Birth and death long-term survival model

Author(s)

Victor Silva Ritter

• Document type: Master's Dissertation
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Instituto de Matemática e Estatística (IME/SBI)
• Defense date: 2014-08-13

Examining board members

Antonio Carlos Pedroso de Lima; Pranab Kumar Sen; Gisela Tunes da Silva

• Advisor: Antonio Carlos Pedroso de Lima

Abstract

This work contributes with a new cure rate survival model developed aiming more realistic formulation and interpretations than the usual long-term survival models. Motivated by studying residual survival times in oncological patients, the model uses birth and death process to allow free variation on the number of latent risk factors during a pre-follow up period, then considers competing risks scenario for accessing the patients survival. Simulations and application to Instituto do Câncer do Estado de São Paulo data showed improvement over the promotion time model. (AU)

• FAPESP's process: 12/07891-5 - Parametric and semiparametric modeling for long-term survival data with destructive components
• Grantee: Victor Silva Ritter
• Support Opportunities: Scholarships in Brazil - Master