EPID 301 Chapter Notes - Chapter 17: Odds Ratio, Confounding, Relative Risk
Document Summary
Chapter 17: other measures of association in epidemiology. Most of these measures are estimates based on risk differences, but they seek to quantify more than just an association (seek to quantify the casual impact of risk factor on a population) If a risk difference is believed to represent a casual effect, you could call it an attributable risk among the exposed- a member of the family of attributable risk parameters. Attributable risk exposed = incidence proportion exposed incidence proportion non- exposed. It may not be justified when the difference between 2 incidence proportions arise from confounding variable. The attributable risk among the exposed reflects the extent of risk in exposed subjects due to the exposure. The risk difference quantifies the extent to which risk in exposed subjects could be diminished if the effects of the exposure were eliminated. Attributable risk among the exposed are the measures of impact as well as measure of association.