Application of Binary Logistic Regression Model to Cancer Patients: a case study of data from Erbil in Kurdistan region of Iraq.

Abstract

In this paper, Binary Logistic Regression technique for fitting the best model for analyzing cancer diseases data is introduced using SPSS software based on forward stepwise procedures using different tests. The main objective is to investigate the last status of patients suffer from various types of cancer in Kurdistan Region of Iraq for the period 2010-2019 based on the main factors that contribute significantly to their last situation. A random sample of size 821 cancer patients from two main public hospitals (Rzgari and Nankali) in Erbil, where the patients take permanent and appropriate treatment, is selected of 619 alive and 202 dead. The results of the study showed that binary logistic regression is an appropriate technique to identify statistically significant predictor variables such as gender, age, cancer site and region to predict the probability of the last status (alive or dead) for each cancer patients. Moreover, it is deduced that despite the higher rate of cancer for female patients than the male; the chance for female cancer patients to be alive is more than male patients.