Predictive models can be applied in different areas, during the emergency of the COVID-19pandemic, in fact, they have proven important in supporting health systems in planningstrategic decisions and in formulating health policies for the containment of thedisease.The Covid-19 pandemic, in particular, has represented a real challenge for our healthcaresystem. In Italy, it was divided into four main waves, each characterized by differenttypes of patients and different therapeutic approaches progressively improved based onnew scientific evidence.The objective is to carry out a study on the data of patients hospitalized for COVID-19at the ASST of Lecco during all four pandemic waves, with different degrees of severityof illness, collecting the data of interest and applying it to they use artificialintelligence to identify recurring patterns of clinical outcome in terms of survival andsecondary infectious complications, so as to build new reliable predictive statisticalmodels that can be used to predict the outcome of the patients themselves.The strong ambition of this project is that the application of artificial intelligence todata of such significant quantity can allow us to build valid statistical models whichcan then be hypothetically applied to any patient to predict, based on anamnesticcharacteristics, blood chemical parameters. at baseline and at the set treatment, theprobability of survival and complications