R/riskRegression: Risk Regression Models and Prediction Scores for Survival Analysis with Competing Risks

CRAN_Status_Badge R-CMD-check

Implementation of the following methods for event history analysis: Risk regression models for survival endpoints also in the presence of competing risks are fitted using binomial regression based on a time sequence of binary event status variables. A formula interface for the Fine-Gray regression model and an interface for the combination of cause-specific Cox regression models. A toolbox for assessing and comparing performance of risk predictions (risk markers and risk prediction models). Prediction performance is measured by the Brier score and the area under the ROC curve for binary possibly time-dependent outcome. Inverse probability of censoring weighting and pseudo values are used to deal with right censored data. Lists of risk markers and lists of risk models are assessed simultaneously. Cross-validation repeatedly splits the data, trains the risk prediction models on one part of each split and then summarizes and compares the performance across splits.

Installation

library(devtools)
install_github("tagteam/riskRegression")

References

The following references provide the methodological framework for the features of riskRegression.

  1. T.A. Gerds and M.W. Kattan (2021). Medical Risk Prediction Models: With Ties to Machine Learning (1st ed.) Chapman and Hall/CRC https://doi.org/10.1201/9781138384484

  2. T.A. Gerds and M. Schumacher. Consistent estimation of the expected Brier score in general survival models with right-censored event times. Biometrical Journal, 48(6):1029–1040, 2006.

  3. T.A. Gerds and M. Schumacher. Efron-type measures of prediction error for survival analysis. Biometrics, 63(4):1283–1287, 2007.

  4. T.A. Gerds, T. Cai, and M. Schumacher. The performance of risk prediction models. Biometrical Journal, 50(4):457–479, 2008.

  5. U B Mogensen, H. Ishwaran, and T A Gerds. Evaluating random forests for survival analysis using prediction error curves. Journal of Statistical Software, 50(11), 2012.

  6. P. Blanche, J-F Dartigues, and H. Jacqmin-Gadda. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statistics in Medicine, 32(30): 5381–5397, 2013.

  7. Paul Blanche, Ce'cile Proust-Lima, Lucie Loube`re, Claudine Berr, Jean- Franc,ois Dartigues, and He'le`ne Jacqmin-Gadda. Quantifying and comparing dynamic predictive accuracy of joint models for longitudinal marker and time-to-event in presence of censoring and competing risks. Biometrics, 71 (1):102–113, 2015.

Functions predict.CauseSpecificCox, predictCox and iidCox: