crmsens {titecrm}R Documentation

Model Sensitivity in the CRM

Description

Returns an object of class ``homesets'' that summarizes the sensitvity (robustness) of the dose-toxicity model used in the CRM.

Usage

crmsens(prior, target, eps=1e-8, maxit=100, incr=0.1, LB=0, UB=Inf, 
detail=FALSE)

Arguments

prior A vector of initial estimates of toxicity probabilities associated the doses.
target The target DLT rate.
detail If TRUE, the details of the home sets will also be shown. Default is set as FALSE.
eps A control parameter for absolute tolerance. Defaults to `1e-8'.
maxit Maximum number of iterations allowed. Defaults to 100.
incr Step size. Defaults to 0.1.
LB Lower bound of the search space. Defaults to 0.
UB Upper bound of the search space. Defaults to `Inf'.

Details

Dose-toxicity relationship is assumed as an empiric power model $a_i^{exp(β)}$ where $a_i$ is the initial estimate of toxicity. Different choice of the initial probabilities results in various level of robustness of the CRM, which can be summarized by the indifference intervals (iint). See Cheung and Chappell (2002) in the reference section.

Value

An object of class ``homesets'' is returned, describing the model sensitivity by indifference intervals and a summary statement.

homeset The home sets of the model parameter $exp(β)$.
iint The indifference intervals of the dose-toxicity model.
prior Initial estimates of toxicity probabilities.
target The target probability of toxicity at the MTD.

References

Cheung, Y. K. and Chappell, R. (2002). A simple technique to evaluate model sensitivity in the continual reassessment method. Biometrics 58:671-674.

Examples

# Create a simple data set
prior <- c(0.05,0.10,0.20,0.35,0.50,0.70)
target <- 0.2
foo <- crmsens(prior,target)
print(foo)

[Package titecrm version 0.1-1 Index]