Figure 2

Post hoc power estimation of for cardiovascular outcomes concerning trials comparing novel glucose-lowering medications with placebo treatment (both on a background of standard of care) in patients with type 2 diabetes for major adverse cardiovascular outcomes (MACE; upper row of panels) and non-fatal myocardial infarction (lower row of panels). Left hand panels display the power calculated regarding the hazard ratio as reported in each individual study. The second through fourth columns of panels illustrate the power to detect differences by 10, 15, 20, or 25% versus the proportion of patients with events reported with placebo treatment with SGLT-2 inhibitors (second column of panels), DPP-4 inhibitors (third column of panels) and GLP-1 receptor agonists (fourth column of panels). The power estimates for study results indicating a negligible difference to placebo treatment (≤ 10%) are highlighted with a dashed rectangle, because small differences are typically associated with low power.