Electronic medical record-based machine learning predicts the relapse of asthma exacerbation

Abstract

The minimization of asthma exacerbation (AE) is a prioritized objective of asthma management given that repeated AE increases the future risk of fixed airway obstruction, adverse effects from corticosteroids and other medications, and even mortality.1,2 Despite the development of novel biologics and supportive strategies proven to reduce AE, the prediction and prevention of exacerbation incidents remain a challenge.3?5 Here, we present a machine-learning model that predicts AE recurrence using EMR in a single tertiary hos