[Conclusion: How should a meaningful meta-analysis of long-term psychodynamic psychotherapy be designed?]

In their answer to our critical evaluation of their meta-analysis about long-term psychodynamic psychotherapy (LTPP), Leichsenring and Rabung claim that all our critical points could be rejected. This is surprising, as the authors already confirmed different shortcomings of their analysis. Our major criticism is that the meta-analysis is based on very few high-quality studies (including follow-up assessments), but on many studies with poor quality and/or observational studies resulting in serious publication bias. Neither the treatment of interest (LTPP) nor the target group is sufficiently circumscribed, but the very few studies are misinterpreted as if confirming efficacy of all psychodynamic interventions for nearly all psychiatric groups. Superior efficacy over other effective psychological interventions is mentioned, although not justified by data. Patient selection bias which is a major issue in long-term treatments is not addressed adequately and led to erroneous conclusions that the more treatment people receive, the higher the benefit. To conclude, the authors try to suggest far-reaching conclusions which are only based on a few studies with poor data quality.