Assignment 3 • This assignment consists of 3 questions with a total of 40 marks and it counts 40% towards your final mark in this unit.

.

• All forms of plagiarism, cheating and unauthorised collusion are regarded seriously by the University and could

Assignment 3
• This assignment consists of 3 questions with a total of 40 marks and it counts 40% towards your final mark in this unit.
• All forms of plagiarism, cheating and unauthorised collusion are regarded seriously by the University and could result in penalties including failure and possible exclusion from the University.
• Table of contents is required.

.

Q1. You are asked to examine about the 11 cases of breast cancer found in the last 8 years in the ABC building in Brisbane. This situation was described as follows:

http://www.smh.com.au/news/National/ABC-investigates-breast-cancer-scare/2005/04/05/1112489489868.html

Some additional information about this situation:
• 10 women fulfilled the case definition (one woman had not invasive breast cancer)

.

• Seven cases had worked in the same newsroom.
• Five were under 40 years of age, 4 were 40-49 years and one older than 50 years.
• Your investigation of employment records indicates that in that time period 550 women had been employed at ABC Toowong.
• Their median length of employment was 11 years.
• The average duration of employment was 3.3 years.
• Distribution of invasive breast cancer cases according age group at diagnosis and person years at risk is shown in the following table. The age-specific incidence is also provided:

.

AGE GROUP in years PERSON YEARS AT RISK Age-specific incidence* Observed cases
15-19 14.8 0.16 0
20-24 217.2 0.94 0
25-29 293.4 6.73 0
30-34 337.1 25.21 1
35-39 283.4 59.13 4
40-44 215.2 112.16 3
45-49 162.5 191.93 1
50-54 141.1 236.24 1
55-59 86.4 287.44 0

.

60-64 53.2 325.41 0
65-69 3.3 323.06 0
70-74 2.0 334.73 0
TOTAL 1809.6

*Of invasive breast cancer in the general Australian population. Per 100,000 population
How would you demonstrate with the existing information that this is likely to be a cancer cluster? Explain all your reasons, and show your calculations (10 marks)

.

What additional information would you require to confirm your conclusions from the previous question (2 marks). Please feel free to include references to support your views.
If you wish, you can get more information in the following links:

http://www.abc.net.au/catalyst/stories/s2056788.htm

http://www.abc.net.au/rn/healthreport/stories/2007/1838331.htm#transcript

Q2. As epidemiologist, your government has asked you to develop a surveillance system for chronic diseases at national level. You follow the STEPS strategy by the World Health Organization (http://www.who.int/ncd_surveillance/en/

.

steps_framework_dec03.pdf, in particular table 4 from page 7).Outline this plan assuming no previous system exists, expecting to spend about 6 years (2 years on each step) to complete the project.

This question is worth 10 marks. Please complete at least one page (1.5 space, any font but 12 size, normal margins) outlining a general rationale for your plan. You can include some flow charts or diagrams in additional page(s). Please feel free to include references to support your views.

Q3. These questions come from the study (attached):

.

Knuiman, M. Serum Ferritin and Cardiovascular Disease: A 17-Year Follow-up Study in Busselton, Western Australia. Am J Epidemiol 2003; 158:144-149

http://aje.oxfordjournals.org/content/158/2/144.full.pdf+html

a. How many years were followed up a person who had an incident stroke the 15 of July 1990 and died the 31 of December 1992? (1 mark)
b. How many years were followed up a person that only reported arthritis on the 30 of August 2000? (1 mark)

.

c. What it is the advantage of using case-cohort sampling ? When it should be used? (2 marks)
d. The target population was all adults on the electoral roll (registration to vote is compulsory in Australia) in the Shire of Busselton; the overall response rate for the 1981 survey was 64 percent. Discuss implications for internal validity (1 mark). Hint: do not get confused with external validity

.

e. A viable blood serum sample was available for about 75 percent of the eligible disease-free cohort. Discuss what authors did to avoid potential bias because of this(2 marks)
f. What was the statistical rationale to select a random subset of 450 individuals (2 marks)
g. If each measurement of ferritin cost UD$10. How much money would be saved in this study, compared to a typical prospective cohort study? (1 mark)

.

h. From table 1, discuss the implications for potential bias and confounding (2 marks)
i. How did the authors check for the quality of the ferritin values from serum taken in 1981? What limitation to this strategy? (2 marks)
j. Explain whether gender may statistically interact in the association ferritin-outcome. Why was not tested in the study? (2 marks)
k. Identify cases and controls in a nested-case control study, instead of this case-cohort’ study. (2 marks)

.