Box 7 Quantitative and qualitative research evidence synthesis—assessment criteria for the level of evidence.

1 An initial level of evidence was assigned to a group of studies that addressed a particular information point as shown
Quantitative studies 46	Qualitative studies
High = randomised trial	High = checklist quality score “++”
Low = observational study	Low = checklist quality score “+”
Very low = any other evidence	Very low = checklist quality score “−“
The initial level of evidence assigned was based on the lowest hierachical type of evidence (ie, study design) in the group of studies.	The initial level of evidence assigned was based on the lowest checklist quality score of any study in the group.*
*The use of the lowest checklist quality score enabled any uncertainty in the quality of the available evidence (in conduct, reporting or both) to be incorporated in the initial level of evidence assigned.
2 The initial level of evidence was modified into one of four levels (high, medium, low and very low) according to several additional considerations
Quantitative studies46	Qualitative studies
Decrease level of evidence if:	Decrease level of evidence if:
Serious (−1) or very serious (−2) limitation to study quality
Important inconsistency (−1)	Important inconsistency (−1)
Some (−1) or major (−2) uncertainty about directness	Some (−1) or major (−2) uncertainty about directness
Imprecise or sparse data (−1)
High probability of reporting bias (−1)
Increase level of evidence if:	Increase level of evidence if:
Strong evidence of association (+1)	Close conformity of findings based on two or more studies rated as ++, directly applicable to the target population and with no major threats to validity (+1)
Very strong evidence of association (+2)
Evidence of a dose response gradient (+1)
All plausible confounders would have reduced the effect (+1)