ABSTRACT
Surrogate endpoints are preferred often because of the convenience of measuring them over the true endpoint in clinical trials. When trial data is available, Rtrial2 denotes the trial-level surrogacy and Rindiv2 denotes the individual-level surrogacy. The package describes the statistical model behind the two stage-approach in its paper and implements the model in code. The unadjusted measure provided in the package is simply the ptrial2, which is the squared correlation between the treatment effects of the two endpoints. The data set ‘gastadv’ in surrosurv package will be used to fit the bivariate Copula Model and return adjusted and unadjusted log-hazard ratio estimates for overall survival and progression-free survival. The gastadv data set contains individual data of 4069 patients with advanced/recurrent gastric cancer from 20 randomized trials of chemotherapy. For the individual-level measure of association, we will consider the estimates of Kendall’s Tau in the ‘allSurroRes’ object.
