We're going to run out of original work

Transparent texts

PDF
Dtsch Arztebl 2008; 105 (40): 675-9; DOI: 10.3238 / arztebl.2008.0675
MEDICINE: From the editorial office
Conflicts of interest on the part of authors are common in scientific articles. The transparent presentation is important.

Since 2005, the Deutsches Ärzteblatt has asked every author of a scientific article to explain their conflicts of interest. We publish this information together with the manuscript, even if the author has not indicated any conflicts of interest. Our readers have welcomed this change and most of the authors support the transparent presentation of conflicts of interest.

Nevertheless, there are always misunderstandings and queries, so this article should explain what constitutes a conflict of interest for an author in the eyes of the medical-scientific editorial staff ("Medicine" section), what editorial balance results after about three years what conclusions a reader should draw from the existence of a conflict of interest - and what not. Before doing this, however, the scientific background to the discussion about conflicts of interest should be discussed.

"Whose bread do I eat, the song I sing"?
There are different definitions of conflicts of interest, the lowest common denominator being that interests conflict when one intention, e.g. to help one's patient or to solve a scientific problem, is negatively influenced by another, e.g. to gain a financial advantage can (1). The selection of the following research results shows the extent to which such antagonistic interests of authors can be reflected as content-related distortions in their publications:
- In a fundamental paper, Stelfox and co-workers (2) examined the relationship between the tenor of articles on calcium antagonists and the financial connections of their authors to the manufacturers. The basis of the study was formed on the one hand by 70 original and review articles as well as letters to the editor on the subject of calcium antagonists and on the other hand by the responses of all authors to a questionnaire on their conflicts of interest sent to them by the research group. Stelfox and his co-authors had divided the texts into positive, neutral and critical of calcium antagonists. It was found that authors of positive articles about calcium antagonists had significantly more often (96%) financial ties to the manufacturers than authors of neutral (60%) or critical papers (37%). The authors of critical texts, on the other hand, had no more material relationships with competing companies than those of positive publications.
- Friedman et al. (3) analyzed all 398 original papers that appeared in the New England Journal of Medicine and JAMA in 2001. There was a conflict of interest in 38.7 percent of the articles. Even after considering possible other influencing factors such as the number of cases, study design and country of origin, there was an association between conflicts of interest and positive study results (odds ratio: 2.35; 95% confidence interval: 1.08–5.09).
- When examining 162 randomized, placebo-controlled studies in the psychiatric literature, a research group led by Roy Perlis found (4) that an existing conflict of interest of at least one author was more than randomly associated with a positive study result (odds ratio 8.4; 95-%) -Confidence interval: 2.6-27.3).

Other studies have confirmed these results, but methodologically mixed up the information on the financing of clinical studies and the authors' conflicts of interest (about [5–10]; one of the few negative studies: [11]; overview in [12]). In the context of the relationship between study results on the one hand and conflicts of interest of authors or source of study funding on the other, it is therefore not possible to determine precisely what proportion is due to the conflicts of interest. Some of these studies are briefly described in Table 1.

It turns out that the association of conflicts of interest with the tenor of an article seems to exist for therapeutic, diagnostic and cost-effectiveness studies and is by no means only valid for cooperation with the pharmaceutical industry. For example, there is evidence that the tobacco industry has also had a successful impact on scientists (7, 8, 13). It also seems plausible to assume that no medical subject is immune to conflicts of interest and their effects. Ultimately, the finding applies not only to original articles, but also to reviews, which is particularly important for the Deutsches Ärzteblatt, because literature reviews currently make up around 80 percent of our articles: An investigation of reviews on the subject of passive smoking revealed that authors collaborate with the tobacco industry was closely associated with a negative outcome (13).

Consequences
One can therefore assume a relationship between the existence of a conflict of interest and the results of a study or the conclusions of an article, although it must be remembered that not every publication by authors with a conflict of interest is unilateral. Nevertheless, the risk of distortion led to the introduction of declarations of conflicts of interest, which were intended to make the links between the authors transparent to the reader. The New England Journal of Medicine was the pioneer in 1984. Since then, it has been followed by many specialist journals, with a larger proportion of journals having rules for dealing with conflicts of interest in medicine than in other subjects (14).

However, declarations of conflicts of interest are not a matter of course in medicine either. Schneider and co-workers (15, 16) found for the field of health services research that only 18 out of 31 journals examined required their authors to issue declarations of conflicts of interest. Among medical periodicals that required such a statement, they published only 60 percent of all articles (17).

Consequences for the Deutsches Ärzteblatt
Since 2002, the authors' notes of Deutsches Ärzteblatt have included the request to notify the editorial team of any conflicts of interest. Since 2005, we have also required a written declaration on possible conflicts of interest. Our approach is based on the definition of a conflict of interest as proposed by the International Committee of Medical Journal Editors (ICMJE) (18) (box). Unlike many other journals, we ask for an explanation from all authors and not just from the author in charge of the correspondence. This rule applies to all types of text, including letters to the editor in the "Discussion" section. All reviewers of manuscripts are also asked about conflicting interests. In the editorial department, we hand over the manuscript to another editor for supervision in the event of a conflict of interests.

