Applications of the Truncated Robust Distance (TRD) Method on Large Aggregated Clinical Lab Data for Safety Monitoring and Assessment

• Activity Number: 194
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 10:30 AM to 12:20 PM
• Sponsor: SSC
• Author(s): Daniel Parks*+ and Xiwu Lin and Lei Zhu and Jie Cheng and Kwan Lee
• Companies: GlaxoSmithKline and GlaxoSmithKline and GlaxoSmithKline and GlaxoSmithKline and GlaxoSmithKline
• Keywords: Outlier Detection ; Hy's law ; Safety Lab data ; Truncated Robust Distance

Abstract

Laboratory safety data are routinely collected in clinical studies for safety monitoring and assessment. Recently, the truncated robust distance (TRD) method was developed for the detection of safety signals. The TRD method can decide a multivariate decision boundary for detecting safety signals using the reference historical data. Here, we show that differences in safety signals between the treatment groups can be determined using the Kaplan-Meier method. We also compare the results of the TRD method with the results of Hy's law, which the FDA has proposed as pre-marketing liver chemistry subject-stopping criteria. For the analysis, data were aggregated from 28 Phase II-IV trials in diseases with normal risk of underlying liver abnormalities.