This R package contains code for a proof of concept study in advanced non-small cell lung cancer (NSCLC) of a meta-analytic framework to adjust external control studies for additional bias and variability due to their non-randomized design. It is not meant to be a general purpose library; instead, it simply aims to facilitate transparent, reproducible, and modular analysis. The analyses estimate the parameters of the meta-analytic model and illustrate use for a new hypothetical single-arm trial.
For a general purpose package, see ecmeta, which describes, assesses, and implements the framework for general use cases.
Details of the meta-analytic framework can be found in our manuscript here.
The framework can only be applied to a new trial after estimation of parameters using relevant historical data from RCTs and external controls. This project estimates parameters relevant for a new trial evaluating treatment of patients with advanced NSCLC and illustrates application of the framework to this new trial. The prespecified statistical analysis plan for parameter estimation can be found here. 14 pairwise comparisons from 11 separate clinical trials are used in the analysis.
The ecmeta.nsclc package can be installed from the R console with:
system("R CMD INSTALL ecmeta.nsclc")