Case-control studies

Question 1

What are case control studies?

Answer 1

Case control studded address some of the issues with doing a cohort study. They’re designed for rare/slow to develop outcomes and can efficiently examine acute or transient exposures
Measured by odds of exposure (cases)/odds of exposure (controls)

Question 2

Cohort vs case-control studies

Answer 2

Cohort: Measure exposure status, follow over time, compare development of outcome (incidence)
Case-control: Identify people with outcome, find people without outcome, compare exposure likelihood beforehand (odds)
Ascertain exposure status then find out outcomes vs. ascertain outcome status, then find out exposures

Question 3

How to complete a case control study

Answer 3

1. Identify source population
   
   2. Identify people with outcome (cases)
   3. Sample people without outcome from same population (control)
   4. Measure exposure prior to outcome in cases and controls
   5. Compare odds of exposure to calculate measure of associated (odds ratio)

Question 4

What are we trying to find in case control studies?

Answer 4

Is the exposure more or less likely with the outcome (case) than without (control)?

Question 5

How to calculate odds in case control studies

Answer 5

To calculate odds use the 2x2 table from the gate cycle
Odds of exposure in cases = a/c (cases with exposure/cases without exposure)
Odds of exposure in controls = b/d (no. of controls with exposure/no. of controls without exposure)

Question 6

What are odds?

Answer 6

Odds are a measure of associated/likelihood. Ratio of odds is used instead of incidence. How many times as likely cases are to have the exposure compared to controls
Null value for odds ratio is 1

Question 7

How do you calculate the odds ratio?

Answer 7

To calculate the odds ratio you divide the odds of exposure in cases (a/c) by the odds of exposure in controls (b/d)
Odds Ratio = (a/c) / (b/d)

Question 8

How do you interpret the odds ratio?

Answer 8

People with outcome are x times as likely to have had the exposure than people without the outcome. When disease is rare, OR approximates the RR hence can interpret OR just like RR: the exposed group were value times as likely to develop the outcome compared to the comparison group

Question 9

Problem with case control

Answer 9

Controls don’t have an outcome so when do you measure exposure for controls
Solution: use index dates, we say that the person develops the outcome the day the case is discovered - relationship between cases and control we recruit
Use cases diagnosis to work out when to measure control

Question 10

How do you select your cases?

Answer 10

Usually try to identify incident cases - sometimes just recruit people with the outcome (prevalent cases)
Case control studies defined by outcome - so only one per study, important clearly define and identifiable (ensure no one can be a case and control)

Question 11

How do you select your controls?

Answer 11

Need to represent the explore distribution of people without the outcome in the source population, must be capable of becoming a case, often select multiple controls per case for statistical power

Question 12

Exposure measurement

Answer 12

Exposure measurement must be comparable: dead cases vs alive controls, interviewers may act differently for cases and controls, case trying to work out what made them sick, outcome may affect recall ability

Question 13

Strength and limitations of case control studies

Answer 13

Strengths: rare outcomes, transient exposure, multiple exposures, temporal sequencing, often comparatively quick and inexpensive
Limitations: usually can only study one outcome, different to select appropriate control group, can be susceptible to selection and recall bias