Sage Journals: Discover world-class research

Abstract

Many time-to-event studies are complicated by the presence of competing risks. Such data are often analyzed using Cox models for the cause-specific hazard function or Fine and Gray models for the subdistribution hazard. In practice, regression relationships in competing risks data are often complex and may include nonlinear functions of covariates, interactions, high-dimensional parameter spaces and nonproportional cause-specific, or subdistribution, hazards. Model misspecification can lead to poor predictive performance. To address these issues, we propose a novel approach: flexible prediction modeling of competing risks data using Bayesian Additive Regression Trees (BART). We study the simulation performance in two-sample scenarios as well as a complex regression setting, and benchmark its performance against standard regression techniques as well as random survival forests. We illustrate the use of the proposed method on a recently published study of patients undergoing hematopoietic stem cell transplantation.

Keywords

Cumulative incidence graft-versus-host disease (GVHD)hematopoietic stem cell transplant machine learning nonproportional treatment heterogeneity variable selection

Get full access to this article

View all access options for this article.

References

Prentice

Kalbfleisch

Peterson A

, et al. The analysis of failure times in the presence of competing risks. Biometrics 1978; 34: 541–554.

Andersen PK, Borgan O, Gill RD, et al. Statistical models based on counting processes. New York, NY: Springer-Verlag, 1993, pp.512–515.

Fine

Gray

. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509.

Klein

Andersen

. Regression modeling of competing risks data based on pseudovalues of the cumulative incidence function. Biometrics 2005; 61: 223–229.

Scheike

Zhang

Gerds

. Predicting cumulative incidence probability by direct binomial regression. Biometrika 2008; 95: 205–220.

Tibshirani

. The lasso method for variable selection in the Cox model. Stat Med 1997; 16: 385–395.

Park

Hastie

. L1-regularization path algorithm for generalized linear models. J Royal Stat Soc: Ser B (Stat Methodol) 2007; 69: 659–677.

Zhang

. Adaptive lasso for Cox’s proportional hazards model. Biometrika 2007; 94: 691–703.

Ridgeway

. The state of boosting. Comput Sci Stat 1999; 31: 172–181.

10.

Luan

. Boosting proportional hazards models using smoothing splines, with applications to high-dimensional microarray data. Bioinformatics 2006; 21: 2403–2409.

11.

Huang

. Clustering threshold gradient descent regularization: with applications to microarray studies. Bioinformatics 2006; 23: 466–472.

12.

Ishwaran

Kogalur

Blackstone

, et al. Random survival forests. Ann Appl Stat 2008; 2: 841–860.

13.

Sparapani

Logan

McCulloch

, et al. Nonparametric survival analysis using Bayesian Additive Regression Trees (BART). Stat Med 2016; 35: 2741–2753.

14.

Van Belle

Pelckmans

Van Huffel

, et al. Improved performance on high-dimensional survival data by application of Survival-SVM. Bioinformatics 2010; 27: 87–94.

15.

Parikh

Zhou

. Penalized variable selection in competing risks regression. Lifetime Data Analys 2017; 23: 353–376.

16.

Ahn

Banerjee

Sahr

, et al. Group and within-group variable selection for competing risks data. Lifetime Data Analys 2018; 24: 407–424.

17.

Ishwaran

Gerds

Kogalur

, et al. Random survival forests for competing risks. Biostatistics 2014; 15: 757–773.

18.

Chipman

George

McCulloch

. BART: Bayesian additive regression trees. Ann Appl Stat 2010; 4: 266–298.

19.

Linero

. Bayesian regression trees for high dimensional prediction and variable selection. J Am Stat Assoc 2018; 113: 626–636.

20.

Gelfand

Smith

. Sampling-based approaches to calculating marginal densities. J Am Stat Assoc 1990; 85: 398–409.

21.

McCulloch R, Sparapani R, Gramacy R, et al. BART: Bayesian additive regression trees, https://cran.r-project.org/package=BART (accessed 24 December 2018).

22.

Bleich

Kapelner

George

, et al. Variable selection for BART: an application to gene regulation. Ann Appl Stat 2014; 8: 1750–1781.

23.

Kapelner A and Bleich J. bartMachine: Bayesian additive regression trees, https://cran.r-project.org/package=bartMachine (2014).

24.

Albert

Chib

. Bayesian analysis of binary and polychotomous response data. J Am Stat Assoc 1993; 88: 669–679.

25.

Fahrmeir

Discrete survival-time models. In: Armitage

Colton

(eds). Encyclopedia of biostatistics, Chichester: Wiley, 1998, pp. 1163–1168.

26.

Friedman

. Greedy function approximation: a gradient boosting machine. Ann Stat 2001; 29: 1189–1232.

27.

Murray

. Log-linear Bayesian additive regression trees for categorical and count responses. arXiv preprint 2017. arXiv: 1701.01503.

28.

Peng

Fine

. Competing risks quantile regression. J Am Stat Assoc 2009; 104: 1440–1453.

29.

Singer

Willett

. It’s about time: Using discrete-time survival analysis to study duration and the timing of events. J Educ Stat 1993; 18: 155–195.

30.

Allison PD. Survival analysis using the SAS system: a practical guide. Cary, NC: SAS Institute Inc., 1995.

31.

Logan

Zhang

. The use of group sequential designs with common competing risks tests. Stat Med 2013; 32: 899–913.

32.

Cox

. Regression models and life-tables (with discussions). Jr Stat Soc B 1972; 34: 187–220.

33.

Aalen

Johansen

. An empirical transition matrix for non-homogeneous Markov chains based on censored observations. Scand J Stat 1978; 5: 141–150.

34.

Ishwaran H and Kogalur UB. Random forests for survival, regression and classification (RF-SRC), https://CRAN.R-project.org/package=randomForestSRC (2018).

35.

Friedman

. Multivariate adaptive regression splines. Ann Stat 1991; 19: 1–67.

36.

Lawrence

Lin

. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989; 45: 255–268.

37.

Eapen

Logan

Horowitz

, et al. Bone marrow or peripheral blood for reduced-intensity conditioning unrelated donor transplantation. J Clin Oncol 2015; 33: 364–369.

38.

Gerds T. pec: Prediction error curves for risk prediction models in survival analysis, https://cran.r-project.org/package=pec (2018).

39.

de Waal

Pannekoek

Scholtus

. Handbook of statistical data editing and imputation, Hoboken, NJ: John Wiley & Sons, 2011.

40.

Daniels

Winterstein

. Sequential BART for imputation of missing covariates. Biostatistics 2016; 17: 589–602.

41.

Daniels M and Singh A. sbart: Sequential BART for imputation of missing covariates, https://CRAN.R-project.org/package=sbart (2018).

42.

Amdahl G. Validity of the single processor approach to achieving large-scale computing capabilities. In: AFIPS conference proceedings Atlantic City, NJ, USA, 18–20 April 1967, pp.483–485.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB

Nonparametric competing risks analysis using Bayesian Additive Regression Trees

Abstract

Keywords

Get full access to this article

References

Supplementary Material