients who met RDC criteria for major, minor or intermittent
depressive disorder, and one included patients who met DSMIV criteria for MDD, dysthymia, or double
depression. Only one study reported withinstudy diagnostic reliability (GallagherThompson &
Steffen, 1994).
The total sample size across PDT conditions was 572, with a mean sample size of 63.6 (SD = 53.22;
range = 10177). The total sample size across CBT conditions was 509.0, with a mean sample size of 56.6
(SD = 51.2; range = 10164). Only three studies assessed treatment adherence using a standardized measure
(Cooper et al., 2003; Gibbons et al., 2016; LeuzingerBohleber et al., 2019). The average number of PDT sessions
was 25.0 (SD = 21.7; range = 1080); the average number of CBT sessions was 22.7 (SD = 15.5; range = 1060). The
average number of months after pretreatment for followup data included was 9 months (SD = 3.8; range = 312).
Four studies reported intentiontotreat data and five reported completersonly data, with no studies reporting
both intentiontotreat and completersonly data. All included studies used an individual treatment format. Except
LeuzingerBohleber et al. (2019), studies assessed various types of shortterm PDT and CBT (see Table 1). Eight
studies included assessorrated measures of depression and one study did not include or did not report assessor
ratings (Shapiro et al., 1994). Lastly, therapists were trained for the therapies in seven studies and were not trained,
or training levels were not reported in two studies.
SMITH and HEWITT
|
11
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2.5 |Measures
Depressive symptoms were measured using three assessorrated measures and five selfreport measures
(see Table 1). Three studies reported independent and blinded assessor ratings of depression (Fonagy et al., 2020;
Gibbon et al., 2016; LeuzingerBohleber et al., 2019).
2.6 |Allegiance
Agreement on the MARS was perfect (Intraclass Correlation Coefficent [ICC] = 1.00). Both raters provided identical
MARS ratings. The mean MARS was 0.2 (SD = 0.7). Five studies were rated as having no clear allegiance towards
either PDT or CBT, three were rated as having an allegiance towards CBT, and one was rated as having an
allegiance towards PDT.
2.7 |Study quality
Agreement on the RCTQR was acceptable (ICC = 0.90) and ratings were averaged (see Table 1). The mean RCTQR
score was 33.4 (SD = 4.5; range = 24.538.5). Given that Gerber et al. (2011) concluded that a quality score of 24
and above implies a reasonably welldone study, the average RCTQR score observed implies that all included
studies were reasonably well done. We speculate that the absence of lowquality studies reflects our stringent
inclusion criteria, which required the use of a manual or manuallike guide for the PDT and CBT conditions, a
minimum of six sessions, and the diagnosis of a depressive disorder as determined by a validated structured
interview.
3|RESULTS
3.1 |Equivalence testing
Equivalency findings are in Table 2. All Q
T
values were nonsignificant (p> .05) and I
2
values ranged from 0.0% to
65.8%. At posttreatment, the weighted pooled SMD was 0.11, indicating a trivialsized difference favoring CBT
over PDT. The 90% CI for this difference ranged from 0.238 to 0.023 and was within the prespecified equivalence
boundary of 0.24 to 0.24. Hence, as hypothesized, there was sufficient evidence to accept the alternative
hypothesis of equivalence (p
equivalence
= .048). This finding held when only samples of patients with MDD
were analyzed (g=0.06; 90% CI: 0.20 to 0.08; p
equivalence
= .019), when only assessor ratings of depression were
analyzed (g=0.03; 90% CI: 0.17 to 0.16; p
equivalence
= .012), and when only HRSD ratings of depression were
analyzed (g=0.04; 90% CI: 0.18 to 0.11; p
equivalence
= .011). Likewise, marginal statistical equivalence was found
when only samples of adults aged 1850 years were analyzed (g=0.08; 90% CI: 0.24 to 0.08; p
equivalence
= .051).
However, when only effects from selfreport measures were included, or when only effects from the BDI were
included, we could not detect statistical equivalence or statistical differences (see Table 2).
At followup, contrary to our hypothesis, the SMD was 0.16 with a 90% CI of 0.31 to 0.02 that fell outside
the prespecified equivalency boundary of 0.24 to 0.24. Hence, there was insufficient evidence to reject the null
hypothesis of nonequivalence at followup (p
equivalence
= .184). Similarly, at followup, there was insufficient
evidence to reject the null hypothesis of nonequivalence when only samples of patients with MDD were included
(g=0.17; 90% CI: 0.33 to 0.01, p
equivalence
= .230) and when only adults between the ages of 18 and 50 with a
diagnosed depressive disorder were included (g=0.14; 90% CI: 0.29 to 0.02, p
equivalence
= .134). Even so, there
12
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
was insufficient evidence to conclude that PDT and CBT differ significantly at followup (p
NHST
= .126). Hence,
results suggest PDT and CBT are neither statistically equivalent nor statistically different in followup levels of
depressive symptoms, suggesting additional data is needed to reach a concrete conclusion (Lakens, 2017a).
3.2 |Sensitivity analyses
No outliers were identified at posttreatment or followup. Even so, removing Fonagy et al. (2020) reduced I
2
from
32.7% to 0.0% for posttreatment depressive symptoms. Findings obtained from metaregression implied that
pretreatment differences did not significantly impact SMDs at posttreatment (B= .15, p= .095) or followup
(B=.01, p= .950), that study quality did not significantly impact SMDs at posttreatment (B= .01, p= .737) or
followup (B= .02, p= .477), that author allegiance did not significantly impact SMDs at posttreatment (B= .06,
p= .621) or followup (B= .17, p= .504), that year of publication did not significantly impact SMDs at posttreatment
(B= .01, p= .508) or followup (B= .01, p= .440), and that data type (intentiontotreat vs. completersonly) did not
significantly impact the SMD at pretreatment (B= .31, p= .051) or followup (B= .04, p= .850). Lastly, we examined
the impact of excluding RCTs of mixed samples with and without diagnosed depressive disorders on our findings.
Namely, we inspected our exclusion table (see Supporting Information S1: Material B) and identified the following
seven RCTs as suitable: Chavooshi et al. (2017), Knekt et al. (2008), Leichsenring et al. (2009), Leichsenring et al.
