Page 26 - Contributed Paper Session (CPS) - Volume 5

CPS657 Folorunso Serifat A. et al.

Right censored observations in a parametric
mixture cure fraction
Models: application to ovarian cancer
2
2
1
Folorunso S. A. , Chukwu A.U , Oluwasola T.A.O , Odukogbe A.A
1
1 Department of Statistics, University of Ibadan, Nigeria
2 Department of Obstetrics and Gynaecology, College of Medicine, University of Ibadan,
Ibadan, Nigeria.

Abstract
The modeling and analysis of lifetime for terminal diseases such as cancer is a
significant aspect of statistical work in a wide variety of scientific and
technological fields. This study focus on the parametric cure model that can
handle survival data such as G-Family link function. Some structural properties
of these models are studied and the method of maximum likelihood was used
to model parameters of the models. The significance of the models in
diagnosis of ovarian cancer is uncovered and a simulation study was done for
assessing the efficiency and capability of the model. Our results show that the
parametric cure fraction model estimates is found to be quite robust.

Keywords
G- Family link function; ovarian cancer; parametric cure model; structural
properties

1. Introduction
The evaluation of cure fractions in oncology research under the well known
cure rate model has attracted considerable attention in the literature (Hsu et.
al., 2016). The benefits of cure rate models over the traditional methods of
survival analysis, including the well-known Cox regression model. However, in
certain types of cancers such as breast cancer, leukemia, a significant fraction
of patients may now be cured through therapy called cured proportion or
immunes or long-term survivors, The population of interest is thus divided
into two groups viz., cured and non-cured and Cure rate models provide
satisfactory models in such cases (Elangovan and Jayakumar, 2016). A patient
who has survived for five years after a cancer diagnosis is not necessarily
medically cured but is considered statistically cured because the five−year
relative survival analysis is considered a good indication that the cancer is
responding to treatment and that the treatment is successfully extending the
life of the cancer patient. The survival figures so obtained are utilized in
choosing treatment types and regimes, doses, in discriminating between the
side effects profiles and cost effectiveness. Cure models are survival models
basically developed to estimate the proportion of patients cured in a clinical

15 | I S I W S C 2 0 1 9