Clinical Trials Comparing Topical Prostaglandins: Potential Industry Bias

Abstract

Purpose: : To evaluate potential bias between industry funded and non-industry funded publications comparing the efficacy of prostaglandin analogues. Several previous studies have found a significant association between funding source and pro-industry conclusions.

Methods: : Systematic review of the literature. MEDLINE was searched until November 2007 using the keywords latanoprost, travoprost and bimatoprost. Inclusion criteria included English articles comparing prostaglandin analogues. Three independent observers reviewed each article. A standardized data collection sheet was used and any discrepancies were resolved by consensus. All studies were scored for methodological quality, statistical significance of study specified main outcome measure, correspondence of results of study specified main outcome to abstract conclusion and source of funding. The chi-square test, the Fisher’s exact test and the odds ratio were used to compare industry funded to non-industry funded studies.

Results: : A total of 180 articles were identified by the original search. After reviewing the abstracts 39 studies published from 2001-2007 met the inclusion criteria. 27 were industry funded and 12 were not industry funded. Significant p-values for study specified main outcome measure were claimed in 8 of 27 industry funded versus 2 of 12 non-industry funded studies, p=0.69. Matching of the results of the main outcome measure claimed by the authors to the published abstract conclusions was found in 8 of 27 industry funded studies versus 10 of 12 non-industry funded studies, p=0.002. This suggests that on average industry-funded results were 92% less likely matching their abstractconclusions (odds ratio=0.08, 95% confidence interval 0.02-0.47) when compared with non-industry-funded studies. 24 of the 27 (88.9%) industry funded studies had positive sponsoring industry conclusions.

Conclusions: : The published conclusions in industry funded studies were less likely to agree with the results of the main outcome measured and were usually pro the sponsoring company.

Keywords: intraocular pressure