Flashcards in Epidemiological study design Deck (14):
Give 6 types of study design
Cohort study, case-control, ecological, cross-sectional, RCT, meta-analysis
describe the design of a cohort study
An observational study
Longitudinal study in similar groups but with different risk factors/treatments
Follow up over time (constantly checking in)
(in contrast to case-control, the outcome has not occurred yet)
Describe case-control studies
An observational study looking at the cause of a disease. Compares similar participants w/ disease and those without (controls). Looks RETROSPECTIVELY for exposure/cause (potential RF data is collected - e.g. questionnaires)
The outcome has already occurred at the time of investigation
important idea: looks backwards in time
describe ecological studies
Observational study of a disease or outcome and exposure of interest are measured in a number of populations/groups (e.g. between electoral wards or different hospitals)
describe a cross-sectional study
Observational study collecting data from a population at a specific point in time (no follow-up) - a SNAPSHOT STUDY
Example: the census (conducted every 10 years) --> other examples include prevalence at a specific time (e.g. number of smokers in a GP practice)
describe an RCT
Similar participants are randomly assigned to an intervention or control group to study effect of intervention
What are the advantages and disadvantages of an RCT?
- Low risk of bias and confounding
- blind - reduces possibility of information bias
- High drop out rate
- Ethical issues
- Time consuming and expensive
- prior knowledge required
What are the advantages and disadvantages of an ecological study?
Advantages: can lookout geographical correlations, good for generating a hypothesis
- subject to bias + confounding
What are the advantages and disadvantages of a cohort study?
- Can follow up rare exposure
- Allows to identify risk factors
- Data on confounders collected prospectively
- Large sample size required
- Impractical for rare diseases
- People drop out
What are the advantages and disadvantages of a case control study?
- Good for rare outcomes
- Difficult finding appropriately matched controls
- Prone to selection and information bias (e.g. recall bias)
- Reverse causality (as the outcome interest has already occurred)
n.b. all the observational studies (case control, ecological, cohort, cross-sectional) are subject to bias and confounding
What are the advantages and disadvantages of a cross-sectional study?
- Large sample size
- Provides data on prevalence of risk factors and disease
- Quick to carry out
- Repeated studies show changes over time
- Risk of reverse causality – which came first?
- Less likely to include those who recover quickly or short recovery
- Not useful for rare outcomes
- subject to bias + confounding
- Temporality: impossible to be certain whether an outcome developed before or after the exposure concerned
What is a meta-analysis?
An evidence synthesis where all the data of similar studies are pooled together and a forest plot is created
What is an ethical dilemma of RCTs?
is it ethical to withhold potentially life-saving treatment form the control group?