Helper function for the connect_baseline argument of nma() to specify how
study baselines (intercepts) are linked, to connect a disconnected network or
single-arm studies.
Usage
con(
type = c("fixed", "random"),
studies,
prior_baseline = NULL,
baseline_trt = NULL
)Arguments
- type
Type of connection, either "fixed" or "random".
- studies
Vector of study names.
- prior_baseline
Prior distribution for the baseline parameters of
studieswhentype = "random", as a nma_prior object (see priors).- baseline_trt
Which treatment the baseline prior distribution applies to, when
type = "random". By default, the random baseline prior will be placed on the network reference treatment. This should be changed if, for example, the prior was obtained from a baseline synthesis of a different treatment.
Details
A random baseline connection is specified with type = "random".
This places an informative prior distribution on the baselines of studies
listed in studies, specified with the prior_baseline argument.
A fixed baseline connection is specified with type = "fixed". This shares
the baseline parameter for studies listed in studies.