Effect of Informative Censoring on the Power of Some Log-Rank Type Tests.

The survival analysis is a set of techniques for analyzing the elapsed time until the occurrence of an event of interest, also called lifetime, which can be completely or partially observed. In practice, a common problem is to compare the survival experiences of at least two groups...

ver descrição completa

Detalhes bibliográficos
Principais autores: Brun González, L. P., Salazar Uribe, J. C.
Formato: Online
Idioma:spa
Publicado em: Universidad Pedagógica y Tecnológica de Colombia 2016
Assuntos:
Acesso em linha:https://revistas.uptc.edu.co/index.php/ciencia_en_desarrollo/article/view/4230
Descrição
Resumo:The survival analysis is a set of techniques for analyzing the elapsed time until the occurrence of an event of interest, also called lifetime, which can be completely or partially observed. In practice, a common problem is to compare the survival experiences of at least two groups under study. The analysis strategy most common is to use Log-Rank tests, but these do not distinguish whether the mechanism of censoring it is informative or not. In this work it was evaluated, under different simulation scenarios, the estimated power of some Log-Rank tests when informative censoring is presented. We found evidence favoring the Wilcoxon test.