Background: Surgical publications are often criticized for methodological
issues. According to evidence-based medicine (EBM) principles this can
lead to low evidence level. The dissemination of quality criteria for
statistical reporting is one of the objectives of the ‘‘Uniform Requirements
for Manuscripts Submitted to Biomedical Journals’’ (URM). Objectives:
Aim of the present study was to analyze statistical reporting requirements
in surgical journals and to explore further the quality of statistical guidance
for authors. Methods: One-hundred and thirty-nine surgical journals were
identified in the Journal Citation Report 2005 (JCR, 2005) applying the
filter ‘surgery’. All journals with impact factor (IF) > =1 (n = 70) were
selected for further analysis. Instructions for authors of these journals were
classified by two independent observers into three categories: URMcriteria
required (URM), other statistical criteria required (other) or no
valuable statistical aspects mentioned (no). Results: Twenty of 70 journals
(28.6%) had no statistical requirements for authors at all. Twenty-seven
journals (39%) required at least URM-standard or further criteria. Twentythree
journals (33%) required only other criteria than URM. Regarding the
spectrum of the journals (18 general surgical journals (e.g. Br J Surg), 39
speciality journals (e.g. J Vasc Surg), 13 sub-speciality-journals (e.g. J
Endovasc Ther)), 67% of the general, 69% of the speciality and 84% of
the sub-speciality journals had at least some statistical information for
authors. Grouped by impact-factor-tertiles there was a tendency of high
IF-journals to require more often URM. The following journals presented -
beyond the URM-criteria - comprehensive and practical guidelines for
authors in order to present correctly statistical data and to avoid common
pitfalls (Br J Surg, J Thorac Cardiov Sur, J Am Coll Surgeons, Ann Surg).
Conclusions: The requirements for statistical reporting quality in the
leading surgical journals are very heterogeneous. It is supposed that
statistical reporting requirements depend more on the composition of
reviewer-groups and journal policy than on impact factor. In the era of
evidence-based medicine it is desirable and would possibly enhance
reporting quality if all surgical journals integrate statistical ebm principles
and guidance in their instructions for authors.