An important instrument in dealing with conflicts of interest scientifically is ensuring best possible transparency 5. Internationally accepted standards for this are set by the Uniform Requirements for Manuscripts Submitted to Biomedical Journals by the International Committee of Medical Journal Editors (ICMJE). Apart from financial interest, problems that may result from non-financial conflicts of interest are explicitly mentioned 10:

" [...] Conflict of interest exists when an author (or the author's institution), reviewer, or editor has financial or personal relationships that inappropriately influence (bias) his or her actions (such relationships are also known as dual commitments, competing interests, or competing loyalties). These relationships vary from those with negligible potential to those with great potential to influence judgment, and not all relationships represent true conflict of interest. The potential for conflict of interest can exist whether or not an individual believes that the relationship affects his or her scientific judgment. Financial relationships, (such as employment, consultancies, stock ownership, honoraria, paid expert testimony) are the most easily identifiable conflicts of interest and the most likely to undermine the credibility of the journal, the authors, and of science itself. However, conflicts can occur for other reasons, such as personal relationships, academic competition, and intellectual passion. All participants in the peer review and publication process must disclose all relationship that could be viewed as presenting a potential conflict of interest. [...]".

Lack of or incomplete disclosure of potential conflicts of interest prevent the reader from applying the appropriate scepticism in the scientific dialogue 11.

There have been no studies so far on the disclosure practice of conflicts of interest in scientific journals within the field of health services research. Therefore, the objective of this study is to examine the current practice using the example of German language journals. In particular, the following research questions should be investigated:

▪ What criteria are set up for the disclosure of conflicts of interests in the instructions for authors?

▪ How transparent is the practice of disclosure for the readers?

Internationally, health services research in Germany¹ is in an early stage of development. However, it has increasingly received attention and support especially within the last few years. Milestones are, for example, the research programs by the Federal Ministry for Education and Research (BMBF) and the compulsory health insurance funds in the year 2000 and the strategies to promote research by the German Medical Association (Bundesaerztekammer) in 2005. Quite recently Pfaff und Kaiser, who are well-known experts in this area, looked critically at the situation of health services research in Germany 12. They state that research activities are characterized by a concentration on topics guided by special interests of stakeholders. The authors attribute this to the high proportion of commissioned research aiming at fast, practical usability and threatening the scientific soundness of health services research.

The instructions for authors were examined in all 31 journals using the freely accessible internet homepages of the journals (all accesses on July 10, 2006):

▪ Is a statement on conflicts of interest demanded?

▪ If so, do the instructions refer to the criteria of the International Committee of Medical Journal Editors (ICMJE criteria)?

▪ If they do not refer to the ICMJE criteria, which other criteria are applied?

Of the articles published in magazines which require an authors' statement on conflicts of interest (n = 71), we examined the following questions using full text articles:

▪ Is the statement explicitly published? We considered a statement as explicitly published if it was presented in the form of a specific paragraph with an explicit title "conflicts of interest" or likewise. We did not consider a statement as explicitly published if information on conflicts of interest could be gained from, e.g., the acknowledgement or unspecificly titled paragraphs (e.g. "annotations").

▪ If the statement is explicitly published, what kind of information is presented in the text?

▪ Are statements on conflicts of interest available from others who are essentially involved in the publication process (reviewers, editors)?

All articles (with and without explicit statements of conflicts of interest; differentiated in the tables 2 and 3, see Additional file 1) were examined as to whether information on possible conflicts of interest could be deducted from the institutional affiliations of the authors (e.g. staff members of for-profit companies) and from the acknowledgements.

Furthermore we wanted to gather information about the relevance of pharmaceutical topics within the included publications, since these topics are particularly important in respect to the research questions. For this purpose the titles and the summaries of all 124 articles were qualitatively analysed: does the article particularly deal with pharmaceutical topics?

The analyses of both the articles and the journals were performed by the principal investigator (NSCH) and the co-author (HL) independently from each other. A medical data assistant crossed-checked the findings. In case of different assessments the investigators came to a decision by consensus.

Overall, the sensitivity for the problem of conflicts of interest seems to be in need of improvement in the circle of the examined journals. Only about half of the journals expect statements from authors on conflicts of interest – and only in a few cases is the reader explicitly informed about the statements. Whether the received statements are published with the articles is in the responsibility of the editors, sometimes in agreement with the authors.

The transparency would be greater if the statements were published on principle, independently whether conflicts of interest are declared or not. This procedure would give the reader the chance to develop an unfiltered opinion about possible conflicts of interest; furthermore, this procedure could be suitable to increasing the sensitivity for the problem of all parties involved (authors, editors, reviewers, readers). Information on possible conflicts of interest of reviewers and editors are not available in any of the articles, although this could also be considered as good scientific practice 10. Particularly little is known about editors' conflicts of interest and how to manage this problem 14.

