Regression modeling of the cumulative incidence function with missing causes of failure using pseudo-observations

Margarita Moreno-Betancur (1) , Aurélien Latouche (2)

1 CESP - Centre de recherche en épidémiologie et santé des populations
2 CEDRIC - Centre d'études et de recherche en informatique et communications

The rationale for the present work is to derive a flexible class of regression models for the the cumulative incidence function when there are missing causes of failure, encompassing key models such as the Fine and Gray and additive (Klein, 2006) models. More precisely, we propose two approaches that extend the Andersen-Klein approach to the missing cause setting. The first approach is grounded on the inverse probability weighting paradigm for dealing with missing data and the second is a multiple imputation method tailored for the Andersen-Klein model. We illustrate both approaches by analyzing the data from the ECOG 1178 breast cancer treatment clinical trial.

Mots clés: competing risks; missing data

Mots clés: risques concurrents; données manquantes

Domaines: Informatique [cs]; Statistiques [math.ST]

Margarita Moreno-Betancur, Aurélien Latouche. Regression modeling of the cumulative incidence function with missing causes of failure using pseudo-observations . 58. Biometrisches Kolloquium, Mar 2012, Berlin, France. ⟨hal-01125964⟩