Respected Researchers,
I am conducting a meta-analysis with respect to pediatric brain trauma, a population with limited literature. In an attempt to gather a large sample of patients, instead of sticking to one outcome measure, GOS/GOS-E, I am also including patients wherein other measures have been reported as well such as CAFAS, BRIEF, VMI etc.
Is it possible to relate all these outcome measures within a single meta-analysis with the sole purpose of identifying whether such patients have a good outcome or bad outcome? If so, can you please tell me the statistical methods required to do this?
Thank you very much to all.
Raj Lavadi , in the absence of a "conversion system/factor" across the outcome measures, I believe you only have few options to get around this, among them:
1. Generate several forest plots, one for each outcome measure.
2. Get in touch with the authors of your included studies to provide you with their raw datasets that led to their determination, so that you can produce a corresponding value for each enrolled patient according to an outcome measure/scale of interest
Aedrian Abrilla , thank you for your response! I will try both options. Hopefully this works out!
In meta-analysis we can compare multiple outcome ,by regression with random effect approch.
Dr. Alok Kumar Bharti , I was not aware of this. Thank you for your answer!
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis