The Lancet in conversation with
Oct 29, 2026 · 23 min · 13 segments
Niall Boyce welcomes Professor Stuart Pocock, author of two Lancet papers, published 35 years apart: 1991's "Can meta-analyses be trusted?" and 2026's "What makes a meta-analysis believable?" They…
Stuart PocockGuest
Niall BoyceHost
So 1991 is the point where I think we're at the beginning, the first stirrings of what these days is called evidence-based medicine, or rather the sort of mass dissemination of that concept.
I remember that I was in medical school in the 90s, and the idea that We were now going to be going back to the literature that a lot of what we learned at medical school was wrong and that the systematic review and the meta-analysis were really held up as the pillars of this movement, at least by some people.
Now, it's interesting that even before that really gained a lot of traction as an idea, you in 1991 were asking, can meta-analysis be trusted? What was the thinking behind that?
It was really that meta-analyses were growing at that time in the extent to which they were being done and they were being promoted very enthusiastically as the way forward for evidence synthesis to come to decisions about whether treatments really worked or not.
But there was not enough skepticism about when a meta-analysis is good and when it's bad.
And we felt it was time to apply some brakes on the process and say, steady on here.
When can we really trust a meta-analysis? And when is it basically flawed and should not be trusted? So that was the thinking behind the articles.
So 1991 is the point where I think we're at the beginning, the first stirrings of what these days is called evidence-based medicine, or rather the sort of mass dissemination of that concept.
I remember that I was in medical school in the 90s, and the idea that We were now going to be going back to the literature that a lot of what we learned at medical school was wrong and that the systematic review and the meta-analysis were really held up as the pillars of this movement, at least by some people.
Now, it's interesting that even before that really gained a lot of traction as an idea, you in 1991 were asking, can meta-analysis be trusted? What was the thinking behind that?
It was really that meta-analyses were growing at that time in the extent to which they were being done and they were being promoted very enthusiastically as the way forward for evidence synthesis to come to decisions about whether treatments really worked or not.
But there was not enough skepticism about when a meta-analysis is good and when it's bad.
And we felt it was time to apply some brakes on the process and say, steady on here.
When can we really trust a meta-analysis? And when is it basically flawed and should not be trusted? So that was the thinking behind the articles.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
