Error message when I conduct Meta analysis

[Nandin-Erdene Bayart]

Dear all

I am having problems with conducting meta-analysis on STATA/SE 16.1 software (My sample data is at the bottom end). The Meta-analysis function doesn't work on my computer. Here are the example data which I am sharing with you now. Can you please help me with this analysis. When I use the functions; this is the syntax displayed:

"meta set LnES_conv LnLo_CI_conv LnHi_CI_conv"

and I get the following error:

confidence intervals not symmetric
CIs defined by variables LnLo_CI_conv and LnHi_CI_conv must be symmetric and based on a normal
distribution. If you are working with effect sizes in the original metric, such as odds ratios
or hazard ratios, with meta set, you should specify the effect sizes and CIs in a normalizing
metric, such as the log metric.

The default tolerance to determine the CI asymmetry is 1e-6. Effect sizes and their CIs are
often reported with limited precision that, after the normalizing transformation, may lead to
asymmetric CIs. In that case, the default of 1e-6 may be too stringent. You may loosen the
tolerance by specifying option civartolerance().

* Example generated by -dataex-. To install: ssc install dataex
clear
input str5 Author int Year float(LnES_conv LnLo_CI_conv LnHi_CI_conv _LCI _UCI _WT)
"Ban "  2021  .007372754   .0011992806     .01350835    1.0012    1.0136    3.811794
"Cai"   2018  .020508876     .00040008     .04144719 1.0004002 1.0423181   1.7498177
"Cai"   2018   .02091725   .0016012813     .04103035 1.0016025 1.0418837   1.8340706
"Cai"   2018  .024907636    .003205131     .04751098 1.0032103 1.0486578    1.593459
"Chen"  2017  .002297359   .0012991558    .003394233    1.0013    1.0034   4.2977204
"Chen"  2018 .0012991558  -.0012007206    .003294567     .9988    1.0033   4.2396116
"Chen"  2021  .002596626  -.0009004052     .00608147     .9991    1.0061    4.138653
"Chen"  2021  -.00060018   -.004008021   .0026963616      .996    1.0027   4.1518097
"Chen"  2021 .0011992806  -.0027036516    .005186527     .9973    1.0052    4.092524
"Dong"  2018 .0022774048 -.00005000125   .0046093604    .99995   1.00462    4.234175
"Luo"   2016 .0041912044    .002297359    .006180859    1.0023    1.0062   4.2582693
"Qian"  2019           0   -.012512188    .013269145  .9875658 1.0133575    2.733701
"Qian"  2019  .003414246   -.007038275    .014858223  .9929864 1.0149691    3.044497
"Qian"  2019  .010046364   -.005251587    .024096886  .9947622 1.0243895    2.466378
"Wu"    2019 -.012477522   -.024395157 -.00050012505     .9759     .9995    2.882573
"Wu"    2019   .00039992   -.011263193     .01212618     .9888    1.0122    2.923192
"Wu"    2019 -.008939842    -.02296161    .004987542     .9773     1.005   2.5687926
"Xie"   2015   .00249688   .0009995003    .003992021     1.001     1.004   4.2807846
"Xie"   2015           0  -.0013008458   .0012991558     .9987    1.0013   4.2889104
"Xie"   2015  .003394233    .001399021    .005286004    1.0014    1.0053     4.25817
"Xu"    2020 .0026963616    .000099995    .005186527    1.0001    1.0052   4.2191806
"Xu"    2020  .001399021   -.001501126   .0041912044     .9985    1.0042    4.195874
"Yan"   2021  .009313234    .007518191    .011111504 1.0075465 1.0111735   4.2662163
"Yan"   2021  .007165612    .005404579    .008929752 1.0054193 1.0089698    4.267997
"Yan"   2021   .00992913     .00798177     .01189041 1.0080137 1.0119613    4.257553
"Zhang" 2018 .0026963616   .0011992806   .0042907814    1.0012    1.0043    4.278561
"Zhang" 2018  .023716526     -3.772261     .02566775      .023     1.026 .0003438366
"Zhang" 2018  .023716526    .022739487     .02469261     1.023     1.025   4.2998614
"Zhou"  2022   .01207258   -.003095205     .02767957  .9969096 1.0280662   2.3655088
end

[Weiwen Ng]

I'm not familiar with these new meta commands, but the error message is self explanatory: somewhere in your data, there's a study where the log effect size appears not to be the midpoint between the log of the lower and upper CIs. Entering your data into Excel, I have a feeling the study is the second version of Zhang 2018, i.e. the 3rd study from the bottom.

Is there a data entry error for that study, particularly for the log effect size? It looks like a possible copy-paste error when you look at the Zhang 2018 directly below it (i.e. the 3rd study that Zhang et al did in 2018, 2nd from bottom). I have a feeling that you may have intended to enter the log effect size as 0.15361641 for Zhang 2018 #2.