Int J Psychoanal 2005;86:841–68

Are psychodynamic and psychoanalytic therapies effective?

A review of empirical data

FALK LEICHSENRING
Clinic of Tiefenbrunn and University of Göttingen, von Sieboldstr.5, D-37075 Göttingen, Germany —
Fleichs@gwdg.de
(Final version accepted 5 January 2005)

There is a need for empirical outcome research in psychodynamic and psychoanalytic

therapy. However, both the approach of empirically supported therapies (EST) and
the procedures of evidence-based medicine (EBM) have severe limitations making
randomised controlled trials (RCTs) an absolute standard. After a critical discussion
of this approach, the author reviews the empirical evidence for the efficacy of
psychodynamic psychotherapy in specific psychiatric disorders. The review aims to
identify for which psychiatric disorders RCTs of specific models of psychodynamic
psychotherapy are available and for which they are lacking, thus providing a basis for
planning further research. In addition, results of process research of psychodynamic
psychotherapy are presented. As the methodology of RCTs is not appropriate for
psychoanalytic therapy, effectiveness studies of psychoanalytic therapy are reviewed as
well. Studies of psychodynamic psychotherapy published between 1960 and 2004 were
identified by a computerised search using Medline, PsycINFO and Current Contents.
In addition, textbooks and journal articles were used. Twenty-two RCTs were identified
of which 64% had not been included in the 1998 report by Chambless and Hollon.
According to the results, for the following psychiatric disorders at least one RCT
providing evidence for the efficacy of psychodynamic psychotherapy was identified:
depressive disorders (4 RCTs), anxiety disorders (1 RCT), post-traumatic stress
disorder (1 RCT), somatoform disorder (4 RCTs), bulimia nervosa (3 RCTs), anorexia
nervosa (2 RCTs), borderline personality disorder (2 RCTs), Cluster C personality
disorder (1 RCT), and substance-related disorders (4 RCTs). According to results of
process research, outcome in psychodynamic psychotherapy is related to the competent
delivery of therapeutic techniques and to the development of a therapeutic alliance. With
regard to psychoanalytic therapy, controlled quasi-experimental effectiveness studies
provide evidence that psychoanalytic therapy is (1) more effective than no treatment
or treatment as usual, and (2) more effective than shorter forms of psychodynamic
therapy. Conclusions are drawn for future research.

Keywords: empirically supported treatments, evidence-based medicine,

psychoanalytic therapy, psychodynamic psychotherapy, psychotherapy outcome and
process research

In these times of evidence-based medicine and empirically supported treatments,

there is a need for empirical outcome research in psychodynamic and psychoanalytic
therapy (Gunderson and Gabbard, 1999). In a first part, this article presents a review
of the available randomised controlled trials of short-term and moderate-length
psychodynamic psychotherapy in specific psychiatric disorders. As the methodology

©2005 Institute of Psychoanalysis

842 FALK LEICHSENRING

of randomised controlled trials is not appropriate for psychoanalytic therapy,

effectiveness studies of psychoanalytic therapy are reviewed in the second part
of the article. Unlike randomised controlled trials (efficacy studies), effectiveness
studies are carried out under the conditions of clinical practice.

Evidence-based medicine and empirically supported treatments

Several proposals have been made to grade the available evidence of both medical
and psychotherapeutic treatments (CTFPHE (Canadian Task Force on the Periodic
Health Examination), 1979; Cook et al., 1995; Guyatt et al., 1995; Nathan and
Gorman, 2002; Chambless and Hollon, 1998; Clarke and Oxman, 2003; National
Institute of Clinical Excellence [internet]). Apart from other differences, all available
proposals regard randomised controlled trials (RCTs) as the ‘gold standard’ for the
demonstration that a treatment is effective. According to these proposals, only RCTs
can provide level I evidence, that is the highest level of evidence. The defining
feature of an RCT is the random assignment of subjects to the different conditions of
treatment (Shadish et al., 2002). Randomisation is regarded as indispensable in order
to ensure that a priori existing differences between subjects are equally distributed.
The goal of randomisation is to attribute the observed effects exclusively to the
applied therapy (internal validity). Thus, randomisation is used to ensure the internal
validity of a study (Shadish et al., 2002).
The exclusive position of RCTs as methods for demonstrating that a treatment
works has recently been queried. The main argument is that it is questionable whether
the results of RCTs are representative for clinical practice (Seligman, 1995; Roth and
Parry, 1997; Beutler, 1998; Henry, 1998; Persons and Silberschatz, 1998; Fonagy,
1999; Leichsenring, 2004; Westen et al., 2004). Thus, the supposed strength of RCTs,
especially randomisation, can turn out to be their central weakness, because RCTs
create artificial conditions that are not representative of clinical practice (Seligman,
1995). This is also true for the use of therapy manuals and the treatment of specific
disorders. The latter aspect promotes the tendency to study isolated disorders that,
according to epidemiological studies, hardly occur in clinical practice (e.g. Kessler
et al., 1994). The reduction of psychotherapy to the use of manuals developed for
the treatment of specific disorders will do serious damage to the breadth of clinical
training (Henry, 1998).
Furthermore, the methodology of RCT is not applicable to long-term
psychotherapy lasting several years or to psychoanalytic therapy (Seligman, 1995;
Wallerstein, 1999). It is neither possible to carry out a treatment according to a
manual for several years, nor to offer patients who seek treatment no therapy for
several years. It is also impossible to create equally credible comparison conditions
over years (Seligman, 1995). Furthermore, patients who decide for a psychoanalytic
therapy differ from those choosing a shorter form of psychodynamic therapy by
specific personality traits (Rudolf et al., 1994). Thus, random assignment of
patients to psychoanalytic therapy vs. a comparison treatment would destroy the
matching between patients and treatments. In other words, randomisation would
destroy the subject of research and produce misleading results. The results of a
randomised study of psychoanalytic therapy would not be valid for the patients who
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 843

are usually treated with psychoanalytic therapy. Uncritically applied in such a way,
randomisation is in danger of becoming a dogma and of losing its function as a useful
method of research. Another debatable aspect of the EST approach is the emphasis
on disorders and on symptoms (Blatt, 1995). As Henry put it (1998, p. 129): ‘EVTs
[empirically validated treatments] place the emphasis on the disorder … and not on
the individual … who seeks our services’.
After all, RCTs serve only a limited function (Roth and Parry, 1997, p. 370):
‘RCTs are … an imperfect tool; almost certainly their results are best seen as one
part of a research cycle’. RCTs can show that a treatment works under controlled
conditions. However, if a method of psychotherapy has been shown to work under
controlled conditions, this does not necessarily imply that it equally works under the
conditions of clinical practice (Leichsenring, 2004). The main reason for this gap
is that psychotherapy is not a drug that extensively works equally under different
conditions, and the medical disease model is inadequate for complex mental illness
conditions. The evidence that a treatment works in the field, can only be provided
by effectiveness studies (Seligman, 1995). Thus, a distinction is necessary between
EST and RCT methodology (Leichsenring, 2004; Westen et al., 2004).
Also in the realm of evidence-based medicine, a critical discussion has begun
as to whether RCTs really represent the gold standard of outcome research (Benson
and Hartz, 2000; Concato et al., 2000; Pocock and Elbourne, 2000). On the basis
of their data, Concato et al. (2000) question the usual hierarchy of research designs
with RCTs ranking at the top. The discussion of effectiveness studies will be taken
up again below.
Unlike RCTs, naturalistic studies (effectiveness studies) are carried out under
the conditions of clinical practice. They are highly clinically representative
(Shadish et al., 2000): Patients with complex (i.e. highly co-morbid) disorders, as
they usually occur in clinical practice, are treated. Therapists apply exactly those
methods of psychotherapy that they usually apply and that they are experienced in.
Patients themselves make a decision for a specific kind of therapy and for a specific
psychotherapist, and the duration of the treatment is determined by the clinical
requirements (Seligman, 1995).
Paradoxically, naturalistic studies are not accepted, for example by the American
Psychological Association (APA) as methods for demonstrating that a therapy
works (TFPDPP (Task Force on Promotion and Dissemination of Psychological
Procedures), 1995; Chambless and Hollon, 1998; Chambless and Ollendick, 2001).
The main argument against naturalistic studies refers to threats to internal validity,
that is, to the reduced possibility of controlling factors influencing outcome apart
from therapy. Measures to improve the internal validity of naturalistic studies will
be discussed below.

Evidence for short-term psychodynamic psychotherapy

The TFPDPP (1995) judged short-term psychodynamic psychotherapy (STPP)
as ‘probably efficacious’. The problem with STPP was that no two studies of
independent research groups were found demonstrating efficacy of the same form of
844 FALK LEICHSENRING

STPP in the same psychiatric disorder. The TFPDPP concluded that further evidence
for STPP in specific psychiatric disorders is required if this clinically validated form
of treatment is to survive in the present market.
For this reason, it is important to know, for which psychiatric disorders RCTs of
specific models of both long-term and short-term psychodynamic psychotherapy are
available and for which they are lacking. The aim of this article is to identify these areas.
Thus, this review will provide a basis for planning further disorder-related research of
psychodynamic psychotherapy. In this review, the criteria proposed by the TFPDPP
(1995) modified by Chambless and Hollon (1998) to define efficacious treatments were
applied. However, this does not imply that these criteria are uncritically accepted. As
described above, they have to be discussed critically with regard to its limitations.
However, the aim of this review was to examine the evidence for psychodynamic
psychotherapy under the requirements of ESTs.
As the method of RCTs is not appropriate for long-term psychodynamic
psychotherapy and psychoanalytic therapy lasting several years, evidence from
effectiveness studies will be included in this review as far as long-term treatments
are concerned.

Method

Definition of psychodynamic psychotherapy

From the results of the Menninger Psychotherapy Research Project, Wallerstein
(1989) has concluded that even the most interpretive therapies include supportive
elements. Thus, psychoanalytic therapy—and psychodynamic psychotherapy in
general—operate on an interpretive–supportive continuum, and the use of more
interpretive or supportive interventions depends on the patient’s needs (Wallerstein,
1989; Gunderson and Gabbard, 1999; Gabbard, 2004). For long-term psychodynamic
psychotherapy the following definition given by Gunderson and Gabbard (1999,
p. 685; Gabbard, 2004, p. 2) may be applied: ‘a therapy that involves careful
attention to the therapist–patient interaction, with thoughtfully timed interpretation
of transference and resistance embedded in a sophisticated appreciation of the
therapist’s contribution to the two-person field’.
With regard to duration, Gabbard (2004, p. 3) has proposed to regard therapies of
more than 24 sessions or lasting longer than six months as long-term. This definition
may apply to the US; in European countries, for example in Germany, therapies of
25 to 50 sessions are not regarded as long-term. This definition would imply, for
example, cognitive-behavioral therapy (CBT) as it is presently applied in clinical
practice in Germany, is long-term therapy. To take these differences into account,
therapies with a duration of 25 to 100 sessions will be called moderate-length
psychodynamic therapy in this review. Certainly, all attempts to define what is short
term and what is long term will set arbitrary boundaries (see also Gabbard, 2004,
p. 2). With regard to the frequency of sessions, psychodynamic and psychoanalytic
therapy includes a range from one session up to five sessions per week.
With regard to short-term psychodynamic psychotherapy (STPP), different
models have been developed, which are reviewed, for example, by Messer and
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 845

Warren (1995). Apart from conceptual and technical differences (e.g. Messer and
Warren, 1995), the different forms of STPP have some therapeutic elements in
common. With regard to formal characteristics, they are time-limited (usually 16–
30 sessions), performed in a face-to-face setting, with one or two sessions a week
(Davanloo, 1980; Messer, 2001; Shapiro et al. 2003). Usually, therapists are more
active, and the therapy is more oriented to achieve predefined goals than in long-term
psychodynamic psychotherapy. STPP focuses on specific conflicts or themes which
are formulated early in therapy. Therapists foster the development of a therapeutic
alliance and a positive transference. The focus is on the patients’ experiences here-
and-now, including their symptoms, but working through pre-existing conflicts is
also included (e.g. Mann, 1973; Horowitz, 1976; Luborsky, 1984). In a review of
empirical studies, Blagys and Hilsenroth (2000) identified seven features that were
significantly more frequently observed in STPP compared to CBT: focus on affect,
on resistance, on identification of consistent patterns (of relationships, feelings and
behaviours), on past experiences, on interpersonal experiences, on the therapeutic
relationship, and on wishes, dreams or fantasies.

