Server-Side Treatment Effect Computation for Federated Studies

Server-side 'DataSHIELD' package that computes local average treatment effect (ATE) statistics from arm-specific regression coefficients provided by the analyst. Implements the server-side component of the G-formula estimator from 'Federated Causal Inference: Multi-Study ATE Estimation beyond Meta-Analysis' (R. Khellaf, A. Bellet and J. Josse, 2025), enforcing 'DataSHIELD' disclosure controls throughout.


dsFedAte

Server-side 'DataSHIELD' package for federated estimation of the average treatment effect (ATE) across multiple studies distributed over separate data nodes. dsFedAte is installed on the server side (Opal or Armadillo); the analyst-side companion package is dsFedAteClient.

Installation

install.packages("dsFedAte")

Example

dsFedAte is called indirectly: the analyst runs dsFedAteClient::ds.EstimateAte() on the client, which dispatches the server-side aggregate function localAteDS() on each data node. A full worked example is provided in the federated_rct vignette of the client package:

vignette("federated_rct", package = "dsFedAteClient")

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("dsFedAte")

0.1.0 by Francisco Andrade, 7 hours ago


Browse source code at https://github.com/cran/dsFedAte


Authors: Francisco Andrade [aut, cre] , Rémi Khellaf [aut] , Julie Josse [aut]


Documentation:   PDF Manual  


AGPL (>= 3) license


Imports cli, dsBase, stats

Suggests testthat


See at CRAN