AGSDest (version 2.3.4)

adapt: Adaptations in group sequential trials

Description

adapt is a function that performs adaptations and plans the secondary group sequential trial. The effect size used for planning the secondary trial is a weighted mean between the interim estimate theta and the initially assumed estimate delta (pT\$delta) of the primary trial.

Usage

adapt(pT, iD, SF, phi, cp, theta = iD$z/(pT$t[iD$T] * pT$Imax), I2min, I2max,
  swImax, delta = pT$delta, weight = 1, warn = TRUE)

Arguments

pT

object of the class GSTobj; primary trial design

iD

interim data; a list with the variables T and z; list(T = stage of interim analysis, z = interim z-statistic)

SF

spending function for the secondary trial

phi

parameter of spending function for the secondary trial when SF=3 or 4 (See below)

cp

conditional power

theta

new effect size (default: estimate from interim analysis)

I2min

minimal total information of secondary trial

I2max

maximal total information of secondary trial

swImax

maximal incremental information per stage

delta

initially assumed effect size for the primary trial (default: estimate from primary trial)

weight

weight of theta when updating the effect size estimate as weighted mean of theta and delta

warn

option if warnings should be printed to the screen (default: true)

Value

adapt returns an object of the class GSTobj; the design of the secondary trial. The adaptation rule is as in the first simulation example of Brannath et al.(2008). If no adaptations are performed, the function returns sT = NULL. An object of class GSTobj is a list containing the following components:

sT

secondary trial

Details

If no adaptation is performed then this indicates that the original plan is kept. In this case sT is set to NULL. If an adaptation is performed sT is a list which contains the following elements:

K	number of stages
a	lower critical bounds of secondary group sequential design(are currently always set to -8)
b	upper critical bounds of secondary group sequential design
t	vector with cumulative information fractions
al	alpha (type I error rate); equal to the conditional type I error rate of the primary trial
SF	spending function
phi	parameter of spending function when SF=3 or 4 (See below)
alab	alpha-absorbing parameter values of secondary group sequential design
als	alpha-values ''spent'' at each stage of secondary group sequential design
Imax	maximum information number
delta	effect size used for planning the secondary trial
cp	conditional power

A value of SF=3 is the power family. Here, the spending function is \(t^{\phi}\), where phi must be greater than 0. A value of SF=4 is the Hwang-Shih-DeCani family, with the spending function \((1-e^{-\phi t})/(1-e^{-\phi})\), where phi cannot be 0.

References

Brannath, W, Mehta, CR, Posch, M (2008) ''Exact confidence bounds following adaptive group sequential tests'', Biometrics accepted.

See Also

GSTobj, print.GSTobj, plot.GSTobj, plan.GST

Examples

# NOT RUN {
##The following performs an adaptation of the sample size and
##number of interim analyses after the first stage of the primary trial.

pT=plan.GST(K=3,SF=4,phi=-4,alpha=0.05,delta=6,pow=0.9,compute.alab=TRUE,compute.als=TRUE)

iD=list(T=1, z=1.090728)

swImax=0.0625

I2min=3*swImax
I2max=3*swImax

sT=adapt(pT=pT,iD=iD,SF=1,phi=0,cp=0.8,theta=5,I2min,I2max,swImax)
# }