Systems and methods for using machine learning models to predict risks posed to patients having various respiratory conditions. According to certain aspects, an electronic device may generate a machine learning model using training data indicating various patient and medication data. The electronic device may access a set of real-world patient data and input the real-world patient data into the machine learning model. An output of the machine learning model may indicate a likelihood of the patients needing intervention care, and the electronic device may facilitate outreach efforts to certain patients in an attempt to decrease this likelihood.