J Dig Cancer Res.  2022 Dec;10(2):99-106. 10.52927/jdcr.2022.10.2.99.

Determining the Optimal Cut-off Point According to the Outcome Variables Using R

• Yang J
• Lee HS

• KMID: 2536427
• DOI: http://doi.org/10.52927/jdcr.2022.10.2.99

Abstract

Clinical research ultimately aimed to promptly diagnose and prevent diseases through precise biomarker development. Finding the optimal cut-off point of a regularly measured biomarker can help its interpretation and ultimately help in disease investigation and diagnosis, more specifically in determining the presence of diseases. Therefore, this study aimed to use the characteristics of outcome variables in clinical research to explain how to determine the optimal cutoff point. The outcome variables can be divided into dichotomous, ordinal, and survival types. The optimal cut-off point can be determined by finding points that maximize the Youden index, extended Youden index, and log-rank statistics. This study will enable clinical researchers to accurately determine the optimal cut-off points for regularly measured biomarkers, thereby enabling prompt disease diagnosis for effective treatment.

Keyword

Optimal cut-off point; Youden index; Log-rank statistics; Biomarkers