Giters

gertvv / gemtc

GeMTC R package: model generation for network meta-analysis

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

UME: baselines in RE data not handled correctly

gertvv opened this issue · comments

commented

When the baseline in an RE study is not the lexicographically first treatment in that study, this may cause trouble in the UME model and consequently in the ANOHE analysis:

  • comparisons may be duplicated in the basic parameters (e.g. both d.A.C and d.C.A)
  • removing the duplicates also removes them from the RE matrix, which means the model can't be estimated
  • reversed parameters aren't handled properly by the ANOHE code, leading to missing values in the I^2 statistics