confIntIndependentDiagnostic: Confidence intervals for the comparison of two diagnostic...

In felix-hof/biostatUZH: Misc Tools of the Department of Biostatistics, EBPI, University of Zurich

Confidence intervals for the comparison of two diagnostic tests from unpaired data

Description

Compute confidence intervals for relative sensitivity, relative specificity and relative likelihood ratios.

Usage

confIntIndependentDiagnostic(
  tp,
  fp,
  tn,
  fn,
  names = c(1, 2),
  conf.level = 0.95,
  adjust = FALSE
)

Arguments

tp	Vector of length 2. Number of true positives of the two tests.
fp	Vector of length 2. Number of false negatives of the two tests.
tn	Vector of length 2. Number of true negatives of the two tests.
fn	Vector of length 2. Number of false negatives of the two tests.
names	Vector of length 2. Names of first and second diagnostic tests. Default to c(1, 2).
conf.level	Confidence level for confidence interval. Default is 0.95.
adjust	Logical of length one indicating whether to compute adjusted CIs or not. Default is FALSE.

Value

A data.frame containing the estimated sensitivity, specificity, LR+ and LR- in the two tests, as well as the ratio (first versus second test) with the corresponding confidence interval.

Author(s)

Leonhard Held, Charlotte Micheloud

References

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

Examples

## Calculate confidence intervals for data from a (hypothetical)
## randomized unpaired study of early amniocentesis (EA)
## versus chorionic villus sampling (CVS)
## for fetal abnormality from
## Pepe (2003)

tp <- c(116, 111)
fp <- c(34, 111)
tn <- c(4844, 4765)
fn <- c(6, 13)
confIntIndependentDiagnostic(tp = tp, fp = fp, tn = tn, fn = fn,
names = c("EA", "CVS"))

felix-hof/biostatUZH documentation built on Sept. 27, 2024, 1:48 p.m.