run_a_noninf_trial: Run a non-inferiority trial
In jatotterdell/automaticsims: Package for automatic simulations

run_a_noninf_trial

R Documentation

Run a non-inferiority trial

Description

Run a single trial using non-inferiority stopping. Stopping for noninferiority requires that:

Usage

run_a_noninf_trial(
  id,
  mu,
  delta_sup = 0.1,
  delta_ctr = delta_sup,
  delta_noninf = delta_sup,
  delta_pair = delta_sup,
  kappa_act_0 = 0.01,
  kappa_act_1 = 0.05,
  kappa_sup_0 = 0.95,
  kappa_sup_1 = 0.8,
  kappa_ctr_0 = 0.95,
  kappa_ctr_1 = 0.95,
  kappa_noninf_0 = 0.6,
  kappa_noninf_1 = 0.6,
  kappa_nonsup_0 = 0.1,
  kappa_nonsup_1 = 0.1,
  brar = FALSE,
  allocate_inactive = FALSE,
  return_all = FALSE,
  ind_comp_ctrl = FALSE,
  ctrl_alloc = 1/13
)

Arguments

`id`	The trial ID
`mu`	The true mean, must be length 13
`delta_sup`	The superiority margin
`delta_ctr`	The superiority margin for control comparisons
`delta_noninf`	The noninferiority margin
`delta_pair`	The margin for pairwise comparisons
`kappa_act_0`	The starting threshold to deactivate poor arms
`kappa_act_1`	The final threshold to deactivate poor arms
`kappa_sup_0`	The starting threshold for superiority
`kappa_sup_1`	The final threshold for superiority
`kappa_ctr_0`	The starting threshold for beating control
`kappa_ctr_1`	THe final threshold for beating control
`kappa_noninf_0`	The starting threshold for non-inferiority
`kappa_noninf_1`	The final threshold for non-inferiority
`kappa_nonsup_0`	The starting threshold for non-superiority
`kappa_nonsup_1`	The final threshold for non-superiority
`brar`	Use Response Adaptive Randomisation?
`allocate_inactive`	Continue to allocate subjects to arms meeting the inactive threshold?
`return_all`	Return all value, or only those from final analysis?
`ind_comp_ctrl`	Should deactivation be based on independent P(max trt arms) and P(beat ctrl) or just global P(max all arms)