(2013), Svensson et al. (2021), Tasca et al. (2006), and Walsh et al. (1997). As with our main analysis, there was
sufficient evidence to reject the null hypothesis of nonequivalence for depressive symptoms immediately at
posttreatment: k= 16, N
PDT
= 1192, N
CBT
= 1135, g=0.09, 90% CI: 0.19 to 0.01, Q
T
= 25.7, I
2
= 41.0,
p
equivalence
= .006, p
NHST
= .161. Yet, unlike our primary analysis, at followup, there was now sufficient evidence
to reject the null hypothesis of nonequivalence: k= 11, N
PDT
= 646 N
CBT
= 646, g=0.01, 90% CI: 0.19 to 0.17,
Q
T
= 47.5, I
2
= 79.0, p
equivalence
= .017, p
nhst
= .909. Crucially, as expected, including mixed samples with and without
depressive disorders increased heterogeneity considerably, particularly for depressive symptoms at followup
(see Supporting Information S1: Table C1).
3.3 |Publication bias
Funnel plots (see Supporting Information S1: Material D) and trim and fill estimates provided evidence of
publication bias for posttreatment and followup depressive symptoms. Adjusting for publication bias resulted in
the inclusion of one trimmedstudy and an adjusted effect size of g=0.10 (90% CI: 0.23 to 0.03),
p
equivalence
= .040, p
NHST
= .203 immediately posttreatment. At followup, adjusting for publication bias resulted in
the inclusion of two trimmedstudies and an adjusted effect size of g=0.14 (90% CI: 0.27 to 0.00),
p
equivalence
= .108, p
NHST
= .100 for depressive symptoms. Hence, the significance of the equivalence findings did not
change substantively after correcting for publication bias with one exception. At followup, adjusting for publication
bias resulted in statistical equivalence for selfreported depression symptoms following the imputation of two
trimmedstudies: g=0.08 (90% CI: 0.21 to 0.05), p
equivalence
= .023, p
NHST
= .311.
4|DISCUSSION
A wealth of evidence has cemented psychotherapy as an effective treatment for depression (e.g., Cuijpers, Noma,
et al., 2020), with PDT and CBT being two of the most used approaches (Solem et al., 2010). However, the question
remains: Is PDT as effective as CBT for treating adults with diagnosed depressive disorders? Despite sustained
metaanalytic inquiry, methodological limitations, such as combining diagnostically heterogeneous samples
SMITH and HEWITT
|
13
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
(e.g., Steinert et al., 2017a; Tolin, 2010) and comparators (e.g., Driessen et al., 2010,2015) have prevented a
concrete answer. Moreover, a metaanalytic test of the equivalence of PDT and CBT for depressive disorders is
absent from the literature. We addressed this by conducting a comprehensive metaanalytic test of the equivalence
of manualized PDT and CBT for posttreatment and followup depressive symptoms using diagnostically
homogenous RCTs of adults with diagnosed depressive disorders.
Results were derived from nine active comparison RCTs of 1081 adults with a depressive disorder as
determined by validated clinical interviews. Heterogeneity was minimal, perhaps due to our narrow inclusion
criteria, increasing clinical relevance and generalizability, but precluding tests of which treatment works best for
whom. Metaanalytic equivalency testing using an exceptionally stringent prespecified MID of g= 0.24 (Cuijpers
et al., 2014) showed that depressive symptoms immediately at posttreatment were statistically equivalent across
manualized PDT and CBT. Sensitivity analyses showed this finding held after accounting for pretreatment
differences, study quality, author allegiance, and the year of publication, as well as when our inclusion criteria were
relaxed to allow mixed samples with and without depressive disorders. Followup depressive symptoms were
neither statistically equivalent nor statistically different across PDT and CBT, which alludes to a need for additional
controlled research evaluating patients beyond posttreatment.
Although we can only speculate on why depressive symptoms immediately at posttreatment are statistically
equivalent between PDT and CBT, some may interpret our findings as yet another paper supporting the dodo bird
verdictthe position that all psychotherapies are equally effective regardless of the population studied or the
specific techniques used (Rosenzweig, 1936). Or, more specifically, the factors shared across PDT and CBT, such as
establishing a therapeutic relationship and patient expectancy, might be far greater determinants of outcomes than
the specific therapeutic techniques used (Wampold, 2015). In support, heterogeneity was small despite the
inclusion of different types of PDT and CBT (see Table 1), suggesting there was little variability left to be explained
by the specific type of PDT and CBT used. Even so, contrary to the dodo bird verdict, it is plausible that PDT and
CBT work equally well for depressive symptoms but for different reasons (see Cuijpers et al., 2019).
4.1 |Clinical and research implications
As hypothesized, results suggest no clinically meaningful differences between PDT and CBT in posttreatment
depressive symptoms for adults with diagnosed depressive disorders. This accords with other metaanalytic
research finding trivial differences between PDT and other psychotherapies for posttreatment depressive
symptoms (Driessen et al., 2010,2015) as well as metaanalytic research (Cuijpers et al., 2023; Leichsenring
et al., 2023) finding that levels of depressive symptoms did not differ significantly across PDT and CBT. In contrast,
we found no evidence supporting Tolin's (2010) conclusion that CBT was superior to PDT for depression. One
possible explanation for this discrepancy is that unlike Tolin (2010), we excluded studies that did not use or report
using a treatment manual and that involved less than six sessions of PDT and CBT (see Supporting Information S1:
Table B1).
Overall, our results accord with the position taken by the American Psychological Association's (2019) clinical
practice guidelines for the treatment of depression in the general adult population, which recommends all
empirically supported psychotherapies, including psychodynamic therapies and cognitive behavioral therapies, as
they appear to have comparable effects(p. 9). In contrast, our results call into question the Canadian Network of
Mood and Anxiety Treatments (see Parikh et al., 2016) recommendation that CBT, but not PDT, be considered a
firstline treatment for adult depression, with one caveat. Though there is enough evidence to conclude there is no
clinically meaningful difference between PDT and CBT in depressive symptoms immediately at posttreatment,
contrary to our hypothesis, we cannot rule out the possibility that PDT and CBT differ meaningfully in followup
levels of depressive symptoms. Accordingly, there remains an urgent need for further highquality RCTs, especially
those that directly compare PDT to CBT at followup and those that look at mechanisms of change.
14
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4.2 |Limitations of the overall literature and future directions
Our study provides novel insights into the state of the literature on the relative efficacy of PDT and CBT for adult
depressive disorders and, by doing so, underscores limitations. One such limitation is that samples were
predominately female and the generalizability of findings to males and nonbinary adults with depressive disorders is
unclear. Similarly, five studies did not report ethnicity and the extent to which findings generalize to ethnic
minorities is unclear. To this end, an RCT directly comparing PDT and CBT for adults with diverse cultural
backgrounds and sexual orientations would make a vital contribution to the literature.
