calc_pwr_conj_test {crt2power}

Calculate statistical power for a cluster-randomized trial with co-primary endpoints using the conjunctive intersection-union test approach.

Description

Allows user to calculate the statistical power of a cluster-randomized trial with two co-primary outcomes given a set of study design input values, including the number of clusters in each trial arm, and cluster size. Uses the conjunctive intersection-union test approach. Code is adapted from "calPower_ttestIU()" from https://github.com/siyunyang/coprimary_CRT written by Siyun Yang.

Usage

calc_pwr_conj_test(
  K,
  m,
  alpha = 0.05,
  beta1,
  beta2,
  varY1,
  varY2,
  rho01,
  rho02,
  rho1,
  rho2,
  r = 1,
  cv = 0,
  deltas = c(0, 0),
  dist = "T"
)

Arguments

K

Number of clusters in treatment arm, and control arm under equal allocation; numeric.

m

Individuals per cluster; numeric.

alpha

Type I error rate; numeric.

beta1

Effect size for the first outcome; numeric.

beta2

Effect size for the second outcome; numeric.

varY1

Total variance for the first outcome; numeric.

varY2

Total variance for the second outcome; numeric.

rho01

Correlation of the first outcome for two different individuals in the same cluster; numeric.

rho02

Correlation of the second outcome for two different individuals in the same cluster; numeric.

rho1

Correlation between the first and second outcomes for two individuals in the same cluster; numeric.

rho2

Correlation between the first and second outcomes for the same individual; numeric.

r

Treatment allocation ratio - K2 = rK1 where K1 is number of clusters in experimental group; numeric.

cv

Cluster variation parameter, set to 0 if assuming all cluster sizes are equal; numeric.

deltas

Vector of non-inferiority margins, set to delta_1 = delta_2 = 0; numeric vector.

dist

Specification of which distribution to base calculation on, either 'T' for T-Distribution or 'MVN' for Multivariate Normal Distribution. Default is T-Distribution.

Value

A numerical value.

Examples

calc_pwr_conj_test(K = 15, m = 300, alpha = 0.05,
beta1 = 0.1, beta2 = 0.1, varY1 = 0.23, varY2 = 0.25,
rho01 = 0.025, rho02 = 0.025, rho1 = 0.01, rho2  = 0.05)