CU Medicine develops severe hypoglycemia predictive AI
Researchers from the Faculty of Medicine at the Chinese University of Hong Kong (CU Medicine) have utilised anonymised big data from the Hospital Authority Data Collaboration Laboratory to develop a new machine learning model that can predict the risk of severe hypoglycemia among older diabetic adults.
They analysed about 1.5 million records of more than 360,000 senior individuals with diabetes from 2013-2018. Based on the XGBoost machine learning algorithm, the risk prediction model uses 258 predictors, including demographics, admissions, and diagnoses, to predict severe hypoglycemia events requiring hospitalisation in the next 12 months.
Besides prolonged hospitalisation, severe hypoglycemia is also associated with an increased risk of falls, cardiovascular disease, dementia, and all-cause mortality, CU Medicine noted.
Achieving an 85% positive predictive value in a study, the model can be potentially integrated into EHR decision support systems for pre-emptive interventions, such as correcting the timing and dosage of insulin injections or changing to diabetes medications with lower hypoglycemic potential.
Content Courtesy- HIMSS