Another area for improvement is that only four studies assessed treatment adherence using standardized
measures. The use of standardized adherence measures is crucial for demonstrating treatment fidelity
(Schoenwald & Garland, 2013) and it is challenging to determine what adherence ratings obtained via a
measure with unknown psychometric properties mean.Moreover,theissueoftreatment adherence is
particularly relevant to the present findings, given that treatment adherence may moderate the efficacy of
CBT such that studies with greater adherence yield superior outcomes (Shafran et al., 2009). Additionally,
most studies analyzed completersonly data and no study reported completersonly and intentiontotreat
data. Relatedly, although some researchers maintain that intentiontotreat data is more appropriate for
NHST and completersonly data more appropriate for equivalence testing (Cristea et al., 2017), the veracity
of this claim is unclear (see Steinert et al., 2017b). To this end, a metaanalysis examining what impact, if any,
prioritizing intentiontotreat data over completersonly data has on the Type 1 error rate when conducting
NHST and equivalence testing would make an essential contribution.
Likewise, although doubleblind designs are not possible in psychotherapy research, singleblind designs, in
which assessors are blind to the patient's condition, are possible and yield higherquality evidence (Lynch
et al., 2010). Even so, only four of the 10 RCTs reported blind assessor ratings of depression. Furthermore,
according to some researchers, clinically meaningful change typically only occurs after the 30th session, irrespective
of the type of therapy (Morrison et al., 2003), and only two RCTs involved over 30 sessions of PDT and CBT, which
may have attenuated the response rate observed (see Shedler & Gnaulati, 2020 for a discussion of this issue). It is
also plausible that an exclusive focus on symptoms is a limited strategy for testing the relative efficacy of
psychotherapies for depression and that a more fruitful approach would involve examining putative underlying
mechanisms and subsequent change in symptom expression and the likelihood of relapse.
Additionally, the HRSD was the most commonly used measure of depressive symptoms across included studies.
However, research suggests that the HRSD's total score is multidimensional, that its factor structure is not invariant
across different populations, and that its conceptualization of depression is several decades out of date (see Bagby
et al., 2004 for review). Hence, future research would likely profit from using a more psychometrically sound
assessorrated measure of depression. Lastly, there is RCT evidence that PDT holds a small advantage over CBT for
patients experiencing severe and persistent depressive symptoms (Driessen et al., 2016). Moreover, metaanalytic
evidence from individual participant data suggests that relative to minimaltonotreatment controls, PDT is
particularly efficacious for patients experiencing longer depressive episodes (Wienicke et al., 2023). To this end, a
metaanalysis of individual participant data from RCTs directly comparing PDT to CBT for depression could advance
the field considerably by testing the potential moderating effect of the chronicity of depressive symptoms.
4.3 |Limitations of the present study
Limitations in the literature translate into limitations in our analysis. One limitation is the small number of studies
reporting change at followup resulted in an underpowered test of equivalence at followup. Relatedly, there was
insufficient data to conduct equivalency tests for secondary outcomes, such as interpersonal problems. Also,
although we attempted to use completersonly data for equivalency tests, this was not possible for four studies and
SMITH and HEWITT
|
15
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
intentiontotreat data had to be used instead. As such, this might have inflated equivalency estimates. Lastly, we
could not test which treatment works best for whomdue to minimal heterogeneity.
4.4 |Concluding remarks
The present study offers a needed metaanalytic test of the equivalence of manualized PDT and CBT for adults with
diagnosed depressive disorders. In synthesizing extant active comparison RCTs, we extended the ever
accumulating evidence base in support of the efficacy of PDT for depressive disorders by showing that there is
no clinically meaningful difference in depressive symptoms immediately at posttreatment relative to the gold
standard psychotherapy for depressive disordersCBT (David et al., 2018). Hence, from a clinical standpoint, our
results underscore the importance of considering patient preference and situational factors when choosing
between PDT and CBT for depressive disorders.
AUTHOR CONTRIBUTIONS
With respect to author contributions, Martin M. Smith and Paul L. Hewitt played equal roles in the
conceptualization. Martin M. Smith played a key role in the data analysis and writing, and Paul Hewitt provided
supervision, editing, and feedback.
ACKNOWLEDGMENTS
The authors received no funding for this study and wish to thank Kaja Bakken and Marcia Moessler for their
assistance with the literature search.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable
request.
ORCID
Martin M. Smith http://orcid.org/0000-0002-4754-3032
Paul L. Hewitt http://orcid.org/0000-0003-4474-8611
PEER REVIEW
The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer
review/10.1002/jclp.23649.
ENDNOTES
1
Equivalence testing and NHST are not mutually exclusive, as they test different hypotheses and convey unique
information (Weber & Popova, 2012). As Lakens (2017a) observed, there are four possible outcomes when equivalence
testing and NHST are used in tandem. First, results could suggest an SMD is neither statistically equivalent nor
statistically different. Second, results could suggest an SMD is statistically equivalent and not statistically different. Third,
findings could show an SMD is not statistically equivalent but is statistically different. Lastly, an SMD can be both
statistically equivalent and statistically different. This occurs when the lower and upper bounds of the 90% confidence
interval are less than and greater than the lower and upper bounds of the equivalence boundary, respectively, and when
the 95% confidence interval does not contain zero.
16
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2
There was insufficient data to report findings related to secondary outcomes (i.e., interpersonal problems, psychosocial
functioning, and satisfaction with life). A second manuscript on differences in rates of response, reliable improvement,
and acceptability is in preparation.
3
Bonafide psychotherapies involve a therapist with a minimum a master's degree, who meets with and develops a
relationship with the patient, who tailors the treatment to the patient's needs, and who provides treatment for an issue
that can reasonably be expected to be amenable to psychotherapy (Wampold et al., 1997). Additionally, the study
reporting on the psychotherapy must satisfy two or more of the following criteria: (a) a citation to established
psychotherapy (e.g., Rogerian) is provided, (b) description of the therapy is provided in text and the process of change
described, (c) a manual for the treatment exists and was used to guide the psychotherapy, and/or (d) the active
ingredients of the treatment was identified, and citations provided for the ingredients.
REFERENCES
Note: Studies marked with an asterisk were included in the metaanalytic review.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd edition, revised).
American Psychiatric Association.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th edition). American
Psychiatric Association.
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th edition text revised).
American Psychiatric Association. https://doi.org/10.1176/appi.books.9780890425787
American Psychological Association. (2019). Clinical practice guideline for the treatment of depression across three age
cohorts. Retrieved from https://www.apa.org/depression-guideline
Bagby, R. M., Ryder, A. G., Schuller, D. R., & Marshall, M. B. (2004). The Hamilton Depression Rating Scale: Has the gold
standard become a lead weight? American Journal of Psychiatry,161, 21632177. https://doi.org/10.1176/appi.ajp.
