Issue #7 Week of Sep 21, 2026 – Oct 05, 2026
37Papers
This issue at a glance
This issue's 37 papers cluster around nurses' readiness for AI and clinical decision support testing. Several studies measure nursing students' and clinicians' AI literacy, anxiety, and attitudes, often finding mixed knowledge but generally positive views. A parallel set of papers tests prediction tools for triage, delirium, pressure injuries, and stroke mortality using nursing and bedside data, usually comparing machine learning against simpler scores. A third thread raises governance and ethics questions, covering intersectionality in AI design, ethics scales, professional organization roles, and frameworks for system-level oversight.
Papers grew from 12 to 37 (+25), with Workforce & Education remaining the leading topic, as it was last issue. AI Ethics & Governance rose in share, now including papers on governance frameworks, human rights advocacy, and concept analysis of AI justification, while Patient-Facing AI fell to a single paper this issue after being a rising thread last time. Clinical Decision Support expanded substantially, from 2 papers to 10, covering triage, delirium, pressure injury, and medication error tools. The prior issue's patient-facing chatbot and telephone follow-up comparisons are not continued this issue.
Nurses' attitudes, anxiety, and readiness for AI 8Testing clinical prediction and decision support tools 9Governance, ethics, and accountability frameworks 8Language models for text analysis and evaluation 6
Compared withIssue #6
Papers37 ▲ 25
Leading topicWorkforce & Education (unchanged)
Rising▲ AI Ethics & Governance
Cooling▼ Patient-Facing AI
Flagged16 ▲ 11
How this digest is made: Papers retrieved bi-weekly from the community-configured sources (PubMed, arXiv, medRxiv, and/or CINAHL — see Digest Settings for this issue's exact sources) using search terms agreed at AINurse-26. Results are restricted to English-language papers; both title and abstract are checked independently, and either being predominantly non-English excludes the paper. Papers are then classified against the community's chosen topics — anything that doesn't clearly match is left out of the issue rather than shown as "Other." Summaries are generated by Claude Sonnet, constrained to report only what the abstract states, with no speculation or inference. The "At a Glance" section above is written under the same constraints — the model synthesizes themes only from the summaries included in this issue, and every number in it (counts, changes since the previous issue) is computed directly from the archive, not by the model. Flagged items indicate incomplete abstracts or unverified peer-review status.
All (37)Clinical Decision Support (10)NLP & Generative AI (6)AI Ethics & Governance (7)Workforce & Education (11)Patient-Facing AI (1)AI Methods & Evaluation (2) Flagged (16)
Clinical Decision Support10 papers
Clinical Decision Support
Development and Temporal Validation of a Logistic Regression-Based Risk Score for Identifying Life-Threatening Chest Pain at Emergency Department Triage: A Retrospective Cohort Study.
Vascular health and risk management · 2026
Researchers in Hefei, China, built a simplified risk score to spot life-threatening chest pain at ED nurse triage, using only information available without lab results. Tested on 2,506 adults, the logistic regression score performed as well as more complex machine-learning models and mapped onto China's four-level triage scale. This gives nurses an interpretable tool to support—not replace—triage decisions, though wider testing is needed before routine use.
Clinical Decision Support
Whose Knowledge Counts in Labour? Obstetric Nurses Negotiating Clinical Judgement, Algorithmic Authority and Woman-Centred Care in Artificial Intelligence-Enhanced Foetal Monitoring.
Nursing inquiry · Oct 2026
This interpretive description study interviewed 18 obstetric nurses in a physician-led Saudi maternity service about using AI-enhanced cardiotocography monitoring. Nurses often already suspected problems before alerts appeared, and when their judgement differed from the system, they faced unequal accountability and had to translate their relational knowledge into algorithm-friendly terms to be heard. This suggests AI monitoring can reshape whose clinical knowledge is trusted during labour, raising questions for nursing informatics about credibility and authority, not just safety.
Clinical Decision Support
An interpretable machine learning framework with clinical nomogram for predicting in-hospital mortality in acute ischemic stroke using high-granularity bedside data.
Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association · Oct 2026
Researchers used nursing, emergency, and lab data from the first 24 hours to build a machine learning model predicting in-hospital death in 5,014 acute ischemic stroke patients across three hospitals. The CatBoost model and a simpler 10-feature nomogram both outperformed standard ICU scoring systems in internal and external testing, with strong accuracy and calibration. This suggests bedside nursing data, combined with early clinical information, can improve early mortality prediction tools for stroke patients.
Clinical Decision Support
Postoperative delirium risk estimation and assessment with machine learning: development and validation of the Open-DREAM Model: A peri-operative dataset analysis.
European journal of anaesthesiology · Oct 2026
Researchers at a Swiss hospital studied 748 surgical patients aged 60 and older to predict postoperative delirium using only preoperative data. A calibrated XGBoost machine learning model reached an AUC of 0.80 and 0.89 sensitivity, using 15 key features, with anaesthesia type as the strongest predictor. This open-access tool could support early delirium risk screening, though external validation is still needed before clinical use.
Clinical Decision Support
Nursing Perceptions of the Intended Use of Artificial Intelligence to Prevent Medication Errors: A Qualitative Descriptive Study.
Journal of clinical nursing · Oct 2026
This qualitative study interviewed 16 emergency and intensive care nurses about using AI tools to prevent medication errors. Nurses saw AI as helpful for decision-making and reducing cognitive load, but raised concerns about training gaps, weak infrastructure, unreliable data, and ethics. The findings suggest successful AI adoption needs participatory design and context-specific training, not just technology, to avoid adding complexity to medication safety work.
Clinical Decision Support
Integrating Artificial Intelligence in Nursing Practice With Decubitus Risk Prediction Alerts: A Pilot Process Evaluation.
Journal of clinical nursing · Oct 2026
This study tested whether nurses on three hospital wards found an AI tool (DRAAI) for predicting pressure ulcer risk acceptable and workable in daily practice. Fifty-five nurses valued the predictions, found DRAAI feasible to use, and nursing care plans followed most at-risk alerts, though some needed help distinguishing risk prediction from detection. The findings suggest AI can support pressure ulcer prevention, but ongoing communication and support are needed, and clinical impact still needs testing.
Clinical Decision Support
Evaluating Large Language Model‐Assisted Emergency Triage: A Comparison of Acuity Assessments by GPT‐4 and Medical Experts.
Journal of Clinical Nursing (John Wiley & Sons, Inc.) · Oct2
Researchers compared GPT-4's Emergency Severity Index (ESI) scores to those of senior ED nurses and physicians for 100 adult emergency department patients, using the same patient data for both. GPT-4 assigned lower median ESI scores than human evaluators, suggesting it rated patients as more severe, and the two approaches weighed age and vital signs differently. This points to a need for careful testing and adjustment before using LLMs in triage decisions.
NLP & Generative AI6 papers
NLP & Generative AI
Understanding Mothers' Experiences of Very Low Birth Weight Infants Across the Childbirth Trajectory in Japan: A Retrospective Analysis Using Natural Language Processing.
Japan journal of nursing science : JJNS · Oct 2026
Researchers in Japan analyzed notebook records from 280 mothers of very low birth weight infants, written during a support class between 2003 and 2018, using natural language processing to study experiences from pregnancy through the infant's first year. Topic modeling found apologetic or self-blaming language around the first meeting with their baby and the NICU stay, shifting to routine caregiving concerns by 6-12 months postpartum. This descriptive overview may support future qualitative and clinical research on maternal experiences during this period.
NLP & Generative AI
Advancing qualitative analysis in nursing research: Comparing text analysis rigor and efficiency of large language models with human-coded analysis for interview data.
International journal of nursing studies · Oct 2026
Researchers tested whether large language models (ChatGPT and DeepSeek) could analyze interview data from 15 postoperative bone tumor patients experiencing kinesiophobia, comparing results to a human researcher's coding. The AI models identified the same four themes as the human analyst, showing strong agreement (Cohen's κ = 0.89) while cutting coding time. This suggests LLMs can support faster qualitative analysis, but human judgment is still needed to interpret emotional nuances and context.
NLP & Generative AI
Developing and Evaluating the Use of ChatGPT as a Screening Tool for Nurses Conducting Structured Literature Reviews: Proof of Concept Study Results.
