Epidemiological Study Design Cheat Sheet 1

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

Case Control Greg Martin’s study design cheat sheet

For research methods teaching

go to www.learnmore365.com Cohort
Studies Exposures Outcomes Studies
Cohort Outcomes
Overview Exposed
Cases
Retrospective Overview

disease
Observational study

Have

Identify cases (with known outcome / disease) Prospective


Identify controls (similar to cases but don’t have the Observational study
outcome / disease) Identify specific exposure of interest (can be rare)
Collect information about history of exposures identify a cohort within which you can distinguish between
Compare the two groups with regards to the odds of each those exposed and those not exposed
exposure Follow over time Not exposed
Controls Compare with regard to multiple possible outcomes
Strengths
Strengths

disease
Quick
Cheap Strong form of evidence

No

Can consider rare diseases Can consider rare exposures


Can consider multiple exposures Can consider multiple outcomes

Weaknesses Weaknesses
Not considered to be a strong form of evidence Expensive
Not great for considering rare exposures
Retrospective Takes a long time to undertake study Prospective
Present future
Past Present

Randomised Ecological
Control Trial Groups Outcomes Studies
Overview Overview

Average life expectancy


Intervention
Prospective Observational study
Interventional study The unit of observation is a group (population or

for countries
Participants are randomly allocated into either a treatment community) instead of an individual
(intervention) or placebo (non-intervention) arm of the
study. Groups followed up over time Strengths
Comparison of the groups with regards to predefined end Sometimes it is the only way to address a particular
points is done in the analysis Random research question
allocation
Good for hypothesis generation
Strengths into groups Usually quick (if secondary data analysis is possible)
Very strong form of evidence Usually cheap (where secondary data is available)
Random allocation means that other than the
intervention itself, there should be no systematic Placebo Weaknesses
differences between the two groups and therefore no The profile of the groups may be different (e.g. age
confounding variables that need to be controlled for. and exposure to risks) making the interpretation of
I.e. RCTs control for known and unknown comparisons difficult
confounding factors Ecological fallacy: If inference about individuals are

Weaknesses
incorrectly made based observations at a group
level, we call this error in logic, “ecological fallacy”
GDP per capita for countries
Expensive Prospective
Takes a long time to undertake study
Present future

You might also like