<![CDATA[Artificial intelligence-driven clinical decision support systems (AI-CDSS) are increasingly used to support diagnosis, risk prediction, medication safety, clinical prioritization, and hospital workflow management. Although AI-CDSS has potential to improve hospital efficiency and patient safety, its real-world value remains uncertain because many studies still emphasize technical performance rather than clinical workflow, governance, and organizational outcomes. This systematic literature review synthesizes recent evidence on the impact of AI-CDSS on hospital management efficiency and patient safety and identifies implementation and governance factors that influence effectiveness in hospital settings. Studies published from 2020 to 2026 were targeted if they examined AI-CDSS in hospital settings using machine learning, deep learning, natural language processing, predictive analytics, or intelligent alert systems. Evidence was organized across four domains: AI-CDSS characteristics, hospital efficiency, patient safety, and implementation governance. AI-CDSS supports diagnostic reasoning, clinical risk prediction, medication safety, early warning, adverse-event detection, incident classification, and workflow prioritization. It may improve hospital efficiency through faster prioritization, better resource allocation, and more coordinated workflows. It may also strengthen patient safety by enabling earlier detection of deterioration, medication errors, falls, pressure injuries, and adverse-event patterns. However, benefits remain conditional on data quality, EMR integration, validation, trust, explainability, clinical workflow fit, monitoring, and governance maturity. AI-CDSS should be understood as socio-technical governance infrastructure rather than a standalone algorithm. Hospitals and policymakers should develop structured governance mechanisms covering multidisciplinary oversight, local validation, workflow simulation, user training, post-deployment monitoring, patient-safety reporting, and accreditation-based accountability. This study contributes to the literature by integrating previous AI-CDSS evidence into a hospital-management framework that explains how algorithmic decision support can generate efficiency and patient-safety value only when supported by data readiness, workflow integration, clinician trust, explainability, local validation, continuous monitoring, and accountable governance.]]>
Copyrights © 2026
