Quantitative Competence and Editorial Gatekeeping: A Cross-Sectional Assessment of Machine Learning Evaluation Capacity at Major US Medical Journals — Data and Code

This deposit contains the complete dataset, analysis code, predictive models, and manuscript supporting a cross-sectional audit of machine learning and artificial intelligence (ML/AI) evaluation capacity at 9 major US medical journals. Key findings: 0 of 14 editors-in-chief hold formal ML/AI training, first-author ML publications, or quantitative doctoral degrees These journals collectively publish 0.115% of global PubMed-indexed ML/AI biomedical output (263 of 227,731 papers, 2023–2025) Predictive models estimate triage error rates of 45–78% for quantitatively intensive manuscripts Contents: 21 raw PubMed Timeline CSV exports (searched February 27, 2026) Consolidated data with EIC credential profiles (JSON) Standalone predictive model (Python, no external dependencies) Data extraction and visualization scripts 4 interactive HTML visualizations Manuscript formatted for BMJ submission (Word) Proposed author experience survey design Full methodological documentation