The role of control groups in non-pharmacological randomised controlled trials of treatment-resistant schizophrenia: A systematic review and meta-analysis

•The psychological interventions were superior to the control groups (active-, inactive-, TAU- and waitlist-control group) in the reduction in overall symptoms in treatment-resistant schizophrenia.•The less active the control group (e.g. waitlist- or treatment as usual control group) is, the bigger is the effect of the psychological intervention.•All control groups were associated with a symptomreduction from the pre- to post-measurement point, with the greatest reduction in the inactive control group. Control groups used in randomised controlled trials investigating psychological interventions for depression and anxiety disorders have effects of their own. This has never been investigated for schizophrenia, in particular treatment-resistant schizophrenia. This systematic review and meta-analysis aimed to examine how control groups in randomised controlled trials on psychological interventions for treatment-resistant schizophrenia behave in their effects on general symptomatology. In a search of various databases until July 2023, 31 eligible studies with 3125 participants were found whose control groups were assigned to four categories: active, inactive, treatment as usual and waitlist. The analyses showed that psychological interventions had a greater effect on symptom reduction to all control groups combined. When separating the control groups, only compared to TAU and waitlist controls the psychological interventions were superior. The difference was larger when less active control groups (e.g. waitlist – or treatment as usual control groups) were used. All control groups were associated with an improvement in symptoms from pre- to post-measurement point, with the greatest improvement observed in the inactive control group. The results are preliminary, but they suggest that the choice of the control group has a considerable impact on study effects as it has been shown in other psychiatric diagnoses.