Evaluation of psychotherapeutic interventions.

Between group ES calculation is unconventional and exaggerates differences
between groups
• Combined between-group ES of overall outcome reported in Leichsenring &
Rabung (2008, p. 5060) was d = 1.8., but none of the 8 studies reported within
group d > 1.45 (Figure 2, p 1558)
• Implausible between-group ES d = 6.9 for personality functioning based on 4
studies, but no study reported d > 2
▪ -> “Leichsenring and Rabung based their meta-analysis on grossly incorrect
calculations.“ (Bhar et al., 2009, p. 209)
▪ Leichsenring and Rabung (2010, p. 48) responded that Bhar et al. “have
ignored our published clarification and raise the question as if it had not
been addressed before. In addition, we even reported the between-group
effect size for overall outcome assessed in the usual way.” d = 0.65* for LTPP
▪ -> “Thus, although comparing within-group effect sizes between treatment
conditions may have been unconventional, this procedure did not
systematically distort the results.“

There are many problems with the meta-analysis of Leichsenring &
Rabung (2008)
• Methodological weaknesses of the studies included
− Small sample size
− No or no adequate control condition (e.g. fewer number of sessions)
• Within sample effect sizes provide information about the amount of
pre-post change
− not controlled for threats of internal validity (e.g. regression to the
mean, spontaneous remission … ),
− Calculation of between groups-ESs (point biserial correlation of within
group ES) leads to unrealistically large differences between LTPP and
control groups
• Due to small number of studies, statistical tests have low power to
detect medium effects. Authors do often not report effect sizes of
their tests, but only the p-values
▪ In order to rule out some of the critiques, the authors updated and
revised their meta-analysis twice

Many patients with chronic mental disorders or personality
disorders do not benefit sufficiently from short-term
psychotherapy
▪ Long-term psychotherapy: higher costs than short-term
psychotherapy
▪ -> Do effects of long-term psychotherapy exceed those of
short-term treatments
▪ Taking critique against 2008 meta-analysis into account
• lack of between-group effect sizes
• ITT analyses
• possible publication bias
• Inclusion of inactive control conditions

Inclusion criteria:
• Studies of individual PPT meeting definition
• ≥ 1 year, or ≥ 50 sessions
• Prospective studies: before-and after or follow-up assessments
• Reliable outcome measures
• Clearly described sample of patients with mental disorders
• Adult patients (≥ 18 years);
• Sufficient data to allow determination of between group effect sizes
• Concomitant treatments were admissible (e.g.,
psychopharmacological)
• Active treatments in the control condition

Studies published between 1960 and April 2010
• MEDLINE, PsycINFO, Current Contents
• Manual searches of articles and textbooks
• Communication with authors and experts in the field
• To reduce file-drawer effect

grundlegend: 23

excluded; 14

dann update research bei dem 2 gefunden + eine gefuert wurde

also letztlich 10 miteingezogen

in ner aktualiseirten form von 2013 hat er dann nochmal 3 dazu genommen also dort insegsamt 13

Between group effect size compares change from pre to post
between the two groups
ppc Hedges’ d (post) =

((M LTPP pre – M LTPP post) – (M control pre – M control post)/SD(pre pooled))*J (= factor corrects for bias due to small sample size)

zusätzlich ne 2te effect size mit gleicher formel wo alle post werte mit follow up statt post ersetzt wurden

also wieder alles sehr random

Intention-to-treat data (ITT) were included
▪ For studies that did not report ITT:
effects for patients who withdrew after randomisation were set
to zero
• Example: 20 completers, 5 drop-outs in a study:
− pre–post treatment difference = 0.40 for completers
− ITT adjusted pre–post difference: 0.40 * (20/25) = 0.32

1095 vpn

7 signi overall effect von .40

overall effectiveness

target problems

psychiatric symptoms

personality functioning

social functioning

Meta-analytic effects were significant and in favor of LTPP, except for
personality functioning

wenn man die 2 später hinzugefügten studies rausgelassen hat war das fazit: Meta-analytic effects were larger and all significant and in favor of LTPP

Spearman correlations of Jadad scale (study quality) with
post treatment between group ES: r = .13 to .53, but not
significant.
• Studies of higher quality yielded larger ES.
In favour of LTPP for psychiatric symptoms: r = 0.53 (n.s.)

The longer the duration and the larger the number of sessions, the larger was
the overall outcome, and the improvement of psychiatric symptoms and social
functioning in the LTPP condition
There were no sig. correlations between number of sessions and outcome in
control conditions

In sum, the meta-analytic update confirmed the results of
previous meta-analyses on LTPP (Leichsenring & Rabung,
2008, 2011b), irrespective of including studies reporting data
of ongoing treatments or not

LTPP was superior to less intensive methods of psychotherapy in complex mental
disorders
• But alternative treatments were heterogeneous
− TAU, CBT, IPT . . .
▪ Positive correlations between outcome and duration or dosage of therapy.
▪ -> Dose–effect relationship for LTPP
▪ Only few studies reported follow-up assessments
• results of previous meta-analysis suggested that the effects of LTPP even increase after the
end of treatment
• Current results suggest stability of effects (-> consequence of Intent to treat analysis)
▪ Methodological quality of meta-analyses and studies included is comparable to that of
many studies of CBT
▪ No significant correlation between outcome and the methodological quality of the
studies
• Size of correlations suggested that studies of higher quality had larger ESs

