Cure rate and statistical inference for the Marshall-Olkin generalized defective Gompertz distribution under censored survival data

Abstract

Cure fraction modeling usually requires the addition of an extra parameter. Defective modeling, alternatively, allows cure fraction modeling with less complexity. A new extension of the defective version of Gompertz distribution is studied in this paper. This distribution is compared to other well-founded distributions and its distinguished performance to fit censored survival data with and without a cure fraction is presented. A computational study is carried out to highlight the correlation between the censorship level and the proportion of long-term survivors. Interesting mathematical properties are established. The practical relevance of the model is illustrated by applying various real and simulated datasets.