Answers are grounded in this paper and cite their sources.
Meta-research protocol reveals a framework to catalogue omitted pathology variables in published multimodal prognostic models, indicating potential bias in reported clinical baseline performance.
Pre-registration of Part E1, the published-baseline catalogue: a meta-research census of the clinical baselines of published multimodal prognostic models (18 venues, 2019–2026), each covariate set scored against a frozen 18-concept reference standard drawn from synoptic pathology reporting. Primary quantity: the distribution of omitted reference concepts per published baseline, reported descriptively; regressing reported gain on omitted-concept count is a pre-declared non-goal. Registered digest (SHA-256 of the frozen constants module, Freeze 2, 2026-09-02): e1a3729f0e9366b2d714cb7af7c1659e0bdee94aaa7136f91a953a88422636c7. The frozen module, the frame record and the known-item test record are attached. At the time of this deposit, no paper has been retrieved, screened or extracted. This filing was planned for OSF; OSF's API was down on the intended filing date, and the author elected to deposit here so the protocol carries an independent timestamp before retrieval begins. Full details, including prior related digests with commit hashes and the prospectivity statement, are in REGISTRATION.md.
No takes yet. Share an insight, caveat, or question.
Farshad Jafarpour (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: