I need some #epitwitter help with this article (https://www.ncbi.nlm.nih.gov/pubmed/29995914 ). Crude rate of secondary outcome (coronary PCI) in marijuana users was 35% vs 32% in controls. Multivariable OR is 0.73 (0.67±0.80), indicating reduced odds of PCI in marijuana users. What am I missing?
@ADAlthousePhD can probably answer better than I, but it looks as if they've thrown everything into a logistic regression model which could come up with ORs that are very different to the crude rate
-
-
Agreed. I've seen OR reduce/increase in intensity due to the influence of confounding in the multivariable model, but just never seen them completely switch direction? Or am I just misunderstanding this...
-
They definitely can switch direction after controlling for confounding, it's not uncommon especially with big confounders like age
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.