Article type
Year
Abstract
Background: Reporting guidelines recommend that reports of trials of pharmacological treatments include disclosure of conflicts of interest (COIs) resulting from pharmaceutical industry funding and author-industry financial ties. A recent study (Roseman etal., JAMA, 2011) found that meta-analyses published in high-impact biomedical journals rarely include information on study funding and other COIs from included trials. The Cochrane Handbook specifies that trial funding sources be collected in all reviews, and it suggests that potential bias related to study sponsorship may be incorporated into risk of bias assessments.
Objective: To investigate whether systematic reviews of pharmacological treatments published in the Cochrane Database of Systematic Reviews report COIs from included trials.
Methods: We searched the Cochrane Database of Systematic Reviews using the MeSH term 'drug therapy’ to identify all systematic reviews of pharmacological interventions published in 2010, with content assessed as up-to-date in 2008 or later and including ≥1 RCT. Two investigators independently extracted data on whether Cochrane reviews reported study funding sources and other COIs from included trials and, if so, where in the review such information was reported.
Results: Of 151 included Cochrane reviews, 30 (20%) reported trial funding information for all included studies; 16 (11%) reported for some, but not all, studies; and 105 (69%) did not report any information. Only 16 reviews (11%) provided any information on trial author financial ties or employment. Of 46 reviews that reported trial funding sources for at least some of the trials reviewed, 29 (63%) noted this in the context of risk of bias, either in the text and in 'Risk of bias' tables. Overall, however, the method of reporting COIs was inconsistent across reviews.
Conclusion: Less than a third (31%) of Cochrane reviews of pharmacological treatments report information on study funding from included trials, and only 11% report on other potential sources of COIs.
Objective: To investigate whether systematic reviews of pharmacological treatments published in the Cochrane Database of Systematic Reviews report COIs from included trials.
Methods: We searched the Cochrane Database of Systematic Reviews using the MeSH term 'drug therapy’ to identify all systematic reviews of pharmacological interventions published in 2010, with content assessed as up-to-date in 2008 or later and including ≥1 RCT. Two investigators independently extracted data on whether Cochrane reviews reported study funding sources and other COIs from included trials and, if so, where in the review such information was reported.
Results: Of 151 included Cochrane reviews, 30 (20%) reported trial funding information for all included studies; 16 (11%) reported for some, but not all, studies; and 105 (69%) did not report any information. Only 16 reviews (11%) provided any information on trial author financial ties or employment. Of 46 reviews that reported trial funding sources for at least some of the trials reviewed, 29 (63%) noted this in the context of risk of bias, either in the text and in 'Risk of bias' tables. Overall, however, the method of reporting COIs was inconsistent across reviews.
Conclusion: Less than a third (31%) of Cochrane reviews of pharmacological treatments report information on study funding from included trials, and only 11% report on other potential sources of COIs.