Machine learning models to determine the risk of death or intubation

Based on data from more than 9,000 patients treated for the SARS-CoV-2 coronavirus in the Madrid hospitals of Ramón y Cajal, Fundación Jiménez Díaz and La Zarzuela, this work will develop machine learning models to predict the risk of death or of being intubated that a person has by analyzing the factors that will determine the prognosis. In addition, it will evaluate the efficiency of a treatment from comparisons of patients who have been treated with one drug or another and the relationship with the death rate. As the last objective, a Bayesian network model will be built that captures all the relationships between various variables -both clinical, as well as treatments and results- to make probabilistic reasoning about the risk of mortality, the success or not of a treatment and the why. This model will also be left open on a web platform for the entire scientific community.