nodesplit_summary-class function

The nodesplit_summary class

The nodesplit_summary class contains posterior summary statistics for node-splitting models, as a result of calling summary() on a nma_nodesplit or nma_nodesplit_df object.

Details

Objects of class nodesplit_summary are tibble data frames, with one row for each node-split comparison and columns:

• trt1, trt2: Treatments forming the comparison
• summary: A list column containing nma_summary objects with the posterior summaries and draws for each of the node-splitting parameters
• p_value: Bayesian p-value for inconsistency
• dic: A list column containing nma_dic objects, giving the model fit statistics

The parameters included in summary are:

• d_net: Network estimate from the corresponding consistency model, if available
• d_dir: Direct estimate from the node-splitting model
• d_ind: Indirect estimate from the node-splitting model
• omega: Inconsistency factor $\omega = d_dir - d_ind$
• tau: Heterogeneity standard deviation from the node-splitting model, if a random effects model was fitted
• tau_consistency: Heterogeneity standard deviation from the corresponding consistency model, if available and if a random effects model was fitted