Journal of Clinical Nursing (John Wiley & Sons, Inc.) · Oct2
This proof-of-concept study tested four versions of ChatGPT against a human researcher for screening 400 nursing education abstracts from four databases. ChatGPT-3.5 Turbo screened fastest and missed fewer relevant studies than the human researcher, while newer versions (GPT-4, 4 Turbo, 4 omni) missed more and struggled with nuanced nursing concepts. This suggests LLMs may speed up literature screening, but nurses need to test and report on their performance before relying on them.
AI Ethics & Governance7 papers
AI Ethics & Governance
Measuring Ethical Perceptions of AI in Nursing: The Development and Validation of the AI-Ren Ethics in Healthcare (RATI) Scale.
International journal of nursing practice · Oct 2026
Researchers developed and tested a new scale to measure Chinese nurses' ethical views on AI use in clinical practice, using a two-phase process with factor analysis and structural modelling. The final 20-item tool covers four dimensions, including cultural values and technological humility, and showed good reliability and validity. Higher ethical perception scores linked to more technology acceptance and less burnout, suggesting the scale could support AI ethics education and culturally aware implementation in nursing settings.
Workforce & Education11 papers
Workforce & Education
Ethical considerations of undergraduate nursing students using artificial intelligence for learning: A qualitative study.
Applied nursing research : ANR · Oct 2026
Researchers interviewed 21 nursing undergraduates at one Eastern China university about ethical experiences using AI for learning, covering academic integrity, privacy, and fairness. Students' trust in AI, verification habits, and views on fairness varied widely, and many valued face-to-face interaction with teachers and peers despite AI's efficiency. These findings may help inform AI ethics curricula, though broader applicability and intervention effectiveness still need study.
Workforce & Education
Nurses' knowledge, attitudes, literacy, anxiety, and readiness towards artificial intelligence in clinical practice: An international scoping review.
Applied nursing research : ANR · Oct 2026
This scoping review examined 33 studies (over 9000 nurses across 10 countries) on nurses' knowledge, attitudes, literacy, anxiety, and readiness regarding AI in clinical practice. Nurses showed low to moderate AI knowledge but generally positive attitudes, with anxiety linked to job loss, dehumanized care, and data security concerns. Since evidence comes from observational studies of varying quality, the review suggests structured education and governance are needed before favorable attitudes translate into safe adoption.
Workforce & Education
Is the Application of Artificial Intelligence-Based Competency Assessments for Selecting Clinical Nurses Acceptable? A Qualitative Study.
The journal of nursing research : JNR · Oct 2026
This qualitative study in South Korea interviewed 10 nursing applicants and 10 nurse educators about AI-based competency assessments used in nurse hiring. Participants raised doubts about evaluation methods, noted efficiency gains, described preparation challenges, and suggested improvements. The findings point to a need for tailored algorithms, clear criteria, and interdisciplinary collaboration to support fair, ethical use of AI in nurse selection.
Workforce & Education
Network structure of artificial intelligence anxiety among nursing students and educational implications: A multicenter network analysis.
Nurse education today · Oct 2026
This multicenter survey of 1113 nursing students across China mapped the network structure of AI-related anxiety using the Artificial Intelligence Anxiety Scale. Learning interaction anxiety and fear of AI misuse emerged as core symptoms, with associate-degree students showing stronger anxiety spillover between fear of job loss and robot autonomy. The findings suggest nursing curricula may need targeted strategies addressing procedural and ethical concerns, particularly for students in shorter degree programs.
Workforce & Education
The mediating role of critical thinking between AI literacy and learning burnout in nursing students using generative AI: A cross-sectional descriptive study.
Nurse education today · Oct 2026
This cross-sectional study surveyed 617 nursing undergraduates across seven Chinese universities to examine how AI literacy, critical thinking, and learning burnout relate to each other. Critical thinking fully explained the link between AI literacy and lower learning burnout, meaning AI literacy alone did not directly reduce burnout without critical thinking skills. This suggests nursing education may need to build critical thinking alongside AI literacy to support student well-being.
Workforce & Education
Generative AI use in instructor-assigned learning activities among nursing students: A cross-sectional survey of perceptions, ethical awareness, faculty restrictions, and continuance intention.
