Machine Learning for the Prediction of Red Blood Cell Transfusion in Patients During or After Liver Transplantation Surgery
Source check: Source checked. Not independent scientific confirmation.
Finding: Recursive feature elimination and XGBoost selected preoperative predictors for red-cell transfusion during or after liver transplantation. The study reports AUROC 0.813 in a 1,193-patient multicenter cohort and 76.9% accuracy in a separate 31-patient prospective validation; this does not establish patient benefit.
Technology: Recursive feature elimination and XGBoost for preoperative transfusion-risk prediction
Publication: Peer-reviewed paper · Test context: prospective or independent check. Catalog categories, not an evidence grade.
- AI contribution
- Recursive feature elimination selected preoperative predictors and XGBoost estimated red-cell transfusion risk.
- Test and result
- The model reported AUROC 0.813 in a 1,193-patient, three-hospital cohort and 76.9% accuracy in a separate prospective 31-patient validation cohort.
- Limitation / still unknown
- The prospective validation was small; different transfusion criteria, selection and confounding limit interpretation, and clinical benefit was not tested.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2021.632210/full
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: Publisher full-text Methods and Results, including the separate prospective validation cohort; date checked..
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 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.