Prediction meets causal inference: the role of treatment in clinical prediction models

• DOI: 10.1007/s10654-020-00636-1
• Publication Date: 2020-05-22T06:02:45Z
• Authors: Nan van Geloven; Sonja A. Swanson; Chava L. Ramspek; Kim Luijken; Merel van Diepen; Tim P. Morris; Rolf H. H. Groenwold; Hans C. van Houwe...; Hein Putter; Saskia le Cessie
• Keywords: Censoring; FOS: Computer and information sciences; Clinical Trials as Topic; Models, Statistical; EMC OR-01; Estimands; 01 natural sciences; 3. Good health; Treatment; Methodology (stat.ME); 03 medical and health sciences; 0302 clinical medicine; Research Design; Clinical Decision Rules; Data Interpretation, Statistical; Methods; Humans; Predictimands; 0101 mathematics; Clinical prediction model; Statistics - Methodology

In this paper we study approaches for dealing with treatment when developing a clinical prediction model. Analogous to the estimand framework recently proposed by European Medicines Agency trials, propose 'predictimand' of different questions that may be interest predicting risk in relation started after baseline. We provide formal definition estimands matching these questions, give examples settings which each is useful and discuss appropriate estimators including their assumptions. illustrate impact predictimand choice dataset patients end-stage kidney disease. argue clearly defining equally important research as causal inference.