Purchase this article with an account.
D.A. Mazzulla, M.P. Rubin, M. Saidel; Comparison of Statistical Significance Between Commercially Supported and Non–Commercially Supported Studies Presented at the 2005 ARVO Annual Meeting . Invest. Ophthalmol. Vis. Sci. 2006;47(13):4410.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare the statistical significance of studies with and without commercial support.
A review of the abstracts in the glaucoma and retina sections from the ARVO 2005 annual meeting was performed. We documented the type of study (basic science or clinical), whether or not a pharmaceutical agent was studied, commercial support, statistical significance and the corresponding p–value. The study was considered to have commercial support only if a commercial entity was listed under the support section of the abstract, regardless of an author’s stated commercial associations. We excluded those studies that did not quantitatively state significance with a p–value.
481 of the abstracts reviewed met the study criteria. 28 (6%) of the abstracts documented commercial support. 87 outcomes were studied in these abstracts. 72 (83%) of the commercially supported outcomes were statistically significant (avg. p=0.06132). 453 of the remaining abstracts not stating commercial support studied a total of 1352 outcomes. 1067 (79%) of the non–commercially supported outcomes were significant (avg. p=0.08915). This difference between significant outcomes was not statistically significant by chi–square analysis (X2=0.393, p=0.53). We subdivided the data to analyze only clinical studies of pharmaceutical agents. 100 abstracts met the criteria. 14 of these abstracts had commercial support and studied 41 outcomes. 31 (76%) were statistically significant (avg. p=0.09166). 86 of the remaining non–commercially supported clinical studies of pharmaceutical agents studied 258 outcomes. 176 (68%) were statistically significant (avg. p=0.1240). This difference in significant outcomes also was not statistically significant by chi–square analysis (X2=0.341, p=0.56).
It has long been conjectured that commercial involvement influences academic research. Here we find that contrary to commonly held belief, commercial support did not appear to have an effect on the significance of outcomes of the studies in the glaucoma and retina sections at the 2005 ARVO annual meeting.
This PDF is available to Subscribers Only