Mock exam
QUESTION 1: STUDY DESIGN (12 points)
The health department in one town conducted a community health assessment by asking all
students at the four primary schools to take a questionnaire for his/her parents to fill in. The
finalized questionnaires were then entered into an SPSS dataset.
a) What type of study design is this and why? (3 points)
This is an example of an analytical cross-sectional study design, because the exposure and
outcome are assessed at one moment in time, so the study is conducted at a specific time.
Besides, the study looks into determinants of health events or states, which corresponds with an
analytical, observational study, and characteristics are not manipulated, but it is merely an
observation of certain characteristics and their relation to an outcome of interest.
→ 3/3
One of the sections was related to asthma. Parents were asked if their children had asthma or
not, in addition to details about the symptoms and duration of illness. Later on, researchers
got interested in investigating a possible association between having asthma and exposure to
chlorine at swimming pools. However, they needed to collect additional information about
swimming habits.
b) How can they use the information from this assessment to recruit students for a
case-control study? Please provide a clear explanation. (3 points)
When you are investigating a possible association between having asthma and exposure to
chlorine at swimming pools, you can use information about swimming habits to collect the cases
for a case-control study. It is retrospective, so therefore it starts with an outcome (having
asthma) and looks into the exposure (exposure to chlorine). To select a relevant study population
of students they need to select students that are actually exposed to chlorine, because they swim
a lot for example. By using the information on swimming habits, they know which students they
should recruit to get a representative sample of the study population. Also, through this they can
select controls that are not too similar and not too different.
→ need to select cases with asthma (outcome of interest) and controls without asthma,
because case-control focuses on outcome → 0/3.
c) List two advantages and two disadvantages of conducting a case-control study (4
points)
Two advantages of a case-control study are that it is (1) inexpensive and (2) it can yield important
findings in a short amount of time.
Two disadvantages of a case-control study are that (1) there is often information and selection
bias, because participants have to recall actions and selection of controls often goes wrong.
Besides, (2) the temporal relation between cause and outcome is often unclear, indicating
causality problems.
→ 4/4
d) What type of study design for an observational study is best to show the association
between exposure to chlorine at swimming pools and having asthma? Please Explain.
(2 points)
A case-control study is the best observational study to show this association, because you start
with the outcome (having asthma) and then look retrospectively back into time whether this is
caused by the exposure (chlorine in swimming pools). Besides, it is an inexpensive study that
does not have to take a long time and they can use objective records, such as the swimming
, records, for information to reduce a recall bias. Besides, it may take a long time before asthma is
developed, which means a cohort study would be less appropriate.
→ cohort study (1 point for naming the design). Have two cohorts, one exposed and one
non-exposed and follow-up into the future to see who will develop the outcome (1 point for
explanation). Temporal direction needs to be drawn → 0/2
QUESTION 2: FREQUENCY, RISK and ASSOCIATION (12 points)
In a case-control study conducted at a hospital in Great Britain in the 1950’s on the
relationship between lung cancer and smoking, researchers recruited 709 controls (patients
admitted to the hospital without lung cancer) to match 709 cases (patients admitted to the
hospital with lung cancer). Of the cases, there were 21 non-smokers, and of the controls there
were 59 non-smokers.
a) What would be the values of a, b, c, d in the 2 x 2 table in this case? (2 points)
Cases (with lung cancer) Controls (without lung cancer) Total
Smokers 688 = a 650 = b 1338
Non-smokers 21 = c 59 = d 80
Total 709 709 1418
→ 2/2
b) Calculate and interpret the odds ratio. (3 points)
𝑎/𝑏 (688/650) 1.058...
𝑂𝑅 = 𝑐/𝑑
= (21/59)
= 0.355...
= 2. 97... = 3. 0
This odds ratio of 3.0 means that people that smoke have 3 times higher odds of developing lung
cancer than people that do not smoke.
→ 3/3
c) Two statements: Which of these statement(s) is/are correct? (3 points)
i) The OR is generally more extreme than RR
ii) When the health outcome is common, the odds ratio provides a reasonable
approximation of the risk ratio.
Statement I is correct, statement II is incorrect.
→ 3/3
The Dutch Ministry of Health estimated that there were 28.500 people in the Netherlands
with lung cancer on January 1st 2018. They also estimated that 13.300 people were diagnosed
with lung cancer in the Netherlands in 2018. In total, 10.362 people died because of lung
cancer in 2018. The Netherlands had an estimated population of 17,18 million people in 2018.
d) Calculate the prevalence on January 1st (2 point)
Total population with lung cancer at Januari 1st 2018 = 28.500 → new cases in 2018 = 13.300 →
10.362 people died in 2018 → total population = 17.180.000
𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑒𝑥𝑖𝑠𝑡𝑖𝑛𝑔 𝑐𝑎𝑠𝑒𝑠 28.500
𝑃= 𝑡𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛
= 17.180.000
𝑥 100. 000 = 165. 9 𝑝𝑒𝑟 100. 000 𝑝𝑒𝑜𝑝𝑙𝑒
So, the prevalence of lung cancer in the Netherlands on January 1st 2018 is 165.9 per 100.000
people in the population.
→ 2/2
e) Calculate the cumulative incidence (2 point)
𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑛𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 2018 13.300
𝐶𝐼 = 𝑡𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑟𝑖𝑠𝑘
= (17.180.000 − 28.500)
𝑥 100. 000 = 77. 4 𝑝𝑒𝑟 100. 000 𝑝𝑒𝑜𝑝𝑙𝑒
