Work

Estimación no paramétrica de la función de supervivencia para datos con eventos recurrentes

Year: 2004

Type: article

Abstract: Recurrent events when we deal with survival studies demand a different methodology from what is used in standard survival analysis. The main problem that we found when we make inference in these kind of studies is that the observations may not be independent. Thus, biased and inefficient estimators can be obtained if we do not take into account this fact. In the independent case, the interocurrence survival function can be estimated by the generalization of the limit product estimator (Peña et al. (2001)). However, if data are correlated, other models should be used such as frailty models or an estimator proposed by Wang and Chang (1999), that take into account the fact that interocurrence times were or not correlated. The aim of this paper has been the illustration of these approaches by using two real data sets.

Source: Revista Española de Salud Pública

DOI: 10.1590/s1135-57272004000200006

Authors: Juan R. González, Edsel A. Peña

Institutions: Institut Català d'Oncologia, University of South Carolina

Language: Spanish

Topic: Statistical Distribution Estimation and Applications

Subfield: Statistics and Probability

Field: Mathematics

Domain: Physical Sciences

Open Access status: diamond

Funder: National Science Foundation

Award: Adaptive Goodness-of-Fit Tests, Recurrent Event Models, and Models with Alternative Time-Scales