161.12.2163
Barlow, D. H. (2010). The dodo birdagainand again. The Behavior Therapist,33,1516.
Barth, J., Munder, T., Gerger, H., Nüesch, E., Trelle, S., Znoj, H., Jüni, P., & Cuijpers, P. (2013). Comparative efficacy of seven
psychotherapeutic interventions for patients with depression: A network metaanalysis. PLoS Medicine,10(5),
e1001454. https://doi.org/10.1371/journal.pmed.1001454
Beck, A. T. (1961). An inventory for measuring depression. Archives of General Psychiatry,4, 561571. https://doi.org/10.
1001/archpsyc.1961.01710120031004
Beck, A. T., & Haigh, E. A. (2014). Advances in cognitive theory and therapy: The generic cognitive model. Annual review of
clinical psychology,10,124. https://doi.org/10.1146/annurev-clinpsy-032813-153734
Beck, A. T., Rush, J., Shaw, B., & Emery, G. (1979). Cognitive therapy of depression. Guilford.
Beck, A. T., Steer, R. A., & Brown, G. (1996). BDIII, Beck depression inventory: Manual. Psychological Corp.
Beck, A. T., Steer, R. A., & Brown, G. K. (1987). Beck depression inventory. Harcourt Brace Jovanovich.
Beck, J. S. (1995). Cognitive therapy: Basics and beyond. Guilford Press.
Bellak, L., & Small, L. (1965). Emergency Psychotherapy and Brief Psychotherapy. Grune & Stratton.
Book, H. E. (1998). How to practice breif dynmaic psychotherapy: The CCRT method. American Psychological Assocaition.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2005). Comprehensive metaanalysis (version 3) [computer
software]. Biostat.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2021). Introduction to metaanalysis. Wiley.
Butler, A., Chapman, J., Forman, E., & Beck, A. (2006). The empirical status of cognitivebehavioral therapy: A review of
metaanalyses. Clinical Psychology Review,26(1), 1731. https://doi.org/10.1016/j.cpr.2005.07.003
Button, K. S., Kounali, D., Thomas, L., Wiles, N. J., Peters, T. J., Welton, N. J., Ades, A. E., & Lewis, G. (2015). Minimal
clinically important difference on the Beck Depression InventoryII according to the patient's perspective.
Psychological Medicine,45(15), 32693279. https://doi.org/10.1017/S0033291715001270
Card, N. A. (2012). Applied metaanalysis for social science research. Guilford.
Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical
Psychology,66(1), 718. https://doi.org/10.1037/0022006X.66.1.7
Chavooshi, B., Saberi, M., Tavallaie, S. A., & Sahraei, H. (2017). Psychotherapy for medically unexplained pain: A randomized
clinical trial comparing intensive shortterm dynamic psychotherapy and cognitivebehavior therapy. Psychosomatics,
58, 506518. https://doi.org/10.1016/j.psym.2017.01.003
Clark, M., DiBenedetti, D., & Perez, V. (2016). Cognitive dysfunction and work productivity in major depressive disorder.
Expert Review of Pharmacoeconomics & Outcomes Research,16(4), 455463. https://doi.org/10.1080/14737167.2016.
1195688
SMITH and HEWITT
|
17
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
*Cooper, P. J., Murray, L., Wilson, A., & Romaniuk, H. (2003). Controlled trial of the shortand longterm effect of
psychological treatment of postpartum depression: I. Impact on maternal mood. British Journal of Psychiatry,182(5),
412419. https://doi.org/10.1192/bjp.182.5.412
Cox, J. L., Holden, J. M., & Sagovsky, R. (1987). Detection of postnatal depression: Development of the 10item Edinburgh
Postnatal Depression Scale. British Journal of Psychiatry,150, 782786. https://doi.org/10.1192/bjp.150.6.782
Cramer, B., RobertTissot, C., Stern, D. N., SerpaRusconi, S., De Muralt, M., Besson, G., PalacioEspasa, F., Bachmann, J.P.,
Knauer, D., Berney, C., & D'Arcis, U. (1990). Outcome evaluation in brief motherinfant psychotherapy: A preliminary
report. Infant Mental Health Journal,11(3), 278300. https://doi.org/10.1002/1097-0355(199023)11:3<278::AID-
IMHJ2280110309>3.0.CO;2-H
Cristea, I. A. (2019). The waiting list is an inadequate benchmark for estimating the effectiveness of psychotherapy for
depression. Epidemiology and Psychiatric Sciences,28(3), 278279. https://doi.org/10.1017/S2045796018000665
Cristea, I. A., Cuijpers, P., & Naudet, F. (2017). Equivalence of psychodynamic therapy to other established treatments:
Limited supporting evidence and clinical relevance. American Journal of Psychiatry,174, 1122. https://doi.org/10.
1176/appi.ajp.2017.17050592
Cuijpers, P. (2016). Metaanalyses in mental health research: A practical guide. Vrije Universiteit. http://bit.do/metaanalysis
Cuijpers, P., Berking, M., Andersson, G., Quigley, L., Kleiboer, A., & Dobson, K. S. (2013). A metaanalysis of cognitive
behavioural therapy for adult depression, alone and in comparison with other treatments. Canadian Journal of
Psychiatry,58(7), 376385. https://doi.org/10.1177/070674371305800702
Cuijpers, P., & Cristea, I. A. (2016). How to prove that your therapy is effective, even when it is not: A guideline.
Epidemiology and Psychiatric Sciences,25(5), 428435. https://doi.org/10.1017/S2045796015000864
Cuijpers, P., Driessen, E., Hollon, S. D., van Oppen, P., Barth, J., & Andersson, G. (2012). The efficacy of nondirective
supportive therapy for adult depression: A metaanalysis. Clinical Psychology Review,32(4), 280291. https://doi.org/
10.1016/j.cpr.2012.01.003
Cuijpers, P., Karyotaki, E., Ciharova, M., Miguel, C., Noma, H., & Furukawa, T. A. (2021). The effects of psychotherapies for
depression on response, remission, reliable change, and deterioration: A metaanalysis. Acta Psychiatrica Scandinavica,
144(3), 288299. https://doi.org/10.1111/acps.13335
Cuijpers, P., Karyotaki, E., Reijnders, M., & Ebert, D. D. (2019). Was Eysenck right after all? A reassessment of the effects of
psychotherapy for adult depression. Epidemiology and Psychiatric Sciences,28(1), 2130. https://doi.org/10.1017/
S2045796018000057
Cuijpers, P., Karyotaki, E., Weitz, E., Andersson, G., Hollon, S. D., & van Straten, A. (2014). The effects of psychotherapies
for major depression in adults on remission, recovery and improvement: A metaanalysis. Journal of Affective Disorders,
159, 118126. https://doi.org/10.1016/j.jad.2014.02.026
Cuijpers, P., Karyotaki, E., de Wit, L., & Ebert, D. D. (2020). The effects of fifteen evidencesupported therapies for adult
depression: A metaanalytic review. Psychotherapy Research: Journal of the Society for Psychotherapy Research,30(3),
279293. https://doi.org/10.1080/10503307.2019.1649732
Cuijpers, P., Miguel, C., Harrer, M., Plessen, C. Y., Ciharova, M., Ebert, D., & Karyotaki, E. (2023). Cognitive behavior therapy
vs. control conditions, other psychotherapies, pharmacotherapies and combined treatment for depression: A
comprehensive metaanalysis including 409 trials with 52,702 patients. World Psychiatry,22, 105115. https://doi.
