Introduction: The aim of this study was to determine the association between the IF of surgery journals and the level of evidence of articles on therapeutic procedures published in these journals.
Material and method: A bibliometric study was performed. Journals listed in the 2002 ISI under the subject heading of Surgery were included. Journals were classified by applying the 33rd percentile of their IF (group 1 from 0.128 to 1.690, group 2 from 1.691 to 2.886, and group 3 from 2.887 to 6.674). Simple randomized sampling was conducted of journals in each group (group 1: Eur J Surg, Am Surg and World J Surg; group 2: J Am Coll Surg, Surgery and Arch Surg; and group 3: Br J Surg and Ann Surg). A crossover analysis was performed by 2 independent teams of reviewers.
Results: There were 751 article on therapy (41.2%). The mean IF was 1.66 +/- 0.48 for group 1, 2.61 +/- 0.15 for group 2, and 4.69 +/- 1.57 for group 3 (P<.001). There were 16 systematic reviews, 82 randomized clinical trials, 96 cohort studies, 474 case series and 79 review articles. Level 1 evidence was found in only 4.8% of the articles. Statistical differences were verified in 15.1% of the study sample, which was represented by systematic reviews and randomized clinical trials (most frequent in group 3 (P<.001).
Conclusions: An association between the IF of scientific surgical journals and the level of evidence in articles published on therapy was verified.