GRADE Outcomes or Studies: How to Use the GRADE Approach Correctly?

Min Zhi,Lihai Zhang,Yaolong Chen,Peifu Tang

DOI: https://doi.org/10.1007/s00264-012-1766-4

IF: 3.479

2013-01-01

International Orthopaedics

Abstract:Dear Editor, We read with deep interest the article by Houwert et al. [1]. We appreciate the work of the authors on the systematic review related to the two major therapeutic choices for displaced mid-shaft clavicle fractures. However, we have some comments about the using of GRADE approach. First, the GRADE approach defines the quality of a body of evidence for outcomes reported in systematic reviews [2, 3]. The level of quality of evidence should be rate separately for each outcome. In Houwert’s research, the authors used the GRADE approach as an evaluation tool for the studies, with similar risk of a bias tool. Second, although the authors described factors which can downgrade or upgrade the level of quality in the methods section, they did not analyse the factors and provide details. Third, high quality means “We are very confident that the true effect lies close to that of the estimate of the effect” [3] in the GRADE system. However, the authors drew conclusions of “high quality evidence” from only one included study without any additional analysis. To summarise, the GRADE approach should be used for grading the quality of evidence for outcomes, but not included studies based on a systematic review. Nonetheless, the above points cannot lessen the overall value of this work and we thoroughly enjoyed reading the article with that in mind.