The Munich Attachment- and Effectiveness Project (MBWP)

Erhardt, I., Mertens, W., Benecke, C., Zehetleitner, M., & Hörz, S. (2010). Klinische Praxis und formalisierte Diagnostik: Lässt sich der Bericht für den Gutachter mit einem OPD-Rating vergleichen? {Clinical practice and formalized diagnostics: Can one compare both?} Psychotherapie & Sozialwissenschaft, 12(1), 79-106.

Hörz, S., Träger, M., Siegl, G., & Mertens, W. (2011). Einsatz der OPD in der Psychotherapieforschung {Implementation of OPD in psychotherapy research}. Psychotherapie in Psychiatrie, Psychotherapeutischer Medizin und Klinischer Psychologie, 16, 41-54.

Hörz-Sagstetter, S., Mertens, W., Isphording, S., Buchheim, A., & Taubner, S. (2015). Changes of Reflective Functioning during psychoanalytic psychotherapies. Journal of the American Psychoanalytic Association, 63(3), 481-509.

Mertens, W. (2009). Ein jedes Leben ist anders - Plädoyer für eine stärkere Feinauflösung in der Psychotherapieforschung {Each life is different - A plea for more fine-tuned psychotherapy research}. In G. Jüttemann (Ed.), Komparative Kasuistik - Die psychologische Analyse spezifischer Entwicklungsphänomene (pp. 131-139). Lengerich: Pabst.


The Munich Attachment- and Effectiveness Project (MBWP) is a naturalistic prospective psychotherapy study examining process and outcome of psychoanalytic psychotherapies.


At baseline, the Operationalized Psychodynamic Diagnostics (OPD-2, OPD Taskforce, 2008) were applied. Then, using the Heidelberg Structural Change Scale (HSCS, Rudolf et al. 2000), five therapeutic foci were chosen, reflecting difficulties in relationship patterns (one focus), psychodynamic conflicts (one to three foci) and impairments in personality structure (one to three foci) (e.g. Hörz et al. 2011). For attachment classification the "Adult Attachment Interview" (AAI) and the "Adult Attachment Projective" (AAP) were employed and the "Reflective Functioning Scale" (RF) was applied to the AAI for the assessment of Reflective Functioning. Furthermore, at baseline and at follow-up, a number of self-reports were used: "Gießen-Test", "Narzißmus-Inventar", "Bielefelder Fragebogen zur Klientenerwartung" (BFKE) and "Symptom-Check-List" (SCL-90).

The 20 psychoanalysts recorded several sessions on audiotape at up to five points in time (three to five sessions around baseline, around the 80th session, the 160th session, the 240th session and the 300rd session).

The study was set up following a quasi-experimental design:  half of the psychotherapists (N=10) received an introduction to attachment research applied to the results from their patient's AAI and AAP (two 90 minute sessions per patient, overall 20 sessions). The other ten psychotherapists were introduced to a dream coding method (Moser & Zeppelin, 1996) and spent 20 sessions discussing the application of this method to the initial dreams of their patients and relating these to the psychodynamic impressions emerging from the first therapy sessions. One of our hypotheses examines the question if in the first group this sensitization for questions relating to attachment research could enhance the understanding of separation traumas during treatment and hence lead to a change of the attachment classification.

However, the main focus of this study is on researching microprocesses and interactions (across the mentioned points in time as well as a follow-up assessment one year after treatment). The following instruments were employed: the plan formulation method (PFM) to assess the patient's unconscious therapy goals, his or her pathogenic beliefs, test situations as well as plan compatibility of therapist  interventions; the psychotherapy process Q-Sort (PQS) to obtain the most and least characteristic items regarding patients' behaviors and experiences, therapists' interventions and features of the interaction. In several cases, the Structural Analysis of Social Behavior (SASB), Core Conflictual Relationship Theme (CCRT) or Verbal Elaboration of Affect Scale (GEVA) were employed. We used AAP, AAI, RF, HSCS and self-report measures as outcome measures. The focus of this research project is set on intensive single case studies combining the mentioned process and outcome instruments at various points in time.

A number of research questions emerge from the MBWP. Combining process research, broadened and differentiated using single cases, and outcome findings at different points in time, the following questions regarding process and outcome can be considered relevant:

How good is the concordance between the analysts' descriptions for the peer reports system for psychotherapy and interview ratings based on the Operationalized Psychodynamic Diagnostic System (OPD-2) (Erhardt et al., 2010)? What correspondence between AAI and AAP assessments in this clinical sample can be found (Hörz et al. in prep.)? Which attachment classification can be found in a patient who takes a good course on the HSCS compared to a patient with a poor course on this scale? How does a patient with secure attachment classification (AAI, AAP) change over time in comparison to a patient with insecure attachment classification or unresolved trauma? What influence do analyst's plan compatibility in the sessions have on the treatment outcome? What are the interrelations between PQS findings and therapy outcome? How much do RF and HSCS results correspond? At which point in time do decisive changes in HSCS take place? Can hints for these changes be found in other instruments, e.g. the Narzissmus-Inventar? Can further changes be found between the end of treatment and follow-up assessment?

Another goal of this project is to encourage the discourse about which research methods could be taught in future psychoanalytic training curricula to reduce the gap between scientists and practitioners. Which of the methods are too cumbersome and need too much training, which of the methods can be improved?


Susanne Hörz-Sagstetter:

Wolfgang Mertens: