A new artificial intelligence study used reinforcement learning to analyze treatment decisions for coronary artery disease, revealing significant variations in clinical practice across patient sexes and hospital sites. Researchers analyzed data from more than 41,000 adults with obstructive coronary artery disease who underwent diagnostic coronary angiography between 2009 and 2019. Evaluation of physician treatment policies showed that female patients received lower expected rewards under physician policies derived from female patient data compared with male patients under male physician policies. Differences were also identified between hospital sites, with policies derived from one site frequently performing differently when applied to patients at another site. The investigators developed reinforcement learning policies and compared their performance with physician behaviour policies, finding that reinforcement learning generally delivered higher expected rewards across sex and site analyses. Reinforcement learning policies tended to recommend coronary artery bypass grafting more frequently than physicians, especially under the least conservative settings. Transfer learning was then applied to adapt models to new target populations; fine tuning a base model with only 10% of target group data produced outcomes that approached those achieved by models trained using the full target dataset. The researchers concluded that reinforcement learning and transfer learning may help optimise treatment, although further validation is required before clinical implementation.