Assignment on Epidemiology

Please put your full name and ID number on this assignment.When you save your file, please use this naming convention – it makes it much faster for us to mark:  Assign 4 First Name Last Name ID number.

1)The population at the North Pole is increasing.Mrs. Claus, an epidemiologist, knows that there are a few different ways that could happen, but, regardless of the reason, one measure must be increasing if there is no immigration.  Which one is it?  [1 point]

2)Santa’s reindeer have been quite prolific, with 100 births in the past year out of a total of 2000 female reindeer.There has only been two deaths, a baby reindeer that died 2 hours after birth and one that died at 2 months old.  None of the mothers died.

 
Santa commonly stops in a remote part of Greenland to give the reindeer a breather.

Last time he was there he was catching up with the reindeer there and they mentioned that they had 10 births in the past year, out of a total of 100 female reindeer.  4 babies died – the first within a day, the second within 2 weeks, the 3rd at 6 weeks and the 4th at 6 months.

a)Compare the infant mortality rate at the north pole and in Greenland.  Show all of your work.

b)Compare the neonatal mortality rate at the north pole and in Greenland.  Show all of your work.

c)Which population had higher rate of infant mortality and neonatal mortality?  Why might this be the case?

d)Why is it important to track measures of infant mortality and neonatal mortality?

3)A number of volunteers are followed for 1 year to see if exercise is associated with cancer.  [2 points]

a)What type of study is this?  (Circle the answer)

i)Cross-sectional

ii)RCT

iii)Ecological study

iv)Cohort

v)Case Control

b)There is a glaring problem with this study.  Which concept is linked to the problem?  (Circle the answer)

i)Selection bias

ii)Latency period

iii)Temporality

4)Suzy learns that a lot of kids are home from school because of gastro-intestinal illness.

The school has 1,500 kids in it, so she doesn’t want to interview everyone. She interviews all of the kids that are sick and randomly selects some children who are well and interviews them too. She collects the following data.  [14 points]

a)What type of study design is this? What features in the description tell you that it is this type of study and not one of the other types?

Question 2

b)Create the appropriate 2x2 table to investigate eating ice cream at the track and field day and illness.

c)Pick the appropriate measures of association based on the study design (eg prevalence & prevalence ratio, incidence & risk ratio, or odds & odds ratio) and calculate. Interpret the results. Show your work.

d)Why did you use those particular measures and not the others?

Can you use any of the other measures? Why or why not? 

5)A member of the public has contacted you asking if there is a causal relationship between wrinkles and heart disease.

You review the literature finding only one study. Dr. Charlatin performed a case control study to examine the relationship between wrinkles and heart disease.

An odds ratio of 5.2 (95% Confidence Interval: 4.7 – 5.8). [18 points]

a)What type of bias could account for this finding (the OR = 5.2), assuming that the true OR = 1.0?  What would have had to have happened in his study for it to have happened?

Be sure to describe the effect on the 2x2 table, the odds in cases and odds in controls, and the OR.

You go back and read the researchers PhD thesis (what this study was based on).  Dr. Charlatin did a sensitivity analysis and found that compared to people with 10-25 wrinkles (ie this is the reference category):

The OR between wrinkles and heart disease for those with 10-25 wrinkles  = 2.7

The OR between wrinkles and heart disease for those with 26-50 wrinkles  = 4.9

The OR between wrinkles and heart disease for those with 50+ wrinkles  = 7.4

b)Could confounding account for these findings?  If so, what factor could it be and how do you know that it’s a confounder?  State all assumptions.

c)Go through the Bradford-Hill criteria.  Which ones are fulfilled?  Why or why not?

d)Based on your evaluation of the evidence in parts a-c, what would you tell the member of the public?
.
6)You want to do a study examining the causal relationship between exposure to tanning beds in youth and leukemia.

Exposure to tanning beds between the ages of 12 and 24 years is relatively common in the general population.

Leukemia is relatively rare form of cancer.  You assume the  latency period for leukemia is 30 years.

The study designs that are of interest are cross-sectional, retrospective cohort, prospective cohort, mixed cohort and case control study.  [26 points]

a)Go through each of the following concepts.

How would each of these guide what study design you choose?

What type of design would be best given that particular concept and why?  State any assumptions you are making.

Example: Temporality. 

Having temporality means that I know that the exposure happens before the outcome.

I want to determine causation, so I want to make sure that temporality is intact.This means that a cross-sectional study would not be as good a choice as a cohort or case control study in respects to temporality.

i)Latency period    
            
ii)Exposure prevalence    
        
iii)Disease prevalence

iv)Likely accuracy of self-report exposure data

v)Availability of pre-existing exposure data.

vi)Availability of a pre-existing cohort.

vii)Likely accuracy of self-report outcome data (Be sure to speak to both study design and data collection method here)

viii)Availability of accessible outcome data

ix)Practical aspects, time and resources

b)Study design is all about making tradeoffs.  Which study design did you decide to pick and why? (i.e. which tradeoffs did you choose and why?)

c)Based on the above, how would you select the participants and gather your data?

d)Who would the results be generalizable to?