odsroc: Nonparametric estimation of AUC and partial AUC under test-result-dependent sampling (R).

Nonparametric Estimation of AUC and Partial AUC under Test-Result-Dependent Sampling. Statistics in Biopharmaceutical Research 4, 313-323 by XF Wang, JL Ma, SL George, HB Zhou (2012).

The area under a ROC curve (AUC) and the partial area under a ROC curve (pAUC) are two important summary measures for assessing the accuracy of a diagnostic test discriminating a binary true disease status. We develop nonparametric estimation of AUC and pAUC under a biased sampling scheme in which a random sample is supplemented with a test-result-dependent sample. Such biased sampling scheme can arise from an existing study with sampling limitation or a study under development for better estimation efficiency.