Fig. 3.

Credibility of individual subgroup effects observed in each review. 1, Bruins et al. [35]; 2, Moran et al. [39]; 3, Lafuente-Lafuente et al. [30]; 4, Heneghan et al. [38]; 5, Salazar et al [31]. Credibility is assessed using the criteria established by Sun et. al. [10]