Assessment and Presentation of Survival Experience in the Danish Breast Cancer Cooperative Group

Abstract

The purpose of this article is to describe some statistical methods usually applied in articles concerning survival data. Some fundamental concepts for survival data will be described and among others a short review of the statistical theory of Kaplan-Meier plot and log-rank test will be given. The theory will be exemplified using DBCG data with examples of increasing complexity of the statistical models. As an advanced statistical model Cox's regression model for survival data is discussed. This model has been applied in a DBCG article concerning histological malignancy grading of invasive ductal breast carcinoma and the results from this will be brought up to date and discussed.