Summarize an exact binomial vaccine or prevention efficacy design

Creates a vaccine or prevention efficacy summary table from an exact binomial design. The table includes event-count bounds, efficacy at each bound, cumulative error spending, and cumulative efficacy-crossing probabilities for selected efficacy assumptions. Optional time-to-event design information adds planned analysis times and expected enrollment.

Usage

VEtable(x, ve, tteDesign = NULL, ratio = NULL)

Arguments

x

An object of class gsBinomialExact, generally created by toBinomialExact.

ve

Numeric vector of vaccine or prevention efficacy assumptions strictly between 0 and 1.

tteDesign

Optional gsSurv object with the same number of analyses as x. When supplied, planned analysis time and expected enrollment are included.

ratio

Experimental-to-control randomization ratio. By default, this is taken from tteDesign or from a design created by toBinomialExact().

Value

A tibble of class gsVETable with one row per analysis. The columns contain analysis number, optional timing and enrollment, total cases, exact efficacy and futility bounds, efficacy at each bound, cumulative alpha and beta spending, and cumulative efficacy-crossing probability under each value in ve. Pass the result to lt() for a formatted table with explanatory footnotes.

Details

Vaccine efficacy (VE), also termed prevention efficacy (PE) for non-vaccine preventive interventions, is translated to the exact binomial probability that an event is in the experimental group using the specified randomization ratio. The argument is named `ve` because the motivating application is a vaccine trial; the same calculation applies to PE. Cumulative alpha is calculated while ignoring non-binding futility, as is required for exact efficacy Type I error control.

See also

toBinomialExact, vignette("VaccineEfficacy")

Examples

x <- gsSurv(
  k = 2, test.type = 4, timing = .6, ratio = 3,
  hr = .3, hr0 = .7, lambdaC = .002, eta = .0001,
  gamma = 10, R = 8, T = 24, minfup = 16
)
exact <- toBinomialExact(x)
VEtable(exact, ve = c(.5, .7), tteDesign = x)
#> # A tibble: 2 × 12
#>   Analysis  Time     N Cases Success Futility ve_efficacy ve_futility   alpha
#>      <int> <dbl> <dbl> <dbl>   <dbl>    <dbl>       <dbl>       <dbl>   <dbl>
#> 1        1  15.9  3569    40      18       26       0.727       0.381 0.00246
#> 2        2  24    3569    67      37       38       0.589       0.563 0.0222 
#> # ℹ 3 more variables: beta <dbl>, `50%` <dbl>, `70%` <dbl>