Using machine learning to predict severe hypoglycaemia in hospital.

Machine-learning approaches can accurately identify patients at high risk of hypoglycaemia in hospital. Future work will involve evaluating whether implementing this model with targeted clinical interventions can improve clinical outcomes.