Industry Support and Correlation to Study Outcome for Papers Published in Spine

In Brief Study Design. Retrospective review of articles published in the journal Spine. Objectives. To evaluate potential correlations between research sponsorship and study outcome. Summary of Background Data. Industry sponsorship has traditionally been associated with more positive results than non-funded or peer-reviewed funded projects in other areas of medicine. The association of such sponsorship and study outcome has not been addressed previously in spine research. Methods. Articles from the journal Spine from January 2002 to July 2003 were reviewed. These were examined for the subject evaluated, type of study design, funding source, and conclusion reached. Results. Of 1143 articles, 527 met inclusion criteria of having abstract, materials/methods, and conclusion sections. Industry support was reported for 84 (15.9%), foundation support for 67 (12.7%), government support for 54 (10.2%), institution support for 17 (3.2%), and no funding was reported for 304 (57.9%). The odds ratio of industry funded reporting positive results was 3.3 times that of studies with any other funding sources (P < 0.001). Conclusions. Industry funded studies demonstrated a statistically greater likelihood to report positive results than studies with other funding sources. Potential explanations for this are biased study design, biased experimental technique, biased result interpretation, or publication bias. Although the expense of research and limited funding sources have forced an increased reliance on industry support for funding basic science and clinical spine research, this does introduce the potential for bias and must be recognized by the reader. This retrospective study of 527 articles published in the journal Spine assessed potential corrections between research sponsorship and study outcome. Positive results were statistically correlated with industry funding in comparison with other funding sources. Although industry funding is often necessary given the expense of clinical research, the potential for bias must be appreciated.