Estimation of Individualized Treatment Effect Using Bayesian Additive Regression Trees with Applications

doi:10.3969/j.issn.1673－5765.2023.07.005

Abstract

Abstract: Individualized treatment effects mainly refer to the differences in outcomes between the treated and non-treated status for a same patient, regardless of whether the patient actually received treatment or did not receive treatment. By evaluating the effects of individualized treatment based on the characteristics of patients, it is possible to assign the treatment plan with the greatest individual benefit to each patient. This paper introduces the estimation and statistical inference of individualized treatment effects based on Bayesian additive regression tree, and introduces the evaluation of the significance of individual treatment effects, the identification of subgroups with heterogeneous treatment effects, and demonstrate the application through a simple case of Alzheimer's disease.

Key words: Individualized treatment effect; Bayesian additive regression tree; Precision medicine; Prediction model

摘要： 个体化治疗效应指对于同一例患者，无论其实际接受治疗与否，在接受治疗和未接受治疗状态下所呈现出的效果差异。整合患者特征开展对个体化治疗效应的评估，可以辅助实现对每一例患者精准化施加获益最大的治疗方案。本文基于贝叶斯加性回归树的个体化治疗效应估计和统计推断，介绍其进行因果效应显著性评价、因果效应异质性亚组识别等要点，并结合阿尔茨海默病案例进行应用展示。

关键词: 个体化治疗效应; 贝叶斯加性回归树; 精准医疗; 预测模型

LIANG Baosheng, ZHOU Jiangjie, WANG Shengfeng. Estimation of Individualized Treatment Effect Using Bayesian Additive Regression Trees with Applications[J]. Chinese Journal of Stroke, 2023, 18(07): 770-779.

梁宝生, 周江杰, 王胜锋. 基于贝叶斯加性回归树的个体化治疗效应估计方法及应用[J]. 中国卒中杂志, 2023, 18(07): 770-779.

