Estimating crude cumulative incidences through multinomial logit regression on discrete cause-specific hazards

Authors:
Federico Ambrogi;Elia Biganzoli;Patrizia Boracchi
Affiliations:
Istituto di Statistica Medica e Biometria "G.A. Maccacaro", University of Milano, Italy;Istituto di Statistica Medica e Biometria "G.A. Maccacaro", University of Milano, Italy and Unití di Statistica e Biometria, National Cancer Institute, Milano, Italy;Istituto di Statistica Medica e Biometria "G.A. Maccacaro", University of Milano, Italy
Venue:
Computational Statistics & Data Analysis
Year:
2009

Citing 4
Cited 0

A general framework for neural network models on censored survival data

Neural Networks
Joint modelling of cause-specific hazard functions with cubic splines: an application to a large series of breast cancer patients

Computational Statistics & Data Analysis
Regression Modeling Strategies

Regression Modeling Strategies
Modern Applied Statistics with S

Modern Applied Statistics with S

Quantified Score

Hi-index	0.03

Visualization

Abstract

In the presence of competing risks, the estimation of crude cumulative incidence, i.e. the probability of a specific failure as time progresses, has received much attention in the methodological literature. It is possible to estimate crude cumulative incidence starting from models defined on cause-specific hazards or to adopt regression strategies modeling directly the quantity of interest. A generalized linear model based on discrete cause-specific hazard is used to obtain the crude cumulative incidence and its asymptotic variance. The model allows inference both on cause-specific hazard and on crude cumulative incidence in the presence of time dependent effects. Standard software can be used to compute all quantities of interest. A trial of chemoprevention of leukoplakia is considered for illustrative purposes, where different patterns of risk are suspected for the different causes of treatment failure.