Question 1 –
You conduct a case-control study of tuberculin skin-test reactivity among migrant farm workers. Your main interest is to examine how sex affects tuberculin skin-reactivity. You obtain the following results
a) What is the most appropriate measure of association between sex and tuberculin skin-test reactivity in this study?
i) Incidence ratio
ii) Relative risk
iv) Attack rate
b) Calculate the appropriate measure of association between sex and tuberculin skin-test reactivity in this study.
You are concerned that country of birth may also be associated with tuberculin skin reactivity. You decide to stratify on country of birth:
Country of birth: Australia (n = 60 subjects)
Country of birth: Non-Australia (n=140 subjects)
c) Based on these data, what do you conclude (you need to calculate Odds ratio for each strata)?
i) Tuberculin skin-test reactivity is associated with male sex in both Australia-born and non-Australia –born subjects.
ii) Tuberculin skin-test reactivity is associated with male sex only among Australia-born subjects.
iii)Tuberculin skin-test reactivity is associated with male sex only among non-Australia –born subjects.
iv) Tuberculin skin-test reactivity is not associated with male sex in both Australia-born and non-Australia –born subjects.
Lymphoma is a relatively rare cancer in the population (incidence ~0.012%). It has been theorised that some blood cancers can be caused by viral infections. Investigators looked for evidence of viral infection with a simian Virus 40 (a virus with previously described cancer-causing potential) in 68 lymphomas and 40 well controls; evidence of viral infection was found in 4 lymphomas and 1 control.
a) What sort of study is this?
b) Draw the 2 x 2 table for this data
c) Calculate the odds ratios for viral exposure and lymphoma, relative to well controls.
d) What is the exact interpretation of this odds ratio in words
e) In order to interpret this as saying that SV40 increases/decreases the chances of developing lymphoma by x fold, what 2 conditions must be satisfied?
f) Given that someone has lymphoma, what is the probability that it was at least partly due to SV40? What measure is this?
g) What is the proportion of lymphoma in the community that could be avoided if SV40 could be eradicated? What measure is this?
A randomised controlled trial addressed the impact of a new drug, nicorandil, in patients with stable angina. Men and women, aged >45 or >55 respectively, with stable angina were recruited from various centres in the UK (n=5126). Randomization was computer generated at a central call-in centre. Patients were assigned to nicorandil or placebo in a double-blind fashion. The primary endpoint was the combined outcome of cardiovascular death, non-fatal MI (heart attack), or hospital admissions for chest pain; mean follow-up was 1.6 years.
The results were as follows:
1o end point 337 398
a) What is the absolute risk of the 1o end pt in each group?
b) What is the absolute risk reduction?
c) What is the relative risk of the 1o end pt in the nicorandil group?
d) What is the relative risk reduction of the 1o end pt in the nicorandil group?
e) What is the NNT? Explain this in words.
f)There were more people who developed side effects and stopped nicorandil (n=1003) than stopped placebo (n=809)
What is the potential bias if these people are not included in the analysis?
g) How can you judge whether randomisation worked properly?
A food borne outbreak of gastroenteritis occurred after the conference dinner at the 2002 World epidemiology conference. You have a list of all 134 people at the dinner and of the foods served and have been asked to investigate the outbreak.
a) What would be your first 3 steps in investigating the cause of the outbreak?
A few days after you start your investigation, you have interviewed some of those who were ill and some who were not, and you have the following information:
|Name||Age ( in yrs)||Sick? (Y = yes; N = No)||Salad||Caviar||Chicken|
b) Calculate the attack rates for those who ate and did not eat salad, caviar and chicken, as well as the difference in attack rates.
c) Can you calculate the risk ratio or the odds ratio (which one is appropriate) for these foods? Explain. Calculate whichever one (s) are appropriate for each food.
d) Can you tell which food is the culprit? What other information do you need and what other analysis could you do?
Select the BEST answers (there is only one correct answer for each).
1.The pipe-fitters’ union has approached you (as you are an epidemiologist!!) with their concern over an apparent increase in lung cancer mortality among retired members. You would like to determine if lung cancer is indeed elevated in this group. You review records showing age, cause of death, and smoking status for all retirees over the past decade. Which statement below is true?
i) You should perform direct adjustment for age and smoking using a standard reference population.
ii) Indirect adjustment for age and smoking will yield a standardised mortality ratio (SMR).
iii)Direct adjustment for age and smoking will provide a ratio of observed to expected deaths.
iv)Indirect adjustment involves applying the observed community rates to a reference population, such as the 1990 US population
2.Which one of the following statements regarding epidemiologic knowledge is true?
i)Experimental epidemiology focuses in characterising the distribution of illness by person, place, and time.
ii)Identification of smoking as a risk factor for lung cancer is an example of analytic epidemiology.
iii)Randomized clinical trials are examples of descriptive epidemiology.
iv)A major strength of epidemiology is its ability to characterise mechanism of disease and causality.
3.To identify risk factors for contracting the condition, the most appropriate population health measure is
i) Disease prevalence
ii) Disease incidence
Disease attack rate
4.The most appropriate study design to examine the natural history of the disease is
i) Prospective cohort study
ii) Randomised Clinical trial
5.Randomisation is the best method to handle:
ii) misclassification bias
iv)none of the above