Study Designs Flashcards
This study measures the prevalence of a certain outcome in a population at a point in time.
Cross-sectional studies
This study measures exposure and outcome at the same time at an individual level.
Cross-sectional studies
This study compares exposure and outcomes across large groups.
Ecological study
What studies can be used to generate hypotheses?
Cross sectional and ecological studies
What are some limitations of cross sectional studies? (3)
1) cannot measure incidence
2) not good for rare outcomes and transient exposures
4) cannot establish temporal sequence (as outcome and exposure measured at the same time)
What are some strengths of cross-sectional studies? (4)
1) used to generate hypothesis
2) can measure multiple exposures and outcomes at a time
3) measures prevalence
4) quick and cheap
This is a concept that describes when characteristics from a group level are attributed/applied to the individual
Ecological fallacy
What are the strengths of ecological studies? (4)
1) can compare between groups
2) good for population level exposures
3) quick and easy to do (data is routinely collected)
4) good for considering hypothesis
What are some limitations of ecological studies? (3)
1) ecological fallacy (also that every data point is a population)
2) confounding
3) cannot show causation
This study is when individuals are defined by their exposure status to a suspected risk factor.
Cohort study
What study would be appropriate to use if you were trying to figure out the effects of living in a decile 10 environment on influenza rates throughout 5 years?
Cohort study
- because it involves an exposure and a follow up time.
- those are the 2 main aspects of a cohort study
What are the 2 types of cohort studies and what is the difference between both?
1) Prospective cohort studies - start with exposure and see if the person will develop the outcome (through follow up)
2) Retrospective cohort studies - start with an outcome of interest and work out from then what kind of exposures (in a person’s life) would have led to the outcome.
- working backwards essentially
- still focused on the exposures
Why might someone not be considered eligible to be in a cohort study?
1) they already have the outcome (you must be free of outcome to be part of prospective cohort study)
2) they have outcome that cannot be developed (like prostate cancer)
What are some strengths of a cohort study? (4)
1) can establish a temporal sequence
2) can examine multiple outcomes from an exposure
3) can calculate incidence (so RR and RD as well)
4) good for studying rare exposures
What are some limitations of cohort study? (4)
1) loss to follow up
2) misclassification of exposure/outcome
3) not good for rare outcomes and transient exposures
4) time consuming and expensive
What are the benefits and limitations of a retrospective cohort study?
Benefits: less time consuming, good for slow to develop conditions, less time consuming
Limitations: may not provide all reliable sources of information
This study is designed for rare outcomes and can efficiently examine acute/transient exposures.
Case-control studies
What’s the difference between cohort study and case-control study?
Cohort study - compares outcome likelihood from exposure. Measures people by their exposure status, participants in sample population must be outcome free when they enter the study. Measures RR/RD
Case-control study - compares exposure likelihood; it measures the odds of exposure in a person who has and does not have the outcome. Is exposure to this more likely in people with the outcome? Participants from sample population are categorised by whether they have the outcome (cases) or they don’t (controls) Measures OR
What are some strengths of case controlled studies? (3)
1) good for rare outcomes and transient exposures
2) good for examining multiple exposures
3) can establish temporal sequencing
What are some limitations of case-controlled studies? (3)
1) can only select one outcome
2) difficult to select control group (remember: control group must be representative of source population)
3) susceptible to selection and recall bias (esp. when trying to figure out exposures)
This is an interventional study that measures the effectiveness of medicines/health interventions
RCTs / randomised controlled trials
In RCTs, what is the main way of ensuring all groups in the study are as similar as possible?
Randomisation/random allocation into case and control.
This concept states that in order to preserve the benefits of randomisation, we must analyse participants as randomised. This can more accurately reflect real world effect of intervention.
Intention-to-treat analysis
What are the 4 potential sources of bias in RCTs?
1) loss to follow up
2) non-adherence
3) lack of concealment of allocation
4) lack of blinding
This describes a genuine uncertainty about benefit or harm of the intervention.
Clinical equipoise
What is the best design to use for determining causation?
RCTs
What is the best design to study rare outcomes and transient exposures?
Case-controlled studies
What is the best design to study multiple outcomes and rare exposures?
Cohort studies
What study would be best to use for outcomes that take a really long time to develop?
Retrospective cohort studies
What is the best study design to use for testing an intervention?
RCTs
What are some limitations of an RCT study? (at least 3)
1) not all exposures can be randomly allocated
2) loss to follow up
3) difficulty in blinding and concealment
4) need to have clinical equipoise
5) highly selective comparison group
6) resource intensive
7) can affect generalisability
What is the best study design for comparing between two groups?
Ecological study
What study design would be best to use for measuring an exposure and outcome at the same time?
Cross sectional studies
What study design is used to measure the prevalence of an outcome in a population?
Cross-sectional studies
What is the main measure of association used in Case Control Studies?
Odds ratio
