Use of E-values for addressing confounding in observational studies-an empirical assessment of the literature.