Disclosing conflicts of interest in German publications concerning health services research

BACKGROUND The influence of the pharmaceutical industry and other stakeholders on medical science has been increasingly criticised. When dealing with conflicts of interest in scientific publications it is important to ensure the best possible transparency. The objective of this work is to examine the disclosure practice of financial and non-financial conflicts of interest in German language publications concerning health services research for the first time. METHODS We performed a systematic literature search in the PubMed data base using the MeSH term "health services research". The review was conducted on July 10, 2006, setting the limits "dates: published in the last 2 years" and "languages: German" (only articles with abstracts). 124 articles in 31 magazines were found. In the magazines the instructions for authors were examined as to whether a statement on conflicts of interest is expected--and if, in which form. Regarding the articles in the journals which require a statement, we examined whether the statement is explicitly published. The results are descriptively represented. RESULTS 13 magazines (42%) do not require any statement on conflicts of interest, whereas 18 journals (58%) expect a statement. Two of these 18 magazines refer explicitly to the uniform requirements of the International Committee of the Medical Journal Editors (ICMJE); the remaining 16 magazines give differently accentuated instructions on how to disclose conflicts of interest, whereby the focus is primarily on financial issues. A statement on conflicts of interest is explicitly published in 11 of the 71 articles (15%) which are found in the magazines that require a statement with the submission of a manuscript. Related to the total number of included articles, this means that the reader explicitly receives information on potential conflicts of interest in 9% of the cases (11 of 124 articles). Statements of others that are involved in the publication process (reviewers, editors) are not available in any of the articles examined. CONCLUSION A better sensitization for possible conflicts of interest in German publications concerning health services research is necessary. We suggest tightening the criteria for disclosure in the instructions for authors in the scientific journals. Among other things the equivalent consideration of financial and non-financial conflicts of interest as well as the obligatory publication of the statements should be part of good practice.