3. Here we will adjust for confounding using stratification on the confounder. Calculate the stratum specific estimates
Posted: Thu Feb 17, 2022 11:06 am
3. Here we will adjust for confounding using stratification on the confounder. Calculate the stratum specific estimates among smokers and nonsmokers from the tables below. Smokers Case (MI+, D+) Control (MI-, D-) Diabetes + (E+) 100 15 Diabetes -(E-) 500 155 Non-smokers Case (MI+, D+) Control (MI-, D-) Diabetes + (E+) 22 16 Diabetes -(E-) 990 1426 4. Is smoking a confounder of the effect of diabetes on the incidence of MI? Is this a good way to check for confounding when we are interested in causal associations? Why/why not? 5. How would you report your results for the association between diabetes and MI, taking smoking into account? How would you interpret this association?