Implications of Using Recurrent Episodes: Statistical Considerations in a Rare Disease Setting

• Activity Number: 511 - In Pursuit of Speed and Efficiency in New Drug Development: Statistical Designs at the Forefront of Pushing the Boundaries of Innovation
• Type: Topic Contributed
• Date/Time: Thursday, August 11, 2022 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Hengrui Sun*
• Companies: FDA
• Keywords: Rare disease; Clinical trials; Recurrent episodes; Correlations

Abstract

One of the major challenges for rare disease clinical trials is the limited number of patients impacted by the disease. For some rare diseases, patients experience recurrent episodes of the outcome or event of interest. Drug developers occasionally consider a design which incorporates multiple episodes when the recurrent events are not life threatening and the responses are not expected to be heterogenous across episodes. However, within patient correlation and its impact on the efficacy evaluation are often not taken into account or are underestimated. A simulation study was conducted to explore the operating characteristics when the design and analysis include recurrent episodes. Several disease characteristics, correlation structures, and recurrence rates were considered, and their impacts were assessed. In this presentation, the gains and the concerns of such design will be discussed.