Studies Flashcards
Cross-sectional study: about?
Descriptive and observational.
Measures both exposure and/or outcomes at a single point in time.
Cross-sectional study: purpose?
- Often used in surveys for large populations (e.g. census, survey)
- Describe and compare prevalence
- Generate hypotheses
- Plan (resource allocation)
Cross-sectional study: what can it calculate?
Prevalence
Cross-sectional study: affected by?
Duration and incidence
Cross-sectional study: strengths and limitations
Strengths:
- assess multiple exposures and outcomes
- can be used to calculate prevalence, distribution of prevalence in the population, and hypothesis generation
- inexpensive and quick
- may suit research question
Limitations:
- no temporal sequence (exposure and outcome measured at the same time)
- can’t measure incidence or measures of association
- not good for rare exposures/outcomes
- not good for assessing transient variable exposures/outcomes
Ecological study: about?
Descriptive and observational.
Compares exposures and outcomes across GROUPS not individuals.
Ecological study: purpose?
- To compare exposure and disease status across populations
- Assess population level factors in disease
- Hypothesis generation
Ecological study: strengths and limitations
Strengths:
- assess population level exposures (e.g. UV light, pollution)
- can be used for hypothesis generation
- inexpensive and easy
- may suit research question
- data is often routinely collected
Limitations:
- ecological fallacy (making assumptions about individuals based on data from the group they belong to)
- cannot control for confounding
- cannot show causation
Cohort study: about?
Analytic and observational.
Observe people’s exposures and what happens to them.
Cohort study: purpose?
- Identify a source population.
- Recruit a sample population who don’t have the outcome of interest.
- Assess their exposure level and categorise participants into appropriate group.
- Follow up over time and see who develops the outcome.
- Calculate your measures of association and occurrence.
Cohort study: what can it calculate?
O: Incidence proportion and incidence rate.
A: Relative risk and relative difference.
Cohort study: considerations?
- The healthy worker effect.
- Ensure our sample population doesn’t have the outcome.
- Ensure our participants have been correctly classified into exposure groups (and haven’t changed exposure groups during the study period).
- Loss to follow up - did any participants leave the study?
- Has the outcome been classified correctly?
Cohort study: strengths and limitations
Strengths:
- temporal sequence
- can look at multiple outcomes
- can calculate incidence and therefore measures of association
- good for rare exposures
Limitations:
- loss to follow up and its associated bias
- potential for exposure/outcome misclassification
- time consuming
- expensive
- not good for rare outcomes
Historical cohort study: about?
Start at the end of the follow-up period.
Historical cohort study: purpose?
Splits study individuals by their exposure status - already knowing the outcome.
Existing data used to recreate study how it would’ve gone as if it were a real one.
Historical cohort study: strengths and limitations
Strengths:
- good for rare outcomes or outcomes that take a long time to develop
- inexpensive
- quick compared to prospective cohort studies
Limitations:
- could be problems with existing data quality
- selection bias
- incomplete information and may not know about all relevant factors
Case-control study: about?
Analytic and observational.
Another example of an analytic and observational study to find an association between exposure and outcome. Looks at outcome status first and the finds out exposures (opposite of cohort studies).
Case-control study: purpose?
- Identify a source population.
- Identify the cases (people with the outcome) and controls (people without the outcome).
- Assess their prior exposure level in cases and controls.
- Calculate your odds ratio (measure of association).
Ideal for rare, slow to develop and transient outcomes.
Case-control study: what can it calculate?
No measures of occurrence!
Only odds ratio - a measure of association.
Can treat as RR when reporting.
Case-control study: affected by?
If pick more people incidence and prevalence goes up, and vice versa.
Case-control study: considerations
- Index dates.
- When selecting cases, needs to be clear definition of the outcome and cases should be newly incident.
- The purpose of having a control group is to estimate the prevalence of exposure in the population from which the cases come from.
- Controls must be capable of becoming a case, can often have multiple controls for better statistical power.
- Several ways we can select controls - one is from the hospital. However, this can cause problems as they aren’t necessarily representative of the population which the cases came from (they may have diseases which are also related to the exposure being studied).
- Need to ensure the exposure is measured before the outcome occurred (so that we have temporal sequencing).
Case-control study: strengths and limitations
Strengths:
- good for rare outcomes and transient exposures
- can assess multiple exposures
- temporal sequencing
- quick
- inexpensive
Limitations:
- can only study one outcome
- difficult to select an appropriate control group
- prone to selection and recall bias
RCT: about?
Analytic and interventional.
At the top of the hierarchy of evidence, an analytic intervention study.
RCT: purpose?
- Identify a source population.
- Randomly select the sample population who don’t have the outcome of interest.
- Randomise the sample to either the intervention or control groups.
- Follow up over time and see who develops the outcome.
- Calculate your measures of association and occurrence.
RCT: what can it calculate?
O: Incidence proportion and incidence rate.
A: Relative risk and relative difference.
RCT: important aspects
- Randomisation (random allocation) = randomly assign to either control or intervention group. NOT the same as random selection. Purpose of this is to control confounding (when another variable like age or sex distorts the relationship between exposure and outcome).
Variants of randomisation:
1. Cluster randomisation - where subgroups are randomised instead of individuals.
2. Stratified (block) randomisation - where participants are randomised blocks.
How do we protect randomisation and its benefits?
- Concealment of allocation - meaning the person randomising participants does not know what the next treatment allocation will be (concealed and unpredictable). This also prevents selection bias from the researchers selecting the treatment they want.
- Intention to treat analysis - analyse as we randomise so we aren’t swapping between groups. Gives us a ‘real world effect’.
- Per protocol analysis - analyse participants who fully complied to the study protocol. This can show the efficacy of the treatment, but we lost the benefits of randomisation.
Bias in RCT
- Lack of blinding
- Loss to follow up
- Non-adherence (when participants are not good at following instructions - this can lead to bias within the results)
RCT: strengths and limitations
Strengths:
- gold standard study design to test an intervention and cause effect associations
- eliminate confounding and bias (if done well)
- calculate incidence and measures of association
Limitations:
- many exposures cannot be randomised and we need to have clinical equipoise
- expensive
- sometimes participants aren’t representative of the general or source population, therefore are not generalisable
- not good for rare outcomes
- exposures must be modifiable
- often funded by big pharma wanting profit bias
