Text this: Response to letter to the editor, “Observational studies on glucocorticoid harm should always consider confounding by indication.”