Task- based analysis of artificial intelligence- driven transformation of clinical work in diabetes care: a Delphi and Fuzzy DEMATEL study

Artificial intelligence (AI) is rapidly becoming embedded in diabetes care, yet workforce impact is still framed mainly in terms of “jobs replaced,” masking the more important reality: clinical work changes at the level of tasks. This study identifies the core task families in diabetes management and maps how AI transforms them through three pathways—substitution, augmentation, and rebundling then explains why these shifts occur using a theoretically grounded, causal framework. Using a mixed qualitative–quantitative design, a focus group of ten experts from India and the United Arab Emirates (UAE) identified twelve clinical task families.

A five-round Delphi process classified each task by its dominant AI pathway. To examine how transformation translates into workforce and care outcomes, six theory-informed constructs were modelled using Fuzzy Decision-Making Trial and Evaluation Laboratory (DEMATEL), capturing expert judgments through linguistic ratings converted into triangular fuzzy numbers and defuzzified to derive prominence and cause–and–effect structure. Across tasks, AI most frequently augments clinical work rather than fully substitutes it, with several tasks showing hybrid patterns.

The expert-informed causal model suggests that Sociotechnical Fit and AI Capability and Utility may function as important upstream drivers, with Task Characteristics acting as an initial trigger. Role Redesign and Workforce Capability emerge as key mediators, while Clinical Outcomes function as a downstream result shaped by these interacting forces. A threshold-based sensitivity analysis further confirmed the stability and robustness of the identified possible causal relationships.

Overall, the exploratory findings suggest that the benefits of AI in diabetes care depend less on automation and more on workflow alignment, governance, and deliberate human–AI collaboration. AI is therefore best understood as a catalyst for task reconfiguration, not job replacement, making sociotechnical alignment and purposeful role redesign central to achieving workforce resilience and improved care quality. The authors of this paper are highly obliged to their respective organizations for the infrastructural support provided.

We are also grateful to the focus group members for providing their invaluable input. Datta Meghe Institute of Higher Education & Research, Wardha, India Thumbay College of Management and AI in Healthcare, Gulf Medical University, Ajman, UAE The authors declare no competing interests. Participation in both the focus group discussion and the Delphi study was voluntary.

All experts were informed about the objectives of the study, the nature of their participation, and the intended use of the findings before data collection. Informed consent was obtained from all participants. To protect privacy, expert responses were anonymized during analysis and reporting, and no personally identifiable information was collected or disclosed.