Optimal Stratification of a Continuous Variable Using Overall C-Index and ROC

• Activity Number: 253
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Sunita Ghosh*+
• Companies: Alberta Health Services
• Keywords: optimal stratification ; overall c-index ; ROC ; cross-validation ; jackknife

Abstract

Studies in clinical or epidemiological setting usually involve categorization of a continuous variable. Several methods have been proposed in literature to determine the optimal cut-point of a continuous variable. In the present paper we will be using overall c-index and receiver operating curve (ROC) to determine the optimal cut-point, and will be comparing both these methods and test for the most unbiased optimal cut-point. Two-fold cross-validation and Jackknife methods will be used to validate the cut-point methods. The cut-point will be determined based on univariable as well as multivariable setting. The cut-point will be determined for binary as well as survival outcome.