Search for studies and inclusion criteria

In addition to the usual search for studies via reviews, meta-analyses and textbooks,
a computerised search was carried out using Medline and PsycINFO with the
following key words: psychotherapy, psychodynamic, psychoanalytically oriented,
study. The search was carried out for the period between 1960 and 2004. Only studies
that fulfilled the following criteria (Chambless and Hollon, 1998) were included in
this review: (a) use of a randomised controlled design, (b) treatment of patients
with a specific psychiatric disorder, (c) use of treatment manuals or manual-like
guidelines, (d) use of reliable and valid measures of diagnosis and outcome, (e)
comparison of treatments with a control group (placebo group, waiting list) or with
another treatment. Studies of interpersonal therapy (IPT; e.g. Elkin et al., 1989;
Wilfley et al., 1993) were not included because the relationship to psychodynamic
therapy is controversial (e.g. Markowitz et al., 1998). According to empirical
results, the interpersonal therapies applied in the National Institute of Mental Health
(NIMH) treatment of depression study (e.g. Elkin et al., 1989) corresponded most
strongly to the ideal prototype of CBT (Ablon and Jones, 2002). As Ablon and
Jones put it: ‘Brand names of therapy can be misleading’ (2002, p. 780). Thus,
this review included only studies for which there seems to be a general agreement
that they represent models of psychodynamic psychotherapy. Comparable, but not
identical inclusion criteria were set up in a recent Cochrane report for short-term
psychodynamic therapy (Abbass et al., 2004). For example, studies in which patient
samples with heterogeneous disorders were included by Abbass et al., but were
excluded by the present review which followed the procedures of Chambless and
Hollon (1998) including only studies of a specific psychiatric disorder.

Results
Twenty-two RCTs of STPP could be included in this review. These studies are
presented in Table 1. In their 1995 report, the TFPDPP mentioned five studies
846 FALK LEICHSENRING

Table 1—Randomised controlled studies of short-term and moderate-length

psychodynamic psychotherapy (PP) in specific psychiatric disorders

Study Disorder N (PP) Comparison group Concept of PP Treatment

duration
Thompson et depression 24 BT: N = 25; Horowitz and Kaltreider 16–20
al., 1987 CBT: N = 27; (1979) sessions
waiting list: N = 19
Shapiro et al., depression 58 CBT: N = 59 Shapiro and Firth (1985) 8 vs. 16
1994 sessions
Gallagher- depression 30 CBT: N = 36 Mann (1973), Rose and 16–20
Thompson and DelMaestro (1990) sessions
Steffen, 1994
Barkham et al., depression 18 CBT: N = 18 Shapiro and Firth (1985) 8 vs. 16
1996 sessions
Bögels et al., social phobia 22 CBT: N = 27 Malan (1976) 36
2003, 2004
Brom et al., PTSD 29 de-sensitisation: N = 31 Horowitz (1976) Mean = 18.8
1989 hypnotherapy: N = 29 sessions
Dare et al., 2001 anorexia nervosa 21 Cognitive-analytic therapy Malan (1976), Dare (1995) Mean = 24.9
(Ryle): N = 22; sessions
family therapy: N = 22;
routine treatment: N = 19
Gowers et al., anorexia nervosa treatment as usual: N = 20 Crisp (1980) 12 sessions
1993 20
Fairburn et al., bulimia nervosa 11 CBT: N = 11 Rosen (1979), Stunkard 19 sessions
1986 (1976), Bruch (1973)
Garner et al., bulimia nervosa 25 CBT: N = 25 Luborsky (1984) 19 sessions
1993
Bachar et al., anorexia nervosa, 17 cognitive therapy: N = 17 Barth (1991), Goodsitt 46 sessions
1999 bulimia nervosa nutritional counselling: N = 10 (1985), Geist (1989)

Svartberg et al., Cluster C personality 25 CBT: N = 25 Malan (1976), 40 sessions

2004 disorders McCullough Vaillant (1976)
Munroe-Blum borderline personality 31 interpersonal group therapy: N Kernberg (1975) 17 sessions
and Marziali, disorder = 25
1995
Bateman and borderline personality 19 Treatment as usual: N = 19 Fonagy (1998), Bateman, 18 months
Fonagy, 1999, disorder (1995)
2001
Woody et al., opiate dependence 31 drug counselling (DC): N = 35 Luborsky (1984) 12 sessions
1983, 1990 CBT + DC: N = 34 + drug counselling

Woody et al., opiate dependence 57 drug counselling: N = 27 Luborsky (1985) 26 sessions

1995 + drug counselling
Sandahl et al., alcohol dependence 25 CBT: N = 24 Foulkes (1964) 15 sessions
1998 (Mean = 8.9
sessions)
Crits-Christoph cocaine dependence 124 CBT + group drug counselling Mark and Luborsky (1992) up to 36
et al., 1999, (DC): N = 97, individual DC: N + group DC individual and
2001 = 92, 24 group
individual DC + group DC: N = sessions;
96 4 months
Guthrie et al., irritable bowel 50 supportive listening: Hobson (1985), Shapiro and 8 sessions
1991 N = 46 Firth (1985)
Creed et al., irritable bowel 59 paroxetine: N = 43 Hobson (1985), Shapiro and 8 sessions
2003 treatment as usual: N = 86 Firth (1985)
Hamilton et al., functional dyspepsia 37 supportive therapy: N = 36 Shapiro and Firth 7 sessions
2000 (1985)
Monsen and somatoform 20 treatment as usual/no therapy: Monsen and Monsen (1999) 33 sessions
Monsen, 2000 pain disorder N = 20
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 847

of STPP. Two of these studies refer to STPP in specific psychiatric disorders:

the study of Thompson et al. (1987) of the treatment of geriatric depression
and the study of Woody et al. (1990) of the treatment of opiate dependence.
Furthermore, the TFPDPP mentioned three studies in which patients with
heterogeneous disorders were treated with STPP (Piper et al., 1990; Winston
et al., 1994; Shefler et al., 1995). Fourteen of the 22 studies identified by the
present review were published in 1995, that is in the year the TFPDPP report
was published, or later. Thus, 64% of the studies that were identified in this
review were not included in the 1995 report of the TFPDPP. In three further
RCTs identified, STPP was combined with pharmacological therapy in the
treatment of depression (de Jonghe et al., 2001; Burnand et al., 2002) or panic
disorder (Wiborg and Dahl, 1996). These three studies showed that STPP was
efficacious in interaction with medication, not that STPP alone was necessarily
efficacious. This also applies to two further RCTs, which studied the treatment
of prolonged or pathological grief by STPP (McCallum and Piper, 1990; Piper
et al., 2001). In these studies, STPP was superior to a waiting list or a supportive
group therapy. However, in both studies a considerable proportion of patients
received psychopharmacological treatments (45% and 55%, respectively).
Thus, the results of these studies only apply to the combined treatment of STPP
and medication.
Therapy duration. In the 22 studies of psychodynamic psychotherapy, between
7 and 46 sessions were conducted (Table 1). According to the definition given above,
14 studies (64%) refer to short-term psychodynamic psychotherapy and 8 (36%) to
moderate-length psychodynamic psychotherapy.
Models of psychodynamic psychotherapy. In the 22 studies, different forms of
psychodynamic psychotherapy were applied (see Table 1). Most frequently, the
models of psychodynamic psychotherapy developed by Horowitz (1976), Luborsky
(1984), and Shapiro and Firth (1985) were applied.

Efficacy of psychodynamic psychotherapy in specific psychiatric disorders

The 22 studies of psychodynamic psychotherapy included in this review will be
presented for the different psychiatric disorders. However, from a psychoanalytic
perspective, the results of a therapy in a specific psychiatric disorder (e.g.
depression, agoraphobia) are influenced by the underlying psychodynamic features
(e.g. conflicts, defences, personality organisation), which may vary considerably
within one category of psychiatric disorder (e.g. Kernberg, 1996).1 The results

1
However, this is true for studies of other forms of therapy as well, e.g. for CBT; for example, patients
included in an RCT of CBT in the treatment of major depression may differ considerably in terms of
their underlying psychodynamic features. However, CBT therapists do not take these features into
account. Although such factors are not taken into account by CBT therapists, they may influence the
results in contributing to the variance of outcome. For this reason, the impact of psychodynamics is
not only a problem of psychodynamic psychotherapy or psychoanalytic therapy. In RCTs, this problem
is usually handled by defining inclusion and exclusion criteria, e.g. by excluding severe personality
disorders. Assessing co-morbid disorders and studying their impact on outcome is another way to deal
with this problem (e.g. Woody et al., 1985).
848 FALK LEICHSENRING

will be presented separately for short-term and moderate-length psychodynamic

psychotherapy.

Major Depression (DSM-IV 269, 300.4; ICD-10 F32, F33)

Four RCTs provided evidence for the efficacy of STPP compared to CBT in
depression (Thompson et al., 1987; Gallagher-Thompson and Steffen, 1994;
Shapiro et al., 1994; Barkham et al., 1996). Different models of STPP were
applied (Table 1). In these studies, STPP and CBT proved to be equally effective
with regard to depressive symptoms, general psychiatric symptoms and social
functioning (Leichsenring, 2001).
STPP achieved large pre–post effect sizes in depressive symptoms, general
psychiatric symptoms and social functioning (Leichsenring, 2001).2 The results
proved to be stable in follow-up studies (Gallagher-Thompson et al., 1990; Shapiro
et al., 1995). These results are consistent with the findings of the meta-analysis of
Wampold et al. (2002) who did not find significant differences between CBT and
‘other therapies’ in the treatment of depression.
Anxiety disorders (DSM-IV 300.XX; ICD-10 F40, F41)
Only one RCT of STPP in the treatment of anxiety disorders was identified that
fulfilled the inclusion criteria, that of Bögels et al. (2003). No further RCT of
psychodynamic psychotherapy alone (i.e. without additional pharmacotherapy)
in the treatment of anxiety disorders was identified. In the RCT of Bögels et al.
(2003) moderate-length psychodynamic psychotherapy of 36 sessions proved to be
as effective as CBT in the treatment of generalised social phobia

Post-traumatic stress disorder (DSM-IV 309.81; ICD-10 F43.1)

In an RCT of Brom et al. (1989), the effects of STPP, behavioural therapy (trauma
de-sensitisation) and hypnotherapy in patients with post-traumatic stress disorder
(PTSD) were studied. STPP proved to be as effective as trauma de-sensitisation in
the reduction of trauma-related symptoms. Both forms of therapy were superior to a
waiting list control group. Results of STPP were not only maintained, but continued
to improve at 3-month follow-up.

Somatoform disorders (DSM-IV 300.81; 307.80; ICD-10 F45).

At present, three RCTs of STPP in somatoform disorders that fulfilled the inclusion
criteria are available (Table 1). In the RCT of Guthrie et al. (1991) patients with
irritable bowel syndrome, who had not responded to standard medical treatment
over the previous 6 months, were treated with STPP in addition to standard medical
treatment. This treatment was compared to standard medical treatment alone.
According to the results, STPP was feasible and effective in two-thirds of the
patients. In another RCT, STPP was significantly more effective than routine care

2
Effect sizes give the amount of change in units of standard deviations. For example, a pre–post effect
size of 1.00 is indicative of a pre–post difference of one standard deviation. According to a convention
proposed by Cohen (1988), an effect size of 0.80 is regarded as a large effect.
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 849

and as effective as medication (paroxetine) in the treatment of severe irritable

bowel syndrome (Creed et al., 2003). During the follow-up period, STPP, but
not paroxetine was associated with a significant reduction in health care costs
compared with treatment as usual. In an RCT of Hamilton et al. (2000) STPP
was compared to supportive therapy in the treatment of patients with chronic
intractable functional dyspepsia, who had failed to respond to conventional
pharmacological treatments. At the end of treatment, STPP was significantly
superior to the control condition on both the gastroenterologists’ and patients’
total symptom score. The effects were stable in the 12-month follow-up. Monsen
and Monsen (2000) compared moderate-length psychodynamic psychotherapy
of 33 sessions to a control condition (no treatment or treatment as usual) in the
treatment of patients with chronic pain. STPP was significantly superior to the
control group on measures of pain, psychiatric symptoms, interpersonal problems
and affect consciousness. The results remained stable in the 12-month follow-up,
or even improved.

