confIntIndependentAUCDiff: Computes confidence interval for the difference in AUC based...
In biostatUZH: Misc Tools of the Department of Biostatistics, EBPI, University of Zurich

Description Usage Arguments Details Value Author(s) References Examples

Computes confidence interval for the difference in the area under the curve based on two independent samples.

1 2	confIntIndependentAUCDiff(casesA, controlsA, casesB, controlsB, type = "Wald", conf.level = 0.95)

`casesA`	Values of the continuous variable from Test A for the cases.
`controlsA`	Values of the continuous variable from Test A for the controls.
`casesB`	Values of the continuous variable from Test B for the cases.
`controlsB`	Values of the continuous variable from Test B for the controls.
`type`	"Wald" or "Logit". Defaults to "Wald".
`conf.level`	Confidence level for confidence interval.

For type="Wald", standard Wald confidence intervals are calculated for AUC of both tests and their difference. For type="logit", the substitution method is used based on the logit transformation for the AUC of both tests. The confidence interval for the difference in AUC is then calculated using Newcombe's method.

A matrix with estimate and confidence limits for AUC of the two tests and their difference.

Leonhard Held

Newcombe, R.G. (1998). Interval estimation for the difference between independent proportions: Comparison of eleven methods. Stat. Med., 17, 873–890.

Pepe, M.S. (2003) The statistical evaluation of medical tests for classification and prediction. Oxford: Oxford University Press.

set.seed(12345)
casesA <- rnorm(200, mean=2.5)
controlsA <- rnorm(100)
casesB <- rnorm(100, mean=1.5)
controlsB <- rnorm(200)

confIntIndependentAUCDiff(casesA, controlsA, casesB, controlsB, type="Wald")
confIntIndependentAUCDiff(casesA, controlsA, casesB, controlsB, type="Logit")