###Causal inference to estimate treatment effects from Real World Evidence
Project ID: 2228bc1239 (You will need this ID for your application)
DTP-CASE project: This project involves industrial collaboration. The student would need to spend a minimum of 3 months (as either a single block, or multiple shorter blocks) at the partner’s premises. Industrial partner: Bayer plc Enhanced stipend: This project will fund the stipend at the Wellcome Trust rates (£22,278 in year 1; £24,093 in year 2; £26,057 in year 3; £26,839 in year 4)
Research Theme: Mathematical Sciences
UCL Lead department: Statistical Science
Lead Supervisor: Gianluca Baio
Project Summary:
This PhD research project will investigate the application of novel causal inference methods to estimate treatment effects from real-world data (RWD) sources. The growing use of RWD from electronic healthcare databases (e.g. claims databases and electronic health records) remains an important trend in healthcare decision-making. RWD offer large sample sizes and heterogeneous populations that are representative of routine clinical practice; aspects that are difficult to achieve with randomized trials alone. Standard methods are often subject to misspecification and so we’ll explore rovide valid statistical inference in conjunction with flexible machine learning methods (e.g. super learning, cross-fit estimation, double/debiased machine learning). These approaches can relax restrictive parametric assumptions, facilitate covariate selection, and may outperform traditional approaches in observational data settings. See some of the references below for further details.