We are happy to announce our eighth online seminar in the Biostatistics Seminar Series on Thursday, May 16th, 16h-17h

This series of Biostatistics seminars targets a broad (bio)statistical audience, in particular PhD-students. Specialists discuss a topic of their interest, paying particular attention to concepts relevant and accessible to a non-specialist audience as well.

Speaker
dr. ir. Nan van Geloven
Leiden University Medical Center
Leiden University, the Netherlands

Title
Prediction under interventions: why, what and how?

Abstract
Prediction algorithms are often used to help decision making, for instance to evaluate whether the prognosis of an individual patient warrants a certain medical treatment. However, a major limitation is that most prediction algorithms ignore the role that treatments already played in the data on which they were developed and evaluated. In some situations using prediction models for decision support will actually do more harm than good. To effectively use prediction models in decision making, they must focus on causal (or counterfactual) predictions, where for each individual we predict what their outcome would be under the treatment options under consideration, also referred to as prediction under interventions. In this presentation, I will give an overview of estimands for prediction under interventions and the causal assumptions needed to estimate these from observational data. I will illustrate how existing causal inference methods can be adapted to provide predictions under interventions and discuss new metrics for assessing their performance.

Teams link: (will follow)