highly significantP =0.001
The overall analysis of the 12 CVOTs showed a highly significant ( P =0.001) 8% reduction of MACE risk associated with no publication bias (Egger test, P =0.210) and significant heterogeneity (I 2 =45.8%, P =0.041), indicating the lack of homogeneity in the whole analysis that grouped 3 different classes of glucose‐lowering drugs.