Methods Core Seminars

Upcoming seminars

Title:   Adaptive Trial Designs

Presenter: Wenjing Zheng, Ph.D.
Postdoctoral Research FellowCenter for AIDS Prevention Studies, UCSF 

Date&time:  Tuesday, October 13, 2015; 2-4pm

(New) McKusick Conference Room
550 16th Street (at 4th street), 3rd floor
San Francisco, CA 94158

Abstract: Adaptive trial designs allow prospectively planned modifications to key aspects of a study during the course of the trial, without undermining its validity and integrity. Such modifications (adaptations) must be planned before study initiation and should be based on the study’s accumulating data. The study aspects to be adapted can be trial procedures (such as eligibility criteria, treatment dose/duration, stopping point) or statistical procedures (such as randomization scheme, sample size, study hypothesis). The promise of greater flexibility and efficiency has sparked growing attention in adaptive designs in recent years. In this talk, we will first review commonly considered types of adaptive designs and their potential utility in public health research. Then, we will focus on the so-called covariate-adjusted response-adaptive (CARA) randomized controlled trial (RCT). Under this design, one can modify the covariate-adjusted randomization schemes based on previous patients’ responses, with the goal of satisfying a given optimality criterion (e.g. maximizing efficiency, or minimizing adverse outcomes). I will also present recent methodological developments on the design and analysis of CARA RCT that provide robust parameter estimates and improve adaptation towards the optimality criterion..

Short bio: Wenjing Zheng is a postdoctoral fellow at the Center for AIDS Prevention Studies at UCSF. She received her Ph.D. in Biostatistics from the University of California, Berkeley, in 2014. Her research centers on development and application of novel methods to problems in HIV prevention and care in resource-limited settings, with a focus on designing and evaluating individualized treatment strategies and population-level interventions. Her technical interests include social network analysis, machine learning, longitudinal data, adaptive RCT designs, causal inference, and mediation analysis. Her current research is focused on using social network data to understand the social dynamics behind HIV-related health outcomes, behavior and beliefs, with the goal of improving efficiency of interventions.


Materials from past seminars



Last modified: September 28, 2015