Metan Command -- Risk Differences with whole numbers?

[Leo Martinez]

Hello!
I am using Stata 12.1.
I am running a code for a meta-analysis of risk differences and have been having trouble with the type of output I have been getting. I would like to have
a risk difference in the forest plots with more than two decimal places. This is because I get results such as 0.12 risk difference which is 12% but I would like a further decimal place or two in the graph so that I can get more clarity in my answer and to potential readers and reviewers of my analysis.

Here is my code for a risk difference with the metan command:

metan casepos caseneg controlpos controlsneg, lcols(author) xlabel(0, 0.20)  by(Agegroup)  random rd nooverall

After this I get the attached graph forest plot which shows only two decimal places.


Can I make this same graph but with 3 decimal places? Or alternatively (but similar idea) as whole numbers so instead of 0.11 I have 11.1 or instead of 0.23 I have 23.1?

Thanks for any help you can provide in advance!

Leo

[David Fisher]

Dear Leo,

Looking at the source code for metan, it looks like there is an undocumented option dp() which will do this for you. The default is 2 (as you can see from your plot), so adding dp(3) to your options should give you want you want.

I can't immediately think of a way to get metan to return risk differences scaled by 100 -- and since RDs are not proportions/percentages, wouldn't that alter the entire analysis?

Thanks,

David,

[Leo Martinez]

Thanks so much David! The dp() command did work and somehow I did not see this when I was going over the tutorials for the metan command.

Thanks again!
Leo

[Brian ER]

Hi all,
I'm using Stata 17
how to change the effect size from proportion meta-analysis to percent format?

[Leonardo Guizzetti]

Hi all,
I'm using Stata 17
how to change the effect size from proportion meta-analysis to percent format?

Please don't spam the same question, especially when posting it to thread asking a different question.

[KHALID MEDANI]

Leo, Your question helped me answer my question too. Thank you. I see you said you watched metan tutorials , were those free? if yes , do you mind share the link for them ?
I really appreciate that .