Abstract
Objectives
Treatment-resistant obsessive-compulsive disorder is a chronic debilitating illness. We conducted a network meta-analysis [NMA] to compare the efficacy of all interventions in SRI-resistant OCD from published Randomised controlled trials [RCT].
Methods
We performed an NMA of RCTs in SRI resistant OCD from all modalities of treatments; pharmacological, psychological, neuromodulation, neurosurgery including deep brain stimulation. The design-by-treatment interaction inconsistency model within the frequentist framework was adopted with a change in Yale-Brown Obsessive-Compulsive Scale score as the primary outcome. We conducted sensitivity analyses excluding studies examining neurosurgical interventions, deep brain stimulation, studies in the paediatric population, and studies from a single geographical region. We also conducted analyses of interventions categorised into treatment groups.
Results
55 RCTs examining 19 treatments or placebo involving 2011 participants were included in the NMA. Ondansetron [Standardised mean difference −2.01 (95% CI: −3.19, −0.83)], deep TMS [– 1.95 (−3.25, −0.65)], therapist administered Cognitive Behavioural Therapy [CBT-TA] [−1.46 (−2.93, 0.01)] and aripiprazole [−1.36 (−2.56, −0.17)] were ranked as the best four treatments on using the Surface Under the Cumulative Ranking [SUCRA] percentage values (85.4%, 83.2%, 80.3%, 67.9% respectively). While all four interventions had large effect sizes, CBT[TA] narrowly missed statistical significance in our analysis. In sensitivity analyses, deep TMS was ranked as the best treatment strategy for SRI-resistant OCD. The small number of subjects in individual studies, higher confidence interval limits, and wider prediction interval for most agents warrant a cautious interpretation.
Conclusions
Considering the principal analysis and sensitivity analyses together, deep TMS, ondansetron, CBT[TA], and aripiprazole may be considered a first-line intervention for SRI-resistant OCD in adults.
Other
This work was not funded. The NMA has been registered with PROSPERO, [Registration number: CRD42020173589]
Acknowledgements
The author acknowledges the contribution of Dr. Manjushree Malkonahalli with data entry.
Author contribution
NPR, S.S. conceptualised the study. NPR, V.K., B.B. and YCJR were involved in the study design. S.S., PKM, TGI, NKC, V.K., NPR were involved in data collection. PKM and B.B. conducted the data analysis. S.S., PKM, V.K., YCJR and NPR were involved in the interpretation of the results. S.S., PKM wrote the first draft of the manuscript which was edited by V.K. and NPR. All authors contributed to revisions. All authors have approved the final manuscript.
Disclosure statement
The authors report no conflict of interest.