Estimate the odds ratio for bladder cancer comparing high-risk to other occupations

Question 1
Cole and MacMahon (1971) present the following data from a case-control study of bladder cancer:
High Risk Occupation?    High Cigarette Consumption?    No. of Cases    No. of Controls
No                                                    No                                           43                     94
No                                                   Yes                                           173                   189
Yes                                                   No                                           26                     20
Yes                                                 Yes                                           111                     72

“Provide evidence”means for you to provide an interpretation of the results in answering the question.
a) Estimate the odds ratio for bladder cancer comparing high-risk to other occupations (ignoring cigarette consumption). Provide a 95% confidence interval around this odds ratio. Is this association significant? (provide evidence)
b) Estimate the odds ratio for bladder cancer comparing high cigarette consumption (yes) to low cigarette consumption (no). Is this association significant? (provide evidence) c) What is the crude odds ratio for bladder cancer comparing high-risk to other occupations for those that have high cigarette consumption? Do the same for low cigarette consumption. Include an interpretation for each.
d) Calculate the Mantel-Haenszel odds ratio for high risk versus other occupations, adjusting for cigarette consumption. Is this association significant (provide evidence
e) Is there evidence of interaction? (again, provided evidence).
f) Is there evidence of confounding? Explain your reasoning as to why or why not?
g) Give an overall interpretation of the results. (e.g. given ALL of the results above, what does it all mean.)

Question 2
As part of your employment with the Newfoundland and Labrador provincial government, you have been asked to conduct a survey examining the association between the number of sexual partners and the risk of contracting a Chlamydia trachomatis infectionin NL youth (15-24). You decide to conduct a sexual health survey of this demographic in NL. Based on your literature search, you decide to code the number of partners as a dichotomous variable with people that have had more than onesexual partner in the past year as “multiple partners”. Recent studies indicate that about 24% of people in that age group report having multiple partners. National chlamydia incidence risk in this age group is about 600 cases per 100,000 persons.
b) With respect to bias, are there any issues regarding your exposure variable?
c) Due to budget constraints, you can only afford to sample students in St. John’s (both high school and university students). How might this bias your results?
d) Briefly outline how you would go about sampling the population for the study. Make note of any limitations you think may impact your study results.

Leave a comment