Reference no: EM132837803

COMH7200 Epidemiology for Health Researchers - Wits University

Section A

Questions 1-5 refer to the following scenario which we discussed in class in the introduction to epidemiology and public health lecture:

Clinicians in New York and San Francisco alerted local public health officials to a cluster young males who have sex with males (MSM) diagnosed with a form of connective-tissue cancer called Kaposi's sarcoma (KS), which was previously known to occur only in older men.

One hypothesis for the cause of the KS outbreak focused on the use of amyl nitrites, or "poppers," a prescription drug thought to enhance sexual experience and commonly used among MSM at the time. In 1982 a study testing the hypothesis was published in The Lancet. In the study, 20 men from New York City with KS who were referred to participating hospitals from clinics for sexually transmitted infections (STIs) were compared to 40 men without KS identified from a local private practice of a physician who was the doctor of two of the individuals with KS in the study. Both those with KS and those without KS were interviewed on past use of "poppers" as well as other factors potentially associated with the disease. The results are presented in the table below.

Table 1. Association between "Poppers" Use and Kaposi's Sarcoma

Poppers use	KS+	KS-	Total
Frequent	12	6	18
Infrequent	8	34	42
Total	20	40	60

1. What type of epidemiologic study design is this?

A. Case-control study
B. Prospective cohort study
C. Randomized control trial
D. Retrospective cohort study
E. Ecological study

2. Calculate the appropriate measure of effect to test the hypothesis that frequent "poppers" use is associated with KS when compared to infrequent use.

A. Odds Ratio (6 x 34) / (12 x 8) = 2.1
B. Risk Ratio (12/18) / (8/42) = 3.5
C. Odds Ratio (12 x 34) / (8 x 6) = 8.5
D. Risk Ratio (12/8) / (6/34) = 8.5

3. Of the choices below, what is the most appropriate conclusion that you can draw from the measure of effect calculated in Question 2, above?

A. There is no association between exposure to "poppers" and the development of KS.
B. Individuals with KS were more likely to report frequent use of "poppers" than were individuals without KS.
C. KS is caused by frequent use of "poppers."
D. No conclusions about association can be drawn from this type of study.

4. Suppose that it was later discovered that individuals with KS were more likely to underreport their use of "poppers" than were individuals without KS. If this occurred in the above study, what can you say about the measure of effect derived in Question 2 in relation to the "true" association?

A. The measured association (derived in Question 2) would be further from the null than the "true" association between frequent use of "poppers" and KS.
B. The measured association (derived in Question 2) would be closer to the null than the "true" association between frequent use of "poppers" and KS.
C. This would have no effect on the measure of association (derived in Question 2).
D. The direction of bias cannot be determined based on the information in this scenario.

5. Critics were concerned that the men without KS who agreed to participate in the study were different from men without KS who refused to participate. ASSUMING that the men without KS who refused to participate were more likely to frequently use "poppers" than were those without KS who were included in the study, what effect would this selection bias have on the calculated measure of association? Assume for this question that the exposure was measured perfectly.

A. The measured association (derived in Question 2) would be further from the null than the "true" association between frequent use of "poppers" and KS.
B. The measured association (derived in Question 2) would be closer to the null than the "true" association between frequent use of "poppers" and KS.
C. This would have no effect on the measure of association (derived in Question 2).
D. The direction of bias cannot be determined based on the information in this scenario.

Section B - Short Answers

1. Outline the design of a case-control study and a cohort study to examine the relationship between a high-fat diet and bowel cancer?

2. A researcher wished to investigate the effect of soda drinking on the development of pancreatic cancer. The researcher decided that a case control study was the best study design to use to study this association.

The hypothetical data from that case-control study are presented below:

All Study Subjects

	Cancer	No Cancer
Soda	689	360
No Soda	311	714

A number of questions based on this scenario are given below

2.1 Calculate and interpret the appropriate measure of effect for these data

2.2 The researcher was concerned about the role of smoking in the development of pancreatic cancer, as a potential modifier so the following stratification was performed to look at the data separately for smokers and non-smokers.

Smokers

	Cancer	No Cancer
Soda	660	264
No Soda	110	46

Nonsmokers

	Cancer	No Cancer
Soda	29	96
No Soda	201	668

Calculate the appropriate measure of effect for these data

2.3 Does smoking modify the effect of soda consumption on the development of pancreatic cancer? How can you make this out from the data?

