Certainty assessment | No of patients | Certainty | Importance |
---|
№ of studies | Study design | Risk of bias2 | Inconsistency3 | Indirectness | Imprecision5 | Publication bias6 | Participants, n | Cases, n |
---|
20 | observational studies | not serious | very serious | not serious | not serious | none | 6071 | 3584 | ⨁◯◯◯ Very low | CRITICAL |
- 1Grading of Recommendations Assessment, Development and Evaluation
- 2Risk of bias based on Newcastle-Ottawa Scale
- 3When I2 was < 30% inconsistency considered as Not serious limitation, > 50 considered as serious and more than 75% considered as very serious limitation
- 5Serious limitations when there was fewer than 400 participants for each outcome and very serious limitations when there was fewer than 300 participants for each outcome
- 6Funnel plot revealed no asymmetry; neither test of publication bias approached P < 0.10