However, potential for bias in the review process and limitations of the evidence (including small numbers of studies and sample sizes, and significant heterogeneity across studies) suggest the findings should be interpreted cautiously as they may not be reliable, particularly for outcomes related to efficacy.