org/10.1002/wps.21069
Cuijpers, P., Noma, H., Karyotaki, E., Vinkers, C. H., Cipriani, A., & Furukawa, T. A. (2020). A network metaanalysis of the
effects of psychotherapies, pharmacotherapies and their combination in the treatment of adult depression. World
Psychiatry,19(1), 92107. https://doi.org/10.1002/wps.20701
Cuijpers, P., Turner, E. H., Koole, S. L., van Dijke, A., & Smit, F. (2014). What is the threshold for a clinically relevant effect?
The case of major depressive disorders. Depression and Anxiety,31(5), 374378. https://doi.org/10.1002/da.22249
David, D., Cristea, I., & Hofmann, S. G. (2018). Why cognitive behavioral therapy is the current gold standard of
psychotherapy. Frontiers in Psychiatry,9,4.https://doi.org/10.3389/fpsyt.2018.00004
de Jonghe, F. (2005). Kort en Krachtig. Kortdurende Psychoanalytische Steungevende Psychotherapie [Short and snappy:
Shortterm psychoanalytical supportive psychotherapy]. Benecke N.I.
Driessen, E., Cuijpers, P., de Maat, S. C., Abbass, A. A., de Jonghe, F., & Dekker, J. J. (2010). The efficacy of shortterm
psychodynamic psychotherapy for depression: A metaanalysis. Clinical Psychology Review,30(1), 2536. https://doi.
org/10.1016/j.cpr.2015.07.004
Driessen, E., Dekker, J. J. M., Peen, J., Van, H. L., Maina, G., Rosso, G., Rigardetto, S., Cuniberti, F., Vitriol, V. G.,
Florenzano, R. U., Andreoli, A., Burnand, Y., LópezRodríguez, J., VillamilSalcedo, V., Twisk, J. W. R., & Cuijpers, P.
(2020). The efficacy of adding shortterm psychodynamic psychotherapy to antidepressants in the treatment of
depression: A systematic review and metaanalysis of individual participant data. Clinical Psychology Review,80,
101886. https://doi.org/10.1016/j.cpr.2020.101886
18
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Driessen, E., Hegelmaier, L. M., Abbass, A. A., Barber, J. P., Dekker, J. J. M., Van, H. L., Jansma, E. P., & Cuijpers, P. (2015).
The efficacy of shortterm psychodynamic psychotherapy for depression: A metaanalysis update. Clinical Psychology
Review,42,115. https://doi.org/10.1016/j.cpr.2015.07.004
Driessen, E., Smits, N., Dekker, J. J., Peen, J., Don, F. J., Kool, S., Westra, D., Hendriksen, M., Cuijpers, P., & Van, H. L. (2016).
Differential efficacy of cognitive behavioral therapy and psychodynamic therapy for major depression: A study of
prescriptive factors. Psychological Medicine,46(4), 731744. https://doi.org/10.1017/S0033291715001853
*Driessen, E., Van, H. L., Don, F. J., Peen, J., Kool, S., Westra, D., Hendriksen, M., Schoevers, R. A., Cuijpers, P.,
Twisk, J. W. R., & Dekker, J. J. M. (2013). The efficacy of cognitivebehavioral therapy and psychodynamic therapy in
the outpatient treatment of major depression: A randomized clinical trial. American Journal of Psychiatry,170(9),
10411050. https://doi.org/10.1176/appi.ajp.2013.12070899
Eaton, W. W., & Kessler, L. G., (Eds.). (1985). Epidemiological field methods in psychiatry: The NIMH epidemiological catchment
area program. Academic Press.
Emery, G. (1981). Cognitive therapy with the elderly. In S. Emery, Hollon & R. Bedrosian (Eds.), New directions in cognitive
therapy. Guilford.
Endicott, J. (1978). A diagnostic interview: The schedule for affective disorders and schizophrenia. Archives of General
Psychiatry,35, 837844. https://doi.org/10.1001/archpsyc.1978.01770310043002
Fabbri, C., Mutz, J., Lewis, C. M., & Serretti, A. (2023). Depressive symptoms and neuroticismrelated traits are the main
factors associated with wellbeing independent of the history of lifetime depression in the UK Biobank. Psychological
Medicine,53(7), 30003008. https://doi.org/10.1017/S003329172100502X
First, M. B., Spitzer, R. K., Gibbon, M., & Williams, J. B. W. (1997). Structured clinical interview for DSMIV Axis I Disorders
(SCIDI)Clinician version. American Psychiatric Press.
Fisher, R. A. (1925). Theory of statistical estimation, Mathematical proceedings of the Cambridge Philosophical Society
(Vol. 22, No. 5, pp. 700725). Cambridge University.
Fonagy, P. (2015). The effectiveness of psychodynamic psychotherapies: An update. World Psychiatry,14(2), 137150.
https://doi.org/10.1002/wps.20235
*Fonagy, P., Lemma, A., Target, M., O'Keeffe, S., Constantinou, M. P., Wurman, T. V., Luyten, P., Allison, E., Roth, A., Cape, J., &
Pilling, S. (2020). Dynamic interpersonal therapy for moderate to severe depression: A pilot randomized controlled and
feasibility trial. Psychological Medicine,50(6), 10101019. https://doi.org/10.1017/S0033291719000928
Furukawa, T. A., Noma, H., Caldwell, D. M., Honyashiki, M., Shinohara, K., Imai, H., Chen, P., Hunot, V., & Churchill, R.
(2014). Waiting list may be a nocebo condition in psychotherapy trials: A contribution from network metaanalysis.
