Discrepancy between Results and Abstract Conclusions in Industry- vs Nonindustry-funded Studies Comparing Topical Prostaglandins TARIQ ALASBALI, MICHAEL SMITH, NOA GEFFEN, GRAHAM E. TROPE, JOHN G. FLANAGAN, YAPING JIN, AND YVONNE M. BUYS ● PURPOSE: To investigate the relationship between industry- vs nonindustry-funded publications compar- ing the efficacy of topical prostaglandin analogs by evaluating the correspondence between the statistical significance of the publication’s main outcome measure and its abstract conclusions. ● DESIGN: Retrospective, observational cohort study. ● METHODS: English publications comparing the ocular hypotensive efficacy between any or all of latanoprost, travoprost, and bimatoprost were searched from the MEDLINE database. Each article was reviewed by three independent observers and was evaluated for source of funding, study quality, statistically significant main out- come measure, correspondence between results of main outcome measure and abstract conclusion, number of intraocular pressure outcomes compared, and journal impact factor. Funding was determined by published disclosure or, in cases of no documented disclosure, the corresponding author was contacted directly to confirm industry funding. Discrepancies were resolved by con- sensus. The main outcome measure was correspondence between abstract conclusion and reported statistical sig- nificance of the publications’ main outcome measure. ● RESULTS: Thirty-nine publications were included, of which 29 were industry funded and 10 were nonindustry funded. The published abstract conclusion was not con- sistent with the results of the main outcome measure in 18 (62%) of 29 of the industry-funded studies compared with zero (0%) of 10 of the nonindustry-funded studies (P  .0006). Twenty-six (90%) of the industry-funded studies had proindustry abstract conclusions. ● CONCLUSIONS: Twenty-four percent of the industry- funded publications had a statistically significant main outcome measure; however, 90% of the industry-funded studies had proindustry abstract conclusions. Both read- ers and reviewers should scrutinize publications carefully to ensure that data support the authors’ conclusions. (Am J Ophthalmol 2009;147:33–38. © 2009 by Elsevier Inc. All rights reserved.) F INANCIAL RELATIONSHIPS BETWEEN PHARMACEUTI- cal companies and researchers and funding of med- ical research by drug companies has increased dramatically during the last two decades. 1–4 This can result in industry bias where the source of funding of clinical trials either affects the results in a systematic way or leads to selective presentation of the results. Industry funding often has been associated with proindustry results 2,5–20 and publication bias, 21–23 which can affect the interpretation and presentation of outcomes resulting in conclusions that overstate results without statistical support. The purpose of this study was to investigate the relationship between industry- vs nonindustry-funded publications comparing ocular hypotensive efficacy of the topical prostaglandin analogs (PGA) latanoprost 0.005%, travoprost 0.004%, and bimatoprost 0.03% by evaluating the correspondence between the statistical significance of the publication’s main outcome measure and its published abstract conclusions. METHODS A MEDLINE SEARCH FROM 1966 TO THE SECOND WEEK OF November 2007 using any combination of the keywords latanoprost, travoprost, and bimatoprost was conducted. The title and abstracts from the initial search were reviewed and those included were English language publications comparing the intraocular pressure (IOP)-lowering efficacy of any combination of latanoprost; travoprost; or bimato- prost. The complete articles were obtained and the refer- ences also were searched to identify relevant publications missed during the initial search. Each publication was reviewed by three independent observers using a standardized data collection sheet eval- uating: source of funding, industry author, study quality, main outcome measure, statistical significance (P  .05) of main outcome measure, abstract conclusion, correspon- dence between statistical significance (P  .05) of main See accompanying Editorial on page 1. Accepted for publication Jul 1, 2008. From the Department of Ophthalmology and Vision Sciences, Uni- versity of Toronto, Toronto Western Hospital, Toronto, Ontario, Canada (T.A., M.S., N.G., G.E.T., J.G.F., Y.M.B.); the Department of Ophthal- mology, King Faisal University, King Fahad Hospital of the University, Riyad, Saudi Arabia (T.A.); the School of Optometry, University of Waterloo, Waterloo, Ontario, Canada (J.G.F.); and the Department of Public Health Sciences, University of Toronto, Toronto, Ontario, Canada (Y.J.). Inquiries to Yvonne M. Buys, Toronto Western Hospital, 399 Bathurst Street, EW6-405, Toronto, Ontario, Canada M5T 2S8; e-mail: y.buys@ utoronto.ca © 2009 BY ELSEVIER INC.ALL RIGHTS RESERVED. 0002-9394/09/$36.00 33 doi:10.1016/j.ajo.2008.07.005