EPI Flashcards
How do you deal with reverse causality?
Try using a latency period and doing sensitivity analysis for the date. Test moving the index date back.
What is reverse causality?
A does not cause B, B causes A
What is a competing risk?
A competing risk is an event that either hinders the observation of the event of interest or modifies the chance that this event occurs.
What is the requirement to create a good propensity score?
A large sample (rule of thumb is ten times more subjects than variables in your PS)
Complete data on important confounders
What is confounding by indication?
Confounding by indication is a special type of confounding that can occur in observational (non-experimental) pharmaco-epidemiologic studies of the effects and side effects of drugs. This type of confounding arises from the fact that individuals who are prescribed a medication or who take a given medication are inherently different from those who do not take the drug.
What is selection bias?
The association between exposure and disease differ between the participants in the study and the non-participants due to the selection of subjects or their participation( i.e. drop-off)..
Information bias
Systematic error in a study due to erroneous information collected about (or from) the study subjects
Define confounding
Open “back door path” between exposure and outcome. The confounding factor is associated with both exposure and outcome but is not a mediator.
How do you deal with confounding by indication?
Adjust for disease severity (disease severity scores).
Propensity score.
Drawbacks of using Propensity scores
Groups might look the same, but unaccounted for confounders may be very important.
If you match on PS, then only those that overlap are included, may not be representative.
Less transparent, what dealt with the confounding
Not possible to assess different exposures at the same time?
What can PS be used for
Matching, adjusting, stratifying, weighting
How do you use a PS
- Model the exposure variable in a regression as a function of potential confounders. This calculates the predicted probability of exposure for every individual as a function of these covariates.
- Apply the propensity score by matching, stratifying, controlling or weighting.
What are the disadvantages with using classification criteria?
Classification criteria generally have a high level of specificity at the expense of somewhat lower sensitivity. Thus using classification criteria might lower the external validity (generalizability)
How do you test for proportional hazards assumptions?
Plot the hazard curves and inspect visually. Stratify on the time-scale used. Introduce an interaction term between the exposure and time
Inverse probability of treatment weighting
Alternative to multivariable regression modelling. Probability of receiving treatment is first modelled with PS (with logistic regression) then this is used to weight the subjects. Key assumptions are that all confounders have been measured and properly modelled in this treatment model.
