ROC analysis as an additional method to characterize time to event data

Pathol Oncol Res. 2005;11(1):50-2. doi: 10.1007/BF03032406. Epub 2005 Mar 31.

Abstract

The receiver operating characteristic (ROC) is predominantly used to assess the discriminant power of diagnostic tests. The present paper proposes this method as an additional alternative to compare survival data of in oncology or related fields. Survival data of brain tumors were analysed with conventional Kaplan-Meier method and ROC. The Area und the Curve (ROC-curve) gives additional illustrative information to distinguish between two therapeutic approaches and low or high grade brain tumors.

Publication types

  • Comparative Study

MeSH terms

  • Adult
  • Area Under Curve
  • Brain Neoplasms / mortality*
  • Data Interpretation, Statistical
  • Female
  • Humans
  • Male
  • Models, Statistical
  • Normal Distribution
  • Predictive Value of Tests*
  • ROC Curve*
  • Sensitivity and Specificity
  • Survival Rate
  • Time Factors