hugaped/MBNMAdose: MBNMAdose: Dose-response Model-Based Network Meta-Analysis Version 0.4.2

Fits Bayesian dose-response model-based network meta-analysis (MBNMA) that incorporate multiple doses within an agent by modelling different dose-response functions, as described by Mawdsley et al. (2016) doi:10.1002/psp4.12091. By modelling dose-response relationships this can connect networks of evidence that might otherwise be disconnected, and can improve precision on treatment estimates. Several common dose-response functions are provided; others may be added by the user. Various characteristics and assumptions can be flexibly added to the models, such as shared class effects, and there are a number of informative plots and outputs to explore model results and assumptions.

• Reference SDs can now be used when modelling using SMDs to avoid using study-specific SDs, which can be problematic.
• Network Meta-Regression: Effect modifiers can now be incorporated using regress.vars argument in <code>mbnma.run()</code>. Various sharing assumptions for effects can be specified in regress.effect.
• Predictions can be estimated for class effect models
• Fractional polynomial powers in <code>dfpoly()</code> can only take numeric values from set defined in Jansen 2015.
• Added <code>calc.edx()</code> to allow easy estimation of different ED values (e.g. ED90 = the dose at which 90% of the
• maximum response (Emax) is reached)
• <code>get.relative()</code> now allows simultaneous comparison of two models in a single league table - can be used to compare MBNMA models with different dose-response functions, or MBNMA and NMA models, or NMA models that assume consistency versus those that use Unrelated Mean Effects.
• Plots of predictions look prettier
• Dose-response parameters that were previously modelled on an exponential scale (ed50, hill, onset) are now on the natural scale and are assigned truncated normal default priors
• Separate prior distributions can be specified for different indices of a parameter - allows for agent-specific prior distributions on dose-response parameters.