Acta Psychiatrica Scandinavica,130(3), 181192. https://doi.org/10.1111/acps.12275
*Gallagher, D. E., & Thompson, L. W. (1982). Treatment of major depressive disorder in older adult outpatients with brief
psychotherapies. Psychotherapy: Theory, Research & Practice,19(4), 482490. https://doi.org/10.1037/h0088461
*GallagherThompson, D., & Steffen, A. M. (1994). Comparative effects of cognitivebehavioral and brief psychodynamic
psychotherapies for depressed family caregivers. Journal of Consulting and Clinical Psychology,62(3), 543549. https://
doi.org/10.1037/0022006X.62.3.543
Gerber, A. J., Kocsis, J. H., Milrod, B. L., Roose, S. P., Barber, J. P., Thase, M. E., Perkins, P., & Leon, A. C. (2011). A quality
based review of randomized controlled trials of psychodynamic psychotherapy. The American Journal of Psychiatry,
168(1), 1928. https://doi.org/10.1176/appi.ajp.2010.08060843
*Gibbons, M. B. C., Gallop, R., Thompson, D., Luther, D., CritsChristoph, K., Jacobs, J., Yin, S., & CritsChristoph, P. (2016).
Comparative effectiveness of cognitive therapy and dynamic psychotherapy for major depressive disorder in a
community mental health setting: A randomized clinical noninferiority trial. JAMA Psychiatry,73(9), 904911. https://
doi.org/10.1001/jamapsychiatry.2016.1720
Goldberg, D. P., Hobson, R. F., Maguire, G. P., Margison, F. R., O'Dowd, T., Osborn, M. S., & Moss, S. (1984). The clarification
and assessment of a method of psychotherapy. British Journal of Psychotherapy,114, 567575. https://doi.org/10.
1192/bjp.144.6.567
Greenberg, P. E., Fournier, A. A., Sisitsky, T., Simes, M., Berman, R., Koenigsberg, S. H., & Kessler, R. C. (2021). The economic
burden of adults with major depressive disorder in the United States (2010 and 2018). PharmacoEconomics,39(6),
653665. https://doi.org/10.1007/s40273021010194
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery & Psychiatry,23(1), 5662. https://
doi.org/10.1136/jnnp.23.1.56
Hasin, D. S., Sarvet, A. L., Meyers, J. L., Saha, T. D., Ruan, W. J., Stohl, M., & Grant, B. F. (2018). Epidemiology of adult
DSM5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry,75(4), 336346. https://doi.
org/10.1001/jamapsychiatry.2017.4602
Hautzinger, M. (2013). Cognitive behavior therapy of depression (7th ed). Beltz Edition.
Hawton, K., Salkovskis, P., Kirk, J., & Clark, D. M. (1989). Cognitive behavioral approaches to adult psychiatric disorders.
Oxford University Press.
SMITH and HEWITT
|
19
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Higgins, J. P. T., Thomas, J., Chandler, J., Cumpston, M., Li, T., Page, M. J., & Welch, V. A. (Eds.). (2023). Cochrane handbook
for systematic reviews of interventions version 6.4. Cochrane. https://training.cochrane.org/handbook
Hobson, R. F. (1985). Forms of feeling: The heart of psychotherapy. Basic Books.
Hsu, L. M. (1989). Random sampling, randomization, and equivalence of contrasted groups in psychotherapy outcome
research. Journal of Consulting and Clinical Psychology,57(1), 131137. https://doi.org/10.1037/0022-006X.57.1.131
IntHout, J., Ioannidis, J. P., & Borm, G. F. (2014). The HartungKnappSidikJonkman method for random effects meta
analysis is straightforward and considerably outperforms the standard DerSimonianLaird method. BMC Medical
Research Methodology,14, 25. https://doi.org/10.1186/147122881425
Jaeschke, R., Singer, J., & Guyatt, G. H. (1989). Measurement of health status. Controlled Clinical Trials,10(4), 407415.
https://doi.org/10.1016/01972456(89)900056
Jakobsen, J. C., Gluud, C., Kongerslev, M., Larsen, K. A., Sorensen, P., Winkel, P., Lange, T., Sogaard, U., & Simonsen, E.
(2014). Thirdwave cognitive therapy versus mentalisationbased treatment for major depressive disorder: a
randomised clinical trial. BMJ Open,4(8), e004903. https://doi.org/10.1136/bmjopen2014004903
John Rush, A., Giles, D. E., Schlesser, M. A., Fulton, C. L., Weissenburger, J., & Burns, C. (1986). The Inventory for
Depressive Symptomatology (IDS): Preliminary findings. Psychiatry Research,18(1), 6587. https://doi.org/10.1016/
01651781(86)900600
Knapp, G., & Hartung, J. (2003). Improved tests for a random effects metaregression with a single covariate. Statistics in
Medicine,22(17), 26932710. https://doi.org/10.1002/sim.1482
Knekt, P., Lindfors, O., Härkänen, T., Välikoski, M., Virtala, E., Laaksonen, M. A., Marttunen, M., Kaipainen, M., & Renlund, C.
(2008). Randomized trial on the effectiveness of longand shortterm psychodynamic psychotherapy and solution
focused therapy on psychiatric symptoms during a 3year followup. Psychological Medicine,38(5), 689703. https://
doi.org/10.1017/S003329170700164X
Kocsis, J. H., Gerber, A. J., Milrod, B., Roose, S. P., Barber, J., Thase, M. E., Perkins, P., & Leon, A. C. (2010). A new scale for
assessing the quality of randomized clinical trials of psychotherapy. Comprehensive Psychiatry,51(3), 319324.
https://doi.org/10.1016/j.comppsych.2009.07.001
Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ9: Validity of a brief depression severity measure. Journal of
General Internal Medicine,16, 606613. https://doi.org/10.1046/j.15251497.2001.016009606.x
Lakens, D. (2017a). Equivalence tests: A practical primer for t tests, correlations, and metaanalyses. Social Psychological and
Personality Science,8(4), 355362. https://doi.org/10.1177/1948550617697177
Lakens, D. (2017b). TOSTER: Two onesided tests (TOST) equivalence testing (Version 3.0). [Computer software]: Retrieved
from https://CRAN.Rproject.org/package=TOSTER
Lakens, D., Scheel, A. M., & Isager, P. M. (2018). Equivalence testing for psychological research: A tutorial. Advances in
Methods and Practices in Psychological Science,1(2), 259269. https://doi.org/10.1177/2515245918770963
Langan, D., Higgins, J., Jackson, D., Bowden, J., Veroniki, A. A., Kontopantelis, E., Viechtbauer, W., & Simmonds, M. (2019).
