PUBH5010 Epidemiology Methods and Uses (End of Semester Exam)

Semester 1 - Main, 2023

Section A (24 marks)

1) The following six studies (Studies A, B, C, D, E and F) all consider the possible

relationship between use of sunscreen and being diagnosed with melanoma of the skin.  For each study, name the study type. You do NOT need to justify your choice. Note that there might be more than one study of the same type. (12 marks)

Study A

The study used information collected from a random sample of twenty local health districts (LHDs) out of 100 LHDs in the country. Each of the LHDs had information on the incidence of melanoma of the skin in the LHD. Each LHD had also conducted a health survey which included information on regular use of sunscreen. The study found that LHDs with a higher level of reported sunscreen use had lower incidence of melanoma of the skin.

Study B

Participants in the study were a random sample of volunteers from the local football competition. The volunteers were divided into two   groups based on the participants’ reported sunscreen use between ages 10 and 29 years – regular use

(Group A) and irregular or no use (Group B).   The two Groups were followed for 10 years. At the end of follow-up the incidence of melanoma of the skin was compared between Group A and Group B.

Study C

Participants in the study were a random sample of volunteers from the local football competition.  The volunteers were divided into two groups by randomly assigning them to regularly use a new form. of sunscreen (Group A) or to take their usual approach to sunscreen use (Group B). The two Groups were followed for 10 years.

At the end of follow-up the incidence of melanoma of the skin was compared between Group A and Group B.

Study D

Participants in the study were volunteers from the local football competition. A random sample of the volunteers was divided into two groups based on the participants’ reported sunscreen use between ages 10 and 29 years – regular use (Group A) and irregular or no use (Group B).  Members in the two Groups were asked if they had ever been diagnosed with melanoma of the skin. The occurrence of melanoma of the skin was compared between Group A and Group B.

Study E

Participants in the study lived in the local town. Participants in Group A were a random sample of people presenting to the town’s main skin clinic over a six-month period with melanoma of the skin.  Participants in Group B were a random sample of people presenting to the town’s main respiratory clinic, which was next door to the skin clinic, over the same six-month period.  Participants in Group B were excluded if they had been diagnosed with melanoma of the skin during the six-month period.

All participants were asked about their use of sunscreen between ages 10 and 29 years.  The reported use of sunscreen between ages 10 and 29 years was compared between the two groups.

Study F

Researchers identified 26 studies which examined the relationship between sunscreen use and the incidence of melanoma of the skin. The data from these studies were combined and showed a clear inverse relationship between sunscreen use and the incidence of melanoma of the skin.

2) A randomized trial (SCREENMAL) was undertaken to examine whether daily use

of sunscreen would decrease the risk of developing melanoma of the skin compared to occasional use of sunscreen. The results are shown in Table 1 below.  Use this information to answers questions 2a), 2b) and 2c).

Table 1: Incidence of melanoma of the skin in persons using sunscreen daily vs occasionally

a)   What is the percentage reduction in risk of developing melanoma of the skin as a result of using daily sunscreen compared with occasional use of sunscreen?

SHOW ALL YOUR WORKING. (4 marks)

b)   How many people would need to use sunscreen daily to prevent ten incident cases of melanoma of the skin? SHOW ALL RELEVANT WORKING? (4 marks)

3) In the SCREENMAL study described above, the researchers interviewed the

study subjects at the end of the study and found out that some of those

randomized to the daily sunscreen group had in fact not used the sunscreen as

often as instructed.  Also, some subjects randomized to the occasional

sunscreen group had in fact used the sunscreen daily. When these subjects were removed from the analysis, the relative risk was found to be 0.45. The researchers argued that this relative risk (rather than the relative risk

calculated as part of Question 2a) above) was a more accurate reflection of the true effect of sunscreen use on the risk of developing melanoma of the skin.

Do you agree with the researchers?  EXPLAIN YOUR REASONING. (4 marks)

Section B (36 marks)

Ariel enjoyed her life under the seas but was always curious about what life on land was really like, despite what Sebastian said.  To stop herself thinking about that,

Ariel decided to conduct a study.  There were lots of naughty prawns living on the

seabed; these prawns were always trying to nip the other sea creatures. Sebastian

lived at Sandy Hollow with his 400 friends. Ursula lived at Caves Catacombs with her 300 friends. Ariel decided to see whether the risk of being nipped by a prawn varied depending on where on the seabed the creature lived.

Ariel arranged for Sebastian and Ursula to contact their friends and ask them to be

involved in the study. With the help of her friend Flounder, 200 of Sebastian’s friends and 200 of Ursula’s friends were randomly selected to be involved in the study.

These 400 participants agreed to be in the study. Fortunately, due to Flounder’s

attention to detail, none of the participants dropped out during the study.  All the participants kept a diary for six months, recording each time they had been nipped by a prawn.

Flounder found that 89 of Sebastian’sfriends and 123 of Ursula’sfriends reported having being nipped by a prawn at least once during the six months of the study.

For all the following questions, use living at Caves Catacombs as the exposure of interest and being nipped at least once by a prawn as the outcome ofinterest.

1) What study type did Ariel use? JUSTIFY YOUR ANSWER. (3 marks)

2) Draw a 2x2 table that adequately summarizes the results of the study. (4 marks)

3) Calculate the relative risk that compares the risk of a sea creature being nipped by a prawn depending on where on the seabed the sea creature lived.  SHOW ALL YOUR WORKING. (5 marks)

4) Explain this result in words suitable for the average resident of Atlantica who has not studied epidemiology. (5 marks)

Ariel and Flounder submitted a paper based on their study to the Annals of Undersea Activities.  In the Limitations section of the paper, the authors noted that there might have been some error because the study participants hadn’t been trained to know

how to decide whether or not they had been nipped by a prawn. The authors argued that even though the sea creature participants were not blind to the exposure, the

sea creature participants didn’t know the study question. This would mean that any error would have been the same for the two study groups and so would have caused a bias towards the null.