Nurse education today · Oct 2026
This cross-sectional survey asked 266 nursing students in northern Taiwan about using GenAI tools like ChatGPT for instructor-assigned learning activities. Perceived academic benefits—like learning enhancement, time savings, and personalized feedback—predicted students' intention to use GenAI, while ethical concerns and faculty restrictions did not. This suggests nursing educators may need to actively guide GenAI use in coursework rather than rely on restrictions alone.
Workforce & Education
Performance of large language models on nursing licensure examinations: A systematic review and meta-analysis.
Nurse education today · Oct 2026
This systematic review and meta-analysis examined how large language models (LLMs) perform on nursing licensure exam questions, pooling 12 studies with 13,870 multiple-choice questions across ten LLMs. Pooled accuracy was 69.6%, with newer or nursing-customised models reaching up to 93.6%, though performance varied widely and models struggled with ethics and psychosocial content. This suggests LLMs may support low-stakes learning activities but are not yet reliable enough for unmoderated, high-stakes licensure decisions.
Workforce & Education
A Scoping Review of Artificial Intelligence Integration in Nursing Education: Challenges, Opportunities, and a Roadmap for the Future.
Nursing Forum · 9/27
This scoping review examined 21 peer-reviewed studies (2019–2024) on AI integration in nursing education, mostly involving nursing students, comparing AI-focused interventions with traditional simulation-based benchmarks. It found AI may enhance engagement and learning when combined with existing simulation methods, but ethical concerns and uneven faculty preparation remain barriers. This suggests AI should extend, not replace, simulation pedagogy, supporting calls for targeted faculty training and clear ethical frameworks.
Workforce & Education
Artificial Intelligence Perceptions and Readiness Among Clinical Nurses: A Latent Profile Analysis With Implications for Nursing Management.
Journal of Nursing Management · 9/28
Researchers surveyed 1,018 clinical nurses at public hospitals in China to identify patterns in AI attitudes, literacy, self-efficacy, and anxiety. They found three distinct profiles: Low (66.8%), Ambivalent (20.8%), and High (12.4%) AI perceptions and readiness, linked to factors like education, training, and hospital setting. This suggests nurse administrators could use these profiles to design targeted AI training and peer support for nursing practice.
Patient-Facing AI1 paper
Patient-Facing AI
The Invisible Workforce: An Algorithmic Audit of Nursing, Primary Care, and Holistic Gaps in AI-Generated Survivorship Plans.
Seminars in oncology nursing · Oct 2026
Researchers audited 50 ChatGPT-5.1 responses using a simulated advanced ovarian cancer survivor to see how survivorship plans credited different healthcare professionals. Medical oncology received full credit (100%), but nursing, primary care, and genetic counseling were rarely named, even though nursing-related tasks appeared in 96% of responses. The findings suggest nurses should review AI-generated plans to fix missing referrals and simplify reading level, since responses averaged an 11th-grade reading difficulty.
Flagged Papers16
Clinical Decision Support 3 Flagged
Clinical Decision Support
Setting unclear
Artificial Intelligence Technologies in Nursing Clinical Decision-Making: An Umbrella Review.
Journal of advanced nursing · Oct 2026
This umbrella review examined 16 literature reviews covering 965 primary studies on artificial intelligence in nurses' clinical decision-making. The review found most research focuses on technology development rather than real-world testing, with gaps in controlled studies, algorithmic bias checks, and nurse involvement in AI design. This matters because nurses need a central role in building AI tools, plus ethics training and privacy guidelines, before AI can reliably support nursing practice.
Clinical Decision Support
Population unclear
Artificial Intelligence Technologies in Nursing Clinical Decision‐Making: An Umbrella Review.
Journal of Advanced Nursing (John Wiley & Sons, Inc.) · Oct2
This umbrella review examined 16 literature reviews covering 965 studies on artificial intelligence in nursing clinical decision-making, published between 2019 and 2024. Most studies focused on building and testing AI tools, but real-world use in nursing settings was rare, and strong comparative evidence was lacking. The authors say nurses should help design AI tools and receive AI literacy training to support safe, effective use in practice.
