Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Forest plot with p from overall z test

From   "Roger B. Newson" <>
Subject   Re: st: Forest plot with p from overall z test
Date   Thu, 03 Nov 2011 22:01:16 +0000

An unpublicized feature of -metan- is that it saves a series of results in -r()-, whose names give an idea regarding what they are.

For instance, if you type in Stata

metan tdeath tnodeath cdeath cnodeath, rd random label(namevar=id, yearid=year) counts;
return list;

then Stata responds with a list of results in r() as follows:

               r(cger) =  .4864029666254636
               r(tger) =  .5012391573729864
               r(tau2) =  .0047751229726653
             r(p_chi2) =  .
               r(chi2) =  .
              r(p_het) =  .0193742235231943
                 r(df) =  19
                r(het) =  33.80513003131778
               r(i_sq) =  43.79551274496503
                r(p_z) =  .260512232288569
                  r(z) =  1.125181260695625
             r(ci_upp) =  .0743049097225629
             r(ci_low) =  -.0201055217834572
               r(seES) =  .0240847363142174
                 r(ES) =  .0270996939695528

            r(measure) : "RD"

These results can be referenced by the user, and presumably added to the -metan- plot in a second execution of -metan-. I do not know why these useful results are not publicized in the on-line help for -metan-.

Another possibility might be to use the -metaparm- module of the -parmest- package, together with the -parmhet- and -eclplot- packages, instead of using -metan-. The -parmest-, -parmhet- and -eclplot- packages are all downloadable from SSC. Together, they should be able to do what -metan- does. However, they require more programming skill than -metan-, so most new users prefer to use -metan-.

I hope this helps.

Best wishes


Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Web page:
Departmental Web page:

Opinions expressed are those of the author, not of the institution.

On 03/11/2011 15:26, Frederico Moreira wrote:

I am a beginner in Stata programming.
I would like to know how(if it is possible) to produce a forest plot with the p-value from overall z test using  the metan routine. I read the help metan and the book Meta-Analysis in Stata:An updated collection from the stata journal but I did not find any indication if it is possible. Normally, this p-value appears in the summary table but not in the graph.
Thank you in advance,
Any help would be much appreciated.

Antonio Silva.

Share photos&  screenshots in seconds...
Works in all emails, instant messengers, blogs, forums and social networks.

*   For searches and help try:
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index