Time Series Deconfounder: Estimating Treatment Effects over Time in the Presence of Hidden Confounders
This addresses a pervasive issue in medicine for researchers and practitioners by enabling more reliable causal inference from observational data, though it is an incremental improvement over existing deconfounding approaches.
The paper tackles the problem of estimating treatment effects from longitudinal observational data in the presence of hidden confounders, which existing methods assume away, and develops the Time Series Deconfounder method that provides unbiased causal effects, as demonstrated with simulated and real data.
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.