Clinical Decision Support
Population unclear Setting unclear
Charting and Predicting Risk: Artificial Intelligence/Machine Learning Pilot Model for Hospital-Acquired Pressure Injuries.
Journal of Nursing Care Quality · Oct2
This pilot study used secondary data to test whether AI/machine learning models could predict hospital-acquired pressure injuries (HAPIs) better than traditional logistic regression. Tree-based AI models predicted HAPIs more accurately than logistic regression, which still gave a reasonable, interpretable fit. This suggests AI/ML tools may help nurses identify at-risk patients more precisely, supporting earlier prevention efforts.
NLP & Generative AI 3 Flagged
NLP & Generative AI
Work in progress
From Sentiment to Signal: Narrative-Physiology Discordance as a Testable Target for Artificial Intelligence in Critical Care.
Medical sciences (Basel, Switzerland) · Sep 2026
This paper addresses a problem in critical care informatics: sentiment analysis of ICU clinical notes predicts mortality only marginally better than existing severity scores. The authors propose a new target instead—a 'narrative-physiology discordance score' that measures gaps between clinician-documented risk and data-driven risk, with formal controls and a staged evaluation plan involving clinician adjudication. This reframing could help nursing informatics researchers design tools that flag disagreement needing review, rather than just predicting death.
NLP & Generative AI
Population unclear
Generative AI at the Bedside: An Integrative Review of Applications and Implications in Clinical Nursing Practice.
Journal of clinical nursing · Oct 2026
This integrative review examined studies on generative AI tools used in nursing clinical workflows, published between 2018 and 2025. Researchers found four key themes: workflow efficiency, AI-supported clinical reasoning, patient communication, and ethics/trust concerns. The review stresses that human oversight, staff training, clear governance, and equity evaluation are needed before wider adoption, as unchecked use could introduce errors and harm patient trust.
NLP & Generative AI
Setting unclear
Mortality Research With Artificial Intelligence Tools: Applying Language Models to the National Violent Death Reporting System Suicide Narratives, 2003–2023.
American Journal of Public Health · Oct2
This essay reviews how AI tools, including language models, have been used in suicide prevention research, then shows an example applying them to 756,645 death narratives from 483,303 suicide decedents in the National Violent Death Reporting System (2003–2023). The authors note these AI tools can help analyze narrative data too large for manual review. This suggests potential for nursing informatics research to process large, unstructured mortality datasets more efficiently.
AI Ethics & Governance 6 Flagged
AI Ethics & Governance
Work in progress Not peer-reviewed
Intersectionality in AI and Machine Learning for Health Care: Protocol for a Scoping Review.
JMIR research protocols · Sep 2026
This protocol describes a planned scoping review of how intersectionality is used to guide AI and machine learning design in health care, including whether patient partners and interdisciplinary teams are involved. The review has not yet been completed; preliminary searches found 4191 records, with results expected by October 2026. This matters for nursing informatics because it will map gaps and offer community-informed guidance on building more equity-oriented AI tools.
AI Ethics & Governance
Population unclear Setting unclear
System-Level AI Governance in Nursing: A Practical Framework for Safe and Scalable Adoption.
Nursing administration quarterly · 2026
This article addresses the challenge nurse leaders face in evaluating and governing AI tools in health care organizations. It presents a practical governance framework covering clinical usefulness, safety, fairness, transparency, privacy, and accountability, illustrated with real-world examples. This gives nursing leaders actionable strategies and a workflow checklist to guide safe, responsible AI adoption and support nursing informatics practice.
AI Ethics & Governance
Population unclear Setting unclear
Conscious justification in using artificial intelligence: A concept analysis for global nursing practice.
International journal of nursing studies · Oct 2026
This concept analysis used Walker and Avant's method to define 'conscious justification'—the deliberate, reasoned explanation of AI-mediated decisions in nursing, addressed to affected stakeholders. The authors identified six key attributes, including normative transparency, relational answerability, and epistemic humility, plus antecedents and consequences like improved trust but also risks like justification theater. This framework helps nursing informatics build global standards for explaining AI decisions responsibly across different healthcare systems and cultures.
AI Ethics & Governance
Population unclear Setting unclear
Nursing Professional Organisations as Human Rights Intermediaries: Towards an Integrated Framework of Stakeholdership for Healthcare AI Governance.
