Acute gastrointestinal bleeding (GIB) requires rapid risk stratification, timely intervention decisions, and accurate triage. Despite validated risk scores (Rockall, Glasgow-Blatchford), substantial variability in clinical decision-making persists. CDSS interest is growing, yet no unified approach integrates prognostic models, endoscopic findings, and clinical parameters.

Key gaps include:

• variability in therapeutic decisions across experience levels;

• limited calibration of scores in high-risk patients;

• absence of a standardized early triage and management pathway;

• limited integration of guideline-based recommendations into routine workflow.

A literature review (PubMed, Embase, Scopus; 2018-2025) assessed:

• prognostic value of Rockall and GBS for rebleeding, mortality, and urgent intervention;

• management algorithms for non-variceal GIB;

• existing digital tools and CDSS components;

• factors driving decision-making variability.

Non-clinical, low-quality, and non-representative studies were excluded.The review highlights the potential of digital CDSS to integrate clinical, laboratory, and endoscopic parameters into a structured and standardized decision pathway.

1. Risk score performance.

• GBS shows high sensitivity for identifying low-risk patients (GBS 0-1 = minimal adverse event risk).

• Rockall score offers acceptable accuracy for rebleeding prediction (AUC ≈ 0.73) but is less reliable for mortality.

• Both tools insufficiently incorporate real-time endoscopy and dynamic hemodynamics.

2. Clinical gaps.

• large decision variability;

• weak calibration in high-risk groups;

• lack of unified triage pathway;

• limited incorporation of guidelines into workflow.

3. Potential of digital CDSS. 

CDSS studies show improved structure of clinical decisions, better adherence to recommendations, greater consistency across clinicians, and reduced assessment time.No validated GIB-specific digital model currently integrates endoscopic, clinical, and dynamic parameters.

Rockall and GBS remain important but fail to incorporate endoscopic stigmata and dynamic patient changes. Persistent gaps - decision variability, absence of standardized triage, and limited implementation of guidelines - underline the need for updated digital solutions.

Digital CDSS represent a promising direction for standardizing GIB management. The literature supports developing multifactorial algorithms that combine laboratory, clinical, and endoscopic data.

Future Directions.

Future work should focus on developing and validating digital CDSS, including web-based and EMR-integrated tools, capable of real-time adaptation and reducing clinician variability. Integration within the ESGE framework could support the creation of a unified standardized pathway for GIB management.