I started my PhD at the group Stochastic Operations Research in 2019 after obtaining my master’s degree in Applied Mathematics (UT). My supervisors are Richard Boucherie (SOR), Jean-Paul Fox (OMD), and my co-supervisor is Aleida Braaksma (SOR). In short, my research deals with the combination of Operations Research and Statistics applied to clinical data.
Part of my research is about the development of efficient Bayesian methods and modeling procedures for clinical data. By generalizing the models usually applied in these settings, it is possible to perform a more detailed assessment of treatment performance in e.g. a clinical trial. For instance, using a multivariate model, inference can be performed on (existence of) treatment/clustering effects in subgroups by looking at outcome correlations. This also has applications in personalized health, where treatment can be tailored to specific persons bases on subgroup membership.
My research also focuses on methods that combine Bayesian analysis and decision making that can be applied in clinical settings. For instance, as Bayesian inference is not based on the data collection procedure (as compared to a frequentist analysis), a clinical trial coupled with a Bayesian analysis allows for a flexible, adaptive design based on intermediate outcome analysis. By specifying an objective, optimal decisions can then be made in Bayesian trials which could include dropping/adding a treatment arm or optimizing treatment allocation for the next patient.
Next to the above, I collaborated with the research group CHOIR (University of Twente) on projects related to COVID-19. In the first project, we forecasted the occupancy by COVID-19 patients at a hospital over time using a queueing network. The resulting forecasting method was implemented in several hospitals in the Netherlands to assess the necessary capacity for COVID-19 patients in the coming days. In the second project, we are looking at fair distribution of COVID-19 patients over hospitals using this forecasting method.