search for


Fitting frailty models for gastric cancer registry data
Journal of the Korean Data & Information Science Society 2018;29:1037-48
Published online July 31, 2018
© 2018 Korean Data and Information Science Society.

Heewon Choi1 · Il Do Ha2 · Maengseok Noh3 · Changhoon Kim4

123Department of Statistics, Pukyong National University
4Pusan National University Hospital
Correspondence to: Professor, Department of Statistics, Pukyong National University, Busan 48513, Korea. E-mail:
Received June 28, 2018; Revised July 16, 2018; Accepted July 18, 2018.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Gastric cancer was the most common carcinoma in Korea, even though incidence has recently decreased and relative survival rate has increased. In particular, the death rate of gastric cancer in Busan was higher than that of the whole country or Seoul during 10 years from 2006 to 2016. In this paper, the survival analysis for the survival time, i.e. time to death from initial diagnosis of gastric cancer, was performed using the Busan cancer registry data. For this, we compared survival rates according to sex and age by Kaplan-Meier survival curves. Especially, because survival rate of gastric cancer may be different according to region, we fitted the frailty models considering Eup-Myon-Dong regional unit as a random effect. Various visualizations and disease mapping of Busan area were presented for the fitting result. In addition, the two inference results were also compared when deprivation index variable was included in the model. As a result, the deprivation index variable decreased the significance for death of gastric cancer among Dongs.
Keywords : Frailty model, gastric cancer, Kaplan-Meier, random effect, survival analysis.