search

dmphillippo / multinma

Network meta-analysis of individual and aggregate data in Stan
https://dmphillippo.github.io/multinma
33 stars 15 forks source link

MD or SMD #8

Closed softloud closed 2 years ago

softloud commented 2 years ago

Thank you for this wonderful package and all the nma tools you created for it.

I'm sure I must've missed something really obvious, does the multinma::nma for continuous data produce estimates of mean difference or standardised mean difference?

dmphillippo commented 2 years ago

Hi @softloud, no conversion is done for continuous outcomes, so the results will be on the same scale as the input data (with an identity link function, or on the log scale for a log link). Does that answer your question?

softloud commented 2 years ago

Thanks for responding. That makes sense for ::set_agd_contrast. I've been using ::set_agd_arm. Now :thinking: I should be using ::set_agd_contrast. I've input means and standard deviations to ::set_agd_arm, what measure does the network output when passed through ::nma? If I set the treatment reference to placebo, what difference are the estimates?

softloud commented 2 years ago

I figured it out -- ::set_agd_arm estimates with mean difference :)