guido-s
commented
5 months ago This is no bug, but a problem with your data: several numbers of events are larger than the group sample sizes, e.g., in the LYK study.
You will get information on the calculation problems if you use argument warn = TRUE in pairwise().
Classification of issue
Please indicate whether you want to submit a
Summary
Was using the netmeta package to calculate treatment effect and its std error using the pairwise function on a long arm form and the most of the Te and seTe was not calculated and there was no error messages stating why they were not calculated.
Input for pairwise: Study treatment r N ZB 2014 DZ+WM 60 200 ZB 2014 WZ 83 200 LYK 2015 ZZ 48.4 32 LYK 2015 ZY 59.9 32 W 2015 ZZ 27.3 43 W 2015 WM 34.2 41 LC 2017 DZ+WM 71.58 53 LC 2017 WM 93.36 38 S 2018 DZ+WM 70.5 53 S 2018 WM 92.5 52 D 2018 ZZ+WM 26.11 32 D 2018 WM 34.31 32 H 2021 DZ+WM 72.6 44 H 2021 WM 117 44 C 2023 ZZ+WM 103.74 43 C 2023 WM 132.58 43
After pairwise function output: studlab treat1 treat2 TE seTE event1 n1 event2 n2 ZB 2014 DZ+WM WZ -0.5039645334 0.210724287 60 200 83 200 LYK 2015 ZZ ZY 48.4 32 59.9 32 W 2015 ZZ WM -1.062077042 0.5259495164 27.3 43 34.2 41 LC 2017 DZ+WM WM 71.58 53 93.36 38 S 2018 DZ+WM WM 70.5 53 92.5 52 D 2018 ZZ+WM WM 26.11 32 34.31 32 H 2021 DZ+WM WM 72.6 44 117 44 C 2023 ZZ+WM WM 103.74 43 132.58 43
Additional information
Silent error when not calculating Te and seTe using the pairwise function on a long arm form study
Environment