All frameworks and cheat sheets

Key figure

How common immortal time bias is in published research

Stat card: 44.0% of 182 studies across 25 topics were affected by immortal time bias, and 5 of 21 topics saw the conclusion flip.
If peer-reviewed studies get this wrong almost half the time, a dashboard built in an afternoon can too. Source: Shin, Kim, Yon et al. with John P. A. Ioannidis, Peer Review Congress, searches to July 2024

A review of systematic reviews by Shin, Kim, Yon and colleagues, with John P. A. Ioannidis, found immortal time bias in 80 of 182 studies across 25 topics. In 5 of the 21 topics that had both biased and unbiased studies, removing the biased ones reversed the pooled conclusion. It is a useful number for anyone who needs to argue for a landmark analysis before a roadmap decision.

Where this appears

What's in this key figure

44.0% of 182 studies across 25 topics were affected by immortal time biasIn 5 of 21 topics the pooled conclusion flipped between significant and non-significant once biased studies were removed.

Use this image

Free to use in your docs, decks and posts under CC BY 4.0, as long as you credit Koji with a link to this page. To embed it, paste this into your page:

html
<a href="https://www.koji.so/visuals/immortal-time-bias-share-of-studies-affected"><img src="https://www.koji.so/visuals/immortal-time-bias-share-of-studies-affected/immortal-time-bias-share-of-studies-affected.png" alt="Stat card: 44.0% of 182 studies across 25 topics were affected by immortal time bias, and 5 of 21 topics saw the conclusion flip." width="1200" height="630" /></a>
<p>Source: <a href="https://www.koji.so/visuals/immortal-time-bias-share-of-studies-affected">How common immortal time bias is in published research</a>, Koji</p>