References

[1] DORRESTEIJN J，VISSEREN F，RIDKER P，et al. Estimating treatment effects for individual patients based on the results of randomized clinical trials[J/OL]. BMJ，2011，343：d5888[2023-03-11]. https://doi.org/10.1136/bmj.d5888.
[2] HENDERSON N C，LOUIS T A，ROSNER G L，et al. Individualized treatment effects with censored data via fully nonparametric Bayesian accelerated failure time models[J]. Biostatistics，2020，21（1）：50-68.
[3] RUBIN D B. Estimating causal effects of treatments in randomized and nonrandomized studies[J]. Epidemiologic Methods，2014，66（1）：688-701.
[4] NGUYEN T L，COLLINS G S，LANDAIS P，et al. Counterfactual clinical prediction models could help to infer individualized treatment effects in randomized controlled trials：an illustration with the international stroke trial[J/OL]. J Clin Epidemiol，2020，125：47-56[2023-03-11]. https://doi.org/10.1016/j.jclinepi.2020.05.022.
[5] MURPHY S A. Optimal dynamic treatment regimes[J/OL]. J R Stat Soc Series B Stat Methodol，2003，65：331-355[2023-02-20]. https://doi.org/10.1111/1467-9868.00389.
[6] QIAN M，MURPHY S A. Performance guarantees for individualized treatment rules[J]. Ann Stat，2011，39（2）：1180-1210.
[7] QI Z L，LIU Y F. D-learning to estimate optimal individual treatment rules[J]. Electron J Stat，2018，12（2）：3601-3638.
[8] MENG H M，QIAO X Y. A robust method for estimating individualized treatment effect[J/OL]. Methodology，2020：1-31[2023-03-11]. https://doi.org/10.48550/arXiv.2004.10108.
[9] BEEMER J，SPOON K，HE L J，et al. Ensemble learning for estimating individualized treatment effects in student success studies[J/OL]. Int J Artif Intell Educ，2018，28：315-335[2023-03-11]. https://doi.org/10.1007/s40593-017-0148-x.
[10] ATHEY S，IMBENS G. Recursive partitioning for heterogeneous causal effects[J]. Proc Natl Acad Sci，2016，113（27）：7353-7360.
[11] WAGER S，ATHEY S. Estimation and inference of heterogeneous treatment effects using random forests[J]. J Am Stat Assoc，2018，113（523）：1228-1242.
[12] ISHWARAN H，KOGALUR U B，BLACKSTONE E H，et al. Random survival forests[J]. J Thorac Oncol，2008，2（12）：841-860.
[13] YOON J，JORDON J，SCHAAR M V D. GANITE：estimation of individualized treatment effects using generative adversarial nets[C/OL]//Vancouver: 6th International Conference on Learning Representations，2018[2023-05-11]. https://openreview.net/pdf?id=ByKWUeWA-.
[14] GOODFELLOW I J，POUGET-ABADIE J，MIRZA M，et al. Generative adversarial nets[M]//Advances in Neural information processing systems. Cambridge：MIT Press，2014：2672-2680.
[15] CHIPMAN H A，GEORGE E I，MCCULLOCH R E. BART：Bayesian additive regression trees[J]. Ann Appl Stat，2017，4（1）：266-298.
[16] HILL J，LINERO A，MURRAY J. Bayesian additive regression trees：a review and look forward[J]. Annu Rev Stat Appl，2020，7（1）：251-278.
[17] HILL J L. Bayesian nonparametric modeling for causal inference[J]. J Comput Graph Stat，2011，20（1）：217-240.
[18] CHIPMAN H，MCCULLOCH R. BayesTree：Bayesian additive regression trees[EB/OL].（2016-01-01） [2023-05-20]. https://CRAN.R-project.org/package=BayesTree.
[19] SPARAPANI R A，LOGAN B R，MCCULLOCH R E，et al. Nonparametric survival analysis using Bayesian additive regression trees（BART）[J]. Stat Med，2016，35（16）：2741-2753.
[20] KAPELNER A，BLEICH J. BartMachine：machine learning with Bayesian additive regression trees[J]. J Stat Softw，2016，70（4）：1-40.
[21] TANG M X，STERN Y，MARDER K，et al. The APOE-epsilon4 allele and the risk of Alzheimer disease among African Americans，whites，and Hispanics[J]. JAMA，1998，279（10）：751-755.
[22] ISHWARAN H，KOGALUR U B. Random forest SRC：random forests for survival，regression and classification（RF-SRC）[EB/OL].（2014-01-01） [2023-05-20]. https://cran.r-project.org/.
[23] THERNEAU T M. A package for survival analysis in R[EB/OL].（2023-03-11）[2023-05-20]. https://rdrr.io/cran/survival/f/inst/doc/survival.pdf.
[24] ZHANG W J，LE T D，LIU L，et al. Mining heterogeneous causal effects for personalized cancer treatment[J]. Bioinformatics，2017，33（15）：2372-2378.
[25] BASAK P，LINERO A，SINHA D，et al. Semiparametric analysis of clustered interval‐censored survival data using soft Bayesian additive regression trees（SBART）[J]. Biometrics，2022，78（3）：880-893.
[26] KAPELNER A，BLEICH J. Prediction with missing data via Bayesian additive regression trees[J/OL]. Can J Stat，2013，43：224-239[2023-03-11]. https://doi.org/10.48550/arXiv.1306.0618.
[27] BICA I，ALAA A M，LAMBERT C，et al. From real‐world patient data to individualized treatment effects using machine learning：current and future methods to address underlying challenges[J]. Clin Pharmacol Ther，2021，109（1）：87-100.
[28] SPANBAUER C，SPARAPANI R. Nonparametric machine learning for precision medicine with longitudinal clinical trials and Bayesian additive regression trees with mixed models[J]. Stat Med，2021，40（11）：2665-2691.
[29] FRENCH S. Explanatory model analysis：explore，explain，and examine predictive models[M]. Boca Raton：CRC Press，2022：1464.
[30] MOLNAR C. Interpretable machine learning：a guide for making black box models explainable[EB/OL]. [2023-03-02]. https://christophm.github.io/interpretable-ml-book/.
[31] RAJKOMAR A，DEAN J，KOHANE I. Machine learning in medicine[J]. N Engl J Med，2019，380（14）：1347-1358.