Did the authors or reviewers comment on this. The lay press certainly didn't.
In summary, the vitamin D study in the news http://www.bmj.com/content/356/bmj.i6583 … is hampered by small study bias. No protective effect in large trialspic.twitter.com/5e1Ra1L2h5
-
-
-
the
@bmj_latest Editors mention it in their decision letter and author response here: http://www.bmj.com/content/356/bmj.i6583/peer-review …pic.twitter.com/2OKfItt6Fa
End of conversation
New conversation -
-
-
Nice work
@mvholmes! Did you try incorporating study size as a covariate in meta-regression?@bmj_latest -
nice idea
@bj_cairns@bmj_latest - the P-value for hetero between subgroups (based on cases >0 to <100, >=100 to <500 and >500) was tiny -
@Richard_D_Riley one step analysis (decent weighting) effect is smaller, should be preferred almost always -
So it's possible to incorporate a study size random or fixed effect into the weights...?
-
weighting known to be suspect in RE models when bias present. IPD one stage, each patient equal weight
-
Ah yes, IPD would be better, though one might still model bias (albeit post hoc in this case)
-
But I guess model uncertainty could then exceed gains in precision, so focus on large studies
End of conversation
New conversation -
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Interesting, will have a read. Though note small study effects may not = bias. Could be genuine heterogeneity
-
Perhaps because the participants in the largest trials weren't deficient?
End of conversation
New conversation -
-
-
is that not the point of metaanalysis?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
is the bmj still considered a journal? Or is it more of a free magazine from the union?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Is that different to small studies being more prone to publication bias as suggested in the funnel plot in the editors comments?
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.