Interval Estimation of Individual-Level Causal Effects Under Unobserved Confounding

We study the problem of learning conditional average treatment effects (CATE) from observational data with unobserved confounders. The CATE function maps baseline covariates to individual causal effect predictions and is key for personalized assessments. Recent work has focused on how to learn CATE under unconfoundedness, i.e., when there are no unobserved confounders. Since CATE may not be identified when unconfoundedness is violated, we develop a functional interval estimator that predicts bounds on the individual causal effects under realistic violations of unconfoundedness. Our estimator takes the form of a weighted kernel estimator with weights that vary adversarially. We prove that our estimator is sharp in that it converges exactly to the tightest bounds possible on CATE when there may be unobserved confounders. Further, we study personalized decision rules derived from our estimator and prove that they achieve optimal minimax regret asymptotically. We assess our approach in a simulation study as well as demonstrate its application in the case of hormone replacement therapy by comparing conclusions from a real observational study and clinical trial.

[1]  Bent Ottesen,et al.  Issues to debate on the Women's Health Initiative (WHI) study. Epidemiology or randomized clinical trials--time out for hormone replacement therapy studies? , 2003, Human reproduction.

[2]  John Duchi,et al.  Bounds on the conditional and average treatment effect in the presence of unobserved confounders , 2018 .

[3]  Nathan Kallus,et al.  Balanced Policy Evaluation and Learning , 2017, NeurIPS.

[4]  Shah Ebrahim,et al.  Commentary: the hormone replacement-coronary heart disease conundrum: is this the death of observational epidemiology? , 2004, International journal of epidemiology.

[5]  Xinkun Nie,et al.  Learning Objectives for Treatment Effect Estimation , 2017 .

[6]  Sören R. Künzel,et al.  Metalearners for estimating heterogeneous treatment effects using machine learning , 2017, Proceedings of the National Academy of Sciences.

[7]  J. Sekhon,et al.  From SATE to PATT : Combining Experimental with Observational Studies to Estimate Population Treatment Effects ∗ , 2013 .

[8]  Masataka Harada,et al.  A flexible, interpretable framework for assessing sensitivity to unmeasured confounding , 2016, Statistics in medicine.

[9]  Fredrik D. Johansson,et al.  Learning Weighted Representations for Generalization Across Designs , 2018, 1802.08598.

[10]  Elizabeth A Stuart,et al.  Improving propensity score weighting using machine learning , 2010, Statistics in medicine.

[11]  Robert P. Lieli,et al.  Estimating Conditional Average Treatment Effects , 2014 .

[12]  D. Green,et al.  Modeling heterogeneous treatment effects in large-scale experiments using Bayesian Additive Regression Trees , 2010 .

[13]  Stefan Wager,et al.  Efficient Policy Learning , 2017, ArXiv.

[14]  E. C. Hammond,et al.  Smoking and lung cancer: recent evidence and a discussion of some questions. , 1959, Journal of the National Cancer Institute.

[15]  Thorsten Joachims,et al.  Counterfactual Risk Minimization , 2015, ICML.

[16]  J. Williamson,et al.  Latest evidence on using hormone replacement therapy in the menopause , 2015 .

[17]  Sören R. Künzel,et al.  Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning , 2017 .

[18]  Stefan Wager,et al.  Estimation and Inference of Heterogeneous Treatment Effects using Random Forests , 2015, Journal of the American Statistical Association.

[19]  S. Murphy,et al.  PERFORMANCE GUARANTEES FOR INDIVIDUALIZED TREATMENT RULES. , 2011, Annals of statistics.

[20]  Dylan S. Small,et al.  Sensitivity analysis for inverse probability weighting estimators via the percentile bootstrap , 2017, Journal of the Royal Statistical Society: Series B (Statistical Methodology).

[21]  Alexandre Poirier,et al.  Identification of Treatment Effects under Conditional Partial Independence , 2017, 1707.09563.

[22]  Susan Athey,et al.  Recursive partitioning for heterogeneous causal effects , 2015, Proceedings of the National Academy of Sciences.

[23]  John Langford,et al.  Doubly Robust Policy Evaluation and Optimization , 2014, ArXiv.

[24]  Stefan Wager,et al.  Policy Learning With Observational Data , 2017, Econometrica.

[25]  Nathan Kallus,et al.  Recursive Partitioning for Personalization using Observational Data , 2016, ICML.

[26]  Qi Li.  Nonparametric econometrics , 2006 .

[27]  Garnet L Anderson,et al.  Statistical Issues Arising in the Women's Health Initiative , 2005, Biometrics.

[28]  Nathan Kallus,et al.  Policy Evaluation and Optimization with Continuous Treatments , 2018, AISTATS.

[29]  D. McCaffrey,et al.  Propensity score estimation with boosted regression for evaluating causal effects in observational studies. , 2004, Psychological methods.

[30]  Uri Shalit,et al.  Estimating individual treatment effect: generalization bounds and algorithms , 2016, ICML.

[31]  Souraya Kheireddine.  On Boundary Correction in Kernel Estimation , 2016 .

[32]  Ben Alamar.  Social Choice With Partial Knowledge of Treatment Response , 2007 .

[33]  Zhiqiang Tan,et al.  A Distributional Approach for Causal Inference Using Propensity Scores , 2006 .

[34]  Dylan S. Small,et al.  Calibrating Sensitivity Analyses to Observed Covariates in Observational Studies , 2013, Biometrics.

[35]  D. V. Lindley,et al.  Randomization Analysis of Experimental Data: The Fisher Randomization Test Comment , 1980 .

[36]  Michael R Kosorok,et al.  Residual Weighted Learning for Estimating Individualized Treatment Rules , 2015, Journal of the American Statistical Association.

[37]  Nathan Kallus,et al.  Confounding-Robust Policy Improvement , 2018, NeurIPS.

[38]  T. Shakespeare,et al.  Observational Studies , 2003 .

[39]  JoAnn E Manson,et al.  Lessons learned from the Women's Health Initiative trials of menopausal hormone therapy. , 2013, Obstetrics and gynecology.

[40]  Donald K. K. Lee,et al.  Interval estimation of population means under unknown but bounded probabilities of sample selection , 2013 .

[41]  C. Manski.  Partial Identification of Probability Distributions , 2003 .