get_confidence_set: Calculates the confidence set.

In OneArmPhaseTwoStudy: Planning, Conducting, and Analysing Single-Arm Phase II Studies

Description

The p-value of Subset Designs depends on two endpoints e.g. the superset endpoint and the subset endpoint. Therefore the confidence interval for the response rate of the subset endpoint depends on the response rate of the superset endpoint and vice versa. This results in a conficence "area" which is called the confidence set. "get_conficence_set" returns a set of points which outline the border of the confidence set.

Usage

get_confidence_set(t, u, r1, n1, n, pc0, pt0, alpha)

Arguments

t	observed responses in subset endpoint.
u	observed responses in the superset endpoint.
r1	critical value for the first stage.
n1	sample size for the first stage.
n	overall sample size.
pc0	the response probability under the null hypothesis for the subset endpoint.
pt0	the response probability under the null hypothesis for the superset endpoint.
alpha	significance level the study was planned for.

References

Kunz, C. U. (2011), Two-stage designs for phase II trials with one or two endpoints. http://d-nb.info/1024218457

See Also

setupSub1Design, plot_confidence_set

Examples

#Setup "sub1"-object
sub1 <- setupSub1Design(pc0 = 0.5, pt0 = 0.6)

#Calculate a  subset design
design <- getSolutionsSub1(sub1, skipN1 = FALSE)$Solutions[4,]

t <- 12
u <- 13
alpha = 0.1

conf_set <- get_confidence_set(t, u, design$r1, design$n1, design$n, design$pc0, design$pt0, alpha)

OneArmPhaseTwoStudy documentation built on May 2, 2019, 9:28 a.m.