MBNMAdose provides a collection of useful commands that allow users to run dose-response
Model-Based Network Meta-Analyses (MBNMA).
MBNMAdose allows meta-analysis of studies that compare multiple doses of different agents in a way that can
account for the dose-response relationship.
Whilst making use of all the available evidence in a statistically robust and biologically plausible framework, this also can help connect networks at the agent level that may otherwise be disconnected at the dose/treatment level, and help improve precision of estimates. It avoids "lumping" of doses that is often done in standard Network Meta-Analysis (NMA). All models and analyses are implemented in a Bayesian framework, following an extension of the standard NMA methodology presented by \insertCitelu2004MBNMAdose and are run in \insertCitejags;textualMBNMAdose. For full details of dose-response MBNMA methodology see \insertCitemawdsley2016;textualMBNMAdose. Within this package we refer to a treatment as a specific dose or a specific agent.
MBNMAdose follow a clear pattern of use:
Load your data into the correct format using
Analyse your data using
mbnma.run(), or any of the available wrapper dose-response functions
Test for consistency at the treatment-level using functions like
Examine model results using forest plots and treatment rankings
Use your model to predict responses using
At each of these stages there are a number of informative plots that can be generated to help understand the data and to make decisions regarding model fitting.
Maintainer: Hugo Pedder firstname.lastname@example.org
Lujin Li [reviewer]
Adil Karim [contributor]
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 26 27 28 29 30 31 32 33 34 35 36
# Generate an "mbnma.network" object that stores data in the correct format network <- mbnma.network(HF2PPITT) # Generate a network plot at the dose/treatment level plot(network, level="treatment") # Generate a network plot at the agent level plot(network, level="agent", remove.loops=TRUE) # Perform "split" NMA to examine dose-response relationship nma <- nma.run(network) plot(nma) # Analyse data using mbnma.run() result <- mbnma.run(network, fun="emax", beta.1="rel", beta.2="rel", method="common") # ...or achieve the same result by using a wrapper function for mbnma.run() result <- mbnma.emax(network, emax="rel", ed50="rel", method="common") # Generate forest plots for model results plot(result) # Rank results and plot rankograms ranks <- rank(result) plot(ranks, params="d.emax") # Predict responses pred <- predict(result, E0=0.2) # Plot predicted response with "split" NMA results displayed plot(pred, disp.obs=TRUE, network=network)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.