Bulimia nervosa (DSM-IV 307.51; ICD-10: F50.2)

Three RCTs of STPP in the treatment of bulimia nervosa are available (Table
1). Significant and stable improvements in bulimia nervosa after STPP were
demonstrated in the RCTs of Fairburn et al. (1986, 1995) and Garner et al. (1993).
In the central disorder-specific measures (bulimic episodes, self-induced vomiting),
STPP was as effective as CBT (Fairburn et al., 1986, 1995; Garner et al., 1993).3 In
another RCT, STPP was significantly superior to both a treatment as usual (TAU)
group (nutritional counselling) and cognitive therapy (Bachar et al., 1999). This was
true for patients with bulimia nervosa and a mixed sample of patients with bulimia
nervosa or anorexia nervosa.

Anorexia nervosa (DSM-IV 307.1; F50.0)

In an RCT of Gowers et al. (1994), STPP combined with four sessions of nutritional
advice yielded significant improvements in patients with anorexia nervosa (Table
1). Weight and body mass index (BMI) changes were significantly better than for
a control condition (treatment as usual). Dare et al. (2001) compared moderate-
length psychodynamic psychotherapy with a mean duration of 24.9 sessions to
cognitive-analytic therapy, family therapy and routine treatment in the treatment
of anorexia nervosa (Table 1). Moderate-length psychodynamic psychotherapy
yielded significant symptomatic improvements, and STPP and family therapy
were significantly superior to the routine treatment concerning weight gain.
However, the improvements were modest, several patients being undernourished
at follow-up.

3
Apart from this, CBT was superior to STPP in some specific measures of psychopathology (Fairburn et
al., 1986; Garner et al., 1993). However, in a follow-up of the Fairburn et al. study (1986), using a longer
follow-up period, both forms of therapy proved to be equally effective and were partly superior to a
behavioural form of therapy (Fairburn et al., 1995). Accordingly, for a valid evaluation of the efficacy of
STPP in bulimia nervosa longer-term follow-up studies are necessary.
850 FALK LEICHSENRING

Borderline personality disorders (DSM-IV 301.83; ICD-10 F60.31)

In an RCT, Munroe-Blum and Marziali (1995) compared STPP to interpersonal
group therapy in the treatment of patients with borderline personality disorder
(Table 1). STPP yielded significant improvements on measures of borderline-related
symptoms, general psychiatric symptoms and depression, and was as effective as
the interpersonal group therapy.
Bateman and Fonagy (1999, 2001) studied the effects of a psychoanalytically
oriented partial hospitalisation treatment for patients with borderline personality
disorder. The treatment lasted a maximum of 18 months, representing moderate-
length psychodynamic psychotherapy by the definition applied in this review.
According to the results, moderate-length psychodynamic psychotherapy was
significantly superior to a standard psychiatric care, both at the end of therapy and
at 18-month follow-up.

Cluster C personality disorders

In an RCT of Svartberg et al. (2004), moderate-length psychodynamic psychotherapy
of 40 sessions was compared to CBT (Table 1). Both psychodynamic psychotherapy
and CBT yielded significant improvements in patients with DSM-IV Cluster C
personality disorders (i.e. avoidant, compulsive or dependent personality disorder).
The improvements refer to symptoms, interpersonal problems and core personality
pathology. The results were stable at the 24-month follow-up. No significant
differences were found between moderate-length psychodynamic psychotherapy
and CBT with regard to efficacy.

Substance dependence (DSM IV Substance Dependence; ICD-10 F1X)

Four RCTs of psychodynamic psychotherapy in the treatment of substance
dependence are available (Table 1). Woody et al. (1983, 1990) studied the effects
of STPP and CBT given in addition to drug counselling in the treatment of opiate
dependence. STPP plus drug counselling yielded significant improvements on
measures of drug-related symptoms and general psychiatric symptoms. At 7-
month follow-up, STPP and CBT plus drug counselling were equally effective,
and both conditions were superior to drug counselling alone. In another RCT,
moderate-length psychodynamic psychotherapy of 26 sessions given in addition
to drug counselling was also superior to drug counselling in the treatment of
opiate dependence (Woody et al., 1995). By 6-month follow-up, most of the gains
made by the patients who had received psychodynamic therapy remained. In the
RCT of Crits-Christoph et al. (1999, 2001), moderate-length psychodynamic
psychotherapy of up to 36 individual sessions was combined with 24 sessions of
group drug counselling in the treatment of cocaine dependence. The combined
treatment yielded significant improvements and was as effective as CBT, which
was combined with group drug counselling as well. However, both CBT and
psychodynamic psychotherapy plus group drug counselling was not more effective
than group drug counselling alone. Furthermore, individual drug counselling
was significantly superior to both forms of therapy concerning measures of drug
abuse. With regard to psychological and social outcome variables, all treatments
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 851

Table 2—Controlled quasi-experimental effectiveness

studies of psychoanalytic therapy (PSA)

Study Disorder N (PA) Comparison group Outcome criteria Results

Dührssen and complex 125 treatment as usual: N = 100 in-patient days PSA > TAU,
Jorswieck, 1965 general population: N = 100 PSA> Controls
Rudolf et al., complex 44 psychodynamic therapy: N = 56 symptoms, interpersonal PSA > PP
1994 psychodynamic in-patient therapy: problems PSA> in-patient
N = 164 treatment
Sandell et al., complex 24 psychodynamic therapy: N = 100 symptoms, PSA> PP
1999, 2001 low-dose treatment: N = 27 sense of coherence, PA> low dose
untreated patients: N = 35 social adjustment PA> untreated
Rudolf et al., complex 32 psychodynamic therapy: N = 27 symptoms, PSA > PP
2004 interpersonal problems, (structural
structural changes changes)
PSA = PP
(symptoms,
interpersonal)

Note:
PSA: psychoanalytic therapy
PP: psychodynamic psychotherapy
TAU: treatment as usual

were equally effective (Crits-Christoph et al., 2001). In an RCT of Sandahl et al.

(1998), STPP and CBT were compared concerning their efficacy in the treatment
of alcohol abuse. STPP yielded significant improvements on measures of alcohol
abuse, which were stable at 15-month follow-up. STPP was significantly superior
to CBT in the number of abstinent days and in the improvement of general
psychiatric symptoms.

Effectiveness of psychoanalytic therapy in patients with

complex psychiatric disorders: evidence from naturalistic studies
As noted above, the method of RCT is not applicable to long-term psychodynamic
therapy of several years or to psychoanalytic therapy. For these methods of treatment,
effectiveness studies (naturalistic studies) are the appropriate method of research
(e.g. Seligman, 1995; Wallerstein, 1999; Leichsenring, 2004; Westen et al., 2004).
The NIMH in the USA has specifically called for more effectiveness research
(Krupnick et al., 1996). With regard to the appropriateness of effectiveness studies as
methods for testing whether a treatment works, the results presented by Shadish at al.
(2000) are relevant. Shadish at al. did not find a significant correlation between the
degree of clinical representativeness (e.g. RCTs vs. naturalistic studies) and the size
of the effects reported in studies of psychotherapy. These results are consistent with
those reported by Benson and Hartz (2000) and Concato et al. (2000) for the realm
of evidence-based medicine. Thus, the conclusion can be drawn that effectiveness
studies do not systematically overestimate the effects of psychotherapy.

Effect sizes of psychoanalytic therapy

With regard to psychoanalytic therapy, several effectiveness studies, which used
reliable and valid outcome measures, have provided evidence that psychoanalytic
852 FALK LEICHSENRING

therapy is effective in the field of clinical practice. The magnitude of change achieved
by a treatment can be assessed, for example, in the form of effect sizes. For example,
the d statistic proposed by Cohen (1988) gives the amount of change in units of
standard deviations. According to a convention proposed by Cohen, an effect size of
d = 0.80 can be regarded as a large effect. It corresponds to a change of 0.80 standard
deviations (see also Kazis et al., 1989).
Large effect sizes (≥ 0.80) for psychoanalytic therapy were reported, for
example, by the studies of Dührssen and Jorswieck (1965), Rudolf, Manz and Öri
(1994), Luborsky et al. (1999), Sandell et al. (1999, 2000, 2001), Brockmann et
al. (2001), Rudolf et al. (2004), and Leichsenring et al. (2005). These effect sizes
refer to symptoms, interpersonal problems, social adjustment, in-patient days and
other outcome criteria. In the study of Leuzinger-Bohleber et al. (2003), patients
retrospectively reported significant improvements in well-being and other aspects
of quality of life (e.g. self-esteem, relationship to others). Furthermore, the authors
showed a significant reduction in both sick leave days and in medical consultations
when comparing a period 1 year before treatment and 1 year after treatment (Beutel
et al., 2004). In a re-evaluation of the Menninger Psychotherapy Research Project
(Wallerstein, 1989), Blatt and Shahar (2004) addressed the question of the unique
nature and effectiveness of psychoanalysis. According to their results, psychoanalysis
contributed significantly to the development of adaptive interpersonal capacities
and to the reduction of maladaptive interpersonal behaviour, especially with more
self-reflective patients. Supportive-expressive therapy, by contrast, only yielded
a reduction of maladaptive interpersonal behaviour and only with dependent,
unreflective patients.

Quasi-experimental studies of psychoanalytic therapy: Superiority to control groups

The main argument against naturalistic studies refers to threats to internal validity,
that is, to the reduced possibility of controlling factors influencing outcome apart from
therapy. However, the internal validity of effectiveness studies can be improved by
quasi-experimental designs (Shadish et al., 2002). By definition, quasi-experimental
studies do not use random assignment; they use other principles to show that alternative
explanations of the observed effect are implausible. These principles include:
a) the identification and study of plausible threats to internal validity;
b) the use of additional design elements (e.g. observation at more pre-test time
points, additional comparison groups) or of statistical controls; and
c) coherent pattern matching, that is, prediction of complex patterns of results (e.g.
of non-equivalent dependent variables or of interactions) (Shadish et al., 2002).
In the schemes grading levels of evidence that regard RCTs as the gold standard
(CTFPHE, 1979; Cook et al., 1995; Guyatt et al., 1995; Chambless and Ollendick,
2001; Nathan and Gorman, 2002), quasi-experimental studies, or effectiveness
studies in general, are regarded as providing level III or, at best, level II
evidence.
However, as RCTs are regarded as the gold standard, these grading schemes
refer to the question of whether a treatment works under controlled (experimental)
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 853

conditions. They cannot be applied to the question of whether a treatment works

under naturalistic conditions. For effectiveness studies, which by definition cannot
use randomisation, levels of evidence must be defined by criteria different from
those of efficacy studies. Recently, a proposal was made to define levels of evidence
of effectiveness studies (Leichsenring, 2004). In analogy to the proposal made
by Nathan and Gorman (2002), six levels of evidence were defined taking the
methodological qualities of a naturalistic study into account (Leichsenring, 2004).
According to this proposal, a level-I naturalistic study which provides highest-level
evidence that a treatment works under the conditions of the field is a prospective
quasi-experimental study of high clinical representativeness, characterised by non-
random comparison groups, the matching or stratifying of groups, clear descriptions
of treatments, patients and their selection, the use of reliable and valid diagnostic
procedures and outcome measures, the use of additional design elements, coherent
pattern matching, the reporting of drop-outs, pre- and post-assessments, follow-up
studies, and the reporting of relevant statistical data. Clinical representativeness
is achieved by the selection of patients, therapists and treatments that are typical
for clinical practice (Shadish et al., 2000). Plausible threats to internal validity are
excluded by the use of additional design elements (e.g. observation at more pre-test
time points, additional comparison groups), statistical controls or coherent pattern
matching, that is, prediction of complex patterns of results (e.g. of non-equivalent
dependent variables or of interactions).
According to this proposal, the gold standard of naturalistic studies (effectiveness
studies) is a prospective quasi-experimental study of high clinical representativeness
that fulfils the aforementioned criteria. Lower-level naturalistic studies differ from
high-level studies in one or more of these aspects (Leichsenring, 2004). In order
to judge the effectiveness of a method of therapy (in a specific disorder) in clinical
practice, the existing studies have to be rated with regard to their levels of evidence
according to defined criteria. Furthermore, definitions must be given similar to
those of the guidelines proposed by Chambless and Hollon (1998) concerning the
number of studies regarded as necessary. For a treatment to be judged as ‘effective’
in clinical practice, at least two independent high-level studies may be regarded as
necessary, to be judged as ‘probably effective’ one high-level study may be regarded
as necessary.
The criteria of the TFPDPP (1995; Chambless and Hollon, 1998) require that a
treatment has proved to be superior to a control condition (placebo, no treatment)
or as effective as an already established treatment.
Several controlled quasi-experimental effectiveness studies of psychoanalytic
therapy fulfil one or other of these criteria. These studies included control groups,
for which comparability with the psychoanalytic treatment group was ensured by
measures of matching, stratifying or statistical control. They represent high-level
(level I) effectiveness studies according to the proposal made by Leichsenring
(2004). We will now review these studies.
Dührssen and Jorswieck (1965) studied the number of days spent in hospital
comparing the 5-year period before and after psychoanalytic therapy. They evaluated
the data of a large health insurance company (Allgemeine Ortskrankenkasse, AOK).
854 FALK LEICHSENRING

