A carregar...
Publicação
Artificial intelligence-mediated clinical judgement in nursing: a sequential multimethod experimental study on pressure injury prevention in intensive care
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 57.1 KB | Adobe PDF |
Orientador(es)
Resumo(s)
Background: Pressure injuries (PIs) remain a frequent and largely preventable complication in intensive care units, where prevention relies on nurses’ clinical judgement under conditions of uncertainty, time pressure and high cognitive load. While artificial intelligence (AI) has been extensively studied for risk prediction, critical gaps remain in understanding how AI-mediated information shapes nurses’ reasoning processes, confidence and cognitive workload during preventive decision-making. Current research predominantly evaluates decision outputs or algorithmic accuracy, overlooking the cognitive and judgement-related mechanisms through which AI may support, or inadvertently disrupt, clinical practice in wound care. Aim: To examine how AI-mediated informational support influences nurses’ clinical judgement processes, decision quality, confidence and cognitive workload during PI prevention in intensive care. Methods & Current Status: This PhD adopts a sequential multimethod design. Phase 1 (completed) comprised two scoping reviews on AI-based PI prediction and AI support for clinical judgement in nursing. Phase 2 (ongoing) focuses on the development and content validation of realistic intensive care PI prevention vignettes. Phase 3 will analyse nurses’ questions during interaction with a clinical chatbot to identify cognitive support needs and develop a taxonomy of judgement-related informational demands. Phase 4 involves a controlled pilot simulation study in which nurses assess clinical vignettes with and without standardised AI support. Outcomes will include judgement quality, decision confidence and objective neurocognitive indicators of cognitive workload derived from electroencephalography. Anticipated Impact: By shifting focus from algorithmic performance to nurses’ clinical judgement processes, this project will generate practice-oriented evidence to inform the design of cognitively aligned, ethically grounded and clinically usable AI decision-support systems for PI prevention. Findings will identify what information, when, and in which format AI support most effectively strengthens nurses’ judgement under high cognitive load. The project is expected to reduce unwarranted variability in preventive decision-making, enhance patient safety, and support responsible AI integration into wound care practice, professional education and evidence-informed clinical guidelines. Ethics: Ethical approval and informed consent will be obtained prior to data collection. All procedures will comply with data protection regulations, including secure handling of neurophysiological, simulation and interaction data.
Descrição
Palavras-chave
Contexto Educativo
Citação
Editora
European Wound Management Association (EWMA)