Title: Novel joint modelling extensions for survival and longitudinal data applied to COVID-19
Authors: Carla Diaz Louzao - Universidade de Santiago de Compostela (Spain) [presenting]
Ipek Guler - KU Leuven (Belgium)
Francisco Gude - Complexo Hospitalario Universitario de Santiago de Compostela (Spain)
Santiago Tome - Complexo Hospitalario Universitario de Santiago de Compostela (Spain)
Carmen Cadarso Suarez - Universidade de Santiago de Compostela (Spain)
Abstract: Since the notification at the end of 2019 of several cases of pneumonia caused by the coronavirus SARS-CoV-2, the expansion of the infection has been very fast worldwide, and nowadays it is considered pandemic. The main affected organ is the lung, but liver failure is frequent during the infection, with an elevation in transaminase levels related with the severity of the disease. Furthermore, numerous studies have reported that higher levels of inflammation markers are associated with higher rates of mortality. It is also suspected that these markers influence significantly in the transaminase levels. With this in mind, the longitudinal inflammation and transaminase levels and the risk of death for patients hospitalized in the University Clinical Hospital (Santiago de Compostela, Spain), during the first wave of the coronavirus disease in Spain, are jointly modelled. The joint modelling approaches for longitudinal and survival data have been widely used in follow-up studies in to explore the relationship between longitudinal biomarkers and survival. The novel extensions on joint models are based on the multivariate longitudinal and survival data. One of these extended models is applied to explore the association between the inflammation markers, transaminase levels and risk of death on SARS-CoV-2.