ROC curve analysis: a useful statistic multi-tool in the research of nephrology

Abstract

In the past decade, scientific research in the area of Nephrology has focused on evaluating the clinical utility and performance of various biomarkers for diagnosis, risk stratification and prognosis. Before implementing a biomarker in everyday clinical practice for screening a specific disease context, specific statistic measures are necessary to evaluate the diagnostic accuracy and performance of this biomarker. Receiver Operating Characteristic (ROC) Curve analysis is an important statistical method used to estimate the discriminatory performance of a novel diagnostic test, identify the optimal cut-off value for a test that maximizes sensitivity and specificity, and evaluate the predictive value of a certain biomarker or risk, prediction score. Herein, through practical examples, we aim to present a simple methodological approach to explain in detail the principles and applications of ROC curve analysis in the field of nephrology pertaining diagnosis and prognosis.

Keywords: Area under the curve; Diagnostic test,; Discriminatory ability; Receiver operator characteristic curve; Sensitivity; Specificity.

Fig. 1

Receiver operating characteristics (ROC) curves. Diagonal, dotted, black line (iii): a diagnostic test with the lowest discriminatory ability, which is no better than chance, area under the curve (AUC) = 0.5. The value of AUC is accompanied by 95% CIs. Gray line (ii): a test with a modest discriminatory ability of about 0.5. Black line (i): a perfect, accurate diagnostic test (highest sensitivity (100%) and specificity (100%)), AUC = 1

Fig. 2

ROC curve showing optimal cut-off value for serum asprosin, with an AUC of 72.5% for cutoff > 369.85 ng/ml (black dot), dotted black lines show sensitivity and specificity for this specific cut-off value, 82.4% and 51.8% respectively

Fig. 3

ROC curve showing optimal cut-off value for UAR, with an AUC of 72% for cutoff > 12.17(black dot). dotted black lines show sensitivity and specificity for this specific cut-off value, 83.21% and 60.81%, respectively

Fig. 4

ROC curves comparing the discriminatory performance of the full and simplified risk model in predicting CV disease in patients with diabetic kidney disease