n.pick.winner: Sample size calculation for Simon's pick winner design

In statapps/power.ctepd: Power and sample size for clinical trials and epidemiology

Sample size calculation for Simon's pick winner design

Description

Calculate sample size of Simon's pick winner design or determine probability of alternative hypothesis to obtain target power.

Usage

n.pick.winner(n = NULL, k = 2, p0=NULL, p1=NULL, power=NULL, 
                tol = .Machine$double.eps^0.25)

Arguments

n	Number of sample size per arm
k	Number of arms, k > or = 2, default is k = 2
p0	Response rate of the lower efficacy drug
p1	Response rate of the best drug, p0 < p1
power	Probability of choosing the winner arm, power = 0.90 or higher is recommended
tol	tol for uniroot function

Details

These calculations use method proposed by R. Simon for pick winner design for randomized phase II clinical trials.

Exactly one of the parameters 'n','power' and 'p1' must be passed as NULL, and that parameter is determined from the others. Notice that the last one has non-NULL default so NULL must be explicitly passed if you want to compute it.

Since pick winner design is not a hypothesis test, there is not formal comparison between arms. Therefore, the type I error rate (sig.level) does not apply in this situation.

'pselect' in package 'clinfun' also provide the probability of slection the winner arm but do not provide a direct solution for sample size.

Value

Object of class 'power.htest', a list of the arguments (including the computed one) augmented with 'method' and 'note' elements.

n	Number of sample size per arm
arm	Number of arms, the same as input k
n.total	Number of total sample size = n*arms

Note

'uniroot' is used to solve power equation for unknowns, so you may see errors from it, notably about inability to bracket the root when invalid arguments are given.

Author(s)

Bingshu E. Chen ([e-mail])

References

Richard Simon, Robert E. Wittes and Susan S. Ellenberg. (1985). Randomized phase II clinical Trials. Cancer Treat Rep. vol. 69, page 1375-1381.

See Also

power.surv.test, power.t.test, pwr.p.test

Examples

## Example: sample size for a three arms study
n.pick.winner(power = 0.95, p0 = 0.2, p1 = 0.35, k =3 )

statapps/power.ctepd documentation built on April 26, 2024, 3:22 a.m.