GRADE in Cochrane systematic reviews: An investigation

Article type
Authors
Xiao J1, Zhen H1, Chen Y1, Yuan J1, Ke W1, Li Y1, Li M1, Zhao X1, Li X1, Liu Y1
1Evidence-Based Medicine Centre, Lanzhou University, China
Abstract
Background: GRADE system aims to evaluate the quality of evidence.

Objectives: To assess the use of GRADE in Cochrane systematic reviews (CSRs).

Methods: We searched The Cochrane Library (2011, 3 issue) for CSRs. Two reviewers independently selected CSRs which used GRADE. Then the data was extracted and analysised.

Results: In total 2,169 CSRs were searched. 307 CSRs have been identified using GRADE to evaluate the quality of evidence. Most (81.11% 249/307) reviews were published after 2008, published by the Cochrane Musculoskeletal Group, Cochrane Neuromuscular Disease Group and Cochrane Schizophrenia Group. There were a total 1,213 outcomes which were assessed by GRADE. Up to 14.59% (177/1,213) outcomes were rated as high quality; 33.8% (410/1,213) outcomes were rated as moderta quality; 28.20% (342/1,213) outcomes were rated as low quality; 12.12% (147/1,213) were rated as very low quality.The outcomes were downgraded mainly because study limitations and imprecision.

Conclusions: Nowadays, few CSR uses GRADE system. Study limitations was the main factor which might decrease the quality of evidence.