149 AI-based risk predictions: Evaluation of pilot DRAAI 5 Dose Temporality Implementation outcome affected Justification 15 informal visits performed (5 up to 30 minutes per ward) 10 educational sessions 55 responses were collected through a survey Informal conversations continuously. Survey during the pilot. Between 3-8 weeks after the start. Acceptability, feasibility The NASSS-Framework was used to categorize the barriers identified from literature and clinical expertise. Barriers were extracted from the literature, extracted from papers implementing prediction models [1] or decision support systems [2]. Barriers from previous implementation efforts in our hospital, which were gathered in questionnaires and interviews after the implementation of smart medical beds. • Evaluation within the project team lasted 5 up to 30 minutes. • Informal evaluations lasted 2 up to 30 minutes with individuals or small groups of nurses/PU nurses/ Nurse managers • Formal evaluation took about 30 up to 60 minutes. • Project team evaluations were conducted frequently throughout the pilot period • Informal evaluations were conducted frequently throughout the pilot period • Formal evaluations were planned between 5 and 8 weeks after the start of the pilot Acceptability, Feasibility To address the determinants Disruption of usual workflow, flexibility and Impact on workload, Every 15-minutes the follow-up of the risk predictions was updated according to the actions taken by nurses (create a nursing care plan) Realized from the start of the pilot, and improved in an update during the pilot Fidelity To address the determinants Clinical utility and User friendliness
RkJQdWJsaXNoZXIy MTk4NDMw