Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: proportion meta-analysis


From   salil deo <salildeo@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: proportion meta-analysis
Date   Mon, 1 Apr 2013 09:14:36 +0530

Hi ,

I have recently conducted a metaanalysis using state ...

if only one arm reports zero rate of an event, you can use the Bates
correction , you can find the reference for that on the net , this
correction involves using 0.5 as the events in the arm with zero so
that a risk ratio or odds ratio can be calculated , actually state
will do that for you if you are going to enter the data as events,
non-events ... the command for that is :

metan Estudy NE study Econtrol NEcontrol, randomi ... and then
followed by all the other commands for the forest plot


if you want to convert the risk ratio or odds ratio to log and SE and
then use those values as the raw data , then use the 0.5 correction
mentioned..

for preparing the forest plot ... you can set the limits of the x-axis
using the command xlab(0,1,10,100) etc. hence you can choose how you
want to prepare the plot

if got all this material from the state journal ... it is available
free on the net ... search for what you want and you will able to
download the pdf ... really helpful .. gives a full list of commands
with examples of how to analyze and prepare forest plots

salil


On 4/1/13, MOHAMMED HASSAN AL-TEMIMI <M.H.temimi@utah.edu> wrote:
> Hello,
>
> I am conducting meta-analysis of case series to identify the rate of
> complications following a surgical procedure that can be conducted in two
> different ways (laparoscopic vs. open). I have the following questions:
> 1.	Some of the case series reported zero rate of complications for the
> laparoscopic procedure. When I am running the analysis using 'metaan'
> command and random effect model, the analysis does not go through because
> the standard error for the case series with 'zero' rate is 'zero'. How can I
> include the result from the latter case series (with zero rate of
> complication) in the meta-analysis?
> 2.	In the other arm (open group), when I do the forest plot, the lower bound
> of the confidence interval is less than zero in some instances, how can I
> truncate the lower bound in the forest plot to zero?
>
> Thanks,
> MHT
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
>
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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