From these data, the authors drew a random sample of patients (N = 125) who were
treated with psychoanalytic therapy and compared them to a randomly drawn sample
of patients who had received no psychoanalytic treatment (N = 100). According
to the results, the patients treated with psychoanalytic therapy spent significantly
fewer days in hospital compared to the sample of patients who had received no
psychoanalytic treatment. The difference between the two groups corresponds to a
large effect size (d = 0.78).4 Furthermore, Dührssen and Jorswieck (1965) showed
that the random sample of patients treated with psychoanalytic therapy spent
significantly fewer days in hospital than a randomly drawn sample of subjects of the
general member population of the AOK health insurance company during a period
of five years after the termination of therapy. Apparently, psychoanalytic therapy led
to a reduction of costs in health services.
Rudolf et al. (1994) studied the outcome of psychoanalytic therapy, moderate-
length psychodynamic therapy (M = 60 sessions) and psychodynamic in-patient
treatment. The three treatment groups were comparable with regard to psychiatric
diagnoses and the severity of the disorder. According to the data of the applied
self-report instruments, 76% of the patients treated with psychoanalytic therapy
fulfilled the criterion of clinically significant improvement compared to only
55% of the patients treated with psychodynamic therapy and 50% of the patients
treated with psychodynamic in-patient treatment. In a recent study, Rudolf et
al. (2004) replicated their earlier results finding psychoanalytic therapy to be
significantly more effective than moderate-length psychodynamic therapy. This
was true concerning the very dimension of outcome for which a superiority of
psychoanalytic therapy is to be expected, that is, concerning structural changes
of personality.
Sandell and co-workers (1999, 2000, 2001) studied the effects of psychoanalytic
therapy and long-term psychodynamic therapy. Differences existing before therapy
were controlled for statistically. With the same base conditions, psychoanalytic
therapy achieved a large effect size of 1.55 concerning the reduction of symptoms. The
corresponding effect size of long-term psychodynamic therapy (M = 233 sessions)
was 0.60 (Sandell et al., 2001).5 The effects of psychoanalytic therapy increased
during the first and the second year after termination of therapy by about one-third;
on the contrary the effects of long-term psychodynamic therapy decreased slightly
during this period (Sandell et al., 1999). Furthermore, the authors investigated
how many patients fulfilled the criteria of clinical cases in the applied self-report
instruments. Accordingly, three years after termination of therapy more than 70% of
the patients who were treated with psychoanalytic therapy were no longer regarded
as clinical cases, in the group of patients treated with psychodynamic therapy the
corresponding percentage was 55%.
The results of these studies can be summarised as follows:

4
Evaluation by the author using the data published by Dührssen and Jorswieck (1965).
5
In two other instruments somewhat lower effect sizes were found for both psychoanalytic and
psychodynamic therapy (Sense of Coherence Scale: 1.18 and 0.40; Social Adjustment Scale: 0.40
and 0.44).
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 855

a) Psychoanalytic therapy yielded effect sizes that significantly exceeded the effects
of untreated or low-dose treated comparison groups (Dührssen and Jorswieck, 1965;
Sandell et al., 1999).
b) There is evidence that psychoanalytic therapy is significantly more effective than
shorter forms of psychodynamic psychotherapy (Rudolf et al., 1994, 2004; Sandell
et al., 1999, 2001).
c) According to the study of Rudolf et al. (2004) psychoanalytic therapy was
significantly more effective than shorter-term psychodynamic psychotherapy
concerning the very dimension of outcome for which a superiority of psychoanalytic
therapy is to be expected, that is concerning structural changes of personality.
d) These results refer to the treatment of patients with complex psychiatric disorders.
Ongoing controlled quasi-experimental studies of psychoanalytic therapy are
being carried out by Huber et al. (2001), Knekt and Lindfors (2004) and Leichsenring
et al. (2005).

Process–outcome relationship: Mechanisms of change

The RCTs presented above have focused on outcome, not on process variables of
psychodynamic psychotherapy. However, studies of psychotherapeutic processes
have provided data concerning mechanisms of change of psychodynamic therapy.
First, there is evidence that outcome of STPP is related to psychotherapeutic techniques
and therapist skilfulness (Crits-Christoph and Connolly, 1999; Messer, 2001):
accuracy of interpretation (Crits-Christoph et al., 1988), adherence of therapists’
interventions to the ‘plan’ (e.g. Messer et al., 1992), and competent delivery of
expressive, but not of supportive techniques (Barber et al., 1996) predicted outcomes
of STPP and of moderate-length psychodynamic psychotherapy. These findings
suggest that specific techniques of psychodynamic psychotherapy, as contrasted to
non-specific factors of psychotherapy, account for the outcome of psychodynamic
psychotherapy (Crits-Christoph and Connolly, 1999). There is less evidence that
frequency of psychodynamic techniques is related to outcome (Crits-Christoph and
Connolly, 1999). Second, there is evidence for an interaction of technique, outcome
and patient variables: frequency of transference interpretations seems to be associated
with both poor outcome and alliance in STPP of patients rated low on quality of
object relations (Hoglend and Piper, 1995; Connolly et al., 1999; Ogrodniczuk et al.,
1999; Piper et al., 2001). Although patients with high quality of object relations may
benefit from low to moderate levels of transference interpretations, results suggest
that they do not benefit from high levels of transference interpretations (Piper et
al., 1991a, 1991b; Connolly et al., 1999). Third, with regard to the common factor
of therapeutic alliance, there is some evidence that alliance is a modest predictor
of treatment outcome in STPP (Stiles et al., 1998; Crits-Christoph and Connolly,
1999; Barber et al., 2000; Messer, 2001). Accuracy of interpretation was found to
correlate significantly with therapeutic alliance in treatments of moderate length
(Crits-Christoph et al., 1993). Thus, one way in which accuracy of interpretation
may exert its effect could be by fostering the therapeutic alliance (Crits-Christoph
et al., 1993). Fourth, with regard to patient process variables, changes in the focus
856 FALK LEICHSENRING

of psychodynamic psychotherapy were shown to correlate with symptom change.

In the Crits-Christoph and Luborsky study (1990), symptom change correlated
with changes in the wish and response-of-self component of the Core Conflictual
Relationship Theme (CCRT). Thus, CCRT changes may mediate changes in
presenting symptoms in psychodynamic therapy. Piper et al. (2003) presented results
that suggest that expression of affect is a mediating variable of outcome in short-
term interpretive group therapy of patients with pathological grief. Fifth, with regard
to patients’ variables, the following variables were found to predict good outcome
of STPP: high motivation, realistic expectations, circumscribed focus, high quality
of object relations, absence of personality disorder (Hoglend, 1993; Messer, 2001;
Piper et al., 2001). Future research should address the question for which forms of
psychodynamic psychotherapy and for which forms of psychiatric disorders these
associations hold, and for which they do not. Barber et al. (2001), for example, did
not find a correlation of alliance to outcome of psychodynamic psychotherapy in the
treatment of cocaine dependence.

Discussion
In the first part of this article, the available evidence for the efficacy of psychodynamic
psychotherapy in specific psychiatric disorders was reviewed. The criteria for
empirically supported therapies proposed by Chambless and Hollon (1998) were
applied. Nevertheless, the limitations of the EST approach, which were described
above, should be kept in mind: RCTs can show nothing more and nothing less than
that a therapy works under controlled (experimental) conditions (Leichsenring,
2004). Effectiveness in the field can only be studied by effectiveness studies. For
the study of the active ingredients of psychotherapy, process studies are required
which link outcome to curative factors. Furthermore, for psychodynamic and
psychoanalytic psychotherapy, it is of interest to study changes beyond symptoms
and manifest behavior (Bond and Perry, 2004; Rudolf et al., 2004).
Under the requirements of ESTs, 22 studies were identified that provided
evidence for the efficacy of psychodynamic psychotherapy in specific psychiatric
disorders. The relatively short duration of the treatments (7 to 46 sessions) that were
applied in these studies reflects the fact that the method of RCT is not appropriate
for long-term psychotherapy or psychoanalysis of several years. According
to the 22 RCTs identified, there is at least one RCT demonstrating efficacy of
various models of short-term to moderate-length psychodynamic psychotherapy
in the following psychiatric disorders: major depression (4 RCTs), social phobia
(1 RCT), post-traumatic stress disorder (1 RCT), somatoform disorder (4 RCTs),
bulimia nervosa (3 RCTs), anorexia nervosa (2 RCTs), borderline personality
disorder (2 RCTs), Cluster C personality disorders (1 RCT) and substance-related
disorders (4 RCTs).
According to the criteria proposed by the TFPDPP (1995), at least two studies of
independent research groups are required for a treatment to be regarded as efficacious
(Chambless and Hollon, 1998). However, it is required that the same method of
therapy has been applied in these studies. In the studies presented above, this is not
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 857

the case: there are no two studies of independent research groups in which the same
form of STPP was applied to the same psychiatric disorder. This is why STPP was
judged as ‘probably efficacious’ by the TFPDPP (1995). Although 14 of the 22
studies included in this review had not been included by the TFPDPP report (1995),
this judgement still seems to hold.6
However, it is important to take psychotherapy process research into account:
the caveat of Ablon and Jones (2002, p. 780)—‘Brand names of therapy can be
misleading’—may also apply to psychodynamic psychotherapy. Ablon and Jones
(2002) recently compared CBT and interpersonal therapies (IPT) as they were applied
in the NIMH treatment of depression study (e.g. Elkin et al., 1989). According to
the results, therapists of both the CBT and IPT treatment adhered most strongly to
the ideal prototype of CBT. In addition, adherence to the CBT prototype yielded
more positive correlations with outcome measures across both types of treatment.
With this finding in mind, it is no longer surprising that IPT and CBT were equally
effective in the NIMH treatment of depression study. In an earlier study, Goldsamt et
al. (1992) compared a demonstration session carried out by Beck, Meichenbaum and
Strupp with the same patient. They found as many significant differences between
Beck and Meichenbaum as between Meichenbaum and Strupp. Most differences
were found between Beck and Strupp. Meichenbaum was somewhere between
Beck and Strupp, and raters could not differentiate between Meichenbaum and
Strupp, although they represent different therapeutic approaches, that is, CBT vs.
psychodynamic therapy.
According to these results, the question of whether the ‘different’ models of
psychodynamic psychotherapy differ among each other empirically is open to
further research. This question cannot be answered by comparing the manuals with
regard to the included interventions. Empirical studies of actual therapy sessions are
required. In a review of empirical studies, Blagys and Hilsenroth (2000) identified
seven features that were significantly more frequently observed in psychodynamic,
psychodynamic-interpersonal or interpersonal psychotherapy than in CBT.7 However,
their review did not address the question of whether different models of STPP differ
among each other and from IPT. The features that were found to discriminate IPT from
CBT were characteristic of STPP as well (see Tables 1–7 of Blagys and Hilsenroth,
2000, pp. 170–84). Comparing prototypical sessions of different (manual-guided)
variants of psychodynamic psychotherapy empirically would be a very interesting
and promising project of research. Other forms of therapy (e.g. IPT, CBT) should be
included. For this kind of research, methods like that used by Ablon and Jones (2002)
can be very useful. Studies addressing the problem of similarity or dissimilarity of

