Article type
Year
Abstract
Background: Cochrane systematic reviews have a wide international audience and aim to facilitate healthcare decision-making by various stakeholders. Therefore, terminology used to describe the quality of evidence from Cochrane reviews should be consistent and understandable. Numerous approaches for rating quality of evidence exist. The Cochrane Collaboration currently recommends the GRADE approach which specifies four levels of quality: high, moderate, low and very low quality evidence.
Objectives: To evaluate the various approaches and terminology used to describe quality of evidence in Cochrane reviews.
Methods: We included new and updated Cochrane reviews published in the 2012 February and March issues of The Cochrane Library. Two reviewers independently extracted information on the approach and terminologies used to rate and describe quality of evidence. We compared findings and resolved disagreements by discussion.
Results: We included 155 reviews. Thirty percent (47/155) mentioned the approach used for rating quality of evidence. All of these used the GRADE approach. Forty-eight of the reviews included a GRADE summary of findings table, three of which were not in the standard format. Ten percent (16/155) of the reviews used GRADE terminology to describe the quality of evidence. The most frequently used non-GRADE terminologies were: ‘There is strong evidence’, ‘There is insufficient evidence’, ‘There is limited evidence’, ‘There is poor evidence’, and ‘There is good evidence’. In the review section on ‘Quality of evidence’, the authors mostly described the risk of bias in the included studies and did not refer to other factors that can influence the quality of evidence.
Conclusions: To reduce the potential for confusion by using various poorly defined terminologies, Cochrane authors should use a common and rational approach to rating and wording quality of evidence that can be understood and interpreted across readers. The GRADE system provides such an approach.
Objectives: To evaluate the various approaches and terminology used to describe quality of evidence in Cochrane reviews.
Methods: We included new and updated Cochrane reviews published in the 2012 February and March issues of The Cochrane Library. Two reviewers independently extracted information on the approach and terminologies used to rate and describe quality of evidence. We compared findings and resolved disagreements by discussion.
Results: We included 155 reviews. Thirty percent (47/155) mentioned the approach used for rating quality of evidence. All of these used the GRADE approach. Forty-eight of the reviews included a GRADE summary of findings table, three of which were not in the standard format. Ten percent (16/155) of the reviews used GRADE terminology to describe the quality of evidence. The most frequently used non-GRADE terminologies were: ‘There is strong evidence’, ‘There is insufficient evidence’, ‘There is limited evidence’, ‘There is poor evidence’, and ‘There is good evidence’. In the review section on ‘Quality of evidence’, the authors mostly described the risk of bias in the included studies and did not refer to other factors that can influence the quality of evidence.
Conclusions: To reduce the potential for confusion by using various poorly defined terminologies, Cochrane authors should use a common and rational approach to rating and wording quality of evidence that can be understood and interpreted across readers. The GRADE system provides such an approach.