mnma.model: Create a data object for multivariate and univariate network...

In vandy10s/mnma: Multivariate Network Meta-Analysis (NMA) within a Bayesian Framework

Description

The function generates a data object for mnma.run or unma.run operation, using outcome 1, outcome 2, and treatments inputs

Usage

mnma.model(outcome1, outcome2, TX, noa = "3arm", ref = 1,
  parameters = "simple")

Arguments

outcome1	A data frame for outcome 1. First 2 (or 3) columns contain the information about number of events for each arm and the rest 2 (or 3) columns contain the sample size for each arm. Number of columns ranges from 2 to 3, depending on the number of maximum arms in the trial data.
outcome2	A data frame for outcome 2, which has the same dimensions as outcome1.
TX	A data frame for tratments, which has maximum 3 treatment arms in 3 columns. The number of rows should coincide with outcome1 and outcome2. Each element needs to be numeric.
noa	Maximum number of arms in the trial studies, which is either 2arm or 3arm. Note that number of arm is different from number of total treatments in the network.
ref	Reference treatment in the analysis in a numeric class.
parameters	List of parameters to infer in the analysis. They are either simple or detail depending on comparison type.

Details

A simple list for parameters includes log odds ratio (comparison) parameters between reference and all other treatments. A detail list for parameters includes log odds ratio parameters between all treatments. Both cases include a correlation parameter and heterogeneity parameters for two outcomes.

Value

An object of class mnma.model that contains data information and parameter lists to estimate for multivariate and univariate network meta-analysis.

Examples

data(acute)
# Split the dataset to create a mnma data object
outcome1 <- acute[,c(1:6)]     # outcome1
outcome2 <- acute[,c(7:12)]    # outcome2
TX    <- acute[,13:15]
# Convert them into log odds ratio inputs
data <- mnma.model(outcome1,outcome2,TX, ref=1, parameters="simple")
data

vandy10s/mnma documentation built on May 19, 2019, 8:24 a.m.