Question
An investigator conducted a one-year case-control study of the relationship between gender and coronary heart disease (CHD). Because of prior work, the investigator believed that the role of a subject’s level of serum ferritin (an iron-binding protein) must also be evaluated in any analyses. The stratified results are below.
a. Calculate and interpret the OR for each table (low and high ferritin).
b. Calculate and interpret the crude ORpooled for the combined data.
c. Calculate and interpret the Mantel Haenszel ORMH adjusted for ferritin levels.
d. Is there statistical evidence of heterogeneity? Calculate the χ2 test of homogeneity and use this information to describe the role of ferritin in the relationship between gender and CHD.
e. Is the adjusted measure of association (the ORMH) useful in this situation? Why or why not?
f. How should the investigator present these findings? Write a one or two sentence abstract of the results.