6
Although it is true that STPP may include ‘different’ models (e.g. according to Luborsky, or Horowitz),
this applies to CBT as well. The forms of CBT applied in the studies that were accepted by the TFPDPP
as providing empirical evidence for the efficacy of CBT in generalised anxiety disorder are similar, but
also not identical to each other (see Chambless and Gillis, 1993, p. 249). Apparently, the question of
heterogeneity or similarity is not only a problem of psychodynamic psychotherapy.
7
However, if two or more methods of therapy can be reliably discriminated or identified on the basis
of these features is open to further research: Significant mean differences are a necessary, but not a
sufficient condition for this purpose.
858 FALK LEICHSENRING

treatments are relevant for the question of whether (some of) the ‘different’ models of
psychodynamic psychotherapy are empirically close enough to be lumped together. If
this is the case, empirical evidence for one model of psychodynamic psychotherapy
is valid for another model that has proved to be similar enough. However, adherence
to a treatment manual can be achieved with considerable differences in the underlying
interpersonal processes, and it is these processes that are related to outcome (Henry
et al., 1990, 1993). Differences between therapists should also be studied. Crits-
Christoph and Mintz (1991), for example, have shown that individual therapists
applying the same form of therapy differed concerning their efficacy. Thus, in a second
step the factors that may be identified to characterise specific forms of psychodynamic
psychotherapy should be related to outcome. At present, an RCT is being performed
comparing psychodynamic psychotherapy (supportive-expressive therapy according
to Luborsky, 1984) to CBT (according to Beck and Emery, 1985) in the treatment of
generalised anxiety disorder (Leichsenring et al., 2002). In this study, the prototypical
psychotherapeutic interventions and their relation to outcome are examined across
models of therapy. Furthermore, changes in the core conflictual relationship theme
(Luborsky, 1984) are studied both in CBT and supportive-expressive therapy, and
their relation to outcome is examined.
As there is at least one RCT of a specific form of psychodynamic psychotherapy
in the psychiatric disorders listed above, only one further study applying one
of the already applied forms of psychodynamic psychotherapy demonstrating
efficacy is required for the respective form of psychodynamic psychotherapy to
be judged as efficacious in the treatment of the respective disorder. This applies,
for example, to psychodynamic psychotherapy according to Horowitz (1976) or
Shapiro and Firth (1985) in the treatment of (geriatric) depression, to Luborsky’s
(1984) supportive-expressive therapy in bulimia nervosa or opiate/cocaine
dependence, or to Shapiro and Firth’s (1985) interpersonal-psychodynamic
therapy in the treatment of irritable bowel syndrome, functional dyspepsia or
depression (see Table 1).
On the other hand, it is important to realise for which psychiatric disorders there
is not even one RCT of psychodynamic psychotherapy. This is true for dissociative
disorder or for some specific forms of personality disorders (e.g. compulsive, avoidant
or narcissistic). Surprisingly, this is also true for some of the anxiety disorders, for
example, for panic disorder (only one study of STPP combined with medication, Wiborg
and Dahl, 1996). This is the more surprising as anxiety is one of the central concepts of
psychoanalytic and psychodynamic theory and therapy (Zerbe, 1990). With regard to
generalised anxiety disorder, the study of Durham et al. (1994) comparing STPP and
CBT did not fulfil the inclusion criteria of this review (no manual for STPP). In that
study, STPP and CBT were not equally carefully carried out (e.g. no specific training
of therapists, no checks of adherence and competence for STPP).8
8
In this study, STPP served as a kind of control group, as a ‘strawman’ as Smith et al. (1980, p. 119)
put it: ‘A comparison therapy might be set up as a kind of strawman over which the favored therapy
would prevail. The comparison therapy (often an ‘insight therapy’) would be treated with fairly obvious
disdain and would be given not much opportunity for success’—the investigator allegiance effect
(Smith et al., 1980; Luborsky et al., 1999).
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 859

With regard to the treatment of borderline personality disorder, an RCT comparing

Transference-Focused Psychotherapy, Dialectical Behavior Therapy and supportive
psychotherapy is presently being carried out (Clarkin et al., 2004). Winston et al. (1994)
compared STPP and brief adaptive psychotherapy in the treatment of predominantly
Cluster C personality disorders. Both forms of therapy were significantly superior to a
waiting list control group concerning target problems, general psychiatric symptoms
and social adjustment. No differences between the two forms of therapy were found.
Results were maintained at follow-up after an average of 1.5 years. In another RCT of
the same research group, STPP yielded significant improvements in target problems,
general psychiatric symptoms and social adjustment, and was as effective as a specific
form of supportive therapy, both at the end of therapy and at 6-month follow-up
(Hellerstein et al., 1998). Again, predominantly patients with Cluster C personality
disorders were treated, but patients with Cluster A (e.g. paranoid) and B (e.g. borderline)
personality disorders were also included. The inclusion of patients with Cluster A and
B personality disorders allows for a less unambiguous interpretation of the results.
This applies to both the study of Winston et al. (1994) and the study of Hellerstein et
al. (1998). However, it can be supposed that patients with Cluster A and B personality
disorders have less favourable results. Thus the results of Winston et al. (1994) and
Hellerstein et al. (1998) can be assumed to represent conservative estimates for the
outcome of patients with Cluster C personality disorders. A separate evaluation of these
studies for Cluster C personality disorders only could be helpful. It is questionable
whether short-term therapies are appropriate for the treatment of (severe) personality
disorders (Bateman and Fonagy, 2000). Perry et al. (1999) estimated the length of
treatment necessary for patients to no longer meet the full criteria for a personality
disorder (recovery). According to these estimates, 50% of patients with personality
disorder would recover after 1.3 years or 92 sessions, and 75% after 2.2 years or about
216 sessions (Perry et al., 1999, p. 1318).
According to the results of this review, further research of psychodynamic
psychotherapy in treating specific psychiatric disorders is necessary, including
assessment of both outcome and active ingredients of psychodynamic psychotherapy
in treating these disorders. Comparison across different models of psychodynamic
psychotherapy and other forms of therapy, for example, CBT or IPT, should be
included as well. Furthermore, the question should be addressed of how effective
the methods of therapy, which have been proved to work in RCTs, are in the field
(effectiveness studies). The NIMH in the USA has specifically called for more
effectiveness research (Krupnick et al., 1996).
Psychotherapy has been shown to be a cost-effective treatment (Gabbard et al.,
1997). This applies to psychodynamic psychotherapy and psychoanalytic therapy as
well (Dührssen and Jorswieck, 1965; Gabbard, 1997; Guthrie et al., 1999; Burnand
et al., 2002; Abbass, 2003; Creed et al., 2003; Beutel et al., 2004).
With regard to psychoanalytic therapy, controlled quasi-experimantal
effectiveness studies provided evidence that psychoanalytic therapy is (1) more
effective than no treatment or treatment as usual, and (2) more effective than shorter
forms of psychodynamic therapy. However, further quasi-experimental controlled
effectiveness studies of psychoanalytic therapy are necessary.
860 FALK LEICHSENRING

Translations of summary
Sind psychodynamische und psychoanalytische Therapien wirksam? Ein Forschungsbericht über
empirische Daten. Es gibt einen Bedarf an empirischer Forschung zu den Ergebnissen psychodynamischer
und psychoanalytischer Therapie. Sowohl der Ansatz empirisch gestützter Therapie (EST) als auch die
Prozeduren der evidence-based-medicine (EBM) weisen allerdings schwerwiegende Einschränkungen auf,
indem sie randomisierte kontrollierte Studien (RCTs) als absoluten Maßstab setzen. Nach einer kritischen
Diskussion dieser Ansätze gibt dieser Artikel zunächst eine Übersicht über die empirische Evidenz zur
Wirksamkeit psychodynamischer Psychotherapie bei spezifischen psychiatrischen Störungen. Das Ziel
dieser Übersicht ist es, zu identifizieren, für welche psychiatrischen Störungen RCTs zu spezifischen
Modellen psychodynamischer Therapie vorliegen und für welche sie fehlen. Auf diese Art und Weise liefert
die vorliegende Übersicht eine Basis für die Planung weiterer Forschung. Zusätzlich werden Ergebnisse
aus der Prozessforschung psychodynamischer Therapie vorgestellt. Da die Methodologie von RCTs nicht
angemessen ist für psychoanalytische Therapie, wird in einem zweiten Teil eine Übersicht über Studien zur
Wirksamkeit psychoanalytischen Therapie gegeben. Studien psychodynamischer Therapie, die zwischen
1960 und 2004 veröffentlicht worden sind, wurden mit einer computerunterstützten Suche über MEDLINE,
PsycINFO and Current Contents identifiziert. Zusätzlich wurden Lehrbücher und Zeitschriftenartikel
herangezogen. Ergebnisse: 22 RCTs zur Wirksamkeit psychodynamischer Therapie wurden identifiziert,
von denen 64% in dem Bericht von Chambless und Hollon (1998) nicht berücksichtigt wurden. Nach
den Ergebnissen dieser Übersicht wurde für die folgenden psychiatrischen Störungen mindestens ein RCT
gefunden, der Evidenz für die Wirksamkeit psychodynamischer Therapie liefert: Depressive Störungen (4
RCTs), Angststörungen (1 RCT), posttraumatische Belastungsstörungen (1 RCT), somatoforme Störungen (4
RCTs), Bulimia nervosa (3 RCTs), Anorexia nervosa (2 RCTs), Borderline-Pesönlichkeitsstörung (2 RCTs),
Cluster C-Persönlichkeitsstörungen (1 RCT) und Substanz-Abhängigkeit (4 RCTs). Nach den vorliegenden
Ergebnissen der Prozessforschung ist der Therapieerfolg bei psychodynamischer Therapie verbunden mit der
kompetenten Anwendung therapeutischer Techniken und der Entwicklung einer hilfreichen Beziehung. Im
Hinblick auf psychoanalytische Therapie liefern kontrollierte quasi-experimentelle naturalistische Studien
Evidenz dafür, dass psychoanalytische Therapie (1.) wirksamer ist als keine Behandlung oder ein treatment-
as-usual, und (2.) wirksamer ist als kürzere Formen psychodynamischer Therapie. Schlussfolgerungen für
weitere Forschung werden gezogen.

¿Son eficaces las terapias psicodinámicas y psicoanalíticas? Una reseña de datos empíricos. Existe una
necesidad de investigar empíricamente los resultados de las terapias psicodinámica y psicoanalítica. Sin
embargo tanto el enfoque de las terapias con apoyo empírico (TAE; empirically supported therapies, EST)
como los procedimientos médicos basados en evidencias (MBE; evidence based medicine, EBM) tienen
graves limitaciones, lo cual convierte a las pruebas de control aleatorio (randomized controlled trials,
RCTs) un elemento indispensable. Tras una discusión crítica de este enfoque, se reseñan pruebas empíricas
sobre la eficacia de la psicoterapia psicodinámica en trastornos psiquiátricos específicos. La reseña
busca identificar qué trastornos psiquiátricos cuentan con RCTs de modelos específicos de psicoterapia
psicodinámica y cuáles no, aportando así una base para futuras investigaciones. Se presentan además
resultados de una investigación sobre el proceso de la psicoterapia psicodinámica. Al no ser apropiada
la metodología de RCTs para la terapia psicoanalítica, se revisan además estudios de efectividad de la
terapia psicoanalítica. Mediante una búsqueda computerizada en Medline, PsycINFO y Current Contents,
se identifican investigaciones sobre psicoterapia psicodinámica publicadas entre 1960 y 2004. Además
se utilizan manuales y artículos de revistas. Se identifican 22 RCTs de los cuales 64% no habían sido
incluidos en el informe de Chambless y Hollon (1998). De acuerdo a los resultados se contó con por
lo menos un RCT, que evidenciaba la eficacia de la psicoterapia psicodinámica, para los siguientes
trastornos psiquiátricos: depresión (4 RCTs), angustia (1 RCT), stress post-traumático (1 RCT), trastornos
somatoformes (4 RCTs), bulimia nerviosa (3 RCTs), anorexia nerviosa (2 RCTs), personalidad borderline
(2 RCTs), trastornos de la personalidad Cluster C (1RCT) y los relacionados con el abuso de sustancias (4
RCTs). Según la investigación realizada, en psicoterapia psicodinámica los resultados están relacionados
con la aplicación competente de técnicas terapéuticas y con el desarrollo de una alianza terapéutica. Los
estudios de efectividad con control quasi-experimental evidencian que la terapia psicoanalítica es (1) más
efectiva que el no tratamiento o el tratamiento habitual, y (2) más efectiva que formas más breves de terapia
psicodinámica. Se extraen conclusiones para futuras investigaciones.

