ci_gs: Determine confidence intervals in a group sequential...
In mjg211/singlearm: Design and analysis of single-arm clinical trials

Description Usage Arguments Details Value See Also Examples

Determines possible confidence intervals at the end of a group sequential single-arm trial for a single binary endpoint, as determined using des_gs(). Support is available to compute confidence intervals using the naive ("naive"), exact ("exact"), and Mid-p ("mid_p") approaches.

1 2	ci_gs(des, k, pi, alpha = des$alpha, method = c("exact", "mid_p", "naive"), summary = F)

`des`	An object of class `"sa_des_gs"`, as returned by `des_gs()`.
`k`	Calculations are performed conditional on the trial stopping in one of the stages listed in vector `k`. Thus, `k` should be a vector of integers, with elements between one and the maximum number of possible stages.
`pi`	A vector of response probabilities to evaluate the expected performance of the point estimation procedures at. This will internally default to be the π₀ and π₁ from `des` if it is left unspecified.
`alpha`	Level to use in confidence interval construction. Defaults to the value of α used in the construction of `des` (i.e., `des$alpha`).
`method`	A vector of methods to use to construct confidence intervals. Currently, support is available to use the naive (`"naive"`), exact (`"exact"`), and Mid-p (`"mid_p"`) approaches. Defaults to all available methods.
`summary`	A logical variable indicating whether a summary of the function's progress should be printed to the console.

In addition, the performance of the chosen confidence interval determination procedures (including their coverage and expected length) for each value of pi) in the supplied vector pi), will also be evaluated.

Calculations are performed conditional on the trial stopping in one of the stages specified using the input (vector) k.

A list of class "sa_ci_gs" containing the following elements

A tibble in the slot $ci summarising the possible confidence intervals at the end of the trial for the supplied design, according to the chosen methods.
A tibble in the slot $perf summarising the performance of the chosen confidence interval determination procedures for each specified value of π.
Each of the input variables as specified, subject to internal modification.

des_gs, opchar_gs, est_gs, pval_gs, and their associated plot family of functions.

# Find the optimal two-stage design for the default parameters
des <- des_gs()
# Determine the performance of all supported confidence interval
# determination procedures for a range of possible response probabilities
ci  <- ci_gs(des, pi = seq(0, 1, 0.01))