Some even have guidelines that prevent being explicit about causal aims and conclusions without an RCT. @_MiguelHernan is good on this.
Medical journals are odd to me. Either they publish rigorous randomized trials or else simplistic correlation-is-causation stuff, but not as much that's in-between (such as regression discontinuity).https://twitter.com/YFeyman/status/1087778711477370882 …
-
-
-
Those guidelines are obviously wrong, but considering the quality and design of many non-RCTs in medical journals, it's probably good to avoid any mention of causation! :)
End of conversation
New conversation -
-
-
The authors don’t make any causal claims though... they’re pretty clear they’re only examining an association—An interesting association nevertheless, and argue for further study.
-
But then why end with the conclusion that policymakers should think about cracking down on marketing?
-
Seems like a sensible recommendation in either case. Whether it’s causal or pharma simply targeting physicians likely to prescribe, this is the most urgent public health crisis in decades. Isn’t a cautious approach better then wait and see?pic.twitter.com/uRUwqYkCru
- 1 more reply
New conversation -
-
-
Medical & public health journals are the final frontier for causal inference methods using observational data. I'm pretty sure those insights have reached all other disciplines by now.
- 1 more reply
New conversation -
-
-
Medical journals are odd in many ways beyond this point. Perhaps there could be a causal study looking at how incentives in publishing influence what gets published?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Not so odd considering clinicians, and not epidemiologists or biostatisticians, run the show.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think appropriately, the journals would prefer analyses that present the data, where the fallacy of inferring causality is understood, rather than present analyses where their readers, and peer-reviewers, will not understand the potential shortcomings.
-
In part this is a failure of methodologists on both sides to allow a wider scope for their outreach and understanding, and in part a failure caused by stovepiping in academia.
End of conversation
New conversation -
-
-
Don't know if that's true or not but if so, I'd guess the denominator (medical studies) isn't much different?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Yeah. It seems like J Clin Epi is where you have to go for that kind of stuff - it has clinical implications (so not purely *methods* work) but it’s a little too exotic for medical journals.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I so agree. -
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Rigorous science gives credibility. Crazy correlation causation stuff gives headlines. They need both.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
They have (re)discovered IV in the form of Mendelian randomization.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.