Les thérapies psychodynamiques et psychanalytiques sont-elles efficaces ? Une revue des données
empiriques. Il y a nécessité pour de la recherche empirique portant sur les résultats des thérapies
psychodynamiques et psychanalytiques. Cependant, tant l’approche des thérapies empiriquement validées
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 861

(empirically supported therapies, EST) que les procédures de la « médecine des faits établis » (evidence-
based medicine, EBM) présentent de sévères limitations, tant et si bien que les essais contrôlés randomisés
(ECR) sont d’une absolue nécessité. Après une discussion critique de cette approche, l’auteur passe en
revue les preuves empiriques de l’efficacité de la psychothérapie psychodynamique dans des troubles
psychiatriques spécifiques. Cette revue de la littérature a pour but d’identifier pour quel trouble psychiatrique
les ECR portant sur des modèles spécifiques de psychothérapie psychodynamique sont disponibles et
pour quel trouble ils manquent, offrant ainsi une base à la planification de nouvelles recherches. L’article
présente en plus des résultats de la recherche actuelle sur la psychothérapie psychodynamique. Étant donné
que la méthodologie des ECR n’est pas adaptée à la thérapeutique psychanalytique, l’auteur a ajouté dans
sa revue des études sur l’efficacité de la thérapie psychanalytique. Les études portant sur la psychothérapie
psychodynamique publiées entre 1960 et 2004 ont été repérées à l’aide d’une recherche informatique utilisant
les base de données MEDLINE, PsycINFO et Current Contents. L’auteur s’est également aidé de manuels
et d’articles de revues. Vingt-deux ECR ont été identifiés, dont 64% n’avaient pas été inclus dans le rapport
de Chambless et Hollon (1998). Selon ces résultats, au moins un ECR ayant fait la preuve de l’efficacité de
la thérapie psychodynamique a été identifié pour les troubles psychiatriques suivants : troubles dépressifs (4
ECR), troubles anxieux (1 ECR), syndrome de stress post-traumatique (1 ECR), troubles somatoformes (4
ECR), boulimie (3 ECR) et anorexie mentale (2 ECR), troubles de la personnalité de type état-limite (2 RCT),
trouble personnalité multiple (1 ECR) et troubles liés à l’abus de substances toxiques (4 ECR). Selon les
résultats de cette recherche bibliographique, l’efficacité de la psychothérapie psychodynamique est liée à la
compétence dans l’utilisation des techniques thérapeutiques et au développement d’une alliance d’aide. En
ce qui concerne la thérapeutique psychanalytique, les études d’efficacité quasi-expérimentales et contrôlées
montrent que la thérapie psychanalytique, premièrement, est plus efficace que l’absence de tout traitement
et d’un traitement standard, et deuxièmement, est plus efficace que les thérapies psychodynamiques plus
brèves. Des conclusions formulées ouvrent sur la perspective de nouvelles recherches.

Le terapie psicodinamiche e psicoanalitiche sono efficaci? Una review dei dati empirici. Nelle terapie
psicodinamica e psicoanalitica si sente l’esigenza di una ricerca sui risultati empirici. Tuttavia sia l’approccio
delle terapie a base empirica (EST) sia le procedure della medicina evidence-based (EBM) comportano
gravi limitazioni, rendendo il controllo casuale delle prove (RTC) un elemento necessario. Dopo aver
discusso criticamente tale approccio, l’articolo prende in esame le prove empiriche dell’efficacia della
psicoterapia psicodinamica in disturbi psichiatrici specifici. Questa review tende a identificare per quali
specifici disturbi gli RCT di modelli specifici di psicoterapia psicodinamica siano disponibili e per quali
manchino, fornendo così una base su cui programmare l’ulteriore ricerca. Inoltre sono presentati i risultati
della ricerca sul processo di psicoterapia psicodinamica. Poiché la metodologia degli RCT non si adatta
alla terapia psicoanalitica, gli studi sull’efficacia di questa terapia sono stati esaminati a parte. Gli studi
sulla psicoterapia psicodinamica pubblicati tra il 1960 e il 2004 sono stati identificati mediante una ricerca
computerizzata, con l’utilizzo di MEDLINE, PsycINFO e Current Contents. Sono stati inoltre utilizzati
libri di testo e articoli apparsi su riviste. Sono stati identificati ventidue RCT, il 64 % dei quali non era stato
incluso nella relazione di Chambless e Hollon (1998). In base ai risultati, si è identificato almeno un RTC
che fornisce la prova dell’efficacia della psicoterapia psicodinamica per i seguenti disturbi psichiatrici:
disturbi di natura depressiva (4 RCT), disturbi da ansia (1 RCT), disturbi da stress postraumatico (1 RCT),
disturbi somatoformi (4 RCT), bulimia nervosa (3 RCT), anoressia nervosa (2 RCT), disturbi borderline
della personalità (2 RCT), disturbi della personalità Cluster C (1 RCT) e disturbi da abuso di sostanze
(4 RCT). Stando ai risultati della ricerca in corso, nella terapia psicodinamica il risultato è correlato alla
somministrazione competente delle tecniche terapeutiche e allo sviluppo di un’alleanza di aiuto. Quanto
alla terapia psicoanalitica, gli studi sulla sua efficacia controllati quasi sperimentalmente forniscono la
prova che la terapia psicoanalitica è (1) più efficace dell’assenza di trattamento o del trattamento consueto,
(2) più efficace di forme più brevi di terapia psicodinamica. Le conclusioni sono tratte in funzione della
ricerca futura.

References
Abbass AA (2003). The cost-effectiveness of short-term dynamic psychotherapy. Expert Rev
Pharmacoecon Outcomes Res 3:535–9.
Abbass AA, Hancock JT, Henderson J, Kisley S (2004). Short-term psychodynamic
psychotherapies for common mental disorders. (Protocol for a Cochrane Review.) In: The
Cochrane Library, Issue 2, 2004. Chichester, UK: Wiley.
862 FALK LEICHSENRING

Ablon JS, Jones EE. (2002). Validity of controlled cinical trials of psychotherapy: Findings
from the NIMH Treatment of Depression Collaborative Research Program. Am J Psychiatry
159:775–83.
Bachar E, Latzer Y, Kreitler S, Berry EM (1999). Empirical comparison of two psychological
therapies: Self psychology and cognitive orientation in the treatment of anorexia and bulimia.
J Psychother Pract Res 8:115–28.
Barber JP, Crits-Christoph P, Luborsky L (1996). Effects of therapist adherence and competence
on patient outcome in brief dynamic therapy. J Consult Clin Psychol 64:619–22.
Barber JP, Connolly MB, Crits-Christoph P, Gladis L, Siqueland L (2000). Alliance predicts patients’
outcome beyond in-treatment change in symptoms. J Consult Clin Psychol 68:1027–32.
Barber JP, Luborsky L, Gallop R, Crits-Christoph P, Frank A, Weiss RD, et al. (2001). Therapeutic
alliance as a predictor of outcome and retention in the National Institute on Drug Abuse
Collaborative Cocaine Treatment Study. J Consult Clin Psychol 69:119–24.
Barkham M, Rees A, Shapiro DA, Stiles WB, Agnew RM, Halstead J, et al. (1996). Outcomes of
time-limited psychotherapy in applied settings: Replicating the second Sheffield Psychotherapy
Project. J Consult Clin Psychol 64:1079–85.
Barth D (1991). When the patient abuses food. In: Jackson H, editor. Using self psychology in
psychotherapy, p. 223–42. Northvale, NJ: Jason Aronson.
Bateman A (1995). The treatment of borderline patients in a day hospital setting. Psychoanal
Psychother 91:3–16.
Bateman A, Fonagy P (1999). The effectiveness of partial hospitalization in the treatment of
borderline personality disorder: A randomized controlled trial. Am J Psychiatry 156:1563–9.
Bateman A, Fonagy P (2000). Effectiveness of psychotherapeutic treatment of personality
disorder. Br J Psychiatr 177:138–43.
Bateman A, Fonagy P (2001). Treatment of borderline personality disorder with psychoanalytically
oriented partial hospitalization: An 18-month follow-up. Am J Psychiatry 158:36–42.
Beck AT, Emery G (1985). Anxiety disorders and phobias: A cognitive perspective. New York:
Basic Books.
Benson K, Hartz AJ (2000). A comparison of observational studies and randomized, controlled
trials. N Engl J Med 342:1878–86.
Beutel M, Rasting M, Stuhr U, Rüger B, Leuzinger-Bohleber M (2004). Assessing the impact of
psychoanalyses and long-term psychoanalytic therapies on health care utilization and costs.
Psychother Res 14:146–60.
Beutler LE (1998). Identifying empirically supported treatments: What if we didn’t? J Consult Clin
Psychol 66:113–20.
Blagys MD, Hilsenroth MJ (2000). Distinctive features of short-term psychodynamic-interpersonal
psychotherapy: A review of the comparative psychotherapy process literature. Clin Psychol:
Science and Practice 7:167–88.
Blatt S (1995). Why the gap between psychotherapy research and clinical practice: A response to
Barry Wolfe. J Psychother Integration 5:73–6.
Blatt S, Shahar G (2004). Psychoanalysis—With whom, for what, and how? Comparisons with
psychotherapy. J Am Psychoanal Assoc 52:393–447.
Bond M, Perry C (2004). Long-term changes in defense styles with psychodynamic psychotherapy
for depressive, anxiety and personality disorders. Am J Psychiatry 161:1665–71.
Bögels S, Wijts P, Sallaerts S (2003). Analytic psychotherapy versus cognitive-behavioral therapy
for social phobia. Paper presented at: European Congress for Cognitive and Behavioural
Therapies, September, Prague.
Brockmann J, Schlüter T, Eckert J (2001). Die Frankfurt–Hamburg Langzeit- Psychotherapiestudie
—Ergebnisse der Untersuchung psychoanalytisch orientierter und verhaltenstherapeutischer
Langzeit-Psychotherapien in der Praxis niedergelassener Psychotherapeuten [The Frankfurt–
Hamburg study of psychotherapy—Results of the study of psychoanalytically oriented and
behavioral long-term therapy]. In Stuhr U, Leuzinger-Bohleber M, Beutel M, editors. Langzeit-
Psychotherapie. Perspektiven für Therapeuten und Wissenschaftler [Long-term psychotherapy.
Perspectives for therapists and researchers]. Stuttgart: Kohlhammer, p. 271–6.
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 863

