Decision makers often want to target interventions so as to maximize an outcome that is observed only in the long-term. This typically requires delaying decisions until the outcome is observed or relying on simple short-term proxies for the long-term outcome. Here we build on the statistical surrogacy and policy learning literatures to impute the missing long-term outcomes and then approximate the optimal targeting policy on the imputed outcomes via a doubly-robust approach. We first show that conditions for the validity of average treatment effect estimation with imputed outcomes are also sufficient for valid policy evaluation and optimization; furthermore, these conditions can be somewhat relaxed for policy optimization. We apply our approach in two large-scale proactive churn management experiments at The Boston Globe by targeting optimal discounts to its digital subscribers with the aim of maximizing long-term revenue. Using the first experiment, we evaluate this approach empirically by comparing the policy learned using imputed outcomes with a policy learned on the ground-truth, long-term outcomes. The performance of these two policies is statistically indistinguishable, and we rule out large losses from relying on surrogates. Our approach also outperforms a policy learned on short-term proxies for the long-term outcome. In a second field experiment, we implement the optimal targeting policy with additional randomized exploration, which allows us to update the optimal policy for future subscribers. Over three years, our approach had a net-positive revenue impact in the range of $4-5 million compared to the status quo.
Living forwards or understanding backwards? A comparison of Inverse Probability of Treatment Weighting and G-estimation methods for targeting hypothetical full adherence estimands in longitudinal cohort studies
University of Exeter(埃克塞特大学);Royal Devon University Healthcare NHS Foundation Trust(皇家德文郡大学医疗保健NHS基金会信托);Novo Nordisk Research Centre (NNRCO)(诺和诺德研究中心)
Difficult Lessons on Social Prediction from Wisconsin Public Schools
University of California, Berkeley(加州大学伯克利分校);Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所);Tübingen AI Center(图宾根人工智能中心);Harvard Society of Fellows(哈佛大学研究员协会)
Causal Diffusion Models for Counterfactual Outcome Distributions in Longitudinal Data
Department of Economics, Northeastern University(东北大学经济学系);Center for Health Policy and Healthcare Research, Northeastern University(东北大学健康政策与医疗研究中心);Department of Public Health and Health Sciences, Bouvé College of Health Sciences, Northeastern University(东北大学布维健康科学学院公共卫生与健康科学系)
Double/debiased machine learning of quantile treatment effects on long-term outcomes in clinical trials
Faculty of Arts and Sciences, Beijing Normal University(北京师范大学文理学院);School of Mathematics and Statistics, Beijing Technology and Business University(北京工商大学数学与统计学院);School of Statistics, Beijing Normal University(北京师范大学统计学院)
Policy learning under constraint: Maximizing a primary outcome while controlling an adverse event
Inria PreMeDICaL, Inserm, University of Montpellier(法国国家信息与自动化研究所PreMeDICaL团队,法国国家健康与医学研究院,蒙彼利埃大学);Elixir Health, Paris(Elixir Health公司);Université Paris Cité, CNRS, MAP5(巴黎西岱大学,法国国家科学研究中心MAP5实验室)
Evaluating treatment effects on longitudinal outcomes with attrition due to death: Methods for a two-dimentional estimand with a case study in Quality of Life
Ghent University(根特大学);Leiden University Medical Center(莱顿大学医学中心);Pfizer Inc(辉瑞公司);Oslo University Hospital(奥斯陆大学医院);European Organisation for Research and Treatment of Cancer (EORTC) Headquarters(欧洲癌症研究与治疗组织(EORTC)总部);University of Freiburg(弗赖堡大学)
Estimation of Time-Varying Treatment Effects in a Joint Model for Longitudinal and Recurrent Event Outcomes in Mobile Health Data
University of Michigan(密歇根大学);Institute for Social Research, University of Michigan(密歇根大学社会研究所);Department of Population Health Sciences and Huntsman Cancer Insti(人口健康科学与亨斯曼癌症研究所)