A comparison of heterogeneity variance estimators in simulated randomeffects metaanalyses. Research Synthesis
Methods,10(1), 8398. https://doi.org/10.1002/jrsm.1316
Lange, S., & Freitag, G. (2005). Special invited papers section: Therapeutic equivalence Clinical issues and statistical
methodology in noninferiority trials: Choice of delta: Requirements and reality Results of a systematic review.
Biometrical Journal,47,1227. https://doi.org/10.1002/bimj.200410085
Leichsenring, F., Abbass, A., Heim, N., Keefe, J. R., Kisely, S., Luyten, P., Rabung, S., & Steinert, C. (2023). The status of
psychodynamic psychotherapy as an empirically supported treatment for common mental disordersAn umbrella
review based on updated criteria. World Psychiatry,22(2), 286304. https://doi.org/10.1002/wps.21104
Leichsenring, F., Leweke, F., Klein, S., & Steinert, C. (2015). The empirical status of psychodynamic psychotherapyAn
update: Bambi's alive and kicking. Psychotherapy and Psychosomatics,84(3), 129148. https://doi.org/10.1159/
000376584
Leichsenring, F., Salzer, S., Beutel, M. E., Herpertz, S., Hiller, W., Hoyer, J., Huesing, J., Joraschky, P., Nolting, B.,
Poehlmann, K., Ritter, V., Stangier, U., Strauss, B., Stuhldreher, N., Tefikow, S., Teismann, T., Willutzki, U., Wiltink, J., &
Leibing, E. (2013). Psychodynamic therapy and cognitivebehavioral therapy in social anxiety disorder: A multicenter
randomized controlled trial. American Journal of Psychiatry,170(7), 759767. https://doi.org/10.1176/appi.ajp.2013.
12081125
Leichsenring, F., Salzer, S., Jaeger, U., Kachele, H., Kreische, R., Leweke, F., Ruger, U., Winkelbach, C., & Leibing, E. (2009).
Shortterm psychodynamic psychotherapy and cognitivebehavioral therapy in generalized anxiety disorder: A
randomized, controlled trial. American Journal of Psychiatry,166(8), 875881.
Lemma, A., Target, M., & Fonagy, P. (2017). Manual for Dynamic Interpersonal Therapy (DIT). Anna Freud National Centre
for Children and Families/Tavistock Clinic.
*LeuzingerBohleber, M., Hautzinger, M., Fiedler, G., Keller, W., Bahrke, U., Kallenbach, L., Kaufhold, J., Ernst, M.,
Negele, A., Schoett, M., Küchenhoff, H., Günther, F., Rüger, B., & Beutel, M. (2019). Outcome of psychoanalytic and
20
|
SMITH and HEWITT
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
cognitivebehavioural longterm therapy with chronically depressed patients: A controlled trial with preferential and
randomized allocation. The Canadian Journal of Psychiatry,64(1), 4758. https://doi.org/10.1177/
0706743718780340
Lewinsohn, P., Hobberman, H., Teri, L., & Hautzinger, M. (1985). An integrative theory of depression. In S. Resiss & R. R.
Bootzin (Eds.), Theoretical issues in behavior therapy (pp. 331359). Academic Press.
Lilliengren, P. (2023). A comprehensive overview of randomized controlled trials of psychodynamic psychotherapies.
Psychoanalytic Psychotherapy,37, 117140. https://doi.org/10.1080/02668734.2023.2197617
Linde, M., Tendeiro, J. N., Selker, R., Wagenmakers, E. J., & van Ravenzwaaij, D. (2023). Decisions about equivalence: A
comparison of TOST, HDIROPE, and the Bayes factor. Psychological Methods,28, 740755. https://doi.org/10.1037/
met0000402
Luborsky, L. (1984). Principles of psychoanalytic psychotherapy: A manual for supportiveexpressive (SE) treatment. Basic Books.
Luborsky, L., Rosenthal, R., Diguer, L., Andrusyna, T. P., Berman, J. S., Levitt, J. T., Seligman, D. A., & Krause, E. D. (2002).
The dodo bird verdict is alive and wellmostly. Clinical Psychology: Science and Practice,9(1), 212. https://doi.org/10.
1093/clipsy.9.1.2
Luyten, P., & Blatt, S. J. (2012). Psychodynamic treatment of depression. Psychiatric Clinics of North America,35, 111129.
Lynch, D., Laws, K. R., & McKenna, P. J. (2010). Cognitive behavioural therapy for major psychiatric disorder: does it really
work? A metaanalytical review of wellcontrolled trials. Psychological Medicine,40(1), 924. https://doi.org/10.1017/
S003329170900590X
Mann, J. (1973). Time limited psychotherapy. Harvard University Press.
Mohr, D. C., Spring, B., Freedland, K. E., Beckner, V., Arean, P., Hollon, S. D., Ockene, J., & Kaplan, R. (2009). The selection
and design of control conditions for randomized controlled trials of psychological interventions. Psychotherapy and
Psychosomatics,78(5), 275284.
Molenaar, P. J., Don, F., Bout, J., van den Sterk, F., & Dekker, J. (2009). Cognitieve gedragstherapie bij depressie [Cognitive
behavioural therapy with depression]. Bohn Stafleu van Loghum.
Morrison, K. H., Bradley, R., & Westen, D. (2003). The external validity of controlled clinical trials of psychotherapy for
depression and anxiety: A naturalistic study. Psychology and Psychotherapy: Theory, Research and Practice,76(2),
109132. https://doi.org/10.1348/147608303765951168
National Institute for Health and Clinical Excellence. (2004). Depression: Management of depression in primary and secondary
care. British Psychological Society and Royal College of Psychiatrists.
