AI, digital twins, and healthcare utilization behaviour: evidence from a Swiss population survey

Abstract

The increasing development of artificial intelligence (AI) and digital twin technologies in healthcare raises important questions about their future acceptance. Investigating how healthcare utilization behaviour influences openness to such future digital innovations can support the design of health systems that align with utilization behaviour and expectations. A nationally representative online survey of 1,486 adults in Switzerland was conducted in 2023. Participants were categorized into five groups based on their healthcare utilization behaviour: conventional medicine users, complementary medicine users, integrative medicine users, self-treatment users, and pharmacy users. The survey assessed digital literacy, current use of digital tools, and attitudes toward AI and digital twins. Digital twin understanding was ensured through an explanatory video and comprehension checks. Weighted analyses and regression models were used to evaluate differences between groups. (no health care intervention on human participants; no requirement for trial registration or ethics approval). Across all healthcare utilization behaviour groups, the majority of participants reported a willingness to use digital twin services, with significantly higher agreement among self-treatment users. This group also demonstrated the highest digital literacy, most frequent use of health-related apps, and strongest familiarity with AI tools such as ChatGPT. They rated the potential benefits of digital twins more highly and expressed greater trust in both public and private institutions offering these services. Complementary and integrative medicine users reported lower digital literacy and were less likely to consider digital twins beneficial across domains such as cost reduction, therapy validation, and disease prediction. Conventional medicine users expressed specific concerns about digital twins replacing physicians, rated predictive features less favourably, and showed the lowest support for independent personal use of digital twins. Pharmacy users showed moderate willingness, with relatively high trust in institutions and greater appreciation for the coordination and predictive functions of digital twins. This study demonstrates that attitudes toward AI-based health technologies are closely tied to existing patterns of healthcare usage. While digital twins are generally well received, differences in acceptance highlight the need to tailor implementation strategies to specific user profiles.

Ähnliche Arbeiten