Sensible Medicine
Sensible Medicine
Yet Another Excellent Explainer About P-Values in Randomized Trials
13
2
0:00
-44:05

Yet Another Excellent Explainer About P-Values in Randomized Trials

John Mandrola's avatar
John Mandrola
Apr 04, 2024
13
2
Share
Cross-post from Sensible Medicine
Excellent conversational overview of clinical trials - how they work, how they SHOULD be powered - signal to noise ratios - how to get them using z-scores and "standard deviations" - an indicative meta analysis that suggests that 80% of all stages of clinical trials/studies) (Phase 1, Phase 2 and Phase 3) are underpowered with a one in ten to one in five chance of being lucky - how to interpret p values ad so on - not too technical! Skip to 28 minutes of the embedded conversation to get a feel for the entire 45 minutes. More than 20,000 RCT's reviewed for tendencies! Free link to the paywalled NEJM article with lots of charts and tables! -
Peter Halligan

Gosh was I lucky to speak with Professor Erik Van Zwet from Leiden University in the Netherlands.

He is the first author on a recent NEJM Evidence paper looking at more than 23,000 trials in the Cochrane Database. (I linked to an URL that should get by the paywall.)

There are technical aspects of this paper. We hit on some (not a lot) of them. The gist of it though is really important when we look at evidence.

Erik did an excellent job of explaining P-values, trial power, and, at the end, we discuss how this work might inform the ability of trials to replicate.

This discussion also pairs well with one I had with computer scientist Ben Recht.

I hope you enjoy the conversation.

Please do consider subscribing or supporting our work as Sensible Medicine remains an ad-free user-supported place to learn about medical evidence. JMM

Sensible Medicine is a reader-supported publication. To receive new posts and support our work, consider becoming a free or paid subscriber.