Non-invasive brain stimulation augmentation therapy for treatment-resistant schizophrenia: a systematic review and network meta-analysis

Background: Non-invasive brain stimulation (NIBS) provides adjunctive therapeutic options for individuals with schizophrenia if medications are insufficient to produce clinical response, but guidelines remain controversial on whether NIBS is effective, and which NIBS methods and targets are preferred. We aimed to compare the efficacy and safety of NIBS for treatment-resistant schizophrenia (TRS). Methods: This systematic review and network meta-analysis of randomised controlled trials investigated NIBS interventions, including electroconvulsive therapy, magnetic seizure therapy (MST), repetitive transcranial magnetic stimulation (rTMS), and transcranial electric stimulation (tES), as adjunctive treatment for TRS. We searched the Cochrane Schizophrenia Group's specialised register from inception to 2025.07.13, and three Chinese databases from inception to 2024.10.30. The primary outcome was overall symptoms, and adverse events were analysed as secondary outcomes. We synthesized the data using random-effects network meta-analysis. Sensitivity analyses examined the robustness of the findings and the effects of the detailed NIBS protocols. The protocol was pre-registered with PROSPERO (CRD42023410645) and published in a scientific journal. Findings: We identified 21710 references and included 78 trials with a total of 3416 participants (1216 women, 1733 men; mean age 37.06 years, range 25.55-48.38; ethnicity data were not recorded). Compared with sham stimulation, rTMS (SMD -0.47, 95% CI -0.62; -0.31) was more efficacious in improving overall symptoms; but not in a sensitivity analysis excluding studies from Chinese mainland (-0.19, -0.38; 0.01). No clear differences between rTMS specific protocols in terms of stimulation targets and protocols were detected. No clear differences were found for electroconvulsive therapy (-0.20, -0.78; 0.37), tES (-0.08, -0.38; 0.22), and MST (-0.30, -1.73; 1.13) compared to sham stimulation. Treatment as usual might be less efficacious than sham (1.13, -0.13; 2.38), based on indirect evidence. NIBS was generally safe (rTMS produced headaches and local reactions), but information about adverse events was rarely reported. Interpretation: rTMS may be efficacious in individuals with TRS, but this finding was driven mainly by studies from Chinese mainland. No clear differences were observed for electroconvulsive therapy, tES, and MST, but the findings were imprecise and inconclusive. Funding: German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung/BMBF; 01KG2206).