Description Usage Arguments Value Examples
Calculates the operating characteristics of futility rules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | futilityOC(
txtype = "norm",
pow.type = "PREDICTIVE",
p.stop = 0.15,
ind = 1,
epts.r = 1,
inF = 0.5,
H1 = TV,
TV = H1,
benef = NA,
ctr.r = 0.5,
NB.k = 0.5,
over.dp = 1.5,
alpha = 0.1,
power = 0.8,
n = 0,
mat = 1,
sd = 1,
LRV = "",
delta = "",
ran.r = 1,
pr.tv = 0.1,
pr.lrv = 0.8,
print = T
)
|
txtype |
The type of treatment (tx) effect "norm", HR","NB","PR","QP","RD", "RR". Default="norm" HR = hazard ratio; NB = Negative binomial; PR = Poisson regression QP = Quasi Poisson; RD = Risk difference; RR = risk ratio |
pow.type |
The type of power for futility rule "PREDICTIVE" or "CONDITIONAL" power. Default="PREDICTIVE" |
p.stop |
power boundary to stop for futility. Default = 0.15 |
ind |
value 1, 2 or 3. 1 = power for statistical significance, 2 = power to meet Go criteria, 3 = power to meet better than NoGo boundary. Default = 1 |
epts.r |
Assumed correlation between the endpoints used at the futility analysis and the final. Default = 1 |
inF |
information fraction at interim. Default = 0.5 |
H1 |
The treatment effect size assumed under the alternative hypothesis. Default = TV |
TV |
Target value for Go No Go decision. Default = H1 |
benef |
A logical value. TRUE indicates a larger effect is beneficial. FALSE indicates a small effect is beneficial. Default values depend on the type of treatment effect: = TRUE for effect types of norm and RD = FALSE for effect types of HR, NB, PR, QP and RR |
ctr.r |
Assumed event rate for the control arm, which is necessary to be specified for effect types of NB, PR, QP, RD and RR. Default = 0.5 |
NB.k |
The shape parameter of negative binomial distribution, which is necessary when effect type is NB. Default = 0.5. |
over.dp |
The over-dispersion estimate for treatment effect type QP. Default = 1.5 |
alpha |
Alpha value that is used for determining sample size n, if n is not specified. Default = 0.1 one-sided. |
power |
Statistical power that is used for determining sample size if n is not specified. Default = 0.8. |
n |
The sample size of the trial. When it is not specified (=0), its value is estimated. Note: n represents the number of events when treatment effect is in HR with time-to-event endpoint and the default mat = 1 is not altered. |
mat |
Data maturity rate of time-to-event endpoint. Default = 1 |
sd |
The standard deviation specified for endpoints following a normal distribution. Default = 1 |
LRV |
The lower reference value specified for its use in Go/NoGo decision rules. Default = 2/3 * TV |
delta |
A vector of assumed/hypothesized treatment effects for which the operating characteristics of futility analysis are to be evaluated. Default = a vector of 4 with values = (TV, LRV, TV/4, 0). |
ran.r |
The randomization ratio. Default = 1 |
pr.tv |
Standard GNG rules. Probability that the treatment effect is equal to or greater than TV. Default = 0.1 |
pr.lrv |
Standard GNG rules. Probability that the treatment effect is equal to or greater than LRV. Default = 0.8 |
print |
TRUE prints the output as well as returning dataframe |
dataframe with Design.specifications, Go.NoGo.parameters, Prob.stop.under.Hypotheses, Futility.stopping.boundary, Conditional.power.at.stopping.boundary, Predictive.power.at.stopping.boundary, Critical.values.for.Go.NoGo.based.on.Interim.data
1 2 3 | futilityOC(txtype = "RD", pow.type = "PREDICTIVE", p.stop = 0.2,
ind = 1, inF = 0.5, TV = 0.2, # target difference
benef = T, ctr.r = 0.12, alpha = 0.025, n = 72)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.