Odds Ratio vs Relative Risk

OR and the RR are both used to express an association between exposure and outcome

jake1.PNG
jake2.PNG

RR isn’t appropriate for a case control study. The denominator is chosen by the investigator (cases/controls). In a cohort study you can calculate RR or OR (RR is better).

jake3.PNG
jake4.PNG

Under some conditions the OR in a case-control design approximates RR that you would have got with a cohort design:

jake5.PNG
jake 6.PNG