122 Chapter 5 Abstract Aims To evaluate the acceptability and feasibility among nurses of Decubitus Risk Prediction Alerts based on Artificial Intelligence, and to assess the feasibility of the implementation plan. Design A process evaluation of a pilot implementation study using mixed methods. Methods The acceptability and feasibility of the AI-application (acronym: DRAAI) among nurses from three general wards in a university hospital was assessed with a questionnaire. The tailored implementation plan included 13 strategies distributed over six domains, e.g. facilitation, continuous evaluation, and educational sessions. Details on adaptations, acceptability and feasibility were recorded in field notes. Results Fifty-five nurses completed the questionnaire and valued DRAAI’s predictions, believing these could contribute to pressure ulcer (PU) prevention. Some initially faced challenges distinguishing between PU risk and PU detection. Most nurses found it feasible to integrate DRAAI into their workflow. Several adaptations were made, for instance adding PU preventive measures to the educational sessions, and sharing frequently asked questions and answers. Overall, implementation efforts were deemed feasible. DRAAI generated PU risk predictions for 428 unique admissions, with nurses identifying 20% of at-risk patients prior to or without a risk prediction. Conclusion Ongoing involvement and clear communication were crucial for successfully integrating AI into nursing workflows. Although some nurses were concerned that DRAAI might miss at-risk patients, they continued to independently identify at-risk patients. Implications for the profession and/or patient care Implementation of DRAAI served as a prompt for nurses to focus more on PU prevention. While DRAAI shows promise in improving PU prevention, future research is needed to evaluate its clinical impact.
RkJQdWJsaXNoZXIy MTk4NDMw