Missing value imputation for clinical prediction models

Missing data are a common problem in the development of prediction models for clinical research. There are ways to handle missing data and some of them are based on variants of multiple imputation by chained equations, while others are based on single imputation. We aimed to investigate by simulation if some of these methods consistently outperform others in performance measures of clinical prediction models. We designed an extensive simulation study following the ADEMP framework to address this question.

The manuscript is currently under review, and a preprint is available on SSRN: here
The simulation study R code and graphical abstract of the manuscript was made available on OSF: here