Evidence based medicine research helper

Estimate the standard error and confidence interval based on the area under the ROC curve (AUC) and related information

AUC value

Number of diseased cases based on gold standard (N1)

Number of non-diseased cases based on gold standard (N2)

Confidence Level

Standard error of AUC

Lower limit

Upper limit

Formula: $$SE(AUC)=\sqrt{\frac{AUC*(1-AUC)+(N1-1)*(Q1-AUC^2)+(Q2)*(Q2-AUC^2)}{N1*N2} }$$ Where: $$Q1=\frac{AUC}{2-AUC}$$ $$Q2=\frac{2*AUC^2}{1+AUC}$$

Reference:

Hanley, J.A. and NcNeil, B.J. 1982. 'The Meaning and Use of the Area under a Receiver Operating Characteristic (ROC) Curve.' Radiology, Vol 148, 29-36