What it is
Many treatments, one connected network, one ranking.
A pairwise meta-analysis answers A versus B. A network meta-analysis joins every trial into one network, so it can compare A, B, C, and D at once, estimate comparisons no trial ever made, and rank the options by effect and by certainty.
What we run
Frequentist and Bayesian, netmeta in R, or gemtc with informative priors when the network is sparse.
Dose and class level, networks by dose, by drug, or by class, and component analyses for combined treatments.
Ranking with honesty, SUCRA and P-scores reported with their intervals, never as a league table alone.
Consistency checks, transitivity assessed, inconsistency tested by node-splitting, certainty rated with CINeMA.
