create_normal_model: Create Normal Emax Model

In bdpopt: Optimisation of Bayesian Decision Problems

Description

Create a simulation model object for optimisation of the dose and sample size for a (or several, parallel) phase III clinical trial given phase II data. There is one efficacy response and one safety response obtained as sample means for each group of patients given a specific dose. These are both normal given the true populations means. The population means for efficacy and safety are in turn modeled using two independent Emax models.

Usage

create.normal.model(theta.mu, theta.tau, eta.mu, eta.tau,
n.II, d.II, YE.II, YS.II, sigmaE, sigmaS, k.III, path.to.package = NA)

Arguments

theta.mu	A numeric, atomic vector with four elements that contains the mean parameters for the prior distributions for the efficacy model.
theta.tau	A numeric, atomic vector with four elements that contains the precision parameters for the prior distributions for the efficacy model.
eta.mu	A numeric, atomic vector with four elements that contains the mean parameters for the prior distributions for the safety model.
eta.tau	A numeric, atomic vector with four elements that contains the precision parameters for the prior distributions for the safety model.
n.II	A numeric, atomic vector of positive integers containing the group sample sizes corresponding to the phase II efficacy and safety responses.
d.II	A numeric, atomic vector containing the group dose levels corresponding to the phase II efficacy and safety responses.
YE.II	A numeric, atomic vector containing the observed sample means for the efficacy responses in the phase II trial.
YS.II	A numeric, atomic vector containing the observed sample means for the safety responses in the phase II trial.
sigmaE	The population standard deviation for an individual efficacy response. The population standard deviation for a group response is then obtained by dividing by the square root of the group sample size.
sigmaS	The population standard deviation for an individual safety response. The population standard deviation for a group response is then obtained by dividing by the square root of the group sample size.
k.III	A positive integer specifying the number of independent phase III trials, each of which consists of a single group. The sample size and dose is the same for all phase III trials.
path.to.package	The search path to the installation directory of bdpopt. For the default value, the function will attempt to find the path using search.

Details

Note that n.II, d.II, YE.II and YS.II must all be of the same length. Using the notation of the JAGS manual, the prior distributions for theta and eta are given by
theta[i] ~ dnorm(theta.mu[i], theta.tau[i]),
eta[i] ~ dnorm(eta.mu[i], eta.tau[i]), for i = 1, 2, and
theta[i] ~ dlnorm(theta.mu[i], theta.tau[i]),
eta[i] ~ dlnorm(eta.mu[i], eta.tau[i]), for i = 3, 4.

Value

A simulation model object of class sim.model, created using the data supplied as the arguments and the JAGS model file ‘normal_model_jags_model.R’.

Author(s)

Sebastian Jobj<f6>rnsson [e-mail]

See Also

create.normal.model.from.file

bdpopt documentation built on May 2, 2019, 9:18 a.m.