Comparative predictive value of AIMS65, Pre-endoscopic Rockall and Glasgow-Blatchford Scoring Systems in Nonvariceal Upper Gastrointestinal Bleeding.

OBJECTIVE

To compare the predictive ability of the AIMS65, pre-endoscopic Rockall score (Pre-RS), and Glasgow-Blatchford score (GBS) for clinical outcomes in patients with non-variceal upper gastrointestinal bleeding (NVUGIB).

METHODOLOGY: This retrospective, single-center, observational study analyzed the clinical records of 314 patients with NVUGIB who presented to Jiashan County First People's Hospital between October 2023 and June 2025. The AIMS65 score, Pre-RS, and GBS were retrospectively calculated for these patients. The predictive ability of the scores for in-hospital mortality, transfusion requirement, endoscopic intervention, and rebleeding was evaluated by calculating the area under the receiver operating characteristic curve (AUC).

RESULTS

Of 314 patients included in the study, 9.6% died during hospitalization, 40.1% required blood transfusion, 76.4% underwent endoscopic intervention, and 6.1% experienced rebleeding. The AIMS65, Pre-RS, and GBS all demonstrated good performance in predicting the clinical endpoints. The AIMS65 showed significantly greater predictive ability for in-hospital mortality than Pre-RS and GBS. The GBS had a markedly superior ability to predict the need for blood transfusion compared with the other scores, and the Pre-RS exhibited a significantly better predictive performance for endoscopic intervention. No significant differences were observed among the three scores in terms of their predictive efficacy for rebleeding.

CONCLUSIONS

The AIMS65, Pre-RS, and GBS demonstrated good predictive value for clinical endpoints in patients with NVUGIB. The AIMS65 was more effective in predicting in-hospital mortality, GBS showed the best performance for predicting transfusion requirements, and Pre-RS performed best for predicting endoscopic intervention. No significant difference was observed among the three scores in predicting rebleeding.