1) In a case-control study of a particular cancer, 166 cancer patients and a matched
sample of 166 healthy controls were each classified as having high or low exposure to
air pollution. The following table shows the results.
Case Control
High exposure 80 44
Low exposure 86 122
Calculate a 9970 confidence interval for the odds ratio for the cancer in high exposure
versus low exposure individuals, and comment briefly. (8 marks)
2) The table below shows the age distribution of an adult population (‘Distribution’),
plus the age-specific death rates per thousand people, per year (‘Pop. Rate’) of a
rare but fatal disease in that population. It also gives the numbers of individuals
(‘Sampled’) out of a sample of 1800 in a particular industrial occupation which it is
thought might be associated with the disease, and the numbers of actual deaths in that
sample (‘Deaths’), due to the disease, over a one-year cohort study.
Age group Distribution Pop. Rate Sampled Deaths
20 to 30 0.305 6.7 502 4
30 to 40 0.269 8.4 563 9
40 to 50 0.233 9.0 457 8
50 to 60 0.194 17.7 278 9
a) Calculate the crude death rate and the age-specific death rates from the industrial
sample. Explain carefully how the age-specific rates from the sample compare with the
population rates (4 marks)
b) Carry out an indirect standardization by calculating the expected number of deaths
in the sample if the age-specific population rates apply and so estimate the standardized
mortality ratio for the sample. (4 marks)
c) Calculate upper and lower 90% confidence limits Ila, pu for the true expected number