Leichsenring and Rabung’s (2011) summary and discussion of results
emphasizes the purpose: To justify long term psychotherapy (with many
sessions) in general and LTPP in particular.
▪ “Long-term psychotherapy, […] is associated with higher direct costs than
short-term psychotherapy. For this reason it is important to know whether
the effects of long-term psychotherapy exceed those of short-term
treatments. In this meta-analysis, LTPP was superior to less intensive
methods of psychotherapy in complex mental disorders.” (p. 20)
▪ “Furthermore, we found positive correlations between outcome and
duration or dosage of therapy.” (p. 20)
▪ “Consequently, we do not claim that LTPP is superior to any specific form of
psychotherapy in complex mental disorders that is carried out equally
intensively, rather that it is superior to less intensive forms of
psychotherapeutic interventions in general.
We expect this to be true for other more intensive approaches of formal
psychotherapy as well, for example that higher-dose CBT is superior to lower-
dose CBT in borderline personality disorder.
For psychodynamic psychotherapy this should also be true.

Some problems of the first meta-analysis of Rabung & Leichsenring (2008)
have been resolved by the second and third meta-analysis.
• Intent to treat (ITT) analyses were conducted (estimation of ITT data)
• ES = standardized difference of change scores of LTPP and comparison group
− Threats to internal validity (e.g. regression to the mean, … ) are now better
controlled.
− However, ES is not directly comparable to ES of other meta- analysis that used
standardized mean difference (Cohen’s or Hedges’ d) of post or follow up scores
▪ Other problems still exist:
• Methodological weaknesses of the studies included
− Small sample size
− Control conditions had less sessions and were briefer. Type of therapy and dose are
therefore confounded
− Replication of effect sizes showed smaller effect sizes
• Large heterogeneity of ES between studies
▪ Woll & Schönbrodt (2020) provided an example for a more balanced and
transparent meta-analysis

The aim of Leichsenring & Rabung (2008, 2011, Leichsenring et al. 2013) was
to demonstrate that LTPP is efficacious – which is legitimate.
▪ They showed that LTPP seems to be more efficacious than other therapies
with shorter duration
▪ However, studies included into the meta-anyalyses were heterogenous and
study quality was frequently low
• “If anything, this suggests that LTPP is often compared against relatively ineffective
“straw man” comparators (Ioannidis, 2008). LTPP comparisons to specialized non-
psychodynamic treatments, like dialectical behaviour therapy and schema-focused
therapy, suggest that LTPP might not be particularly effective.” (Smit et al., 2012, p.
89)
• “The effect sizes of the individual trials varied substantially in direction and
magnitude. In contrast to previous reviews, we found the evidence for the
effectiveness of LTPP to be limited and at best conflicting.” (Smit et al., 2012, p. 81)
▪ -> Leichsenring & Rabung’s (2008, 2011, Leichsenring et al. 2013) meta-
analysis provide arguments for the proponents of LTPP -> Lobbyism:
Promoting LTPP in the health care system
- Justification of LTPP practice
- Justification of long term therapies

psychotherapy

counseling entspricht beratung per se

psychosocial + health prevention programms(e.g. aids prevention programs)

-interventions why modify mental disorders systematically

-interventions which follow a certain order

-interventions which have a theoretical framework

-systematisches und durchgägiges shaping of the therapist-patient-relationship

valere(latin)--> value--> wert

determining the merit, worth,
or value of something, or the product of that process

identification of relevant standards of merit, worth, or value--> was ist bei dem thema überaupt relevant

investigation of the performance of the evaluands on these
standards--> performt die methode die man grad anschaut

integration or synthesis of the results to achieve an overall
evaluation or set of associated evaluations--> gesamtbewertung bzw. einordnung in bisherige ergebnisse

-summative evaluation: z.b. gibts nen einfluss einer intervention und falls ja wie stark ist dieser?

 --> new mthod vs. old methd vs. placebo

-formative evaltion: based on a given method, how can we adjust that one to increase its performance

--> wie kann man ne expo bei archnophobikern angepasst werden, damit diese noch erfolgreicher ist

-soll zu credible knowledge(gegenteil fake news) führen, that can be translated into practice; Ziel Evidenz im treatment finden

im speziellen:

-testing whether a treatment really produces the positive effects it
intends to produce

- Testing whether a treatment produces negative side effects

-Identifying conditions under which a treatment works
(moderators)

-Testing assumptions how a treatment works: Identifying causal
pathways (mediators)?

plaussibility hoch + efficacy hoch--> expected

plausabillity niedrig + efficacy niedrig --> expected --> ich stinke also bin ich

jeweils unterschiedlich --> not expected

lt. vl. weder plasubiel noch gibts ne efficacy

plausible aber lt. peters studie ned efficiant, sogar rather harmful

nicht plausible but under some circumstances efficiant

nicht plausibel, aber hat sich bei Personen mit shhweren Depressionen als effektiv erwiesen

the patients/ clients --> are often waek + have impaired judgement: deswegen wichtig das ne behandlungsmethode save ist

ourself --> ich studier hier ja ned zum spass ...of science und ned ...of arts

society + healt care system: beitragszahler wollen auch, dass das geld für was gutes ausgegeben wird und dass kein kokolores damit angestellt wird

-little Albert --> aquiration of fear via classical conditioning

-the wolf man(Anna Freud)--> central for the development of psychoanalysis

-Vivid(=lebendid) description of the psychotherapeutic process and its principals

-Example is usually highly convincing

-Necessary for documentation of very rare cases

-Useful for the generation of hypotheses