Fabio Ynoe de Moraes, Associate Professor at Queen’s University, shared a post on Facebook:
“9 truths about high-impact meta-analysis (that nobody told you)
1) You wasted months because the question was weak
It wasn’t the software. It wasn’t RevMan.
It was a generic question that doesn’t resolve any real clinical decision.
2) ‘There are many studies’ doesn’t mean it’s worth meta-analyzing
If the answer is already obvious, the editor knows it too.
High-impact meta-analyses are born from uncertainty, not volume.
3) The reviewer killed your paper because of the outcome — and you didn’t notice
Surrogate outcome, wrong scale, mixed follow-up.
Result: ‘clinically not meaningful.’
4) A high I² wasn’t the problem. Not knowing how to explain it was
Heterogeneity isn’t a statistical error.
It’s poorly thought-out clinical reasoning.
5) Random-effects didn’t save your meta-analysis
It just hid the fragility.
With few studies, the model doesn’t work miracles.
6) The funnel plot didn’t convince anyone
The editor knows this doesn’t ‘fix’ bias.
It only shows you didn’t think about the source of the problem.
7) Your method wasn’t wrong — it was shallow
No protocol, no duplication, no GRADE.
Today, that’s the minimum requirement, not a differentiator.
8) You did statistics, but you didn’t do medicine
Pretty odds ratio.
No practical answer for the clinician.
9) A meta-analysis that changes guidelines always delivers something usable
NNT. Clear subgroup. A decision that can be made tomorrow.
If it doesn’t change practice, it doesn’t change impact.
Follow me and share this with anyone trying to publish a meta-analysis in a high-impact journal (10.0 or higher) and only collecting rejections.
Here you learn real research — no shortcuts and no illusions.”
Proceed to the video attached to the post.
More posts featuring Fabio Ynoe de Moraes.