Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: difficulty using META command and exporting output in Stata


From   Roger Harbord <Roger.Harbord@bristol.ac.uk>
To   statalist <statalist@hsphsun2.harvard.edu>, "Steichen, Thomas" <STEICHT@RJRT.com>
Subject   st: RE: difficulty using META command and exporting output in Stata
Date   Thu, 17 Oct 2002 12:49:16 +0100

--On 16 October 2002 15:26 -0400 "Steichen, Thomas" <STEICHT@RJRT.com> wrote:


John Hopkins writes (in part):

I am working on a meta analysis of studies looking at the
risk of cancer in asthma patients.  We are using the META
command to generate our summary estimates (we have adjusted
odds ratios, with standard errors).  I am trying to take
pooled estimate and confidence limits out of the table as
variables, without manually typing them in.

Here is what I typed into stata to generate the output
tables: meta loggeneff seloggeneff, eform graph(f) cline
xline(1) xlab(.1,1,10)
id(studygen) b2title(gender-specific estimates) print

Here are the output tables:

Meta-analysis (exponential form)

       |  Pooled      95% CI         Asymptotic      No. of
Method |     Est   Lower   Upper  z_value  p_value   studies
-------+----------------------------------------------------
Fixed  |   1.068   1.038   1.099    4.489    0.000      2
Random |   1.073   0.989   1.165    1.689    0.091

Test for heterogeneity: Q=  8.025 on 1 degrees of freedom (p=
0.005) Moment-based estimate of between studies variance =  0.003

                   |      Weights      Study       95% CI
             Study |   Fixed  Random     Est   Lower   Upper
-------------------+----------------------------------------
  Vesterinen Males | 2013.93  280.27    1.12    1.07    1.17
Vesterinen Females | 2646.46  289.91    1.03    0.99    1.07


How can I save the summary estimate and confidence limits
from the analysis (random effects) to variables?  I need to
take these values and do some data management and additional
analysis with them, and I do not want to hand-enter every
estimate generated from the command.

The help file for -meta- indicates that it saves the
following results (copied directly from the help file):

Saved values
------------

S_1    Theta (fixed)
S_2    SE Theta (fixed)
S_3    Lower CI Limit (fixed)
S_4    Upper CI Limit (fixed)
S_5    Asymptotic Z-value (fixed)
S_6    Asymptotic p-value (fixed)
S_7    Theta (random)
S_8    SE Theta (random)
S_9    Lower CI Limit (random)
S_10   Upper CI Limit (random)
S_11   Asymptotic Z-value (random)
S_12   Asymptotic p-value (random)
S_13   Between-studies Variance, tau^2

<snip>

Which version of -meta- do you have, or maybe more to the point, which version of the help file do you have and where did you get it from? The above doesn't appear in the STB43 (sbe16.2) version:

. net from "http://www.stata.com/stb/stb43/";
Click on sbe16_2
In the Viewer, click on sbe16_2/meta.hlp

- there's nothing about Saved values. The above global macros are created by the sbe16.2 version of -meta- though (except I get S_0 rather than S_10).

I also just noticed that the v7 Stata manuals have an entry for [R] meta, which lists the various user-written programs. However, if you slavishly follow the example instructions given for installing
-meta-, you'll end up with an out-of-date version of -metap- instead!

Roger.
----------------------------------------------------
Roger Harbord mailto:roger.harbord@bristol.ac.uk
Department of Social Medicine, University of Bristol


*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index