Propensity Score Methods for Merging Observational and Experimental Datasets

Rosenman, Evan T. R.; Owen, Art B.; Baiocchi, Michael; Banack, Hailey R.

doi:10.48550/arxiv.1804.07863

Cited by 13 publications

(22 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When the RCT data is treated as the anchoring site, the target site propensity score model is known, so the target site estimator for the TATE is consistent, and the global adaptive estimator is likely to be more reliable. Our FACE framework can thus be viewed as a contribution to recent work on using observational studies to reduce the variance associated with treatment effect estimates from experimental studies (Rosenman et al 2018, Athey et al 2020. For greater generalizability, participants for whom there is only observational data can be taken to be the target population.…”

Section: Discussionmentioning

confidence: 99%

Federated Adaptive Causal Estimation (FACE) of Target Treatment Effects

Han¹,

Hou²,

Cho³

et al. 2021

Preprint

View full text Add to dashboard Cite

Federated learning of causal estimands may greatly improve estimation efficiency by aggregating estimates from multiple study sites, but robustness to extreme estimates is vital for maintaining consistency. We develop a federated adaptive causal estimation (FACE) framework to incorporate heterogeneous data from multiple sites to provide treatment effect estimation and inference for a target population of interest. Our strategy is communication-efficient and privacy-preserving and allows for flexibility in the specification of the target population. Our method accounts for site-level heterogeneity in the distribution of covariates through density ratio weighting. To safely aggregate estimates from all sites and avoid negative transfer, we introduce an adaptive procedure of weighing the estimators constructed using data from the target and source populations through a penalized regression on the influence functions, which achieves 1) consistency and 2) optimal efficiency. We illustrate FACE by conducting a comparative effectiveness study of BNT162b2 (Pfizer) and mRNA-1273 (Moderna) vaccines on COVID-19 outcomes in U.S. veterans using electronic health records from five VA sites.

show abstract

Section: Discussionmentioning

confidence: 99%

Federated Adaptive Causal Estimation (FACE) of Target Treatment Effects

Han¹,

Hou²,

Cho³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The idea of using statistical surrogates dates back to at least the 1980s [26,7,17,11]. More recent work provides different variations of the problem with different identifying assumptions [28,4,3,27,21]. I build on and extend the framework of [4], where treatment is only observed in the experimental sample and outcome is only observed in the observational sample.…”

Section: Related Workmentioning

confidence: 99%

Generalized Kernel Ridge Regression for Long Term Causal Inference: Treatment Effects, Dose Responses, and Counterfactual Distributions

Singh¹

2022

Preprint

View full text Add to dashboard Cite

I propose kernel ridge regression estimators for long term causal inference, where a short term experimental data set containing randomized treatment and short term surrogates is fused with a long term observational data set containing short term surrogates and long term outcomes. I propose estimators of treatment effects, dose responses, and counterfactual distributions with closed form solutions in terms of kernel matrix operations. I allow covariates, treatment, and surrogates to be discrete or continuous, and low, high, or infinite dimensional. For long term treatment effects, I prove √ n consistency, Gaussian approximation, and semiparametric efficiency. For long term dose responses, I prove uniform consistency with finite sample rates. For long term counterfactual distributions, I prove convergence in distribution.Preprint. Under review.

show abstract

“…We note that there is a growing interest in combining observational data with RCTs [Athey et al, 2016, Kaizar, 2011, Rosenman et al, 2018 to improve power. In our work, we focus on only using the RCT data, and leave it to future work on how to incorporate observational data in our framework.…”

Section: Related Workmentioning

confidence: 99%

Covariate Balancing Sensitivity Analysis for Extrapolating Randomized Trials across Locations

Nie¹,

Imbens²,

Wager³

2021

Preprint

View full text Add to dashboard Cite

The ability to generalize experimental results from randomized control trials (RCTs) across locations is crucial for informing policy decisions in targeted regions. Such generalization is often hindered by the lack of identifiability due to unmeasured effect modifiers that compromise direct transport of treatment effect estimates from one location to another. We build upon sensitivity analysis in observational studies and propose an optimization procedure that allows us to get bounds on the treatment effects in targeted regions. Furthermore, we construct more informative bounds by balancing on the moments of covariates. In simulation experiments, we show that the covariate balancing approach is promising in getting sharper identification intervals.

show abstract

Propensity Score Methods for Merging Observational and Experimental Datasets

Cited by 13 publications

References 14 publications

Federated Adaptive Causal Estimation (FACE) of Target Treatment Effects

Federated Adaptive Causal Estimation (FACE) of Target Treatment Effects

Generalized Kernel Ridge Regression for Long Term Causal Inference: Treatment Effects, Dose Responses, and Counterfactual Distributions

Covariate Balancing Sensitivity Analysis for Extrapolating Randomized Trials across Locations

Contact Info

Product

Resources

About