Complex AI-nursing-care projects as a challenge
Artificial intelligence (AI) already supports medical diagnostics, optimizes clinical processes or indicates complex risk constellations for individual patients and suggests healthcare interventions. Nursing could also benefit from increased use of AI in the future. Research & Development (R&D) projects, however, often encounter hurdles. Projects fail, AI solutions fall short of their potential, and nurses are reluctant to welcome the new technology into their daily work.
The participation of stakeholders is both a normative and a functional requirement for R&D projects that want to set the course for success from the planning phase onward. However, nurses or nursing scientists are often not involved in projects at all or they only play a marginal role in fundamental project decisions. Knowledge of the needs for AI in nursing care and of the challenges R&D projects have to overcome helps researchers and practitioners to plan and conduct successful projects that achieve real-world impact.
How can we support nurses with assistive technologies in order to provide the right information at the right time in the right place?
In preparation for a new, large-scale German public funding program for AI in nursing, we conducted a mixed-methods study to explore needs, use cases, requirements, facilitators, and barriers to R&D. We invited nurses, nursing directors, digitization officers, informal caregivers, and researchers with backgrounds in nursing science or education, informatics and AI research, or ethics to discuss and prioritize needs and challenges for AI in nursing care.
We were particularly interested in requirements for and characteristics of successful R&D projects. Specific needs for AI solutions in nursing care were also collected, as addressing specific problems and demands of the users is one important factor for a successful product in itself. We sifted through literature, conducted workshops, interviews, an online survey, and a datathon. Topics of particular interest for research funding included AI-based care needs and risk assessment, evidence-based decision support systems, and care planning in complex care settings.
These deep learning systems, you can use them as tools to try to understand complex issues for which you don’t yet have a model, for which you don’t yet have a theory
We summarized requirements and characteristics of successful research projects and identified five overarching topics. Regulatory requirements addressed data analysis and models of data sharing in compliance with privacy regulations. Process and translational requirements included aspects of project planning, execution, and management.
Success criteria include ethical and legal aspects, supportive communities and ecosystems, and incorporation or reflection of existing frameworks and tools for the development and implementation of technologies and interventions in care and healthcare.
What does this mean for R&D on AI in nursing care?
The use of AI methods shows diverse opportunities to support nursing care. A prerequisite for the use of AI is access to representative and high-quality data. When implementing AI in nursing processes, all stakeholders such as nursing professionals and nursing science must be closely involved. This should happen as soon as possible – preferably upon ideation – and partners should be equipped with reasonable resources for meaningful collaboration.
By implementing communities and eco-systems that go beyond the boundaries of individual projects, translation and dissemination of practical findings that may contribute to nursing education and the scientific community could be supported.
Participatory and demand-driven development remains a key element in the design of R&D projects, which should ideally aim to bring AI solutions out of the lab and into nursing practice.