Fake medical trials cluster around a few repeat authors — one person wrote 104 of them
Medical trials built on invented patient data mostly came from small groups that faked results again and again. Researchers examined 236 such withdrawn trials; a single author accounted for 104 of them.
Why it matters
When doctors choose a treatment, they usually trust summaries that pool many trials together. A trial with invented data can sit inside that pool for years and push doctors toward the wrong choice. Here, authors who faked data repeatedly went a median of 13.6 years before their work was withdrawn, compared with under 3 years for one-time offenders. The hopeful part is that the fakery looks concentrated in a few places rather than spread everywhere, so checking harder in those places could catch a lot of it.
Who's behind it: T. Jajieh, F. Naudet, J. Ioannidis and colleagues, A French-led academic research team with international collaborators.
Summary by the Lemma AI · how we grade