Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
When, where, who, what, and why? The five Ws of workflow analysis for implementing an AI decision support tool at the intensive care
2
Zitationen
9
Autoren
2025
Jahr
Abstract
There has been a surge in the development of clinical artificial intelligence (AI) decision support tools. Yet, few of these tools have been implemented into clinical practice. A lack of understanding of the clinical workflow may hamper implementation. This study develops and illustrates a method for the analysis of a clinical decision-making workflow at the ICU to facilitate the safe and efficient implementation of an AI decision support tool. A method is proposed that uses observations to study the context in which clinical decisions occur (physical workflow), and the key factors influencing decision-making (cognitive workflow). This approach provides a comprehensive understanding of the ‘when’, ‘where’, ‘who’, ‘what’, and ‘why’ of the decision-making process. The method was applied to a use case involving the decision to discharge a patient from the ICU to the regular ward. The results showed that the proposed methodology effectively provided a thorough understanding of the physical workflow, in terms of time, location, actors, and materials for receiving decision support with AI. Additionally, it identified the key factors influencing the cognitive workflow. We presented a method for the workflow analysis of clinical decision-making at the ICU before AI implementation. A thorough understanding of the clinical workflow is essential for implementation of an AI decision support tool. This method forms a blueprint and further validation is needed for other clinical contexts.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.539 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.426 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.921 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.586 Zit.