Observational Studies in CVD Flashcards
What is a cross sectional study?
A study of a sample population at one point in time that provides a descriptive output, particularly prevalence. It is cheap and easy but only provides weak evidence for causality.
What is a case control study?
An analytical study that compares previous exposure between cases and controls. Controls are matched with cases. Retrospective - point of recruitment is after outcome of interest.
What is the main advantage of case control studies?
Can study rare diseases
What is the output of a case control study?
Odds ratio - an approximation of relative risk
What is an odds ratio?
(people with disease who were exposed/people without the disease who were exposed)/(people with the disease who weren’t exposed/people without the diease who weren’t exposed)
What does an odds ratio of 2 mean?
That the exposure doubles the likelihood of the outcome
What is a cohort study?
A longitudinal study that collects incidence data and compares outcomes between subgroups who were exposed to a risk factor and subgroups who were not
What is the outcome of cohort studies?
Relative risks
What is a prospective cohort study?
The study starts after exposure but before the outcome occurs
What is a retrospective cohort study?
The study starts after exposure and outcome occur
What is the benefit of a cohort study?
Can look at multiple outcomes and exposures
What are the difficulties of a cohort study?
it is difficult to study rare outcomes, it is expensive and difficult
What is active outcome ascertainment?
explicitly undertaking surveillance for outcome
What is passive outcome ascertainment?
database linkage done retrospectively
What is bias?
A systematic difference between or among groups that leads to an under or over estimation of true results
What is a selection bias?
A systematic difference in the characteristics of people selected for the study and those not selected or within groups being compared or between groups being compared or between those who stay in the study and those who don’t
How is selection bias minimised?
careful recruitment, maximise response, minimise lost follow up
What is information bias?
A systematic difference in the way information is collected between/among groups
How is information bias minimised?
ensure that methods for collecting information are the same between/among groups
What is confounding?
Where a 3rd variable independently predicts the outcome and is related to the exposure
What are common confounders?
Age and sex
How is confounding minimised?
By matching for age and sex
Can confounding be dealt with in analysis?
yes
Can bias be dealt with in analysis?
no