Open schuemie opened 4 months ago
preliminary steps might be as straight-forward as reusing the current EvidenceSynthesis
meta-analytic tools given likelihood-profiles from Poisson (incidence) and logistic (prevalence) Cyclops
models
A (developing)(best) practice in OHDSI is to provide the prediction interval, not just the confidence / credible interval around mu
Currently this package only supports synthesis of causal effect estimates. We could extend to support synthesis of incidence (and prevalence) rates (or proportions).