Journal of Biopharmaceutical Statistics, 22: 988–1000, 2012 Copyright © Taylor & Francis Group, LLC ISSN: 1054-3406 print/1520-5711 online DOI: 10.1080/10543406.2012.703603 QUALITATIVE CONSISTENCY OF TREATMENT EFFECTS IN MULTIREGIONAL CLINICAL TRIALS Yoko Tanaka 1 , Gang Li 2 , Yining Wang 3 , and Josh Chen 4 1 Eli Lilly and Company, Indianapolis, Indiana 2 LifeScan Inc. a Johnson and Johnson company, West Chester, Pennsylvania 3 Johnson and Johnson, Titusville, New Jersey 4 Merck & Co., Inc., Rahway, New Jersey Consistency of treatment effects across different regions in multiregional clinical trials (MRCTs) has been an important question for the regulatory authorities. Many consistency definitions are proposed in literature. One of the definitions of consistency is expressed as qualitative consistency, whereas inconsistency is defined as qualitative treatment by region interaction. This article focuses on the qualitative consistency and extends Gail-Simon and Sasabuchi’s one-sided multivariate likelihood ratio tests. Simulations are used to evaluate operating characteristics of these qualitative consistency assessment approaches. For a given number of regions, the guideline for setting significance level, and consistency cut-off are explored. Key Words: False-positive; Multiregional clinical trial; Multivariate likelihood ratio test; Probability for claiming consistency; Qualitative consistency. 1. INTRODUCTION Multiregional clinical trials (MRCTs) have become a standard in drug development for many reasons. However, the interpretation of MRCTs can be a challenge because results may not be consistent due to differences in ethnicity, culture, and clinical practice across regions, as such differences may potentially have an impact on the treatment administration or dosing regimen, as well as on the observed treatment efficacy and safety. In 1998, the ICH E5 guideline (ICH, 1998) was adopted to recommend a framework for evaluating the impact of ethnic factors on drug effects. Basically, for compounds that might be sensitive to ethnic factors, if a sponsor can show evidence of similarity of treatment effects across regions through either a multiregional clinical trial (MRCT) or a bridging study, results can be extrapolated across regions. However, while ICH E5 recommended strategies for accepting foreign clinical data for drug approval in a region, the guideline did not provide either a specific standard for the design and conduct of multiregional and bridging studies, or specific criteria for the assessment of similarity of treatment effect across regions. Received October 22, 2011; Accepted May 31, 2012 Address correspondence to Yoko Tanaka, Eli Lilly and Company, Lilly Corporate Center DC1542, Indianapolis, IN 46285, USA; E-mail: yokot@lilly.com 988