Another critical issue is the understanding of conflicts of interest. The fact that they can result from direct financial allowances is comparatively self-evident. However, the various other causes for possible conflicts of interest are easily underestimated 11. Thus the International Committee of Medical Journal Editors (ICMJE) describes the characteristics of non-financial interest conflicts in detail 10.

Of the German language journals included in this study, only two journals explicitly refer to the "uniform requirements for manuscripts submitted to biomedical journals". In the instructions for authors in the remaining 29 journals the topic of non-financial conflicts of interest is in no case dealt with as elaborately as the ICMJE criteria suggest. Instead, non-financial conflicts of interest are addressed in a very general way, and partially completely left blank.

The three journals with the most contributions on health services research in this study proceed differently (see table 1, see Additional file 1): Bundesgesundheitsblatt demands statements on conflicts of interest from the participants of the publication process (explicitly mentioned: authors, editors, reviewers), whereby financial aspects are addressed clearly and non-financial rather generally ("activities"). However, information on the statements is available only in one of the examined cases, in form of an acknowledgement (article no. 29). For the reader, this could mean that in all other cases no conflicts of interest were declared; it could also mean that conflicts of interest were declared but the responsible persons did not see the necessity to publish the statements. However, the readership has no chance to form an unfiltered opinion.

The journal Das Gesundheitswesen provides no information at all on the disclosure of conflicts of interest in its instructions for authors. Psychiatrische Praxis is one of the two journals which explicitly refer to the ICMJE criteria. However, in none of the 13 articles examined are the statements presented to the readership.

The extent of ambiguity on the question of which issues should be declared as possible conflicts of interest may be illustrated by an example outside of the focus of this study. In the Journal of the American Medical Association, Kurt et al. published an article on "Migraine and risk of cardiovascular disease in women"; the authors declared that they have no relevant conflicts of interest 15.

In fact, the authors have numerous connections to pharmaceutical companies which sell products for the treatment of migraine and cardiovascular diseases, as Kurt et al. had to reveal later in a letter to the publishers. The authors state that they assumed that in the case of the article these connections are not conflicts of interest. They close with the sentence: "We have learned that it is best to disclose all relationships with for-profit companies and allow the editor(s) to decide what is relevant"16. In the editorial it is pointed out clearly that the editors of JAMA expect the disclosure of all possible conflicts of interest 17.

From the fact that only a fifth of the articles in our study deal with pharmaceutical topics in a closer or wider sense, it should not be concluded that conflicts of interest which may result from the influence of for-profit companies are less important in health services research than, e.g., in medical research. Besides the pharmaceutical industry, other for-profit companies could influence science, for example companies that sell technical equipment 18. Furthermore, within the health system the influence on science of social insurance institutions, political institutions, various associations and numerous lobby groups has to be critically reflected – in particular against the background of the high proportion of commissioned studies that characterize health services research in Germany 12.

As in no other discipline health services research is concerned with the framework requirements of health care on different levels of the system. Thereby, various stakeholders could be very interested in using health services research to shape specific issues within the health care system in their favour. For example, one should reflect on the phenomenon of "disease mongering" which can be described as the attempt to define the population which allegedly suffers from a certain disease as large as possible and to extend the need for medical care 9. A quotation from a critical editorial in Zeitschrift für Allgemeinmedizin makes clear how attentive health services research should be designed to avoid being used for the objectives of third parties: "develop a topic field, and prove that there is undersupply; everything else finds itself" 13.

Pfaff and Kaiser recommend for future health services research in Germany that topics not in the immediate interest of financially potent stakeholders should specifically be promoted by the state 12. Our data suggest that it is necessary to increase the transparency in the practice of scientific publication, not only related to direct financial allowances but also related to non-financial conflicts of interest. The "uniform requirements for manuscripts submitted to biomedical journals" 10 provide a good orientation for its realisation.

It is important that the claim for disclosure of all possible conflicts of interest contains no allegation against authors, editors or reviewers. An explicit comment seems to be useful at this point, e.g. as it is done in the instructions for authors of the journal Der Nervenarzt (table 1, see Additional file 1).

However, even if the disclosure practice is optimised, achieving full transparency in scientific publication is not a realistic ambition. In her editorial on the influence of money on science, DeAngelis outlined the problem as follows 1: "There is simply no way to guarantee that all financial relationships and arrangements of all authors are disclosed. It is not feasible to independently investigate the financial relationships of every author, as no comprehensive, up-to-date source of this information exists". Finally the appeal for honesty remains the elementary component of good science.