Dear All,
I am conducting a meta-analysis and I'm being given the trials, number of patients on each trial, mean, standard deviation and sampling number for both treatment and control. I also collected some background information for each trial. these information include both categorical factors and continuous factors such as experimental position and experimental duration. I would like to use a mixed model in our analysis to estimate the total mean effect size, and test the statistical significance of each explanatory factor. I will very much appreciate if you are able to guide me through on how write this SAS code. Thanks.
Here is a template of my dataset:
trial exp_posit exp_durat mean_of_treat stand_dev_of_treat numb_on_treat mean_of_contr stand_dev_of_contr numb_on_contr
1 Germany 3 44.14 8.33 5 47.93 10.89 5
2 USA 5 80.14 25.62 5 64.14 7.68 5
3 USA 16 140.18 39.2 8 144.14 73.06 8
4 Germany 4 35.53 9.19 8 37.12 12.67 8
5 USA 8 8.94 5.31 3 9.87 3.95 3
6 UK 12 19.43 16.32 3 10.92 8.83 3
7 Germany 4 20.86 2.65 4 0.71 0.81 4
8 UK 15 23.16 11.47 3 49.57 24.26 3
9 Germany 9 25.64 21.61 7 32.02 20.29 7
10 UK 7 10.5 7.93 3 8.56 3.97 3
Kind regards,
Roy
You should read:
http://apsjournals.apsnet.org/doi/abs/10.1094/PHYTO-03-10-0069
This article is now open access. The on-line supplement has sas code and instructions.
http://apsjournals.apsnet.org/doi/suppl/10.1094/PHYTO-03-10-0069
Also, I would like to estimate the treatment effect by calculating the natural log of the response ratio (RR), a metric commonly used in meta-analyse.
RR = ln(mean_of_treat/mean_of_contr) = ln(mean_of_treat) –ln(mean_of_contr)
If you don't know enough to write your code I would very HIGHLY suggest hiring a statistician.
You'll get some help here, but you need someone who understands the intracacies of your data and the analysis you're trying to do.
For sample code and methodologies search on LexJansen.com
http://www.lexjansen.com/search/searchresults.php?q=meta%20analyis
You should read:
http://apsjournals.apsnet.org/doi/abs/10.1094/PHYTO-03-10-0069
This article is now open access. The on-line supplement has sas code and instructions.
http://apsjournals.apsnet.org/doi/suppl/10.1094/PHYTO-03-10-0069
I fixed a link in my original post. THe one I took out was for a more complex problem. This one should work for you to learn the topic.
Thanks, Ivm. I have successfully applied your SAS code in my meta-analysis.
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.
Find more tutorials on the SAS Users YouTube channel.