Incorporating a New Summary Statistic into the Min–Max Approach: A Min–Max–Median, Min–Max–IQR Combination of Biomarkers for Maximising the Youden Index

Linearly combining multiple biomarkers is a common practice that can provide better diagnostic performance. When the number of sufficiently high, computational burden problem arises. Liu et al. proposed distribution-free approach (min–max approach) linearly combines minimum and maximum values biomarkers, involving only single coefficient search. However, combination alone may not be sufficient in terms discrimination. In this paper, we propose new extends by incorporating summary statistic, specifically, median or interquartile range (min–max–median min–max–IQR approaches) order to find optimal maximises Youden index. Although more computationally intensive than one al, it includes information parameters estimated remains reasonable. We compare performance approaches min–max–IQR) with min–max logistic regression. For purpose, wide different simulated data scenarios were explored. also apply two real datasets (Duchenne Muscular Dystrophy Small for Gestational Age).