ThreeArmedTrials (version 1.0-4)

power_RET: Power related calculations for three-arm clinical trials

Description

Compute power, sample size, or level of significance for Wald-type test for non-inferiority or superiority of the experimental treatment versus reference treatment with respect to placebo.

Usage

power_RET(
  experiment,
  reference,
  placebo,
  Delta,
  sig_level = NULL,
  power = NULL,
  n = NULL,
  allocation = c(1/3, 1/3, 1/3),
  distribution = NULL,
  ...
)

Value

A list with class "power.htest" containing the following components:

n

The total sample size

power

A numeric value specifying the target power

Delta

A numeric value specifying the non-inferiority or superiority margin.

sig.level

A character string specifying the significance level

type

A character string indicating what type of Wald-type test will be performed

allocation

A vector with the sample size allocation (nExp/n, nRef/n, nPla/n)

sig.level

The significance level (Type I error probability)

nExp

A numeric value specifying the number of sample in the experimental treatment group

nRef

A numeric value specifying the number of sample in the reference treatment group

nPla

A numeric value specifying the number of sample in the placebo treatment group

Arguments

experiment

a numeric vector specifying the parameters of the experimental treatment group in the alternative hypothesis

reference

a numeric vector specifying the parameters of the reference treatment group in the alternative hypothesis

placebo

a numeric vector specifying the parameters of the placebo treatment group in the alternative hypothesis

Delta

a numeric value specifying the non-inferiority/superiority margin

sig_level

A numeric value specifying the significance level (type I error probability)

power

A numeric value specifying the target power (1 - type II error probability)

n

The total sample size. Needs to be at least 7.

allocation

A (non-empty) vector specifying the sample size allocation (nExp/n, nRef/n, nPla/n)

distribution

A character specifying the distribution of the endpoints. Must must be either of "binary", "poisson", "negbin", "exponential", "normal"

...

Further arguments. See details.

Details

If the individual group sample sizes, i.e. n*allocation are not natural number, the parameters n and allocation will be re-calculated.

The additional parameter var_estimation is a character string specifying how the variance for the Wald-type test statistic is estimated in the Poisson and negative binomial model. Must be RML for restricted maximum-likelihood, or ML for unrestricted maximum-likelihood

Examples

power_RET(experiment = 15, reference = 17, placebo = 20,
         Delta = 0.8, sig_level = 0.025, power = 0.8,
         allocation = c(1, 1, 1) / 3,
         var_estimation = "RML",
         distribution = "poisson")