1. The results of the study of the occurrence of tuberculosis (TB) in a village in Malawi are given in the table below. All individuals in the village were examined during two surveys made two years apart. Assume that there were no deaths in the village during these two years. Assume that new cases of TB were distributed evenly over time. Children born after the first survey were excluded from the secondary survey. There was no immigration and no emigration. There were two religious groups in the village.
|
|
First survey
|
Second Survey
|
|
Religion Group
|
Number of cases of TB observed
|
Number of persons free of TB
|
Number of new cases of TB
|
|
A
|
100
|
5000
|
10
|
|
B
|
75
|
10000
|
25
|
i. Two surveys were carried out to determine incidence of TB. Why were two surveys required? What would one call the first and second surveys? What are the appropriate measure for each survey and the measure of disease association in each survey?
ii. What was the prevalence proportion of tuberculosis in the village at the first survey overall and in each religion group? What is the most appropriate measure to compare disease occurrence between the two groups?
iii. Assuming a steady state, give an estimate of the average duration of cases of TB in the village. Show units in your calculations. Give the conditions that are required for steady state to be present.
iv. What was the incidence rate of TB in the two religion groups?
v. What was the relative risk of acquiring TB in the religion group A relative to religion group B?
vi. Consider the exposed group to be religion group A. What was the population attributable risk of tuberculosis?
vii. Incidence and prevalence are different in the two groups. Are the differences in the same or opposite directions and speculate why this might be.
2. The table below gives prevalence estimates for selected cancers in the US in 1997. It can be seen that the prevalence of breast cancer is about five times that of lung cancer. What can we then infer about the incidence of these two cancers: will the incidence of breast cancer also be approximately five times larger than that of lung cancer? Explain (4-5 sentences)
|
Disease Condition
|
Prevalence (thousands)
|
|
Breast cancer
|
2000
|
|
Lung cancer
|
400
|
|
Prostate cancer
|
1000
|
3. The following shows AIDS patients who are exposed and not exposed to IV drug use and mortality after a follow-up of certain years.
Exposed subjects (n = 289):
Person-years observation = 2138.5 and deaths = 41
Unexposed subjects (n = 347):
Person-years observation = 2586.2 and deaths = 11 Calculate and interpret the following:
i. Cumulative incidence for each group
ii. Incidence rate for each group (incidence density)
iii. Risk ratio and Risk difference
iv. Incidence rate ratio Incidence rate difference
v. Attributable fraction for the exposed (use the incidence density measure)
vi. Attributable fraction for the population (use the incidence density measure)
4. The table shows relative and attributable mortality associated with heavy smoking.
|
Cause of death
|
Death rate ratio
|
Attributable death rate/1000 per year
|
|
Chronic bronchitis
|
21.2
|
1.01
|
|
Lung cancer
|
32.4
|
2.20
|
|
Cardiovascular disease
|
1.4
|
2.61
|
i. For which disease is a causal association the most likely? Explain with 3-5 sentences.
ii. For which disease could the most deaths be prevented (assuming that it is possible to eliminate smoking and that smoking is causal)? Explain with 3-5 sentences.