Article type
Year
Abstract
Background: GRADE approach is a useful tool in the evaluation of quality of evidence, but until now we were unclear whether it could be applied into systematic reviews of etiological studies.
Objectives: To evaluate whether GRADE approach could be applied into systematic reviews of etiological studies.
Methods: The Cochrane Library, PubMed, Embase, ISI Web of Knowledge, China Academic Journal Network Publishing Database, Chinese Scientific Journals Full text Database and Chinese Biomedical Literature Databasewere searched at April 2011 to include systematic reviews about factors that were related to gastric cancer. For the evaluation of quality of evidence,GRADE approach were used to specify high, moderate, low, and very low levels.
Results: Fifty-five systematic reviews were included. The biggest challenge of using GRADE approach into systematic reviews of etiological studies was lack of evaluations of scientific quality of included studies. Only 18.18% assessed the scientific quality of the included studies and 7.27% used them in formulating conclusions. The second challenge was inconsistency between different study designs. Only 12 systematic reviews only included one study type, and therewere inconsistent results between the pooled results of cohort studies and case control studies in seven out of the remaining 43 systematic reviews. There were also other challenges, such as lack of pooled results.
Conclusions: GRADE approach might be used in systematic reviews of etiological studies. However, before applying GRADE approach, we should reach a consensus about how to solve these challenges and how to do in the future systematic reviews of etiological studies.
Objectives: To evaluate whether GRADE approach could be applied into systematic reviews of etiological studies.
Methods: The Cochrane Library, PubMed, Embase, ISI Web of Knowledge, China Academic Journal Network Publishing Database, Chinese Scientific Journals Full text Database and Chinese Biomedical Literature Databasewere searched at April 2011 to include systematic reviews about factors that were related to gastric cancer. For the evaluation of quality of evidence,GRADE approach were used to specify high, moderate, low, and very low levels.
Results: Fifty-five systematic reviews were included. The biggest challenge of using GRADE approach into systematic reviews of etiological studies was lack of evaluations of scientific quality of included studies. Only 18.18% assessed the scientific quality of the included studies and 7.27% used them in formulating conclusions. The second challenge was inconsistency between different study designs. Only 12 systematic reviews only included one study type, and therewere inconsistent results between the pooled results of cohort studies and case control studies in seven out of the remaining 43 systematic reviews. There were also other challenges, such as lack of pooled results.
Conclusions: GRADE approach might be used in systematic reviews of etiological studies. However, before applying GRADE approach, we should reach a consensus about how to solve these challenges and how to do in the future systematic reviews of etiological studies.