terogeneity across studies and the high risk of bias.