Some methods for heterogeneous treatment effect estimation in high dimensions.

Pubmed ID: 29508417

Pubmed Central ID: PMC5938172

Journal: Statistics in medicine

Publication Date: May 20, 2018

MeSH Terms: Humans, Algorithms, Randomized Controlled Trials as Topic, Propensity Score, Regression Analysis, Treatment Outcome, Computer Simulation, Causality, Decision Making, Biostatistics, Observational Studies as Topic, Precision Medicine, Electronic Health Records, Machine Learning, Patient-Specific Modeling

Grants: R01 EB001988, T15 LM007033, U54 EB020405, R25 CA180993

Authors: Qian J, Powers S, Jung K, Schuler A, Shah NH, Hastie T, Tibshirani R

Cite As: Powers S, Qian J, Jung K, Schuler A, Shah NH, Hastie T, Tibshirani R. Some methods for heterogeneous treatment effect estimation in high dimensions. Stat Med 2018 May 20;37(11):1767-1787. Epub 2018 Mar 6.

Studies:

Abstract

When devising a course of treatment for a patient, doctors often have little quantitative evidence on which to base their decisions, beyond their medical education and published clinical trials. Stanford Health Care alone has millions of electronic medical records that are only just recently being leveraged to inform better treatment recommendations. These data present a unique challenge because they are high dimensional and observational. Our goal is to make personalized treatment recommendations based on the outcomes for past patients similar to a new patient. We propose and analyze 3 methods for estimating heterogeneous treatment effects using observational data. Our methods perform well in simulations using a wide variety of treatment effect functions, and we present results of applying the 2 most promising methods to data from The SPRINT Data Analysis Challenge, from a large randomized trial of a treatment for high blood pressure.