Risk controlled decision trees and random forests for precision Medicine.

Publisher: Wiley Country of Publication: England NLM ID: 8215016 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1097-0258 (Electronic) Linking ISSN: 02776715 NLM ISO Abbreviation: Stat Med Subsets: MEDLINE

Original Publication: Chichester ; New York : Wiley, c1982-

Diabetes Mellitus, Type 2*/drug therapy ; Precision Medicine*/methods; Algorithms ; Computer Simulation ; Decision Trees ; Humans

Statistical methods generating individualized treatment rules (ITRs) often focus on maximizing expected benefit, but these rules may expose patients to excess risk. For instance, aggressive treatment of type 2 diabetes (T2D) with insulin therapies may result in an ITR which controls blood glucose levels but increases rates of hypoglycemia, diminishing the appeal of the ITR. This work proposes two methods to identify risk-controlled ITRs (rcITR), a class of ITR which maximizes a benefit while controlling risk at a prespecified threshold. A novel penalized recursive partitioning algorithm is developed which optimizes an unconstrained, penalized value function. The final rule is a risk-controlled decision tree (rcDT) that is easily interpretable. A natural extension of the rcDT model, risk controlled random forests (rcRF), is also proposed. Simulation studies demonstrate the robustness of rcRF modeling. Three variable importance measures are proposed to further guide clinical decision-making. Both rcDT and rcRF procedures can be applied to data from randomized controlled trials or observational studies. An extensive simulation study interrogates the performance of the proposed methods. A data analysis of the DURABLE diabetes trial in which two therapeutics were compared is additionally presented. An R package implements the proposed methods ( https://github.com/kdoub5ha/rcITR).
(© 2021 John Wiley & Sons Ltd.)

Stat Med. 2011 Sep 20;30(21):2601-21. (PMID: 21786278)
J Comp Eff Res. 2014 Jan;3(1):63-78. (PMID: 24266593)
Diabetes Metab. 2000 Nov;26(5):337-51. (PMID: 11119013)
J Am Stat Assoc. 2017;112(517):169-187. (PMID: 28943682)
J Diabetes Sci Technol. 2008 Sep;2(5):831-8. (PMID: 19885269)
Ann Stat. 2011 Apr 1;39(2):1180-1210. (PMID: 21666835)
Int J Methods Psychiatr Res. 2013 Sep;22(3):185-94. (PMID: 23956114)
Sisli Etfal Hastan Tip Bul. 2019 Aug 28;53(3):215-220. (PMID: 32377086)
Biometrics. 2014 Mar;70(1):53-61. (PMID: 24400912)
J Mach Learn Res. 2012 Nov;13(Nov):3253-3295. (PMID: 23741197)
Biostatistics. 2011 Apr;12(2):270-82. (PMID: 20876663)
J Am Stat Assoc. 2013 Jan 1;108(502):527-539. (PMID: 24058223)
Stat Med. 2011 Oct 30;30(24):2867-80. (PMID: 21815180)
Diabetes Care. 2009 Jun;32(6):1007-13. (PMID: 19336625)
Biometrika. 2015 Mar 1;102(1):151-168. (PMID: 25937641)
J Am Stat Assoc. 2018;113(521):1-13. (PMID: 30034060)
J Am Stat Assoc. 2012 Sep 1;107(449):1106-1118. (PMID: 23630406)
J Comput Graph Stat. 2018;27(4):849-860. (PMID: 32523325)

K01 DK106116 United States DK NIDDK NIH HHS; R01 HG006139 United States HG NHGRI NIH HHS; R21 HL150374 United States HL NHLBI NIH HHS; R35 GM141798 United States GM NIGMS NIH HHS

Keywords: decision trees; precision medicine; random forests; risk control; variable importance

Date Created: 20211117 Date Completed: 20220331 Latest Revision: 20230501

20230501

PMC8863134

10.1002/sim.9253

34786731