Classes #11-#12: Confounding & Effect Modification Flashcards Preview

Epidemiology -- Zach H. > Classes #11-#12: Confounding & Effect Modification > Flashcards

Flashcards in Classes #11-#12: Confounding & Effect Modification Deck (15)
Loading flashcards...
1

What is a confounding variable?

A 3rd variable that distorts an observed relationship (association; RR/OR/HR) between the exposure and the outcome (disease).

***Looking at exposure and outcome***

2

What characteristics must a 3rd variable have to be a confounder?

The 3rd variable must be associated (related/correlated) with the exposure and the outcome of interest, yet independent of both...
-but not directly in the hypothesized causal-pathway between the exposure and the outcome

3

What are the 2 main ways that confounders impact the study?

1) Intensity/Magnitude/Strength
-produces an association more or less extreme than true association

2) Direction
-produces an association that moves true association in a positive or negative direction
>towards or away from a null association (RR/OR/HR = 1.0)

4

How do you go about knowing if confounding is present?

Step 1) Calculate crude outcome measure of association (OR/RR) between exposure and outcome
-commonly called "unadjusted" association

Step 2) Re-calculate outcome measure of association (OR/RR) between exposure and outcome while statistically controlling the effects of the confounder
-commonly called "adjusted" association

Step 3) compare the crude vs. adjusted measures of association between the exposure and outcome
-the crude and adjusted estimate (RR/OR) of the association will be different by 20% if there is confounding present

5

What is the purpose for controlling for confounders?

To get a more precise (accurate) estimate of the true association between the exposure and the disease/outcome.

6

What are the ways that confounding can be controlled for?

1) Study Design Stage
-randomization (blocked or stratified)
-restriction
-matching
2) Analysis of Data Stage
-stratification
-multivariate statistical analysis

7

What are the strength and weakness for randomization?

Strength:
-with sufficient sample size (N), randomization will likely be successful in serving its purpose (making groups 'equal')

Weakness:
-sample size may not be large enough to control for all known and unknown confounders
-randomization process does not guarantee successful, equal allocation between all intervention groups for all known and unknown confounders
-practical only for interventional studies

8

What is randomization in controlling for confounders?

Randomization technique hopefully allocates an equal number of subjects with the known (and unknown) confounders into each intervention group.

9

What is restriction in controlling for confounding?

Study participation is restricted to only subjects who do not fall within pre-specified category(ies) of confounder.

10

What is matching in controlling for confounding?

Study subjects selected in matched-pairs related to the confounding variable to equally distribute confounder among each of the study groups.

11

What is stratification in controlling for confounding?

Statistical analysis of the data by evaluating the association between the exposure and disease within the various strata (layers) within the confounding variable(s).

12

What is multivariate analysis in controlling for confounding?

Statistical analysis of the data by mathematically factoring out the effects of the confounding variable(s).

13

What is an effect modification?

a 3rd variable, that when present, modifies the magnitude of effect of an association by varying it within different levels or a 3rd variable (effect modifier).
-if an interaction is present, the researcher must report the measures of association for each strata individually.
>so, unlike confounding, an effect modifying variable should be described and reported at each level of the variable, rather than controlled-for.

14

How do you figure out if an effect modification is present?

Step 1) Calculate crude outcome measure of association between exposure and outcome (OR/RR).

Step 2) Calculate crude outcome measure of association (OR/RR) between exposure and outcome for each strata (layers) of the effect-modifying variable.

Step 3) Compare the stratum-specific measure of associations (for each strata of the rd variable between the exposure and outcome (OR/RR)
- the point estimate (RR/OR) for the association will be different by 20% between the lowest and highest strata (layers) of the effect-modifying variable if there is effect modification (interaction) present.

15

How can researchers assure we have found a true association between 'exposure' and 'outcome'?

Researchers evaluate 3 aspects of their study before declaring a statistical association:
1) Check for confounding or effect modification (interaction)
2) Check for bias
3) Check for statistical significance