Construction and Validation of a Machine Learning-Based Risk Prediction Model for Recurrent Upper Gastrointestinal Bleeding in Patients with Liver Cirrhosis: A Prospective Study
Source check: Source checked. Not independent scientific confirmation.
Finding: A machine-learning risk model for recurrent upper-GI bleeding in cirrhosis was assessed in a prospective development cohort and an independent external cohort. The selected logistic model achieved external AUC 0.827 (95% CI 0.747–0.907). This supports prediction in the studied cohorts; it does not establish improved outcomes or causal effects.
Technology: Four machine-learning classifiers; logistic regression was selected
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Classifiers estimated individual recurrence risk from patient predictors.
- Test and result
- A prospective 505-patient development cohort and an independent 148-patient external cohort were used to predict six-month recurrent upper-GI bleeding in cirrhosis. The selected logistic model had external AUC 0.827 (95% CI 0.747–0.907), with decision-curve analysis.
- Limitation / still unknown
- No impact trial or evidence of improved patient outcomes; findings are bounded to the studied cohorts and setting.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.researchsquare.com/article/rs-11469264/v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.