1/24
Looks like no tags are added yet.
What is Confounding?
The systematic difference between the groups being compared that distorts the true association between an exposure and the disease.
Mixing of effect between and exposure, an outcome, and a third variable known as a confounder.
“An Alternate Explanation for Association”
What is the Source of Confounding Variables?
When exposed & unexposed/cases & controls differ by more than just exposure status.
Can Occur in ALL Types of Epidemiological Studies!
What is the Effect of Confounding Variables on a Study?
Distortion of True Association Between Exposure & Disease

What does Directive Acyclic Graph Represent?
An illustration of a Confounder
True or False: You can “Fix” an Analysis After Study Conduction if a Confounder is Present?
True
What is the Counterfactual Ideal?
An ideal comparison group would be the exact same people who are in the exposed group, had they not been exposed.
Epidemiologists aim to select different sets of people who are as similar as possible!
When does Confounding Occur?
This occurs when the risk of disease in the unexposed group does not equal the risk of disease the exposed group had they been unexposed.
The difference in disease risk between the exposed and unexposed group is due to factors other than just the exposure!
What are the (3) Criteria Confounders Must Meet?
1) Independent Predictor of the Outcome: The confounder is a risk factor for disease among people who are unexposed, confounder is present in both exposed and unexposed populations.
2) Associated with the Exposure: The confounder occurs more or less often among the exposed than the unexposed.
3) Cannot be an Intermediate on the Casual Pathway Between Exposure & Disease: Evaluate these by examining the biological mechanisms for which the exposure is thought to cause the disease. (the confounder cannot be caused by the exposure)
How to Determine Potential Confounders?
1) Understand Your Subject Ares
2) Complete a Comprehensive Literature Review
3) Be Aware of Historical Confounders
If the exposed and non-exposed have unequal amount of influence via a confounder, what can we do during the initial stages of the study to counteract confounding?
Randomization!
What (3) Methods Makeup the Design Phase?
1) Randomization
2) Restriction
3) Matching
What (3) Methods Makeup the Analysis Phase?
1) Standardization
2) Stratified Analysis
3) Multivariate Analysis
What is the Overall Goal of Both Design & Analysis Methods?
The overall goal is to “break” the association between the confounder and the exposure.
What is Randomization?
A method of randomly allocating study subjects to treatment groups so each subject has an equal change of being assigned to the treatment or comparison group.
This only works when 1) the study is large enough and 2) the treatment assignment is not influenced by the investigator.
What is Restriction?
The idea of limiting the study to people who are within one category of the confounder.
Example: If sex is a confounder, limit your study groups to only men or only women.
What is Matching?
A type of method which selects study subjects so that confounders are distributed identically among the exposed and unexposed OR cases and controls.
What is Stratification?
A method of separating your study population into subgroups where one group has the confounder characteristic and one group does not. Within each group, you are then able to calculate measure of association.
What is Multivariate Regression?
This method involves construction of statistical model, requiring a computer, that describes the association between exposure, disease, and confounder.
Strength: Simultaneously adjusts for several variables.
Weakness: Difficult to conceptualize & data needs to fit into an available statistical model.
What is Residual Confounding?
A type of confounding that persists despite efforts to control or adjust for confounding variables.
Sources: Confounders for which no data were collected, inaccurate data on a confounder, use of broad categories of a confounder in your analysis.
This type of confounding should be acknowledged and addressed in the discussion section of a published paper.
What Method of Confounding can we use to Negate the Implications Before the Study?
Randomization
What Method of Confounding can we use to Negate the Implications After the Study (During Analysis)?
Stratification
What are the (5) Steps of Stratified Analysis?
1) Calculate Crude Measure of Association
2) Divide Subjects into the Strata of Confounder
3) Calculate Strata-Specific Measurements of Association
4) Calculate Adjusted Measure of Association
5) Determine if Crude Measure & Adjusted Measure Differ & By How Much (Magnitude)
Is Confounding Present if Crude Rate = Adjusted Rate?
NO!
Is Confounding Present if Crude Rate Does NOT Equal Adjusted Rate?
YES!
What is the 10% Rule?
When confounding occurs, epidemiologists make a judgment call to decide whether it is important.
If rates differ by more than 10% then confounding is deemed important.