reliability is a relative concept .. so it depends on the methodology and avoidance of bias which is very difficult to be achieved
Yes in case of including only the large and high quality randomized control trials
But most of the meta-analyses fail to give a clear message or mandate about the proposed subject and just end up saying more RCS, large and multicentric trials are required!!!
Meta-analyses in general are more indicative than reliable ( which is a relative concept as mentioned by the first responder ). One still has to look carefully at the methodological details and of course at the limitations declared by the authors themselvs.
Evidence based medicine depends on three parts: RCT-Meta/system review, patients need, and doctors' clinical experience and judgement.
The evidence strong or not through Meta is risky sometime, because Meta depends on RCTs, if the info from RCTs have bias or other methodology issues, it will cause Meta with problem.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis