Confounding and Effect Modification Flashcards
What is Confounding?
A 3rd variable (characteristic related to study subjects) that distort an association (RR/OR/HR) between exposure and outcome.
An alternative explanation of the association.
What is a second way to state confounding?
A mixing of effects; an association is distorted due to the 2 being mixed with another factor that is also associated with the outcome at the same time
What is a third way to state confounding?
A confusion of effects, where effect of exposure is distorted because the effect of an extraneous factor is mistaken for the effect of the exposure
What are the three requirements of confounders?
- Independently associated with the exposure
- Independently associated with the outcome
- Not directly in the casual-pathway linking exposure to outcome (independent)
What is the first step in testing for confounding?
Calculate crude measure of association (OR/RR) between exposure and outcome
AKA – unadjusted
What is the second step in testing for confounding?
Calculate outcome measure of association (OR/RR) between Exposure and Outcome for each individual strata of the 3rd variable (potential confounder)
Create a weighted-average of all strata
AKA – adjusted
NOTE: Authors must disclose this
What is the third step in testing for confounding?
Compare crude vs. adjusted RR/OR of measure of association between exposure and outcome
At what measure of difference between crude and adjusted estimates (RR/OR) is confounding present?
15%
10-20%
What are the two main impacts of confounders?
- Magnitude of Association
2. Direction of Association
Define Magnitude of Association
Association is either more or less extreme than the adjusted crude association
Define Direction of Association
May produce an association in an opposite direction
Towards or away from a null
What is the purpose of controlling for confounders?
To get a more precise (accurate), truer-estimate of the measure of association between exposure and outcome
What are the two main ways to control confounding?
- Study Design Stage
2. Analysis of Data Stage
In the Study Design Stage, what are the three ways to control confounding?
- Randomization (Simple or Stratified)
- Restriction
- Matching
In the Analysis of Data Stage, what are the two ways to control confounding?
- Stratification (w/ weighting)
2. Multivariative statistical analysis (Regression analyses)
What is Randomization?
Hopefully allocates an equal number of subjects with the known (and un-assessed) confounders into each intervention group
What are the strengths of Randomization?
- With sufficient sample size, will likely be successful in serving its purpose to make groups equal
- Stratified version more precisely assures equal-ness
What are the weaknesses of Randomization?
- Sample size may not be large enough to control for all unknown or unassessed confounders
- Process doesn’t guarantee successful, equal allocation between all intervention groups for all known and unknown confounders
- Practical only for interventional studies
What is Restriction?
Study participation is restricted to only subjects who do not fall within pre-specified category/ies of the confounder
What are the strengths of Restriction?
- Straight forward, convenient and inexpensive
- Does not negatively impact internal validity
What are the weaknesses of Restriction?
- Sufficiently narrow restriction criteria may negatively impact ability to enroll subjects (reduced sample size)
- If restriction criteria is not sufficiently narrow it will allow the introduction of residual confounding effects
- Eliminates researchers’ ability to evaluate varying levels of the factor being excluded
- Can negatively impact external validity (generalizability)
What is Matching?
Study subjects selected in matched-pairs related to the confounding variable, to equally distribute confounder among each study group
What are the strengths of Matching?
- Intuitive
- Some feel it gives greater analytic efficiency
What are the weaknesses of Matching?
- Difficult to accomplish, can be time consuming, and potentially expensive
- Doesn’t control for any confounders other than those matched on – over-matching is possible and will mask blunt findings
