AI in hospital medical coding: Performance and uncertainty

The deployment of artificial intelligence (AI) in healthcare organisations is not limited to the domains of research and treatment. Medical coding, an essential administrative stage in hospital operations, particularly in the context of activity-based pricing, is another arena where AI plays its part. Drawing on a survey carried out through interviews and observations in a French hospital centre, we analyse the methods and effects of introducing AI into a Medical Information Department (MID). Given that AI currently has only a marginal impact on professional communication practices and methods within this MID, we show that the main reason for adopting the system is the financial efficiency of coding in the context of activity-based pricing. The article establishes not only that AI does not reduce uncertainty in coding, but also that it is likely to generate uncertainties about the inherent quality of the data produced, data that are essential resources for medical research.