Identification of predictive models of COVID-19 severity in a multi-state setting for its use in risk stratification (ID-CoV)

During the course of the COVID-19 outbreak, a wealth of data has been accumulated
from the efforts of the health systems to overcome the pandemic. Months of patient
encounters with primary to tertiary care systems are leaving an affluence of valuable
information reflecting the real impact of the virus in people’s health and lives. These
real-world data [RWD] offer an unparalleled opportunity to understand COVID-19 but
also an important analytical challenge due to the dissimilar and heterogeneous nature
of this information. Using complementary data sources from primary health care and
hospitals, this project aims to set up a methodological framework for data
harmonization, linkage, and analytical development of a novel tool for multi-state risk
prediction identifying the role of comorbidities, among other factors, in predicting
COVID-19 progression into severity, and subsequent recovery or death. This research
will afford a unique instrument for risk stratification and resource allocation in the
face of current and future epidemics and will serve as a proof of concept of the
usefulness of RWD and the feasibility of the adaptation of novel this methodological
framework to other countries/settings based on local data.

Project type:PhD project