The review involved a diverse range of interventions, participants and outcomes and in light of this heterogeneity we judged meta-analysis to be inappropriate. This made direct comparisons between studies more difficult. Nevertheless, despite this, we were still able to make judgements about the strength and consistency of the findings.