Flashcards in L18. Observational Studies in CVD Deck (17):
What is a cross sectional study?
Where a sample of a population is taken at one point in time and contribute only once to data (no follow up)
What are the advantages and disadvantages of cross-sectional studies?
A: cheap and easy
D: require a representative sample, can only explore associations (no temporal relationships), only weak evidence of causality
What is a case-control study?
A comparison between a previous exposure status and finding out whether there was an outcome or no outcome as a result
Often matching controls with cases by confounders
Why is a case-control study retrospective?
A retrospective study because we ask people already with the outcome about whether they were exposed or not
What are the advantages and disadvantages of case-control studies?
A: get explicit knowledge of temporal relationships, useful in studying rare outcomes (because this is the starting point)
What is an odds ratio?
The equivalent of risk ratio but an estimate because it is not a longitudinal study.
= Odds of exposure to non-exposure in cases divided by the odds of exposure to non-exposure in control cases.
Table of values
What is a cohort study?
LONGITUDINAL (follow up) studies where we collect incidence data by comparing the outcomes of exposed to not exposure people to find outcomes later in time.
What are the differences between prospective and retrospective cohort studies?
Prospective: start at the exposure (before an outcome has occurred) and follow subjects over time and measure the incidence as it occurs
Retrospective: Start after outcomes have occurred and look back on whether or not exposure occurred or not
What are the advantages and disadvantages of cohort studies
A: give explicit info about temporal relationship between exposure and outcome, can include multiple exposures and outcomes (post-hoc research hypothesis)
D: difficult in studying rare outcomes, not cheap or east
What are two was outcomes are ascertained in cohort studies?
Active: asking them to participate in a second encounter
Passive: looking up their date (eg. hospital records)
What is Selection Bias?
Systematic differences in the characteristics of people selected for studies and those not selective and thus reduces the generalisability of the study results.
It is also a result of knowing what groups are receiving what treatments
How can selection bias be minimised?
Careful recruitment, maximising the response (sample size) and minimising lost to follow up numbers
What is information bias?
Systematic differences in the way information (results) are collected and tends to be a result in subjective results.
What is recall bias?
When subjects have subjectivity in recalling information based on what they think the 'correct' answer is or because they've forgotten.
How can information bias be minimised?
Standardising the ways information is collected (objective collection if possible)
What is confounding?
What are some universal confounders?
When there is an external factor that is causing a relationship between two variables (not the cause effect, but a commonality or conincidental factor)
Age and Gender