Brom D, Kleber RJ, Defares PB (1989). Brief psychotherapy for posttraumatic stress disorders.
J Consult Clin Psychol 57:607–12.
Bruch H (1973). Eating disorders: Obesity, anorexia nervosa, and the person within. New York:
Basic Books.
Burnand Y, Andreoli A, Kolatte E, Venturini A, Rosset N (2002). Psychodynamic psychotherapy
and clomipramine in the treatment of major depression. Psychiatr Serv 53:585–90.
Chambless DL, Gillis MM (1993). Cognitive therapy of anxiety disorders. J Consult Clin Psychol
61:248–60.
Chambless DL, Hollon SD (1998). Defining empirically supported treatments. J Consult Clin
Psychol 66:7–18.
Chambless DL, Ollendick TH (2001). Empirically supported psychological interventions:
Controversies and evidence. Ann Rev Psychol 52:685–716.
Clarke M, Oxman AD (2003). Cochrane Reviewer’s Handbook 4.1.6 (updated January 2003). In
The Cochrane Library Issue 1. Oxford: Update Software. Updated quarterly.
Clarkin JF, Levy KN, Lenzenweger MF, Kernberg OF (2004). The Personality Disorders Institute/
Borderline Personality Disorder Research Foundation randomized control trial for borderline
personality disorder: Rationale, methods, and patient characteristics. J Personality Disord
18:52–72.
Cohen J (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence
Erlbaum.
Concato J, Shah N, Horwitz RI (2000). Randomized, controlled trials, observational studies, and
the hierarchy of research designs. N Engl J Med 342:1887–92.
Connolly MB, Crits-Christoph P, Shappell S, Barber JP, Luborsky L, Shaffer C (1999). Relation
of transference interpretation to outcome in the early sessions of brief supportive-expressive
psychotherapy. Psychother Res 9:485–95.
Cook DJ, Guyatt GH, Laupacis A, Sacket DL, Goldberg RJ (1995). Clinical recommendations
using levels of evidence for antithrombotic agents. Chest 108:227S–230S.
Creed F, Fernandes L, Guthrie E, Palmer S, Ratcliffe J, Read N, et al. (2003). The cost-effectiveness
of psychotherapy and paroxetine for severe irritable bowel syndrome. Gastroenterology 124:
303–17.
Crisp AH (1980). Anorexia nervosa: Let me be. London: Academic Press.
Crits-Christoph P, Connolly B (1999). Alliance and technique in short-term dynamic therapy. Clin
Psychol Rev 19:687–704.
Crits-Christoph P, Luborsky L (1990). Changes in CCRT pervasiveness during psychotherapy.
In Luborsky L, Crits-Christoph P, editors. Understanding transference: The CCRT method,
p. 133–46. New York: Basic Books.
Crits-Christoph P, Mintz J (1991). Implications of therapist effects for the design and analysis of
comparative studies of psychotherapies. J Consult Clin Psychol 59:20–26.
Crits-Christoph P, Cooper A, Luborsky L (1988). The accuracy of therapists’ interpretations and
the outcome of dynamic psychotherapy. J Consult Clin Psychol 56:490–5.
Crits-Christoph P, Barber JP, Kurcias J (1993). The accuracy of therapists’ interpretations and the
development of the therapeutic alliance. Psychother Res 3:25–35.
Crits-Christoph P, Siqueland L, Blaine J, Frank A, Luborsky L, Onken LS, et al. (1999). Psychosocial
treatments for cocaine dependence: National Institute on Drug Abuse Collaborative Cocaine
Treatment Study. Arch Gen Psychiatr 56:493–502.
Crits-Christoph P, Siqueland L, McCalmont E, Weiss RD, Gastfriend DR, Frank A, et al. (2001)
Impact of psychosocial treatments on associated problems of cocaine-dependent patients.
J Consult Clin Psychol 69:825–30.
CTFPHE (Canadian Task Force on the Periodic Health Examination) (1979). The periodic health
examination. Can Med Assoc J 121:1193–254.
Dare C (1995). Psychoanalytic psychotherapy (of eating disorders). In: Gabbard GO, editor. Treatment
of psychiatric disorders, p. 2129–51. Washington, DC: American Psychiatric Press.
864 FALK LEICHSENRING

Dare C, Eisler I, Russell G, Treasure J, Dodge L (2001). Psychological therapies for adults with
anorexia nervosa: Randomised controlled trial of out-patient treatments. Br J Psychiatr 178:216–
21.
Davanloo H (1980). Short-term dynamic psychotherapy. New York: Jason Aronson.
Dührssen A, Jorswieck E (1965). Eine empirisch-statistische Untersuchung zur
Leistungsfähigkeit psychoanalytischer Behandlung [An empirical-statistical study of the
effectiveness of psychoanalytic treatment]. Nervenarzt 36:166–9.
Durham RC, Murphy T, Allan T, Richard K, Treliving LR, Fenton GW (1994). Cognitive therapy,
analytic psychotherapy and anxiety management training for generalised anxiety disorder. Br
J Psychiatr 165:315–23.
de Jonghe F, Kool S, van Aalst G, Dekker J, Peen J (2001). Combining psychotherapy and
antidepressants in the treatment of depression. J Affect Disord 64:217–29.
Elkin I, Shea T, Watkins J, Imber S, Stosky S, Collins J, et al. (1989). National Institute of Mental
Health Treatment of Depression Collaborative Research Program: General effectiveness of
treatments. Arch Gen Psychiatr 46:971–82.
Fairburn C, Kirk J, O’Connor M, Cooper PJ (1986). A comparison of two psychological treatments
for bulimia nervosa. Behav Res Ther 24:629–43.
Fairburn CG, Norman PA, Welch SL, O’Connor ME, Doll HA, Peveler RC (1995). A prospective
study of outcome in bulimia nervosa and the long-term effects of three psychological
treatments. Arch Gen Psychiatr 52:304–12.
Fonagy P (1999). Process and outcome in mental health care delivery: A model approach to
treatment evaluation. Bull Menninger Clin 63:288–304.
Foulkes SH (1964). Therapeutic group analysis. London: Allen & Unwin.
Gabbard GO (2004). Long-term psychodynamic psychotherapy. Washington, DC: American
Psychiatric Publishing.
Gabbard GO, Lazar SG, Hornberger J, Spiegel D (1997). The economic impact of psychotherapy:
A review. Am J Psychiatry 154:147–55.
Gallagher-Thompson DE, Steffen AM (1994). Comparative effects of cognitive-behavioral and
brief psychodynamic psychotherapies for depressed family caregivers. J Consult Clin Psychol
62:543–9.
Gallagher-Thompson DE, Hanley-Peterson P, Thompson LW (1990). Maintenance of gains versus
relapse following brief psychotherapy for depression. J Consult Clin Psychol 58:371–4.
Garner DM, Rockert W, Davis R, Garner MV, Olmsted MP, Eagle M (1993). Comparison of
cognitive-behavioral and supportive-expressive therapy for bulimia nervosa. Am J Psychiatry
150:37–46.
Geist RA (1989). Self psychological reflections on the origins of eating disorders. J Am Acad
Psychoanal 17:5–27.
Goldsamt LA, Goldfried MR, Hayes AM, Jerr S (1992). Beck, Meichenbaum and Strupp: A
comparison of three therapists on the dimension of therapist feedback. Psychother 29:167–
76.
Goodsitt A (1985). Self psychology and the treatment of anorexia nervosa. In: Garner DM,
Garfinkel DE, editors. Handbook of psychotherapy for anorexia nervosa and bulimia, p. 55–
82. New York: Guilford.
Gowers D, Norton K, Halek C, Crisp AH (1994). Outcome of outpatient psychotherapy in a random
allocation treatment study of anorexia nervosa. Int J Eating Disord 15:165–77.
Gunderson JG, Gabbard G (1999). Making the case for psychoanalytic therapies in the current
psychiatric environment. J Am Psychoanal Assoc 47:679–704.
Guthrie E (2000). Psychotherapy for patients with complex disorders and chronic symptoms: The
need for a new research paradigm. Br J Psychiatr 177:131–7.
Guthrie E, Creed F, Dawson D, Tomenson B (1991). A controlled trial of psychological treatment
for the irritable bowel syndrome. Gastroenterology 100:450–57.
Guthrie E, Moorey J, Margison F, Barker H, Palmer S, McGrath G, et al. (1999). Cost-effectiveness
of brief psychodynamic-interpersonal therapy in high utilizers of psychiatric services. Arch
Gen Psychiatr 56:519–26.
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 865

Guyatt GH, Sacket DL, Sinclair JC, Hayward R, Cook DJ, Cook R (1995). User’s guides to the
medical literature. IX. A method for grading health care recommendations. J Am Med Assoc
275:1800–4.
Hamilton J, Guthrie E, Creed F, Thompson D, Tomenson B, Bennett R, et al. (2000). A randomized
controlled trial of psychotherapy in patients with chronic functional dyspepsia. Gastroenterology
119:661–9.
Hellerstein DJ, Rosenthal RN, Pinsker H, Wallner Samstag L, Muran JC, Winston A (1998). A
randomized prospective study comparing supportive and dynamic therapies: Outcome and
alliance. J Psychother Pract Res 7:261–71.
Henry WP (1998). Science, politics, and the politics of science: The use and misuse of empirically
validated treatment research. Psychother Res 8:126–40.
Henry WO, Schacht TE, Strupp HH (1990). Patient and therapist introject, interpersonal process
and differential psychotherapy outcome. J Consult Clin Psychol 58:768–74.
Henry WO, Strupp HH, Butler SF, Schacht TE, Binder JL (1993). The effects of training in time-
limited dynamic psychotherapy: Changes in therapist behavior. J Consult Clin Psychol 61:
434–40.
Hobson RF (1985). Forms of feeling: The heart of psychotherapy. London: Tavistock.
Hoglend P (1993). Suitability for brief dynamic psychotherapy: Psychodynamic variables as
predictors of outcome. Acta Psychiatr Scand 88:104–10.
Hoglend P, Piper WE (1995). Focal adherence in brief dynamic psychotherapy: A comparison of
findings from two independent studies. Psychother 32:618–28.
Horowitz M (1976). Stress response syndromes. New York: Aronson.
Horowitz M, Kaltreider N (1979). Brief therapy of the stress response syndrome. Psychiatr Clin N
Am 2:365–77.
Huber D, Klug G, von Rad M (2001). Die Münchner-Prozess-Outcome Studie: Ein Vergleich
zwischen Psychoanalysen und psychodynamischen Psychotherapien unter besonderer
Berücksichtigung therapiespezifischer Ergebnisse [The München process-outcome study:
A comparison between psychoanalysis and psychotherapy]. In: Stuhr U, Leuzinger-
Bohleber M, Beutel M, editors. Langzeit-Psychotherapie. Perspektiven für Therapeuten und
Wissenschaftler [Long-term psychotherapy. Perspectives for therapists and researchers],
p. 260–70. Stuttgart: Kohlhammer.
Kazis LE, Anderson JJ, Meenan RF (1989). Effect sizes for interpreting changes in health status.
Med Care 27:MS178–89.
Kernberg OF (1975). Borderline conditions and pathological narcissism. New York: Aronson.
Kernberg OF (1996). A psychoanalytic model for the classification of personality disorders.
In: Ackenheil M, Bondy B, Engel R, Ermann M, Nedopil N, editors. Implications of
psychopharmacology to psychiatry. New York: Springer.
Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S, et al. (1994). Lifetime
and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Arch Gen
Psychiatr 51:8–19.
Knekt P, Lindfors O, editors (2004). A randomized trial of the effect of four forms of psychotherapy
on depressive and anxiety disorders. Helsinki: Edita Prima. (Studies in social security and
health 77).
Krupnick JL, Sotsky SM, Simmens S, Moyer J, Elkin I, Watkins J, et al. (1996). The role of the
therapeutic alliance in psychotherapy and pharmacotherapy outcome: Findings in the National
Institute of Mental Health Treatment of Depression Collaborative Research Program. J Consult
Clin Psychol 64:532–9.
Leichsenring F (2001). Comparative effects of short-term psychodynamic psychotherapy and
cognitive-behavioral therapy in depression: A meta-analytic approach. Clin Psychol Rev
21:401–19.
Leichsenring F (2004). Randomized controlled vs. naturalistic studies. A new research agenda.
Bull Menninger Clin 68:137–51.
866 FALK LEICHSENRING

Leichsenring F, Winkelbach C, Leibing E (2002). Die generalisierte Angststörung. Krankheitsbild,

