HELP! confidence intervals for meta-analysis STATA v16

[Jae Song]

I am trying to do a meta-analysis and calculated odds ratios and Upper and Lower 95% CI. On STATA v16, I go to the "setup" tab for meta-analysis, selection "Declare generic, precomputed effect sizes" and select my precalculated "odds ratio" for the Effect Size. Then I click, "CI variables." In the dropdown menu, I select "LowCI" and "UppCI" variables for the Lower and Upper tabs. However, under "CI symmetry tolerance" I'm not sure what to type in. When I try to run the analysis, I get an error saying they are not "normal" values. So if I go back and type in "1," for example, the analysis runs but the calculated upper and lower 95% CI are not shown in the forest plot. I'm really confused.

How can I get either (1) my precalculated Upper and Lower CIs to show in my forest plot? And (2) What do I type in for "CI symmetry tolerance" in the selection box in the STATA v16 meta-analysis function?

[Jae Song]

To rephrase, my question is, in my forest plot, I’d like to have the ODDS RATIOS plotted, not the log-odds ratios (and similarly the actual 95% Cis not, the log-95%CIs). How can I do this?

[Houssein Assaad (StataCorp)]

Hi Jae, If you can share your data or a portion of it, I would be happy to help solve your problem. Basically, whenever you are working with per-computed odds ratios (or risk ratios, hazard ratios, etc.), you have to provide your effect size and CI variables in the log metric. That is, in the control panel, under the -Setup- tab, you have to provide variables that contain log odds-ratios (NOT odds ratios) and CIs for the log odds-ratios. You can generate these variables for example by typing:

. generate double logor = log(<or_var>)
. generate double cil     = log(<or_cilvar>)
. generate double ciu    = log(<or_ciuvar>)

where you need to replace <or_var>, <or_cilvar>, and <or_ciuvar> by the names of the variables containing odds ratios, lower CI limit, and upper CI limit in your data set. You can then use variables, logor, cil, and ciu in the -Setup- tab. Alternatively, you can type:

. meta set logor cil ciu

The CI defined by variables cil and ciu should be symmetric. That is logor - cil should be almost the same as ciu - logor. If you have checked that the CIs are now symmetric but the above command produces an error, then you may relax the CI symmetry tolerance under the Setup tab (try values like .01 or .1 before you try 1). Alternatively, you can type:

. meta set logor cil ciu, civartolerance(.1)

If you type help meta set, check the last example in the help file that walks you through the same steps I did here (the ex works with risk ratios instead of odds ratios but the concept is the same). Finally, to construct a forest plot for the odds ratios, you can use option eform() in meta forestplot:

. meta forestplot, eform(Odds ratios)

I hope this solves your problem. If you still have an issue, please check the description of option civartolereance() in the documentation of the meta set command. Also, feel free to e-mail our tech support and they would be happy to help you further.

[yasar sattar]

Hi Jae, If you can share your data or a portion of it, I would be happy to help solve your problem. Basically, whenever you are working with per-computed odds ratios (or risk ratios, hazard ratios, etc.), you have to provide your effect size and CI variables in the log metric. That is, in the control panel, under the -Setup- tab, you have to provide variables that contain log odds-ratios (NOT odds ratios) and CIs for the log odds-ratios. You can generate these variables for example by typing:

. generate double logor = log(<or_var>)
. generate double cil = log(<or_cilvar>)
. generate double ciu = log(<or_ciuvar>)

where you need to replace <or_var>, <or_cilvar>, and <or_ciuvar> by the names of the variables containing odds ratios, lower CI limit, and upper CI limit in your data set. You can then use variables, logor, cil, and ciu in the -Setup- tab. Alternatively, you can type:

. meta set logor cil ciu

The CI defined by variables cil and ciu should be symmetric. That is logor - cil should be almost the same as ciu - logor. If you have checked that the CIs are now symmetric but the above command produces an error, then you may relax the CI symmetry tolerance under the Setup tab (try values like .01 or .1 before you try 1). Alternatively, you can type:

. meta set logor cil ciu, civartolerance(.1)

If you type help meta set, check the last example in the help file that walks you through the same steps I did here (the ex works with risk ratios instead of odds ratios but the concept is the same). Finally, to construct a forest plot for the odds ratios, you can use option eform() in meta forestplot:

. meta forestplot, eform(Odds ratios)

I hope this solves your problem. If you still have an issue, please check the description of option civartolereance() in the documentation of the meta set command. Also, feel free to e-mail our tech support and they would be happy to help you further.

Hi Houssein,
I have similar issue, my data have predefined or ciu cil, when I put code "meta set or cil ciu" I get this error: effect size variable or must be <= CI upper limit, ciu r(459); please help me I checked my data multiple times. I will be waiting for reply.

[yasar sattar]

my data only have odds ratio, uci, lci, so as per my understanding I have to use meta set or ciu cil, and the state will generate se for me, and then i will be able to make forest plot. but the above error I mentioned I dnt know how to fix.

[Sarah Marchina]

Hi Houssein, I have exactly the same problem as Jae. I followed your guidance which worked out perfectly and I was able to produce the forest plot with Odds ratios, however, I have subgroup analyses and when I enter the last bit of code (eform(Odds ratios) I get only the overall group, but not the subgroups. I would like to mention that I am NOT familiar with coding hence I was hoping to use the GUI, but that obviously doesn't work. My questions are: What code would I have to use to be able to get a forest plot with 2 or more subgroups (after going through your steps above) Also, how do I add a reference line in a plot? (works smoothly with GUI, but no idea about coding it) I tried to do add it in the actual plot, but the function "add reference line" is always greyed out and I cannot use it. Please let me know if you need more information to help me.

[Barsha Gadapani]

HELP! confidence intervals for meta-analysis STATA v16

I am trying to do a meta-analysis by using precomputed effect size. I initially generated plots using a data set I had. However, after a recent re-installation of Stata v16 I am getting error messages i.e. "confidence intervals not symmetric." I used the following command: meta set es lci uci ( where es= effect size, uci/lci= upper and lower confidence interval) but is followed by this message i.e. "CIs defined by variables lci and uci must be symmetric and based on a normal distribution. If you are working with effect sizes such as odds ratios, risk ratios, or hazard ratios, the CIs should be specified on the log scale." I followed the instruction given in the Stata document to resolve the issue by using commands: generate double relative_diff= reldiff(es-lci, uci-es) followed by summarizing relative difference. Then got the relative difference value ranging from (0-0.0033). I then tried this i.e., .meta set es lci uci,civartolerance(0.0033), however ,the following command pops up again and again "option civartolerance not allowed". Can you please help me in solving this issue and generating the plots.? I am sharing a plot created with the same data set (which now shows the CI error message).

Thanks and regards
Barsha
plots for meta-analysis_precomputed data.docx