21 publications: baking soda paste removed more plaque in single-brushing studies, only partly confirmed in longer studies; gum bleeding fell slightly versus control pastes, with no difference in gingival index. Moderate risk of bias, high heterogeneity.