1

I am looking for some real clinical examples for variables that are NOT confounders:

1) variables only impact treatment, but not outcome.

2) variables only impact outcome, but not treatment

So the variables in the two sets are not confounders. Any good examples (and citation) will be welcome! Thanks,

Vincent
  • 421
  • You need to be more specific to get a good answer. These things are pretty context specific. – dimitriy Sep 25 '14 at 23:39
  • "Confounder" also has many definitions. One of the common ones taught to intro epidemiology students is that the confounder must merely be associated with, but not necessarily cause the treatment (or exposure), while also being a causal antecedent of the outcome. When one gets into DAGs, one begins to understand that "confounder" is perhaps too blunt a concept for the complexities of causal inference. – Alexis Sep 26 '14 at 00:53

0 Answers0