Exact Permutation and Bootstrap Distribution of Generalized Pairwise Comparisons Statistics

To analyze multivariate outcomes in clinical trials, several authors have suggested generalizations of the univariate Mann–Whitney test. As statistic compares subjects’ outcome pairwise, are known as generalized pairwise comparisons (GPC) statistics. For GPC statistics such net treatment benefit, win ratio, and odds, asymptotic based or re-sampling tests been literature. However, methods require a sufficiently high sample size to be accurate, come with computational burden. We use graph theory notation obtain closed-form formulas for expectation variance permutation bootstrap sampling distribution statistics, which can utilized develop fast accurate inferential each A simple example simulation study demonstrate accuracy exact methods, even very small samples. time complexity is O(N2), where N total number patients, fast. In situations used these matrices, new will more equally has used, both much faster.