University of Hertfordshire

From the same journal

By the same authors


View graph of relations
Original languageEnglish
Article number152339
Number of pages9
JournalComprehensive Psychiatry
Early online date18 Jul 2022
Publication statusE-pub ahead of print - 18 Jul 2022


Despite promising results from several randomized controlled trials (RCTs) and meta-analyses, the efficacy of r-TMS as a treatment for OCD remains controversial, at least in part owing to inconsistency in the trial methodologies and heterogeneity in the trial outcomes. This meta-analysis attempts to explain some of this heterogeneity by comparing the efficacy of r-TMS in patients with or without resistance to treatment with selective serotonin reuptake inhibitors (SSRI), defined using standardised criteria.

We conducted a pre-registered (PROSPERO ID: 241381) systematic review and meta-analysis. English language articles reporting blinded RCTs were retrieved from searches using MEDLINE, PsycINFO, and Cochrane Library databases. Studies were subjected to subgroup analysis based on four stages of treatment resistance, defined using an adaptation of published criteria (1=not treatment resistant, 2=one SSRI trial failed, 3=two SSRI trials failed, 4=two SSRI trials failed plus one or more CBT trial failed). Meta-regression analyses investigated patient and methodological factors (age, duration of OCD, illness severity, stage of treatment-resistance, or researcher allegiance) as possible moderators of effect size.

Twenty-five independent comparisons (23 studies) were included. Overall, r-TMS showed a medium-sized reduction of Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) scores (Hedge’s g: -0.47; 95%CI: - 0.67 to -0.27) with moderate heterogeneity (I2=39.8%). Assessment of publication bias using Trim and Fill analysis suggested a reduced effect size that remained significant (g: -0.29; 95%CI: -0.51 to -0.07). Subgroup analysis found that those studies including patients non-resistant to SSRI (stage 1) (g: -0.65; 95%CI: -1.05 to -0.25, k=7) or with low SSRI-resistance (stage 2) (g:-0.47; 95%CI: -0.86 to -0.09, k=6) produced statistically significant results with low heterogeneity, while studies including more highly resistant patients at stage 3 (g: -0.39; 95%CI: -0.90 to 0.11, k=4) and stage 4 (g: -0.36; 95%CI: -0.75 to 0.03, k=8) did not. Intriguingly, the only significant moderator of the effect size found by meta-regression was the severity of baseline depressive symptoms. All trials showed evidence of researcher allegiance in favour of the intervention and therefore caution is required in interpreting the reported effect sizes.

This meta-analysis shows that r-TMS is an effective treatment for OCD, but largely for those not resistant to SSRI or failing to respond to only one SSRI trial. As a consequence, r-TMS may be best implemented earlier in the care pathway. These findings would have major implications for clinical service development, but further well-powered RCTs, which eliminate bias from researcher allegiance, are needed before definitive conclusions can be drawn.


© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (

ID: 27903294