Application of Logistic Regression for Incidence of Breast Cancer among Sudanese Females (2005-2008)

Abstract

This paper aimed at application of logistic regression model to the data of  Sudanese females living with breast cancer with reference to Gezira State, the logistic regression method has  been used to study breast cancer situation through a sample of 110 patients. The study basically depended on secondary data collected from the National Cancer Institute (NCI) , the data covered the period (2005–2008). Research methodology included chi-square test to know characteristics of the variables of the study population. The results that the tribe, the stage of illness and marital status are the most statistically significant variables. Multiple logistic regression models are applied in two parts, first part contains all the variables in this research, and the second contains significant variables. The study showed that single women who have breast cancer are more likely to die of breast cancer than women who are married, and women in the late stage are more likely to die of breast cancer than those who are in early stage. The research recommended that, there must be assessment for the current situation of the breast cancer so as to construct strategic plan to stop or eradicate the spread among people. Also case-control data used in this research, together with the cancer registry data, can be used to construct such models of absolute risk.