Exploratory Scientific Data Analysis

• Activity Number: 663
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Gary Cline*+
• Companies: ICON Clinical Research
• Keywords: Type I ; error ; exploratory ; market

Abstract

A key component of statistical market support is to explore the relationship between an intervention and disease with respect to scientific, safety and efficacy endpoints. Physicians, reimbursement decision makers and researchers have many post-approval questions related to an intervention's performance with respect to subgroups, perceived mechanism of action, new safety and efficacy endpoints, new population identification markers or disease progression and other confounding factors. Development programs should prospectively plan to address some of these needs where possible early in the process. The analyses become well defined and conducted after the intervention is approved. It is important to highlight the level of evidence achieved with these analyses as suggestive/investigate rather than confirmatory. The Type I error is inflated due to multiplicity for these additional analyses. This is somewhat mitigated through the high to moderate correlation of these additional analyses to the primary endpoint in the trials being used. The results of these analyses can be used as the basis for confirmatory or epidemiology studies and additional scientific understanding.