Public reproducibility repository for the manuscript:
What time does a FAST score represent? AST timing, patient selection, and category stability in routine VCTE
Authors: Hongyu Ying, Ze Zhang, Chunping Qiao, Chang Dong, and Hong-wei Yu.
This repository contains numbered analysis scripts, aggregate tables, model outputs, mathematical definitions, figure/table legends, and frozen release metadata for the v8 primary analysis and the v9 external-validation extension.
external_validation/ additionally contains redistributable public source data,
frozen source checksums, analysis scripts, and derived validation outputs for
two NHANES cycles, survival::pbcseq, and five PLOS ONE supplementary datasets.
The v9 extension adds same-calendar-day CORONASH, single-visit Singapore T2DM,
and prespecified-short-window German CAD cohorts. Source-specific attribution
and reuse terms are recorded in external_validation/PUBLIC_DATA_NOTE.md and
external_validation/SOURCE_CITATIONS.md.
Participant-level VCTE and laboratory data are governed clinical data and are not posted publicly because unrestricted disclosure could compromise patient privacy and would not comply with institutional data-governance requirements. They may be requested from the corresponding authors and are subject to institutional approval.
The study was approved by the Ethics Committee of Shanghai Pudong Hospital (approval no. 2024-IIT-013-E01; approval date 14 August 2024; valid through 13 August 2026). The requirement for individual informed consent for this retrospective analysis was waived.
All reported numbers in the manuscript and figures are fixed to the v8 release identifiers and hashes listed in documentation/RELEASE.json and documentation/FROZEN_MANIFEST.tsv.
Run analysis/00_run_all.py from an authorized v8 workspace containing the two governed study-data Parquet files. Aggregate results supplied here permit direct checking of manuscript values without access to participant-level data.
Run python external_validation/06_scripts/09_run_external_validation.py to reproduce the public-data analyses from the source files included in that directory. The pipeline keeps cohort-specific estimates separate and performs no pooled cross-cohort inference.
Public reproducibility release v9, 22 July 2026.