Offline and online evaluation of medical treatment decision strategies based on large scale electronic health registers

01 January 2016 → 31 December 2019
Research Foundation - Flanders (FWO)
Research disciplines
  • Natural sciences
    • Analysis
    • Applied mathematics in specific fields
    • General mathematics
    • History and foundations
    • Other mathematical sciences
intensive care medicine time-dependent confounding electronic health records Big Data G-estimation sal inference
Project description

The aim of this project is to come up with flexible, robust and efficient statistical methods for evaluating the effect of time-varying treatment strategies or policies, which match the complexities of large electronic health registers. The results will be broadly useful to policy makers and companies who wish to use ‘big data’ to evaluate interventions.