Abstract/Summary

Background: Selective mutism (SM) is an anxiety disorder that often starts in early years with serious and lasting consequences. Nonpharmacological interventions are commonly seen as the preferred first treatment. This systematic review identifies outcome measures used and outcomes achieved for nonpharmacological interventions for children and adolescents with SM. Methods: Systematic searches were conducted using thirteen electronic databases and hand searches, including peer-reviewed and grey literature since 1992. Results: Twenty-five studies were identified. While specific measures varied, all studies reported an outcome measure for speaking behaviour and eighteen used a measure of anxiety. Few studies reported measures of SM remission (k=6), well-being (k=6), academic impact (k=2), or quality of life (k=1). Within subject outcomes for nonpharmacological interventions were variable for improvements in speaking behaviours (very small to large positive effects) and reduction in anxiety symptoms (very small negative to large positive effects). Only five randomised controlled trials (RCTs) were included in the meta-analysis. Three studies compared a combined systems/behavioural approach with waitlist controls indicating a significant and large effect (Hedges g = 1.06, p < .0001, 95% CI: 0.57-1.56) on improved speaking behaviour. Two of these RCTs showed a large effect for SM remission 5 favouring the intervention (Risk Ratio = 4.25, p = .1774, 95% CI: 0.52 – 34.84) but this did not reach statistical significance. Non-significant outcomes for two RCTs with active controls (Hedges g = 0.55, p < .2885, 95% CI: -0.47 – 1.57) showed considerable heterogeneity in approach and outcomes, one with large and one with negligible effects. Conclusion: Despite the considerable impairment caused by SM, there has been little systematic evaluation of non-pharmacological interventions. Although combined systems/behavioural interventions are promising, further systematic evaluations are urgently needed to inform treatment approaches. Cross-study measurement harmonisation is required to promote learning from all studies, including wider clinical and economic outcomes.