An Extended Weibull Regression for Censored Data: Application for COVID-19 in Campinas, Brazil

This work aims to study the factors that increase risk of death hospitalized patients diagnosed with COVID-19 through odd log-logistic regression model for censored data two systematic components, as well provide new mathematical properties this distribution. To achieve this, a dataset individuals residing in city Campinas (Brazil) was used and simulations were performed investigate accuracy maximum likelihood estimators proposed model. The provided properties, such stochastic representation, identifiability, moments, among others, can help future research since they important information about distribution structure. simulation results revealed consistency estimates different censoring percentages show empirical modified deviance residuals converge standard normal proved be efficient identifying determinant variables survival study, which find more opportune treatments medical interventions. Therefore, considered an interesting alternative works evaluate lifetimes.