Evidence Briefs · Project Lattice, Phase 0

Most “failed” trials didn’t fail. The registry just stopped telling the story.

ClinicalTrials.gov records why 52,408 trials stopped early. Most of those reasons have nothing to do with whether the science worked — funding ran out, a site closed, a sponsor changed course. This is what happened when we went looking for the trials worth a second look, then checked our own answer against reality.

52,408

trials classified

22,557

stopped for a non-scientific reason

7

internal audit rounds

0.2%

external-checked error rate

The pilot finding

The mechanical ranking put these 10 cases at the top of the list. Real web research resolved every single Category-A case to something other than "still simply open" — and confirmed 2 of 5 Category-B cases as genuinely unused results.

0 of 5 top-ranked “unanswered” candidates were actually still just open2 of 5 “good result” candidates confirmed as genuinely unused

The 10 Evidence Briefs

verified 2026-08-02