Neyman, J., & Pearson, E. S. (1933). IX. On the problem of the most efficient tests of statistical hypotheses. Philosophical
Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character,231,
289337. https://doi.org/10.1098/rsta.1933.0009
Parikh, S. V., Quilty, L. C., Ravitz, P., Rosenbluth, M., Pavlova, B., Grigoriadis, S., Velyvis, V., Kennedy, S. H., Lam, R. W.,
MacQueen, G. M., Milev, R. V., Ravindran, A. V., & Uher, R., CANMAT Depression Work Group. (2016). Canadian
Network for Mood and Anxiety Treatments (CANMAT) 2016 clinical guidelines for the management of adults with
major depressive disorder: Section 2. Canadian Journal of Psychiatry,61(9), 524539. https://doi.org/10.1177/
0706743716659418
Rose, J., & DelMaestro, D. (1990). Separationindividuation conflict as a model for understanding distressed caregivers:
Psychodynamic and cognitive case studies. The Gerontologist,30, 693697. https://doi.org/10.1093/geront/30.5.693
Rosenzweig, S. (1936). Some implicit common factors in diverse methods of psychotherapy. American Journal of
Orthopsychiatry,6(3), 412415. https://doi.org/10.1111/j.19390025.1936.tb05248.x
Rush, A. J., Trivedi, M. H., Ibrahim, H. M., Carmody, T. J., Arnow, B., Klein, D. N., Markowitz, J. C., Ninan, P. T., Kornstein, S.,
Manber, R., Thase, M. E., Kocsis, J. H., & Keller, M. B. (2003). The 16Item Quick Inventory of Depressive
Symptomatology (QIDS), Clinician rating (QIDSC), and Selfreport (QIDSSR): A psychometric evaluation in patients
with chronic major depression. Biological Psychiatry,54(5), 573583. https://doi.org/10.1016/s00063223(02)
018668
Dos Santos, É. N., Molina, M. L., Mondin, T., Cardoso, T. A., Silva, R., Souza, L., & Jansen, K. (2020). Longterm effectiveness
of two models of brief psychotherapy for depression: A threeyear followup randomized clinical trial. Psychiatry
Research,286, 112804. https://doi.org/10.1016/j.psychres.2020.112804
Schoenwald, S. K., & Garland, A. F. (2013). A review of treatment adherence measurement methods. Psychological
Assessment,25(1), 146156. https://doi.org/10.1037/a0029715
Shafran, R., Clark, D. M., Fairburn, C. G., Arntz, A., Barlow, D. H., Ehlers, A., Freeston, M., Garety, P. A., Hollon, S. D.,
Ost, L. G., Salkovskis, P. M., Williams, J. M. G., & Wilson, G. T. (2009). Mind the gap: Improving the dissemination of
CBT. Behaviour Research and Therapy,47(11), 902909. https://doi.org/10.1037/a0029715
*Shapiro, D. A., Barkham, M., Rees, A., Hardy, G. E., Reynolds, S., & Startup, M. (1994). Effects of treatment duration and
severity of depression on the effectiveness of cognitivebehavioral and psychodynamicinterpersonal psychotherapy.
Journal of Consulting and Clinical Psychology,62(3), 522534. https://doi.org/10.1037/0022006x.62.3.522
SMITH and HEWITT
|
21
10974679, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.23649 by Cochrane Canada Provision, Wiley Online Library on [07/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Shedler, J., & Gnaulati, E. (2020). March/April. The tyranny of time: How long does effective therapy really take? Psychotherapy
Networker. Retrieved from https://www.psychotherapynetworker.org/article/tyrannytime
Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., & Dunbar, G. C.
(1998). The MiniInternational Neuropsychiatric Interview (M.I.N.I): The development and validation of a structured
diagnostic psychiatric interview for DSMIV and ICD10. The Journal of Clinical Psychiatry,59 Suppl 20(Suppl. 20),
2233. https://doi.org/10.1037/t18597000
Soares, M. C., Mondin, T. C., Silva, G., Barbosa, L. P., Molina, M. L., Jansen, K., Souza, L., & Silva, R. (2019). Comparison of
clinical significance of cognitivebehavioral therapy and psychodynamic therapy for major depressive disorder:
retraction. Journal of Nervous & Mental Disease,207(11), 993. https://doi.org/10.1097/NMD.0000000000001113
Soares, M. C., Mondin, T. C., Silva, G. D. G., Barbosa, L. P., Molina, M. L., Jansen, K., Souza, L. D. M., & Silva, R. A. (2018).
Comparison of clinical significance of cognitivebehavioral therapy and psychodynamic therapy for major depressive
disorder: A randomized clinical trial. Journal of Nervous & Mental Disease,206(9), 686693. https://doi.org/10.1097/
NMD.0000000000000872
Solem, S., Vogel, P. A., & Hofmann, S. G. (2010). An international comparison between different theoretical orientations of
psychotherapy: A survey of expert opinions. The Behavior Therapist,33,17.
Spitzer, R., & Endicott, J. (1977). The SADSChange interview. New York State Psychiatric Institute.
Spitzer, R. L. (1978). Research diagnostic criteria: Rationale and reliability. Archives of General Psychiatry,35, 773782.
https://doi.org/10.1001/archpsyc.1978.01770300115013
Spitzer, R. L. (1992). The Structured Clinical Interview for DSMIIIR (SCID). I: History, rationale, and description. Archives of
General Psychiatry,49(8), 624629. https://doi.org/10.1001/archpsyc.1992.01820080032005
Steinert, C., Munder, T., Rabung, S., Hoyer, J., & Leichsenring, F. (2017a). Psychodynamic therapy: As efficacious as other
empirically supported treatments? A metaanalysis testing equivalence of outcomes. The American Journal of
Psychiatry,174(10), 943953. https://doi.org/10.1176/appi.ajp.2017.17010057
Steinert, C., Munder, T., Rabung, S., Hoyer, J., & Leichsenring, F. (2017b). Different standards when assessing the evidence
for psychodynamic therapy? Response to cristea et al. The American Journal of Psychiatry,174(11), 11231124.
https://doi.org/10.1176/appi.ajp.2017.17050592r
Stern, D. (1995). The motherhood constellation. Basic Books.
Svensson, M., Nilsson, T., Perrin, S., Johansson, H., Viborg, G., Falkenström, F., & Sandell, R. (2021). The effect of patient's
choice of cognitive behavioural or psychodynamic therapy on outcomes for panic disorder: A doubly randomised
controlled preference trial. Psychotherapy and Psychosomatics,90(2), 107118. https://doi.org/10.1159/000511469
Tasca, G. A., Ritchie, K., Conrad, G., Balfour, L., Gayton, J., Lybanon, V., & Bissada, H. (2006). Attachment scales predict
outcome in a randomized controlled trial of two group therapies for binge eating disorder: An aptitude by treatment
interaction. Psychotherapy Research,16(1), 106121. https://doi.org/10.1080/10503300500090928
Taylor, D. (2010). Das TavistockManual der psychoanalytischen Psychotherapieunter besonderer Berücksichtigung der
chronischen depression [The Tavistock Manual of Psychoanalytic Psychotherapy for chronic Depression]. Psyche,
64(9), 833861.
*Thompson, L. W., Gallagher, D., & Breckenridge, J. S. (1987). Comparative effectiveness of psychotherapies for depressed
elders. Journal of Consulting and Clinical Psychology,55(3), 385390. https://doi.org/10.1037//0022006x.55.3.385
Tolin, D. F. (2010). Is cognitivebehavioral therapy more effective than other therapies? A metaanalytic review. Clinical
Psychology Review,30(6), 710