What is it about?
Accurate diagnosis of diseases is important for medical practice. This article breaks new ground by proposing methods for biomarker performance evaluation that do not suffer from key issues that affect the performance of most applied approaches in practice, such as AUC (Area under the Curve) and the Youden Index. The proposed methods are particularly tailored for medical and genomic studies.
Featured Image
Photo by Ani Kolleshi on Unsplash
Why is it important?
The most widely used metrics of biomarker performance evaluation---the AUC (Area under the Curve) and the Youden Index---suffer from key issues that question their reliability. This article proposes for the first time novel statistical methods for biomarker performance evaluation that do not suffer from those issues.
Perspectives
Writing this article made me re-think how 'fragile' are the state-of-the-art methods for assessing the performance of biomarkers. Can you imagine the number of decisions that are being made on the basis of these methods while you are reading this? It is extremely worrying. Our paper may be a small step towards more solid ground, but I invite others to join us on this inquiry.
Miguel de Carvalho
University of Edinburgh
Read the Original
This page is a summary of: Affinity-based measures of biomarker performance evaluation, Statistical Methods in Medical Research, May 2019, SAGE Publications,
DOI: 10.1177/0962280219846157.
You can read the full text:
Resources
Contributors
The following have contributed to this page
