In assessing the quality of evidence using the GRADE approach, with regards to inconsistency/heterogeneity domain, how to know that I should downgrade the quality of a body of evidence by 1 level or 2 level?
In other words, how would I determine that the inconsistency is:
- 'not' serious (no change in quality,
- 'serious' (downgrade quality of evidence 1 level), or
- 'very serious' (downgrade quality of evidence 2 level)?
As with most things evidence-based medicine, this is a judgment call.
The GRADE Handbook (Section 5.2.2) and the JCE publication on GRADE and rating down for Inconsistency (both links below) provide guidance and examples:
https://gdt.gradepro.org/app/handbook/handbook.html#h.g2dqzi9je57e
https://www.jclinepi.com/article/S0895-4356(11)00182-X/fulltext
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis