Determination of Optimal Cut Point for Competing Risks Data

• Activity Number: 75
• Type: Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: ENAR
• Author(s): Jeong Youn Lim*+ and Jong-Hyeon Jeong
• Companies: University of Pittsburgh and University of Pittsburgh
• Keywords: breast cancer ; competing risks ; improper random variable ; log-rank test ; survival data

Abstract

In clinical or epidemiological research, investigators are often interested in dichotomizing the continuous variable to quantify the prognostic effect. Median, mean or quantile is commonly used as a cut point to split the continuous variable. "Outcome-oriented" approach is also considered, which is to seek an optimal cut point that provides the maximum difference in prognostic effect between the splits. Contal and O'Quigley (1999) suggested a method to estimate the optimal cut point of age effect on the survival in breast cancer, and to calculate its significance after the dichotomization. We modify their approach to analyze competing risks data, using the notion of the improper random variable (Gray, 1988). We present the simulation results of the operating characteristics of the proposed method. A real dataset from a clinical trial on breast cancer is exemplified.