Week 10 Flashcards
What are the three elements of the general rule for defining the presence of confounding?
The confounding variable is causally associated with the outcome, non-causally or causally associated with the exposure, and not in the causal pathway between exposure and outcome.
What is residual confounding?
Residual confounding can occur when adjustment does not completely remove the confounding effect due to a given variable or set of variables. It is a form of confounding that remains even after adjustment.
How much different does the adjusted OR have to be from the crude OR for there to be potential confounding?
10% (crude OR - adjusted OR)/ (Crude OR)
Confounding is different from bias because confounding is
an existing relationship, rather than an error introduced into the study by the investigators.
In a case control study assessing the relationship between being an airplane pilot and risk of brain cancer, the average age of controls was 45.9 years and the average age of cases was 53.9 years. This is an example of
possible confounding
what is confounding?
a situation in which a non-causal association between a given exposure and an outcome is observed as result of the influence of a third variable (or group of variables); the real association is distorted because of the third variable
how is distinguishing confounders important for primary prevention?
a causal association between risk factor and outcome must exist, or modification of risk factor won’t lead to reduction in outcome
how is distinguishing confounders important for secondary prevention?
an unbiased (true) yet confounded association allows the identification of high-risk groups - important for screening
should confounding be assessed by statistical tests?
No, it should be assessed by common sense
If adjusted estimate changes from crude, confounding is present. True or false?
true
What is positivity?
for every level of a potential confounder, all possible levels of the exposure must have the possibility of being present
what can result in residual confounding?
any variable that has multiple strata
confounding is more likely to be found in observation studies rather than RCTs, becayse
RCTs have randomization
how can a researcher check to see whether randomization was successful in an RCT?
if the p-values of the characteristics between treated and nontreated groups are comparable
Do colliders block a pathway?
yes
