How Does Therapy Harm? A Model of Adverse Process Using Task Analysis in the Meta-Synthesis of Service Users' Experience

How does therapy harm? A model of adverse process using task analysis in the synthesis of service users’ experience. Background Despite repeated discussion of treatment safety, there remains little quantitative research directly addressing the potential of therapy to harm. In contrast, there are numerous sources of qualitative evidence on clients’ negative experience of psychotherapy, which they report as harmful. Objective To derive a model of process factors potentially leading to negative or harmful effects of therapy, from the clients’ perspective, based on a systematic narrative synthesis of evidence on negative experiences and effects of psychotherapy from a) qualitative research findings and b) participants’ testimony. Method We adapted Greenberg’s (2007) task analysis as a discovery-oriented method for the systematic synthesis of qualitative research and service user testimony. A rational model of adverse processes in psychotherapy was empirically refined in two separate analyses, which were then compared and incorporated into a rational-empirical model. This was then validated against an independent qualitative study of negative effects. Results Over 90% of the themes in the rational-empirical model were supported in in the validation study. Contextual issues, such as lack of cultural validity and therapy options together with unmet client expectations fed into negative therapeutic processes (e.g. unresolved alliance ruptures). These involved a range of unhelpful therapist behaviours (e.g. rigidity, over-control, lack of knowledge) associated with clients feeling disempowered, silenced or devalued. These were coupled with issues of power and blame. Conclusions Task analysis can be adapted to extract meaning from large quantities of qualitative data, in different formats. The service user perspective reveals there are potentially harmful factors at each stage of the therapy journey which require remedial action. Implications of these findings for practice improvement are discussed..