Article type
Year
Abstract
Background: Network meta-analysis can combine information from all randomized controlled trials when many treatment regimens already exist, in order to rank the regimens to determine the best option for patients. But is is not clear what are the reporting and methodological qualities of network meta-analysis.
Objectives: To evaluate the reporting and methodological qualities of network meta-analyses.
Methods: We searched Pubmed, the Cochrane library, Embase, ISI Web of Knowledge, and Google Scholar using 'Network meta analysis' OR 'mixed treatment comparisons meta analyses' OR 'multiple treatments meta analysis' in title/abstract without language, time, or publication type restrictions up to 2010 October 21 and updated up to 2011 February21. We included any network meta-analysis, regardless of the conditions or interventions. We evaluated the reporting qualities of network meta-analysis using the PRISMA statement and methodological quality was assessed using OQAQ as a network meta-analysis was considered as a meta-analysis. Two trained authors independently selected studies, abstracted data, and assessed the qualities, and disagreement was resolved by a third author.
Results: We found 37 network-meta analyses. Thirteen items were reported in less than 75% articles, in which seven items were reported in less than 50%. Fifty-one percent tried to avoid bias in the selection, and 54% reported criteria used for assessing validity, but 24% were assessed appropriately.
Conclusions: Some necessary methodological and reporting quality items were not performed or reported adequately, so there is much room to improve.
Objectives: To evaluate the reporting and methodological qualities of network meta-analyses.
Methods: We searched Pubmed, the Cochrane library, Embase, ISI Web of Knowledge, and Google Scholar using 'Network meta analysis' OR 'mixed treatment comparisons meta analyses' OR 'multiple treatments meta analysis' in title/abstract without language, time, or publication type restrictions up to 2010 October 21 and updated up to 2011 February21. We included any network meta-analysis, regardless of the conditions or interventions. We evaluated the reporting qualities of network meta-analysis using the PRISMA statement and methodological quality was assessed using OQAQ as a network meta-analysis was considered as a meta-analysis. Two trained authors independently selected studies, abstracted data, and assessed the qualities, and disagreement was resolved by a third author.
Results: We found 37 network-meta analyses. Thirteen items were reported in less than 75% articles, in which seven items were reported in less than 50%. Fifty-one percent tried to avoid bias in the selection, and 54% reported criteria used for assessing validity, but 24% were assessed appropriately.
Conclusions: Some necessary methodological and reporting quality items were not performed or reported adequately, so there is much room to improve.