2.4 List when and how you handle confounding when you are doing a study.

2.5 Residents of a community in the United States were asked to participate in a study to determine whether cigarette smoking was associated with age-related maculopathy (ARM). At a baseline examination, participants were asked to report their smoking habits. After 5 years, participants had an examination to determine whether they had developed age-related maculopathy. Table 1 presents the number of cases of age-related maculopathy in the study measured at the follow-up examination among the 1682 male participants, aged 43-86 years, who did not have age-related maculopathy (ARM) at the baseline examination:

Table 1: Cases of age-related maculopathy by selected exposures, Flesher Bay, United States

	Smoking	N	Case of ARM
Alcohol Drinkers	Ever smoke	864	200
	Never smoke	368	56
Non-alcohol Drinkers	Ever smoke	250	30
	Never smoke	200	20

Residents of a community in the United States were asked to participate in a study to determine whether cigarette smoking was associated with age-related maculopathy (ARM). At a baseline examination, participants were asked to report their smoking habits. After 5 years, participants had an examination to determine whether they had developed age-related maculopathy. Table 1 presents the number of cases of age-related maculopathy in the study measured at the follow-up examination among the 1682 male participants, aged 43-86 years, who did not have age-related maculopathy (ARM) at the baseline examination:

a) What is the study design? State a reason to support your answer

b) Does this study show any association between smoking and maculopathy? Interpret your result.

c) In Table 1, the data for alcohol drinkers and non-alcohol drinkers were presented separately. What was the investigator trying to assess?
d) What conclusion did the investigator draw after the assessment in (c)?

2.6. What Study Design could one use for the research question below examining ‘The prevalence of malnutrition among children under five in the community.

Please list the advantages and disadvantages of the study design below?

Section C:

In 2019, Akpalu and colleagues sought to explore a gap in the literature of sex differences in stroke in the Sub-Saharan African region. Their publication was published in the journal Stroke and has been posted for you on SAKAI. Theeffect sizes of risk factors of stroke by sex among West Africans were compared. Stroke was the outcome determined by clinical evaluation and brain neuro imaging (CT scan or MRI), electrocardiography (ECG), transthoracic echocardiography and carotid Doppler ultrasound performed according the standard operating procedures (SOP). Ischemic Stroke was sub-typed clinically according to the Oxfordshire Community Stroke Project (OCSP) criteria.

The team employed rigorous statistical methods and assembled a case control study known as

The Stroke Investigative Research and Educational Networks(SIREN). SIREN was conducted at 15 sites in Ghana and Nigeria. Cases were adults over the age of 18 years with CT/MRI confirmed stroke and controls were age-and sex-matched stroke-free adults. Comprehensive evaluation for vascular, lifestyle and psychosocial factors was conducted using validated tools. The team used conditional logistic regression (which you will learn more about in your biostatistics classes) to estimate odds ratios (OR) and reported risk factor specific and population attributable risks (PAR) with 95% Confidence Intervals.

Please answer the questions below based on SIREN

3.1. In the methods pasted below the authors state:

‘Controls were consenting stroke-free adults recruited via robust control recruitment from the community, and participating hospitals. Stroke-free status was confirmed using the 8-item questionnaire for verifying stroke-free status (QVSFS) validated in 3 major languages spoken in West Africa (Ashanti, Yoruba and Hausa).26 Controls were matched by age (+/- 5 years), sex and ethnicity to minimize the potential confounding effect of these variables on the relationship between stroke and the main environmental risk factors ‘

3.1.1. Please comment on how disease free status was ascertained

3.1.2. Please comment on how disease status was ascertained.

3.1.3. Please detail how the controls were matched and provide any limitations for the variables on which they were matched.

3.1.4 In Table 2 at the end of the document, a range of confounding variables are presented. Please comment on the Odds Ratio for Age, education and hypertension in relation to the outcome Stroke and specify for females and males.

3.1.5. Please interpret the PAR for education and hypertension in relation to the outcome Stroke and as per above specify for males and females

Section D

In 2020 Haywood and colleagues examined the relationship between Covid 19 hospitalization and in hospital death amongpatients seen within an integrated-delivery health system (Ochsner Health) in Louisiana, USA between March 1 and April 11, 2020, who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2, the virus that causes Covid-19) The Ochsner Health population within which they work is 31% black non-Hispanic and 65% white non-Hispanic. The paper has been loaded for you on SAKAI

As an epidemiologist working in South Africa please write a mini-essay where you consider this scenario.
Please present the following in your 1 page answer. You will be marked on clarity and coherence of your writing:

1. What study design you will use to examine Covid 19 death and hospitalization?

2. A range of variables operating on the causal pathway including mediators and effect measure modifiers.

3. Examples of potential confounders with motivation for why they are confounders

4. A coherent explanation for race as a social determinant of health and how it is conceptualized in this causal relationship.

Attachment:- Epidemiology for Health Researchers.rar