Cluster RCT with Machin Eq 12.10 (DE + Student t correction)

Returns the total number of subjects N and the number of clusters per arm. Rounding is done ONCE, at the cluster level (Machin Ch.12 "Numerical Accuracy"): each extra cluster adds a further m subjects, so the cluster count must not be derived from an already-rounded subject count.

Usage

cluster_design_from_indiv(
  n_per_arm_indiv,
  cluster_size,
  icc,
  cv_m = NULL,
  alpha = 0.05,
  power = 0.8,
  phi = 1
)

Arguments

n_per_arm_indiv

Subjects per arm from the corresponding individually-randomised formula (unrounded).

cluster_size

Cluster size m.

icc

Intra-cluster correlation rho.

cv_m

Optional coefficient of variation of cluster size (Eq 12.7).

alpha, power

Type I and type II error.

phi

Allocation ratio (Machin's phi); default 1.