Im doing a meta-analysis for the estimation of delays in treatment seeking at different levels for neonatal morbidity, The analysis is region-wise e.g. sub-Saharan Africa, south Asia, Latin America and Caribbean. Three levels of delay are considered.this would result in 9 forest plots. How can i combine the results of all three levels of delay into one figure (region-wise)?. This would result in just 3 plots which will be compact and easier to comprehend. Kindly help
You can use R with the library 'meta'. When typing the command line to create the forest plot, enter the option "byvar = x". "x" is the stratification variable. For example:
forest(your.meta.object, byvar = your.dataset$region)
Depends on what software you are using and if it has the ability to create subgroups
You can use R with the library 'meta'. When typing the command line to create the forest plot, enter the option "byvar = x". "x" is the stratification variable. For example:
forest(your.meta.object, byvar = your.dataset$region)
How about using stem-and-leaf plots? In my recent meta-analysis publication at Public Health Nutrition journal, I tried 5 different outcome measures and plotted against each other.
Use Review Manager and make subgroups. Easy-peasy. Download RevMan for free from this site: http://ims.cochrane.org/revman.
in Review Manager 5, add use subgroups for each level of delay under each region. You can then choose to calculate the measure for subtotals only or for subtotals and total. This would make a lot of sense by the way.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology