Welcome to the website for:

Clinical Prediction Models

A practical approach to development, validation
and updating; by Ewout Steyerberg

Welcome to this site that supplements the book Clinical Prediction Models.

Prediction models are increasingly relevant in the medical field. We witness an increase in biological knowledge on potential predictors of outcome, e.g. on biomarkers; increasing access to large data sets; and popularity of Machine Learning methods. Applications for prediction models include targeted early detection of disease, and individualized decision-making for diagnostic testing and treatment.

The first edition of Clinical Prediction Models was published in 2009. A fully revised version was published in 2019, again by Springer. It includes more material and updates of every chapter; most graphics are in color now. 

This site aims to provide additional material for the book, references and presentations

RMarkDown files to repeat simple and advanced statistical analyses

Extra material per
book chapter


Chapter Topics

2 Power of subgroup analyses in RCTs

4 Competing risks

5 Exercise overfitting: between center differences

6 Exercise non-linearity and interactions: Medicare surgical mortality

7 Missing values: simulation study

12 Interactions: smart coding and penalized estimation

15 Predictive performance: scaled Brier score and other measures
for binary and survival outcomes

16 Clinical usefulness: Net Benefit

17 Internal validation: sample size and interpretation

18 Presentation formats: score chart development and alternatives

19 Generalizability: change of setting scenarios

20 Updating with recalibration and shrinkage

21 Updating with renal transplant illustration

22 Case study: prediction of mortality in GUSTO-I

23 Case study: prediction of survival in SMART

Exercises with R code and data sets

Checklist: 7 steps for model development