3/23/2016- Laura Balzer- Targeted Learning in the SEARCH trial and HIV prevention in East Africa
Publication information:
Abstract
Title: Targeted Learning in the SEARCH trial and HIV prevention in East Africa
Abstract:
Evaluation of community-based interventions presents significant methodological challenges. In this talk, we describe the design and analysis of the SEARCH trial, an ongoing community randomized trial to evaluate the impact of early HIV diagnosis and immediate treatment with streamlined care in rural East Africa. We focus on 3 choices to optimize study power: adaptive pair-matching over complete randomization, targeting the sample average treatment effect instead of a population average parameter, and data-adaptive adjustment through a pre-specified targeted maximum likelihood estimator (TMLE). These choices are compared theoretically and with finite sample simulations. We demonstrate each choice improves efficiency relative to standard practice, while maintaining nominal confidence interval coverage.
[[{"fid":"3148011","view_mode":"default","type":"media","attributes":{"height":"360","width":"480","alt":"Applied Stats 3/23/16 - Laura Balzer on YouTube","title":"Applied Stats 3/23/16 - Laura Balzer","class":"media-element file-default"}}]]
Full text
Title: Targeted Learning in the SEARCH trial and HIV prevention in East Africa
Abstract:
Evaluation of community-based interventions presents significant methodological challenges. In this talk, we describe the design and analysis of the SEARCH trial, an ongoing community randomized trial to evaluate the impact of early HIV diagnosis and immediate treatment with streamlined care in rural East Africa. We focus on 3 choices to optimize study power: adaptive pair-matching over complete randomization, targeting the sample average treatment effect instead of a population average parameter, and data-adaptive adjustment through a pre-specified targeted maximum likelihood estimator (TMLE). These choices are compared theoretically and with finite sample simulations. We demonstrate each choice improves efficiency relative to standard practice, while maintaining nominal confidence interval coverage.
[[{"fid":"3148011","view_mode":"default","type":"media","attributes":{"height":"360","width":"480","alt":"Applied Stats 3/23/16 - Laura Balzer on YouTube","title":"Applied Stats 3/23/16 - Laura Balzer","class":"media-element file-default"}}]]