Also less likely than government-, nonprofit-funded trials to be published within two years
TUESDAY, Aug. 3 (HealthDay News) -- Drug clinical trials supported by industry are more likely to produce favorable results than trials supported by government or nonprofit/nonfederal organizations, and they are less likely to be published within two years of the study being completed, according to research published in the Aug. 3 issue of the Annals of Internal Medicine.
Florence T. Bourgeois, M.D., of Children's Hospital Boston, and colleagues reviewed 546 clinical trials on five drug classes (anticholesteremics, antidepressants, antipsychotics, proton-pump inhibitors, and vasodilators) conducted during 2000 to 2006 and listed on ClinicalTrials.gov. The researchers determined publication date, study design, and study outcomes and compared these data by funding source: industry, government agencies, and nonprofit/nonfederal organizations.
Allowing for a three-year lag time after trial completion, the reviewers found 66.3 percent of the trials published results, with industry-funded trials reporting positive results 85.4 percent of the time and government-funded trials and nonprofit/nonfederal organization-funded trials reporting positive results 50 and 71.9 percent of the time, respectively. Moreover, the nonprofit/nonfederal trials that received some industry contribution were more likely to report positive outcomes than those that received no industry contribution (85.0 versus 61.2 percent). The rates of trial results being published within two years of study completion ranged from 32.4 percent for industry-funded trials to 56.2 percent among nonprofit/nonfederal organization-funded trials that did not receive industry funding.
"While we cannot specifically point to which factors contribute to the association between funding source and positive result reporting, our findings speak to the need for more disclosure of all elements of a study," Bourgeois said in a statement. "Publication bias is likely a contributing factor, but there may be many more, including biases in study design, patient selection, data analysis and results reporting."
Full Text (subscription or payment may be required)