In mjg211/multiarm: Design of single- and multi-stage multi-arm clinical trials

knitr::opts_chunk$set(echo = F, results = "asis")

K          <- params$K
alpha      <- params$alpha
beta       <- params$beta
delta1     <- params$delta1
delta0     <- params$delta0
sigma      <- params$sigma
ratio_type <- params$ratio_type
ratio_init <- params$ratio_init
ratio      <- params$ratio
correction <- params$correction
power      <- params$power
integer    <- params$integer
N          <- params$large_N
n          <- params$small_n
opchar     <- params$opchar
gamma      <- params$gamma
gammaO     <- params$gammaO
plots      <- params$plots

if (correction == "bonferroni") {
  correction_text     <- "Bonferroni's correction"
} else if (correction == "dunnett") {
  correction_text     <- "Dunnett's correction"
} else if (correction == "none") {
  correction_text     <- "no multiple comparison correction"
} else if (correction == "sidak") {
  correction_text     <- "Sidak's correction"
} else if (correction == "benjamini_hochberg") {
  correction_text     <- "the Benjamini-Hochberg correction"
} else if (correction == "benjamini_yekutieli") {
  correction_text     <- "the Benjamini-Yekutieli correction"
} else if (correction == "hochberg") {
  correction_text     <- "Hochberg's correction"
} else if (correction == "holm_bonferroni") {
  correction_text     <- "the Holm-Bonferroni correction"
} else if (correction == "holm_sidak") {
  correction_text     <- "the Holm-Sidak correction"
} else if (correction == "step_down_dunnett") {
  correction_text     <- "the step-down version of Dunnett's correction"
}

if (power == "marginal") {
  power_type_text     <- "marginal power for each null hypothesis"
  power_scenario_text <-
    "each of their respective least favourable configurations"
} else {
  power_scenario_text <- "the global alternative hypothesis"
  if (power == "conjunctive") {
    power_type_text   <- "conjunctive power"
  } else if (power == "disjunctive") {
    power_type_text   <- "disjunctive power"
  }
}

if (ratio_type == "equal_all") {
  ratio_text          <- "the same as the control arm"
} else if (ratio_type == "equal_exp") {
  ratio_text          <- paste0(ratio_init[1], " times that of the control arm")
} else if (ratio_type == "unequal") {
  ratio_text          <- paste0(paste(ratio_init[1:K], collapse = ", "),
                                " times that of the control arm")
} else if (ratio_type == "root_K") {
  ratio_text          <- paste0(round(1/sqrt(K), 3),
                                " times that of the control arm")
} else {
  ratio_text          <- paste0("chosen for ", ratio_type, "-optimality")
}

if (integer) {
  integer_text        <- "will"
} else {
  integer_text        <- "will not"
}

if (plots) {
  plot_text           <- "will"
} else {
  plot_text           <- "will not"
}

FWER                  <- round(opchar$FWERI1[1], 3)
if (correction %in% c("benjamini_hochberg", "benjamini_yekutieli", "hochberg",
                      "holm_bonferroni", "holm_sidak", "step_down_dunnett")) {
  punc_text_3         <- "s"
  gamma_text             <- paste0("(", paste(round(gammaO, 3), collapse = ", "), ")")
} else {
  punc_text_3         <- ""
  gamma_text             <- round(gamma, 3)
}

if (power == "marginal") {
  power_type_text     <- "minimum marginal power"
  power_value         <-
    round(min(diag(as.matrix(opchar[-(1:2), (K + 3):(2*K + 2)]))), 3)
} else {
  power_scenario_text <- "the global alternative hypothesis."
  if (power == "conjunctive") {
    power_type_text   <- "conjunctive power"
    power_value       <- round(opchar$Pcon[2], 3)
  } else if (power == "disjunctive") {
    power_type_text   <- "disjunctive power"
    power_value       <- round(opchar$Pdis[2], 3)
  }
}

Inputs

The following choices were made:

$K = \boldsymbol{r K}$ experimental treatments will be included in the trial.
A significance level of $\alpha = \boldsymbol{r alpha}$ will be used, in combination with r correction_text.
The standard deviations of the responses will be assumed to be: $(\sigma_0,\dots,\sigma_K) = \boldsymbol{(r paste(round(sigma, 3), collapse = ", "))}$.
The r power_type_text will be controlled to level $1-\beta = \boldsymbol{r 1 - beta}$ under r power_scenario_text.
The interesting and uninteresting treatment effects will be: $\delta_1 = \boldsymbol{r round(delta1, 3)}$ and $\delta_0 = \boldsymbol{r round(delta0, 3)}$ respectively.
The target allocation to each of the experimental arms will be: r ratio_text.
The sample size in each arm r integer_text be required to be an integer.
Plots r plot_text be produced.

Outputs

The total required sample size is: $N = \boldsymbol{r round(N, 3)}$.
The required sample size in each arm is: $(n_0,\dots,n_K) = \boldsymbol{(r paste(round(n, 3), collapse = ", "))}$.
Therefore, the realised allocation ratios to the experimental arms are: $(r_1,\dots,r_K) = \boldsymbol{(r paste(round(ratio, 3), collapse = ", "))}$.
The maximum familywise error-rate is: $\boldsymbol{r FWER}$.
The r power_type_text is: $\boldsymbol{r power_value}$.
The following critical thresholdr punc_text_3 should be used with the chosen multiple comparison correction: $\boldsymbol{r gamma_text}$.