Response to “Association between metformin use prior to admission and lower mortality in septic adult patients with diabetes mellitus: beware of potential confounders”