Mapping Economic Considerations in Clinical Artificial Intelligence Research

Abstract

Artificial intelligence (AI) has been rapidly adopted in clinical research over the past decade, yet the extent to which economic considerations are integrated into this literature remains unclear. This study presents a large-scale bibliometric analysis of clinical AI research indexed in the Web of Science. Temporal analyses span 2000–2024 (54,219 clinical AI studies), while network mapping, citation overlay, and density analyses focus on the 2024 snapshot (N = 14,995). A ratio-based indicator was used to track the relative prominence of economic considerations over time. The results show a sharp acceleration in clinical AI publications after 2015, while studies explicitly addressing cost, cost-effectiveness, or economic burden remained persistently rare, accounting for less than 1% of annual output in most years. Structural analyses indicate that economic terms are closely linked to modeling and decision-oriented keywords but do not form independent thematic clusters. Although economics-focused studies achieve moderate normalized citation impact when present, their low frequency limits structural influence. The findings reveal a persistent imbalance between rapid methodological innovation and limited economic evaluation in clinical AI research, highlighting the need for more systematic integration of economic perspectives to support sustainable clinical deployment.

Ähnliche Arbeiten