A Taxonomy and Archetypes of AI-Based Health Care Services: Qualitative Study

Background: To cope with the enormous burdens placed on health care systems around the world, from the strains and stresses caused by longer life expectancy to the large-scale emergency relief actions required by pandemics like COVID-19, many health care companies have been using artificial intelligence (AI) to adapt their services. Nevertheless, conceptual insights into how AI has been transforming the health care sector are still few and far between. This study aims to provide an overarching structure with which to classify the various real-world phenomena. A clear and comprehensive taxonomy will provide consensus on AI-based health care service offerings and sharpen the view of their adoption in the health care sector.

Objective: The goal of this study is to identify the design characteristics of AI-based health care services.

Methods: We propose a multilayered taxonomy created in accordance with an established method of taxonomy development. In doing so, we applied 268 AI-based health care services, conducted a structured literature review, and then evaluated the resulting taxonomy. Finally, we performed a cluster analysis to identify the archetypes of AI-based health care services.

Results: We identified 4 critical perspectives: agents, data, AI, and health impact. Furthermore, a cluster analysis yielded 13 archetypes that demonstrate our taxonomy's applicability.

Conclusions: This contribution to conceptual knowledge of AI-based health care services enables researchers as well as practitioners to analyze such services and improve their theory-led design.