Using real-time machine learning to prevent in-hospital hypoglycemia: a prospective study - Internal and Emergency Medicine

We conducted a prospective analysis of a machine learning algorithm to predict hypoglycemia. The algorithm was trained, validated, and tested using data from 2013 to 2019. The details of the machine learning methods have been published, but in brief we employed multiple supervised machine learning techniques (e.g., extreme gradient boosting) to predict inpatient hypoglycemia and severe hypoglycemia using a wide-range of patient-level data (i.e., features) including medications, labs, nursing notes, comorbid conditions, among others. Our deployed model was an extreme gradient boosting model. The pre-implementation period for the model was Jan 1, 2018, to May 31, 2020, and the model was implemented on the cardiovascular surgery and vascular surgery ward at St. Michael's Hospital of Unity Health January 1, 2021 and evaluated until April 30, 2022.