confIntPairedDiagnostic: Compute confidence intervals for the comparison of two...

Description Usage Arguments Value Author(s) References Examples

Description

Compute confidence intervals for relative fractions. Specifically, confidence intervals for the relative true positive, true negative, false positive and false negative fraction are provided.

Usage

1
confIntPairedDiagnostic(Diseased, nonDiseased, conf.level = 0.95, adjust=FALSE)

Arguments

Diseased

Frequency table with results from both tests in the diseased population. First row and first column refers to negative test results.

nonDiseased

Frequency table with results from both tests in the non-diseased population. First row and first column refers to negative test results.

conf.level

Confidence level for confidence intervals.

adjust

Adjusted CIs? By default FALSE.

Value

A dataframe containing the estimated confidence intervals for the measures of relative accuracy (column test vs. row test).

Author(s)

Leonhard Held

References

Section 3.3 in Pepe, M.S. (2003) The statistical evaluation of medical tests for classification and prediction. Oxford: Oxford University Press.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
## Calculate confidence intervals for data from a study performing the 
## Exercise stress test (EST) and the determination of chest pain
## history (CPH) for patients with suspected or probably coronary
## heart disease (CHD).

Diseased <- matrix(c(25, 183, 29, 786), ncol=2, nrow=2, byrow=TRUE)
nonDiseased <- matrix(c(151, 176, 46, 69), ncol=2, nrow=2, byrow=TRUE)
colnames(Diseased) <- colnames(nonDiseased) <- c("CPH=0", "CPH=1")
rownames(Diseased) <- rownames(nonDiseased) <- c("EST=0", "EST=1")

print(confIntPairedDiagnostic(Diseased=Diseased, nonDiseased=nonDiseased))

biostatUZH documentation built on May 2, 2019, 6:06 p.m.