Bridging the Gap Between Potential and Practice: An Integrative Review of Generative Artificial Intelligence in Nursing

Clinical Snapshot

Key Findings

• P Value: Not reported for the synthesised findings; the single experimental clinical trial's quantitative outcomes are not separately described in the abstract

• Effect Size: Not applicable — predominantly qualitative synthesis; no pooled quantitative effect size reported

• Primary Outcome: Four thematic barriers to GAI adoption in nursing practice identified: (1) educational gaps in AI literacy among nurses; (2) ethical concerns regarding patient privacy, consent, and accountability; (3) data transparency issues related to LLM explainability and bias; (4) workflow misalignment between GAI tool design and actual nursing workflows

• Nnt Or Sensitivity: Not applicable — no diagnostic, therapeutic, or prognostic quantitative outcomes are synthesised; the review is descriptive and thematic in nature

• Confidence Interval: Not applicable — no meta-analytic pooling performed

Clinical Application

The recommendations — nurse involvement in GAI design, mandatory AI curricula in nursing education, and governance models fostering trust — are feasible in principle but require institutional investment, regulatory frameworks, and interdisciplinary collaboration. Implementation feasibility varies significantly by healthcare system, resource availability, and existing digital infrastructure. The evidence base is insufficient to guide specific implementation protocols. In the Australian context, GAI integration in nursing intersects with several regulatory and policy frameworks. The TGA has begun developing guidance on software as a medical device (SaMD), which may encompass clinical decision-support LLMs. The Australian Digital Health Agency's National Digital Health Strategy 2023–2028 emphasises safe and ethical AI adoption in healthcare. AHPRA and the Nursing and Midwifery Board of Australia (NMBA) have not yet issued specific guidance on nurse use of GAI tools, creating a governance gap consistent with the 'data transparency' and 'ethical concerns' barriers identified in this review. The RACGP has published preliminary AI guidance for general practice, but equivalent nursing-specific frameworks are absent. Australian nursing education bodies (ANMF, ANMAC) would need to incorporate AI literacy into accreditation standards to address the educational gap identified. The journal's publication by the Australian College of Perioperative Nurses signals particular relevance to perioperative nursing settings, where EHR-integrated LLMs for documentation and handover are increasingly being piloted. Registered nurses, nurse practitioners, nursing educators, and nursing informatics specialists in acute care, perioperative, primary care, and community settings where GAI tools and LLM-integrated EHR systems are being introduced or evaluated

Abstract

BACKGROUND: Generative artificial intelligence (GAI) has emerged as a transformative tool in health care, particularly with the integration of large language models (LLMs) into electronic health record (EHR) systems. These tools have the potential to enhance clinical decision-making, increase efficiencies, and decrease time in the EHR. However, despite expanding availability, adoption in nursing practice remains limited. This integrative review aimed to synthesize the literature and explore barriers related to the integration of GAI tools into nursing practice. METHODS: We conducted an integrative search of peer-reviewed literature published between November 2022 and July 2025 to examine the current evidence while assessing for methodological quality. The retrieved articles were screened for title, abstract, and full text eligibility. We synthesized themes related to GAI adoption by nurses, focusing on ethical, educational, and workflow-related factors that influence use. RESULTS: A total of 10 studies met criteria including 7 qualitative descriptive design, 2 case reports, and 1 experimental clinical trial. Four key barriers emerged: educational gaps, ethical concerns, data transparency issues, and workflow misalignment contributing to nurses' hesitancy to engage with GAI tools. Recommendations include involvement of nurses in the design, implementation, and evaluation of GAI tools and mandatory AI curricula in nursing education. CONCLUSIONS: Overcoming barriers requires nurse involvement in GAI design efforts, targeted education, and governance models that foster trust and usability. Nurse-centered integration of GAI tools has the potential to advance workforce efficiencies while preserving patient safety.

References

1. 1.Campbell, L., & Tiase, V. L. (2026). Bridging the gap between potential and practice: An integrative review of generative artificial intelligence in nursing. Computers, Informatics, Nursing: CIN. Advance online publication. https://doi.org/10.26550/2209-1092.1366