PERFORMANCE GUARANTEES FOR INDIVIDUALIZED TREATMENT RULES

By Min Qian, Susan A. Murphy. 2011

615 citations recorded by the source corpus.

Because many illnesses show heterogeneous response to treatment, there is increasing interest in individualizing treatment to patients [11]. An individualized treatment rule is a decision rule that recommends treatment according to patient characteristics. We consider the use of clinical trial data in the construction of an individualized treatment rule leading to highest mean response. This is a difficult computational problem because the objective function is the expectation of a weighted indicator function that is non-concave in the parameters. Furthermore there are frequently many pretreatment variables that may or may not be useful in constructing an optimal individualized treatment rule yet cost and interpretability considerations imply that only a few variables should be used by the individualized treatment rule. To address these challenges we consider estimation based on l1 penalized least squares. This approach is justified via a finite sample upper bound on the difference between the mean response due to the estimated individualized treatment rule and the mean response due to the optimal individualized treatment rule.

Subfields: Statistics and Probability, Causal Inference & Discovery, Medical & Health AI

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

In Causal Inference & Discovery

In Medical & Health AI

Citations in the corpus

Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data (2017, methodology): “Reference [22] adopted Q-learning [15], [23] to study a two-stage decision making problem in GVHD prevention and treatment, it simplified the problem of our motivating example to consider only two treatment options and two stages.”
Balanced Policy Evaluation and Learning (2017, neutral): “A policy is learned either by ⇡̂direct(x) = argmint2[m] μ̂t(x) or by minimizing ⌧̂ direct(⇡) over ⇡ 2 ⇧ [33].”
Estimating treatment effect heterogeneity in randomized program evaluation (2013, methodology): “When the derivation of an optimal treatment regime rather than the description of treatment effect heterogeneity is of interest, we can replace the GCV statistic with the average effect size of the optimal treatment rule [Imai and Strauss (2011), Qian and Murphy (2011)].”
Using Decision Lists to Construct Interpretable and Parsimonious Treatment Regimes (2015, background): “Thus, a goal of many regression-based estimators is to ensure correct model specification under a large class of generative models (Zhao et al., 2009; Qian and Murphy, 2011; Moodie et al., 2013; Laber et al., 2014; Taylor et al., 2014).”
Efficient Policy Learning (2017, methodology): “, 2011; Swaminathan and Joachims, 2015), outcome-weighted learning (Chen et al., 2016; Qian and Murphy, 2011; Zhang et al., 2012; Zhao et al., 2012; Zhou et al., 2015), and empirical welfare maximization (Kitagawa and Tetenov, 2015).”
Augmented outcome-weighted learning for estimating optimal dynamic treatment regimens (2018, methodology): “…if we let P denote the probability measure generated by ( X k , A k , R k ) for k = 1 , … , K , then according to the work of Qian and Murphy, 12 it can be shown that P  is dominated by P and Consequently, the goal is to find the optimal treatment rule  ∗ = (  ∗ 1 , … ,  ∗ K ) that…”
Using Reinforcement Learning to Personalize Dosing Strategies in a Simulated Cancer Trial with High Dimensional Data (2017, methodology): “A third class of methods maximize mean responses across patients and subgroups either by using penalized regression (Qian and Murphy 2011), or by treating the identification of subgroups as a classification problem with each participant weighted by their outcome (Yingqi Zhao et al. 2012; Zhang et…”
Optimal trees for prediction and prescription (2018, background): “This approach has been followed historically by several authors in clinical research (for example [45]), and more recently by researchers in statistics [102] and operations research [19].”
A Robust Method for Estimating Optimal Treatment Regimes (2012, methodology): “In Web Appendix E, we present the results of simulations comparing our methods to the marginal structural mean model approach of Robins et al. (2008) and Orellana et al.”
Adaptive Interventions Treatment Modelling and Regimen Optimization Using Sequential Multiple Assignment Randomized Trials (Smart) and Q-Learning (2018, methodology): “Another method of describing optimal treatment is derived from different econometrics and bio-medical literatures is known as welfare contrast [22-27].”
Policy Learning With Observational Data (2017, methodology): “…learning from observational data has also been considered in parallel literatures developed in both statistics (Luedtke and van der Laan, 2016; Qian and Murphy, 2011; Zhang, Tsiatis, Davidian, Zhang, and Laber, 2012; Zhao, Zeng, Rush, and Kosorok, 2012) and machine learning (Beygelzimer and…”
Learning Cost-Effective and Interpretable Treatment Regimes (2017, positive): “Regression-based methods [27, 31, 28, 32, 43, 27, 25] model the conditional distribution of the outcomes given the treatment and characteristics of patients and choose the treatment resulting in the best possible outcome for each individual.”
Estimating Individualized Treatment Rules Using Outcome Weighted Learning (2012, methodology): “We also note that l 1 -PLS has comparatively larger MSE, resulting from high variance of the method, which may be explained by the conﬂicting goals of maximizing the value function and minimizing the prediction error (Qian and Murphy 2011).”
Statistical Inference in Dynamic Treatment Regimes (2010, neutral): “Penalized estimation and Q-learning in one-stage decision problems are discussed in (Qian and Murphy 2009) and in multi-stage problems in (Song et al., 2010).”
Learning to Personalize from Observational Data (2016, background): “For example, in medicine, there is a vast literature on predicting patient-specific responses to treatment [6, 7] and picking the best by comparing predictions [8, 9], although recent work [10] has looked at alternative approaches for the case of randomized, experimental data.”
A Theoretical Analysis of Deep Q-Learning (2019, background): “ﬁnd sequential decision rules for individual patients that adapt to time-evolving illnesses. There is a huge body of literature on this line of research. See, e.g., Murphy (2003); Zhao et al. (2009); Qian and Murphy (2011); Zhao et al. (2011); Zhang et al. (2012); Zhao et al. (2012); Goldberg and Ko”
Policy Evaluation with Latent Confounders via Optimal Balance (2019, methodology): “One standard approach to this is the direct method [38], which given an estimate ⇢̂t of ⇢t predicts the”