Ethical Dimensions and Patient Safety Impacts of Artificial Intelligence- Assisted Clinical Decision-Making in Critical Care Nursing: A Qualitative Systematic Review and Meta-Synthesis

Abstract

The rapid integration of Artificial Intelligence (AI) and clinical decision-support systems in critical care units has profoundly transformed nursing workflows. While these advanced technologies offer real-time predictive insights to optimise patient care, they also introduce complex ethical challenges and significantly alter the dynamics of clinical judgment at the bedside. This qualitative systematic review and meta-synthesis aims to aggregate, appraise, and synthesise empirical qualitative evidence on critical care nurses' firsthand experiences with AI-assisted decisionmaking, with a specific focus on its underlying ethical dimensions and direct impacts on patient safety outcomes. A systematic literature search will be executed across major electronic databases, including PubMed, CINAHL, Scopus, and ScienceDirect, for relevant qualitative studies. Methodological quality will be rigorously evaluated using the Critical Appraisal Skills Programme (CASP) Qualitative Checklist, and metaaggregation will be employed to synthesise findings into overarching thematic categories. Preliminary scoping indicates that the synthesis will yield critical themes surrounding automation bias versus clinical intuition, accountability for AI-driven clinical errors, and digital workflow friction that compromises timely bedside interventions. Aligning with the imperative for ethical scholarship, this study transcends purely technical evaluations by highlighting the essential human and ethical guardrails necessary for digital health innovation. The resulting synthesis provides a robust framework to inform institutional clinical policies, minimise technological friction, and guide the responsible, patient-centric integration of AI in high-acuity healthcare environments.