Power Estimation for Clearance of Live Vaccine Virus Using Weibull Survival Models with Interval Censoring

• Activity Number: 501
• Type: Contributed
• Date/Time: Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Daniel Zaccaro*+ and Barry Eggleston and Susan Lieff and Dennis Wallace and Jon Hanifin and Donald Leung and Marshall Plaut and Erica Brittain and Robert Holliday and Mark Slifka
• Companies: Rho, Inc. and Rho, Inc. and Rho, Inc. and Rho, Inc. and Oregon Health & Science University and National Jewish Medical and Research Center and National Institute of Allergy and Infectious Diseases and National Institute of Allergy and Infectious Diseases and Rho, Inc. and Oregon Health & Science University
• Keywords: vaccine ; Weibull model ; survival analysis ; atopic

Abstract

A clinical study of an attenuated live vaccine is planned to evaluate immunity of atopic dermatitis (AD) and nonatopic subjects (40 subjects per group), with 5 assessments of viremia up to 14 days postvaccination. Due to interval censoring, comparison of duration of viremia between groups is planned using a Weibull survival model. Power for group comparisons was estimated using simulation studies, assuming Weibull and Cox Proportional Hazards models. A hazard ratio of 0.5 was assumed, but varying sample sizes and extra time points were considered. With 52 subjects per group, an extra postvaccination assessment increased anticipated power to 90%. Without the extra assessment, 70 subjects per group are required for 90% power. The Weibull models had greater power and estimates of the true hazard ratio were unbiased when compared to Cox models. (NIH/NIAID Contract HHSN266200400033C)