Overcoming Medical Overuse with Ai Assistance: An Experimental Investigation

Abstract

This study examines the role of Artificial Intelligence (AI) in reducing medical overtreatment, a costly and risky issue in healthcare. Two lab-in-the-field experiments at a medical school and a hospital tested a novel medical prescription task under three incentive schemes: Flat (constant pay), Progressive (pay increases with treatment quantity), and Regressive (penalties for overtreatment). AI assistance significantly reduced overtreatment, achieving reductions of up to 62% (medical students) and 82% (physicians) under specific schemes. Diagnostic accuracy improved with AI across all treatments, with gains between 14% and 37%. AI adoption was notable, with half of medical students and one-third of physicians incorporating AI suggestions. Policy analysis revealed that AI mitigates non-monetary incentives for overtreatment and promotes social welfare. These findings offer practical insights for integrating AI into healthcare systems to enhance decision-making and reduce unnecessary interventions.

Note:
Funding Information: Lijia Wei acknowledges support from National Science Foundation China (72433003,72173093), and the Center for Behavioral and Experimental Research (CBER) at Wuhan University. Lian Xue acknowledges support from the Research Funds for Youth Academic Team in Humanities and Social Sciences of Wuhan University (413000425).

Conflict of Interests: All authors declare no conflict of interest. Lijia Wei, Ziyi Wang, Lian Xue

Ethical Approval: The Medical Students’ Experiment (Experiment 1): IRB provider: Economics and Management School of Wuhan University; IRB Approval Number:EM240014. The Physicians’ Experiment (Experiment 2): IRB provider: Economics and Management School of Wuhan University; IRB Approval Number:EM240021.

Keywords: medical overuse, overtreatment, artificial intelligence (AI), lab-in-the-field experiment