Case control study designs Flashcards
What are case control studies
They are observational
where natural events occuring indivduals with diease /condition of interest (cases) are passivley observed and compared with people who don have conidition of interest (Controls )
Group assignement based on dieased status
“randomly selected = observational study”
Why would we do a case control study?
Unable to force a group allocation (randomize)
Limited reasources
DIease of interest is rare in occurrence and little is known about its assocation and causes
How do retrospective and prospective studies compare?
Retrospective - where our outcome is already known
- Case studies are customarlily conductive in a retrospective fashion
Prospective - the outcome is NOT ALREADY known
- All intervential studies are prospective
What are the strengths of case control study?
- Multiple exposures of one outcome
- rare dieases
- determining assocations NOT CAUSATION
- less expensive than interventional trials and prospective cohort studies
- long induction / latent peroid
- useful when Ethical issues limit intervential studies
What are weaknesses of case control studies ?
Cant show Causation
Impacted but unassessed confounders
retrospective cant control for other exposures or potential changes in amount of study exposure during the study frame
can be impacted by various biases and specificalyy SELECTION AND RECALL/ASSESSMENT BIASES
limited by available data (retrospective design flaw)
how selerction of cases done in a case control study?
defined by investigator using accurate, medically reliable, efficent data sources
objectively, consistantly, accurately, and w/ validity applied to all particpants
classifying patients correctly is ideal but present is the risk of misclassifying subjects into wrong group
Que es Counterfactial theory?
All else being equal (in the same group) the outcome if something didnt occur
When comparing to groups remember the assumption of exchangeability and that exchangeability = comparability
How do you select you controls?
- Most difficult part
- Controls repersent the baseline risk of exoisure in general pop.
- The way the controls are slected is a major determinant in whether any conclusion is valid
- Internal validity and selection bias
- Groups showl be as equal of possible except the presense of disease outcome of interest
- If the exposure truly has no effect the odds will be exactly the same for both groups or will be 1.0 (no difference)
- CONTROLS SHOULD ONLY BE SELECTED IF THE OUTCOME IS PRESENT NOT THE EXPOSURE STATSUS
What sources should your controls be selected from?
- can be slected from several
- population (state/community,neighborhood)
- can be obtained via numerous avenues, even randomly
- Institutional/organizational/provider
- illness(es) of controls should be unrelated to exposures being studied
- Spouses/Relatives.Friends
- Genetic environmental socio-economic
What is control and what is a case?
A case is a presence of dieases or condition
A control is people who do not have the conditon of interest
What is case cross over design?
In a situation of a brief acute change in risk of outcome in interest (hazard peroid) subjects are their own controls during the other times they dont have the acture change in risk ( this quality can also be assocated with out breakw ith multiple exposures)
It is the only design abie to adequately attempt to address the issue of temportality
What is a nested case control study?
Case control studies conducted after or our of a prospective previous study type (cohort or intervential)
- Subjects that eventually develope diease or outcome (outside previous studies peroid) are dfeined as cases for the next case-control study
How are controls sampled in a nested case-contol study?
- Survivor sampling
- Sample of non dieased indivduals (survivors at end of study peroid)
- Base sampling
- Sample of non dieased indivduals at start of study peroid
- Risk set sampling
- sample of non diseased indivduals during study peroid at same time when case was diagonsed
What are common bias related to the way subjects are chosen for study?
Selection bias
- related to the way subjects are chosen for study (usually more of a concern for contorl selections)
recall bias
- is related to the amount/specificity that cases or control recall past events differnetly
- its more common that cases will remember compared to controls
What the matching schemes involved in studies?
indivdual matching
- Indivduals are matched based on specific characteristics
- controlls confounding
Group matching
- Proportion of cases and proportion of controls with idnetrical characterisitcs are matched
- ex: 41% cases are male so 41% controls are male
- requries cases be selected first
NO MATTER WHAT DONT MATCH ON ANYTHING THAT MIGHT BE A RISK FACTOR