5) Do you think the authors are correct in their argument and conclusion? Why or why not? EXPLAIN YOUR ANSWERS. (5 marks).

One of the reviewers of the article criticized the study because during the study some of Sebastian’s friends had actually lived at Caves Catacombs near Ursula instead of at Sandy Hollow near Sebastian.   The reviewer said this meant the study would have underestimated  the  risk  of being  nipped  if  a  creature lived  at  Caves  Catacombs (compared to if they lived at Sandy Hollow).

6) Do you agree with the reviewer? EXPLAIN YOUR ANSWER. (5 marks)

Ariel later conducted another study to test the accuracy of self-reporting of being nipped by a prawn. Fortunately, Ariel had discovered that any time a sea creature was nipped by a prawn it caused the sea creature to make an antibody (the Prawn- Nipped Antibody or PNA) which stayed in the blood. So, Ariel could work out how many sea creatures had truthfully been nipped by a prawn by doing a blood test and checking for the antibody. She enrolled  500 sea creatures in the study, followed them for six months and again asked the sea creatures to keep a diary to note if they had been nipped by a prawn. At the end of the six months, Ariel took some blood from each sea creature and checked for the Prawn-Nipped Antibody. The results of the study are shown below in Table 2.

Table 2: Accuracy of self-report of being nipped by a prawn

7) In this population, what proportion of sea creatures who were truthfully

nipped by a prawn reported that they had been nipped by a prawn? SHOW ALL YOUR WORKING (2 marks)

8) If self-report had been the only way the outcome had been measured in the study shown in Table 2, what proportion of the sea creatures would appear to have been nipped by a prawn? SHOW ALL YOUR WORKING (2 marks)

Ariel did some analyses of other studies that produced a range of results. Answer the

following questions about these results. You do NOT need to show working.

9)

a) The Population Attributable Fraction for the Exposure is 0.16. If there are 1,000 cases in the whole community, how many of the cases would still

have occurred even if the Exposure was removed? (1 mark)

b) The Incidence in the Exposed is 0.004 and the Incidence in the Unexposed is 0.009.  What is the Number Needed to Treat? (1 mark)

c) The Relative Risk is 0.74. What is the Relative Risk Reduction in the study? (1 mark)

d) The Relative Risk is greater than one, the Risk Difference is 0.10 and the Incidence in the Unexposed group is 0.30.  What is the Attributable

Fraction? (1 mark)

e) There is no cure for the disease and the disease is not fatal. There are

1,000 people in the population.  If 20 have the disease at the beginning of

the year and 30 develop the disease during the year, what is the prevalence of the disease (expressed as a proportion) at the end of the year? (1 mark)

Section C (40 marks)

A university student residence, Grayson College, had 200 residents.  After an

outbreak of vomiting and diarrhoea amongst students at the College from 1st March to 7th March 2023, an investigation was undertaken to try to determine what the relevant exposure might have been. It was assumed that the cause would be something the students had eaten or drunk.

All 200 students were reviewed by the College’s health team in the initial stages of the planned study.  The planned study and its purpose was explained to all the students at this initial review.  Eighty of the students reported having had vomiting and/or diarrhoea in the previous seven days. The remaining 120 did not have any vomiting or diarrhoea during that time.

The eighty students who reported vomiting or diarrhoea were asked to take part in the study.  Sixty agreed and 20 declined.  There was no information as to why these 20 declined to be involved.

Sixty comparison students were randomly chosen from the 120 students who didnot have vomiting or diarrhoea.  All but two of these students agreed to participate in the study.

Information was collected from the participants by any one of four experienced interviewers.  The interviewers used a questionnaire that had been previously validated on a university student population and found to be reliable and valid. Participants were randomly allocated to any one of the four interviewers. The interviewers had been trained; testing prior to the study had shown the

interviewers had good intra- and inter-rater reliability and good validity.

Interviewers were aware of whether a student had reported vomiting and/or

diarrhoea.  They were also aware of the purpose of the study. The questionnaire

began with general demographic questions and then asked questions about food and drink intake in the three days before the symptoms developed. Comparison

persons were randomly assigned a ‘symptom’ date within the week of interest and were asked questions about their food and drink intake in the three days prior to

this ‘symptom’ date.

Based on this study description, answer the questions below. MAKE SURE YOU EXPLAIN YOUR ANSWERS.

1) Explain why this is a case-control study. In your explanation, describe the study factor(s), the outcome factor and the study base. (5 marks)

2) Comment on the approach to selection of cases for the study. (5 marks)

3) Comment on the approach to selection of controls for the study. (5 marks)

4) Comment on the measurement of exposure in this study. (15 marks)

5) Comment on the measurement of the outcome in this study. (5 marks)

6) The study found that the main causes of the vomiting and diarrhoea were

contaminated mayonnaise and contaminated garlic.  Twenty per cent of the participants who ate mayonnaise also ate garlic and 60% of the participants who did not eat mayonnaise ate garlic.  If the analysis did not take account of the effect of garlic, would this have meant the estimate of effect of

mayonnaise would be biased upwards, downwards or be unchanged? EXPLAIN YOUR ANSWER. (5 marks).