Psychological and psychosocial interventions for treatment-resistant schizophrenia: a systematic review and network meta-analysis

Many patients with schizophrenia have symptoms that do not respond to antipsychotics. This condition is called treatment-resistant schizophrenia and has not received specific attention as opposed to general schizophrenia. Psychological and psychosocial interventions as an add-on treatment to pharmacotherapy could be useful, but their role and comparative efficacy to each other and to standard care in this population are not known. We investigated the efficacy, acceptability, and tolerability of psychological and psychosocial interventions for patients with treatment-resistant schizophrenia. In this systematic review and network meta-analysis (NMA), we searched for published and unpublished randomised controlled trials (RCTs) through a systematic database search in BIOSIS, CINAHL, Embase, LILACS, MEDLINE, PsychInfo, ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform for articles published from inception up to Jan 31, 2020. We also searched the Cochrane Schizophrenia Group registry for studies published from inception up to March 31, 2022, and PubMed and Cochrane CENTRAL for studies published from inception up to July 31, 2023. We included RCTs that included patients with treatment-resistant schizophrenia. The primary outcome was overall symptoms. We did random-effects pairwise meta-analyses and NMAs to calculate standardised mean differences (SMDs) or risk ratios with 95% CIs. No people with lived experience were involved throughout the research process. The study protocol was registered in PROSPERO, CRD42022358696. We identified 30 326 records, excluding 24 526 by title and abstract screening. 5762 full-text articles were assessed for eligibility, of which 5540 were excluded for not meeting the eligibility criteria, and 222 reports corresponding to 60 studies were included in the qualitative synthesis. Of these, 52 RCTs with 5034 participants (1654 [33·2%] females and 3325 [66·8%] males with sex indicated) comparing 20 psychological and psychosocial interventions provided data for the NMA. Mean age of participants was 38·05 years (range 23·10–48·50). We aimed to collect ethnicity data, but they were scarcely reported. According to the quality of evidence, cognitive behavioural therapy for psychosis (CBTp; SMD –0·22, 95% CI –0·35 to -0·09, 35 trials), virtual reality intervention (SMD –0·41, –0·79 to –0·02, four trials), integrated intervention (SMD –0·70, –1·18 to –0·22, three trials), and music therapy (SMD –1·27, –1·83 to –0·