Journal of clinical nursing · Oct 2026
This discursive paper proposes a framework to guide nursing professional organisations in shaping artificial intelligence governance in healthcare. The framework has three parts: a domain axis for areas of engagement, a modality axis for aligning actions with organisational functions, and a human rights axis defining roles toward rights claimants and duty bearers. It offers nursing leaders and policymakers a practical tool to advocate for equitable, rights-based AI use in healthcare.
AI Ethics & Governance
Population unclear
Ethical and Regulatory Challenges of Generative Artificial Intelligence in Healthcare: A Chinese Perspective.
Journal of Clinical Nursing (John Wiley & Sons, Inc.) · Oct2
This discussion paper examines ethical and regulatory challenges of using generative AI in healthcare, using China as an example. It finds that issues like data use, privacy protection, cost-effectiveness, and regulation remain unclear, and calls for strict ethical guidelines and ongoing oversight. This matters because it highlights concerns nursing informatics professionals should consider when evaluating GenAI tools, especially in resource-limited settings.
AI Ethics & Governance
Population unclear
Ethical Use of Artificial Intelligence in Public Health: Mitigating Risks and Advancing Health Equity.
American Journal of Public Health · Oct2
This paper examines how AI is being used in public health, including surveillance and disease prevention, and finds that weak regulation and oversight create ethical and equity risks, especially for marginalized communities. The authors propose steps like equity impact assessments, community validation, and data sovereignty principles. For nursing informatics, this highlights the need to build workforce capacity and governance frameworks before adopting AI tools in practice.
Workforce & Education 2 Flagged
Workforce & Education
Population unclear Setting unclear
Implementing America's AI Action Plan in Health Care: Impact on Student and Practice-Based EHR Education.
Nursing administration quarterly · 2026
This conceptual article examines the gap between fast-moving U.S. AI policy and health care organizations' readiness, focusing on EHR-embedded AI tools and clinician education. It finds that many organizations already use AI features like ambient documentation and clinical decision support, but clinicians often lack consistent training to safely oversee these tools. The authors argue workforce education and EHR governance are essential to ensure AI adoption improves care rather than increasing risk.
Workforce & Education
Population unclear Setting unclear
Emerging Technologies in Oncology Nursing: Integrating Informatics Competencies to Advance Safe, High-Quality, and Compassionate Care.
Seminars in oncology nursing · Oct 2026
This narrative review examines emerging digital technologies, such as AI, telehealth, virtual nursing, remote monitoring, and device integration, used in oncology nursing care. These tools show benefits like better symptom management and fewer emergency visits, but gaps remain in accountability, equity, and nurse involvement in design. Nurse-led frameworks can help oncology nurses build informatics skills so technology supports, rather than replaces, compassionate care.
AI Methods & Evaluation 2 Flagged
AI Methods & Evaluation
Setting unclear
Cloud-Based and Locally Deployed Language Models in Nursing and Health Care: An AI Act-Aligned Framework.
JMIR medical informatics · Sep 2026
Researchers tested 17 language models (15 large, 2 small) using a new evaluation framework aligned with EU AI Act rules, applying clinical nursing prompts judged against expert panel answers. A strict safety check ruled out 11 models for failing ethical or evidence-based standards; only Anthropic's Sonnet models consistently met recommended safety and accuracy levels, while a locally run model showed promise for staffing tasks. This suggests structured, expert-reviewed evaluation frameworks may help guide safer LLM use in nursing education and clinical decision support.
AI Methods & Evaluation
Population unclear
Optimized Diagnostic Classification of Coronary Artery Disease: Integrating Genetic Algorithm for Feature Selection With an Improved AdaBoost Model.
Health & Social Care in the Community · 9/27
Researchers tested a diagnostic model combining genetic algorithm feature selection with an Improved AdaBoost classifier to detect coronary artery disease, using a public dataset and a local cohort from Noor Heart Center. The model identified chest pain type, ST depression, slope, and sex as key predictors, reaching 93-95% accuracy and outperforming standard AdaBoost and other classifiers. This suggests potential for decision-support tools that could help clinicians classify CAD risk more accurately.