Frank Harrell

f2harrell.bsky.social

17 October 2026 18:32

Spearheaded by David van Klaveren & with co-authors Peter Austin, Patrick Serruys, @benvancalster.bsky.social our new article in Stat in Med provides a minimal-assumption way to estimate the relationship between survival probability and a continuous variable: onlinelibrary.wiley.com/doi/10.1002/...

Andrew Vickers

vickersbiostats.bsky.social

17 August 2026 02:58

Can everyone stop using the term "real world data" and just be precise about setting, participants and interventions? grandroundsinurology.com/the-reality-...

Lars Mølgaard Saxhaug

load-dependent.eurosky.social

12 July 2026 20:44

The aim of New Public Management is to make public services work as badly as if they were for-profit services (rather than services for the common good), so that people won’t see the downside of privatisation into actual for-profit services.

Darren Dahly

statsepi.bsky.social

10 July 2026 15:22

I would love for anyone to explain the clinical utility of this, or to point to any evidence that placing multimorbid patients into statictially derived clusters has ever improved care or helped elucidate mechanisms leading to novel treatment options.

Frank Harrell

f2harrell.bsky.social

10 July 2026 15:21

Good luck finding utility in clustering patients, other than padding CVs. Cluster variables, not people.