140 Chapter 5 Discussion The pilot implementation of DRAAI was well-received among nurses, demonstrating its potential to enhance PU prevention. Most nurses found DRAAI predictions helpful, particularly for early risk detection and reducing the registration burden compared to the Waterlow score. However, the need for clearer communication regarding the distinction between PU risk prediction and PU detection, as well as additional explanations on how DRAAI generates its predictions, became apparent. Concurrently, a notable portion expressed neutrality or uncertainty about its predictions and ease of use. The implementation strategies were generally feasible, though attending all daily joint sessions was challenging for the project team. Continuous evaluation during the pilot phase led to several adaptations to both DRAAI and the implementation strategies. These adjustments were crucial for addressing practical challenges and improving DRAAI’s usability. Overall, high fidelity in follow-up actions for PU prevention was observed, with most of the at-risk predictions resulting in timely preventive PU measures. This pilot underscores the importance of ongoing engagement, proactive adaptations to the implementation plan, and iterative refinements to fully integrate DRAAI into clinical practice. Ultimately, these efforts aim to better support nurses in delivering improved patient care. Integrating AI-based prediction models into daily practice can potentially assist nurses in clinical decision-making, enabling more evidence-based and personalized care [43, 44]. Nevertheless, significant challenges remain in successfully integrating AI into healthcare. Ronquillo, Peltonen [44] conducted an international think-tank and identified three key priorities for AI integration: (a) Nurses must understand the relationship between the data they collect and AI technologies they use; (b) Nurses should be actively involved in all stages of AI development and implementation; and (c) there is substantial untapped potential for nursing to contribute to the development of AI technologies for global health and humanitarian efforts. Our study addressed the first two topics during the development and implementation phases of DRAAI. In the development phase, we co-developed DRAAI with nurse champions and wound care nurses, which provided valuable insights into nursing practice and the quality of nurses’ documentation in patient records regarding PU risk [28]. The findings from a cross-sectional survey that explored the relationship between nurses' digital health literacy and their educational levels [45]. While we did not collect demographic data from the nurses involved in our study, we believe that factors such as educational level and frequency of using digital technologies substantially influenced
RkJQdWJsaXNoZXIy MTk4NDMw