Time Series Deconfounder: Estimating Treatment Effects over Time in the Presence of Hidden Confounders

Abstract

The estimation of treatment effects is a pervasive problem in medicine. Existing methods for estimating treatment effects from longitudinal observational data assume that there are no hidden confounders, an assumption that is not testable in practice and, if it does not hold, leads to biased estimates. In this paper, we develop the Time Series Deconfounder, a method that leverages the assignment of multiple treatments over time to enable the estimation of treatment effects in the presence of multi-cause hidden confounders. The Time Series Deconfounder uses a novel recurrent neural network architecture with multitask output to build a factor model over time and infer latent variables that render the assigned treatments conditionally independent; then, it performs causal inference using these latent variables that act as substitutes for the multi-cause unobserved confounders. We provide a theoretical analysis for obtaining unbiased causal effects of time-varying exposures using the Time Series Deconfounder. Using both simulated and real data we show the effectiveness of our method in deconfounding the estimation of treatment responses over time.

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Causal Effect Inference with Deep Latent-Variable Models
• Reliable Decision Support using Counterfactual Models
• Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representations
• A Bayesian Nonparametric Approach for Estimating Individualized Treatment-Response Curves
• Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventions
• Multiple Causal Inference with Latent Confounding
• Estimation of Individual Treatment Effect in Latent Confounder Models via Adversarial Learning
• Multiple Causes: A Causal Graphical View
• Identifiability of causal effects with multiple causes and a binary outcome
• Accounting for hidden common causes when inferring cause and effect from observational data

Cited by in corpus (5)

• A Survey on Causal Inference
• Causal Inference for Time series Analysis: Problems, Methods and Evaluation
• The Medical Deconfounder: Assessing Treatment Effects with Electronic Health Records
• Estimating Individual Treatment Effects with Time-Varying Confounders
• Weighted Tensor Completion for Time-Series Causal Inference