Diagnostik, Therapie [Generalized anxiety disorder: Symptoms, diagnosis and treatment].
Z Psychosomat Med Psychother 48:235–55.
Leichsenring F, Biskup J, Kreische R, Staats H (2005). The Göttingen study of psychoanalytic
therapy: First results. Int J Psychoanal 86:433–56.
Leuzinger-Bohleber M, Stuhr U, Rüger B, Beutel M (2003). How to study the ‘quality of
psychoanalytic treatments’ and their long-term effects on patients’ well-being: A representative,
multiple-perspective follow-up study. Int J Psychoanal 84:263–90.
Luborsky L (1984). Principles of psychoanalytic psychotherapy: Manual for supportive-expressive
treatment. New York: Basic Books.
Luborsky L, Diguer L, Seligman DA, Rosenthal R, Krause E, Johnson S, et al. (1999) The
researcher’s own allegiances: A ‘wild’ card in comparison of treatment efficacy. Clin Psychol:
Science and Practice 6:95–106.
Luborsky L, Stuart J, Friedman S, Diguer L, Seligman DA, Bucci W, et al. (2001). The Penn
Psychoanalytic Treatment Collection: A set of complete and recorded psychoanalyses as a
research resource. J Am Psychoanal Assoc 49:217–34.
Malan DH (1976). Toward the validation of dynamic psychotherapy. New York: Plenum.
Mann J (1973). Time-limited psychotherapy. Cambridge, MA: Harvard Univ. Press.
Mark D, Luborsky L (1992). A manual for the use of supportive-expressive psychotherapy in the
treatment of cocaine abuse. Philadelphia, PA: Univ Pennsylvania, Department of Psychiatry.
Markowitz JC, Svartberg M, Swartz HA (1998). Is IPT time-limited psychodynamic psychotherapy?
J Psychother Prac Res 7:185–95.
McCallum M, Piper WE (1990). A controlled study of effectiveness and patient suitablility for short-
term group psychotherapy. Int J Group Psychother 40:431–52.
McCullough Vaillant L (1997). Changing character. New York: Basic Books.
Messer SB (2001). What makes brief psychodynamic therapy time efficient? Clin Psychol 8:5–
22.
Messer SB, Warren CS (1995). Models of brief psychodynamic therapy: A comparative approach.
New York: Guilford.
Messer SB, Tishby O, Spillman A (1992). Taking context seriously in psychotherapy research:
Relating therapist interventions to patient progress in brief psychodynamic therapy. J Consult
Clin Psychol 60:678–88.
Monsen K, Monsen TJ (1999). Affects and affect consciousness—A psychotherapy model of
integrating Silvan Tomkins’ affect and script theory within the framework of self-psychology.
Progress in Self-psychology 15:287–306.
Monsen K, Monsen TJ (2000). Chronic pain and psychodynamic body therapy. Psychother
37:257–69.
Munroe-Blum H, Marziali E (1995). A controlled trial of short-term group treatment for borderline
personality disorder. J Personality Disord 9:190–8.
National Institute of Clinical Excellence [internet]. Clinical Guideline 1. www.nice.org.uk
Nathan PE, Gorman JM, editors (2002). A guide to treatments that work, 2nd edn. New York:
Oxford Univ. Press.
Ogrodniczuk JS, Piper WE (1999). Use of transference interpretations in dynamically orientied
individual psychotherapy for patients with personality disorders. J Personality Disord 13:297–
311.
Ogrodniczuk JS, Piper WE, Joyce AS, McCallum M (1999). Transference interpretations in short-
term dynamic psychotherapy. J Nerv Ment Dis 187:571–8.
Perry JC, Banon E, Floriana I (1999). Effectiveness of psychotherapy for personality disorders.
Am J Psychiatry 156:1312–21.
Persons JB, Silberschatz G (1998). Are results of randomized trials useful to psychotherapists? J
Consult Clin Psychol 66:126–35.
Piper WE, Azim HFA, McCallum M, Joyce AS (1990). Patient suitability and outcome in short-term
individual psychotherapy. J Consult Clin Psychol 58:475–81.
 ARE PSYCHODYNAMIC AND PSYCHOANALYTIC THERAPIES EFFECTIVE? 867

Piper WE, Azim HFA, Joyce AS, McCallum M (1991a). Transference interpretations, therapeutic
alliance, and outcome in short-term individual psychotherapy. Arch Gen Psychiatr 48:946–
53.
Piper WE, Azim HFA, Joyce AS, McCallum M, Nixon GWH, Segal PS (1991b). Quality of
object relations versus interpersonal functioning as predictors of therapeutic alliance and
psychotherapy outcome. J Nerv Ment Dis 179:432–8.
Piper WE, McCallum M, Joyce AS, Ogrodniczuk J (2001). Patient personality and time-limited
group psychotherapy for complicated grief. Int J Group Psychother 51:525–52.
Piper WE, Ogrodniczuk JS, McCallum M, Joyce AS, Rosie JS (2003). Expression of affect as a
mediator of the relationship between quality of object relations and group therapy outcome for
patients with complicated grief. J Consult Clin Psychol 71:664–71.
Pocock SJ, Elbourne DR (2000). Randomized trials or observational tribulations? N Engl J Med
342:1907–9.
Poey K (1985). Guidelines for the practice of brief dynamic group therapy. Int J Group Psychother
35:331–54.
Rose JM, DelMaestro SG (1990). Separation-individuation conflict as a model for understanding
distressed caregivers: Psychodynamic and cognitive case studies. Gerontologist 30:693–7.
Rosen B (1979). A method of structured brief psychotherapy. Br J Med Psychol 52:157–62.
Roth AD, Parry G (1997). The implications of psychotherapy research for clinical practice and
service development: Lessons and limitations. J Mental Health 6:367–80.
Rudolf G, Manz R, Öri C (1994). Ergebnisse psychoanalytischer Therapie [Outcome of
psychoanalytic therapy]. Z Psychosomat Med Psychother 40:25–40.
Rudolf G, Dilg R, Grande T, Jakobsen T, Keller W, Krawietz B, et al. (2004) Effektivität und Effizienz
psychoanalytischer Langzeittherapie: Die Praxisstudie Analytische Langzeitpsychotherapie
[Effectiveness and efficiency of long-term psychoanalytic therapy: The practice study of long-
term psychoanalytic therapy]. In: Gerlach A, Springer A, Schlösser A, editors. Psychoanalyse
des Glaubens [Psychoanalysis of religious belief]. Gießen: Psychosozial Verlag.
Sandahl C, Herlitz K, Ahlin G, Rönnberg S (1998). Time-limited group psychotherapy for
moderately alcohol dependent patients: A randomized controlled clinical trial. Psychother
Res 8:361–78.
Sandell R, Blomberg J, Lazar A (1999). Wiederholte Langzeitkatamnesen von Langzeit-
psychotherapien und Psychoanalysen [Repeated long-term follow-up studies of of long-
term psychotherapies and psychoanalyses]. Z Psychosomat Med Psychother 45:43–56.
Sandell R, Blomberg J, Lazar A, Carlsson J, Broberg J, Schubert J (2000). Varieties of long-
term outcome among patients in psychoanalysis and long-term psychotherapy: A review of
findings in the Stockholm Outcome of Psychoanalysis and Psychotherapy Project (STOPP).
Int J Psychoanal 81:921–42.
Sandell R, Blomberg J, Lazar A, Carlsson J, Broberg J, Schubert J (2001). Unterschiedliche
Langzeitergebnisse von Psychoanalysen und Langzeitpsychotherapien. Aus der Forschung
des Stockholmer Psychoanalyse- und Psychotherapieprojekts [Differential long-term outcome
of psychoanalyses and long-term psychotherapy. Results from the Stockkolm research project
of psychoanalysis and psychotherapy]. Psyche 55:273–310.
Seligman MEP (1995). The effectiveness of psychotherapy: The Consumer Reports study. Am
Psychol 50:965–74.
Shadish WR, Matt G, Navarro A, Phillips G (2000). The effects of psychological therapies under
clinically representative conditions: A meta-analysis. J Consult Clin Psychol 126:512–29.
Shadish WR, Cook TD, Campbell DT (2002). Experimental and quasi-experimental designs for
generalized causal inference. Boston, MA: Houghton Mifflin.
Shapiro DA, Firth JA (1985). Exploratory therapy manual for the Sheffield Psychotherapy Project
(SAPU Memo 733). University of Sheffield, Sheffield, England.
Shapiro DA, Barkham M, Rees A, Hardy GE, Reynolds S, Startup M (1994). Effects of treatment
duration and severity of depression on the effectiveness of cognitive-behavioral and
psychodynamic-interpersonal psychotherapy. J Consult Clin Psychol 62:522–34.
868 FALK LEICHSENRING

Shapiro DA, Rees A, Barkham M, Hardy GE (1995). Effects of treatment duration and severity
of depression on the maintenance of gains after cognitive-behavioral and psychodynamic-
interpersonal psychotherapy. J Consult Clin Psychol 63:378–87.
Shapiro DA, Barkham M, Stiles WB, Hardy GE, Rees A, Reynolds S, et al. (2003). Time is the
essence: A selective review of the fall and rise of brief therapy research. Psychol Psychother
76:211–35.
Shefler G, Dasberg H, Ben-Shakar G (1995). A randomized controlled outcome and follow-up
study of Mann’s time-limited psychotherapy. J Consult Clin Psychol 63:585–93.
Smith ML, Glass GV, Miller TI (1980). The benefits of psychotherapy. Baltimore, MD: Johns
Hopkins Univ Press.
Stiles WB, Agnew-Davies R, Hardy GE, Barkham M, Shapiro DA (1998). Relations of the
alliance with psychotherapy outcome: Findings in the second Sheffield Psychotherapy
Project. J Consult Clin Psychol 66:791–802.
Stunkard AJ (1976). The pain of obesity. Palo Alto, CA: Bull.
Svartberg M, Stiles T, Seltzer MH (2004). Randomized, controlled trial of the effectiveness of
short-term dynamic psychotherapy and cognitive therapy for Cluster C personality disorders.
Am J Psychiatry 161:810–17.
TFPDPP (Task Force on Promotion and Dissemination of Psychological Procedures) (1995).
Training and Dissemination of empirically-validated psychological treatments. Report and
recommendations. Clin Psychol 48:3–23.
Thompson LW, Gallagher D, Breckenridge JS (1987). Comparative effectiveness of
psychotherapies for depressed elders. J Consult Clin Psychol 55:385–90.
Wallerstein R (1989). The Psychotherapy Research Project of the Menninger Foundation: An
overview. J Consult Clin Psychol 57:195–205.
Wallerstein R (1999). Comment on Gunderson and Gabbard. J Am Psychoanal Assoc 47:728–
34.
Wampold B, Minami T, Baskin TW, Tierney SC (2002). A meta-(re)analysis of the effects of
cognitive therapy versus ‘other therapies’ for depression. J Affective Disord 68:159–65.
Westen D, Novotny CM, Thompson-Brenner H (2004). The empirical status of empirically
supported psychotherapies: Assumptions, findings, and reporting in controlled clinical trials.
Psychol Bull 130:631–63.
Wiborg IM, Dahl AA (1996). Does brief dynamic psychotherapy reduce the relapse rate of panic
disorder? Arch Gen Psychiatr 53:689–94.
Wilfley DE, Agras WS, Telch CF, Rossiter EM, Schneider JA, Cole AC, et al. (1993). Group
cognitive-behavioral therapy and group interpersonal psychotherapy for the nonpurging
bulimic individual: A controlled comparison. J Consult Clin Psychol 61:296–305.
Winston A, Laikin M, Pollack J, Samstag LW, McCullough L, Muran JC (1994). Short-term
psychotherapy of personality disorders. Am J Psychiatry 151:190–94.
Woody GE, McLellan T, Luborsky LL, O’Brien CP (1985). Sociopathy and psychotherapy outcome.
Arch Gen Psychiatr 42:1081–6.
Woody GE, Luborsky L, McLellan AT, O’Brien CP (1990). Corrections and revised analyses for
psychotherapy in methadone maintenance patients. Arch Gen Psychiatr 47:788–9.
Woody GE, Luborsky L, McLellan AT, O’Brien CP (1995). Psychotherapy in community methadone
programs: A validation study. Am J Psychiatry 152:1302–8.
Woody GE, Luborsky L, McLellan AT, O’Brien CP, Beck AT, Blaine J, et al. (1983). Psychotherapy
for opiate addicts: Does it help? Arch Gen Psychiatr 40:639–45.
Zerbe KJ (1990). Through the storm: Psychoanalytic theory in the psychotherapy of the anxiety
disorders. Bull Menninger Clin 54:171–83.