Please join us on Thursday, October 25 for a CDUHR Methods Core Presentation featuring Daniel J. Feaster, PhD from the University of Miami, who will give a talk titled "Methods for Identifying Subgroups and Heterogeneity in Treatment Response". This talk will provide a non-technical explanation of two different statistical methods for identifying subgroups and heterogeneity in treatment effects. Latent class analysis can be used to identify homogeneous subgroups with similar patterns of experiences or characteristics (for example patterns of substance use). This approach will be illustrated using substance use data and examining the relationship of the subgroups to sexually transmitted infections. Random forests are a non-parametric statistical learning (aka machine learning) approach which can be used to identify how individuals respond to different treatments. This approach is behind much of what is now called precision medicine. An application based on HIV risk reduction counseling will illustrate the approach.
Daniel J. Feaster, PhD, is a Professor in the Division of Biostatistics, Department of Public Health Sciences at the University of Miami Miller School of Medicine. Dr. Feaster’s research focuses on how best to address substance abuse and other vulnerabilities to improve access to and retention in HIV care and HIV prevention services. He is also developing methods to predict individual-specific treatment effects to uncover and understand heterogeneity of outcomes in HIV and substance abuse clinical trial, observational and registry data. He has been the statistician on 22 federally funded projects, including center grants, RO1s, cooperative agreements, and multi-site trials.
Registration is not required. For more information about this event, reach out to Chris Hilliard at crh218@nyu.edu.