Introduction to Cohort Study Designs Flashcards
What is a cohort study?
Get exposed/unexposed outcome-free individuals – time – then record: person-years at risk, how many devel outcome, incidence rate, incidence rate ratio (IRR)
What is the difference between a historical vs concurrent cohort study?
H = exposed/unexposed identified from the past. C = exposed/unexposed identified now
What is a cross-sectional survey?
data collected from a population, or a representative subset, at a specific point in time
What is the incidence rate ratio?
Allows the comparison of incidence rates between two groups of individuals within a cohort
What would be an ideal study?
Compare ‘like with like’ – where groups are only differing in the exposure of interest
What is relative risk?
Risk level in regards to another = males are twice as likely as females = ratio (with no units)
What is absolute risk?
Risk alone
What are the advantages on a cohort study over routinely available data?
You can study exposures and personal characteristics that are not routinely collected. You can obtain more detailed information on outcomes and exposures. You can collect additional data on potential confounding factors
Why use a prospective cohort study?
Wait to establish for certain that the individual does not already have the outcome in question
What is a historical/retrospective cohort study?
longitudinal cohort study that studies a cohort of individuals that share a common exposure factor to determine its influence on the development of a disease, and are compared to another group of equivalent individuals that were not exposed to that factor.
Why are cohort studies better than case-control studies?
studying a range of different outcomes, studying a rare exposure, establishing that exposure(s) precede outcome(s)
What are the disadvantages of cohort studies?
Large, expensive, long-time, risk of high number of losses to follow up = survivor bias, results take a long time, difficulty with confounding
Why use a historical/retrospective cohort study?
Conducted on a smaller scale, require less time, better for analysing multiple outcomes, they can potentially address rare diseases, which would necessitate extremely large cohorts in prospective studies.