First experiences with Deutsches Ärzteblatt
In 2006 and 2007, 65 of 207 original and review articles (31.4%) in the medicine section of Deutsches Ärzteblatt had a conflict of interest with at least one of the authors. Of the 746 authors of all articles in this period, almost one in five reported a conflict (n = 140; 18.8%). The vast majority of these were financial ties to the pharmaceutical industry. In the years 2002 and 2003 used for comparison, there were no reviews or original articles in the Deutsches Ärzteblatt with a declaration of an existing conflict of interest (Table 2).
Our current figures are of a similar dimension to those of the Deutsche Medizinische Wochenschrift (DMW), which documented the existence of conflicts of interest in a third (34.3%) of its overview and original articles between 2006 and 2007 (Table 2). As can be seen from the low 9.2 percent percentage of the DMW from 2002 and 2003, the information on conflicts of interest has increased overall, although the DMW has not changed its survey practice in the meantime (personal communication from Dr. Hirschel, deputy editor-in-chief of DMW). She, too, asks all authors to provide a written declaration on possible conflicts of interest. The increase in declarations of conflicts of interest could be due to an actual increase, but also to the fact that authors are now more willing than in the past to provide transparent information. The definition of a conflict of interest is also likely to have become clearer in recent years and public awareness has developed further. At Deutsches Ärzteblatt, the increase in conflicts of interest since 2002/2003 is certainly also related to the introduction of a mandatory, written declaration.

Authors who give false information
Readers are more critical of articles with a conflict of interest than articles without (19). However, we believe that the transparent presentation of conflicts of interest should be part of every scientific article. This is also a legitimate claim of the society that finances the health system. Appropriate handling of conflicts of interest has meanwhile also been included in the “Charter on Medical Professional Ethics” (20).

We depend on the cooperation of our authors. We are aware that the information provided by the authors is usually not verifiable and that conflicts of interest can be concealed. For example, Goozner (21) found for four American magazines that about every tenth to twentieth article lacked a conflict of interest that should have been reported. One reason for these findings may be that some authors believe that the benefits they received did not affect their medical judgment. However, this subjective assessment is not decisive for answering the question of whether there is a conflict of interest. The only important thing is whether there are any financial or immaterial connections at all (see Professor Klemperer's article in this issue). For journals there is only the possibility of asking the authors openly and hoping that they will voluntarily provide correct information. However, the basis for a trusting future collaboration is undoubtedly shaken if an author has already provided incorrect information.

What does a conflict of interest mean?
For some readers of Deutsches Ärzteblatt, there was irritation despite the generally positive reception of the declarations on conflicts of interest. A resident colleague wrote to us in 2006: “It is a step forward that you have introduced the names of the authors' industrial financiers. Also take the second step and leave out the articles by sponsored ‘authors."

Is the reader right? Does a conflict of interest invalidate the statements of an article? In the medical-scientific editorial office we mean: No. To give an example from clinical practice: A doctor can get into a conflict if he offers a medical service that is not necessary for his patient but lucrative for the doctor. Some probably offer this service. However, many colleagues resolve the conflict by limiting themselves to their medical role and refraining from taking additional measures. This example makes it clear that the existence of a conflict of interest by no means always goes hand in hand with misconduct.

Applied to the field of scientific medical journalism, this means: A conflict of interest does not mean a distorted judgment and a bad article. So as important as it is to be aware of the dangers of conflicts of interest, it is also important not to forget that many works by authors with conflicts of interest are nonetheless balanced in terms of content. This is also a result of the studies cited. A conflict of interest in and of itself is therefore not dishonorable. It is important, however, that the readers can get a transparent picture of the actual conflict of interests. Not least because of this, the specialist journals try to prevent distortions in articles by means of an editorial check and the peer review process.

When evaluating conflicts of interest, one should keep in mind that cooperation between clinicians and scientists, for example with the pharmaceutical industry or the manufacturers of medical products, is often desirable and sometimes indispensable. The participation of scientists or doctors working in clinical practice can be very important, for example in the development of new drugs. At the same time, it is only fair to adequately reward those who make a substantive contribution.

Conflicts of interest are not uncommon not only in specialist journals and academic medicine, but also in medicine as a whole: In a US study, 94 percent of all doctors reported financial relationships with the pharmaceutical industry in some form (from drug samples and dinner invitations to lecture fees or consultancy contracts [ 22]). Even if it is plausible not to transfer the situation in the USA one-to-one to Germany, one must assume that conflicts of interest are ubiquitous. In a way, they cannot be avoided entirely in a healthcare system, as the clinical example should show. One should also not forget that there are conflicts of interest other than material that threaten the seriousness of an article. Unfortunately, such conflicting interests are stated far less often than the easier to operationalize financial conflicts of interest.

Due to the omnipresence of conflicts of interest, the New England Journal of Medicine had to relax its formerly strict rules for the authors of editorials and review articles: The editorial team had realized that they could no longer find authors without conflicts of interest for important topics. The New England Journal of Medicine therefore distinguishes between different interrelationships (23). The editors consider fees for lectures to be less important than shareholdings. The British magazine The Lancet proceeds in a similar way.

What to do?
Since it is not justified to reject articles simply because of their authors' conflicts of interest, there is no other way than critical reading to find out when an article is really skewed. This applies to the readers of Deutsches Ärzteblatt as well as to our editorial team when examining manuscripts. The existence of a conflict of interest should in any case be a reminder to adopt a strenuous, but also sensible, skeptical attitude when reading scientific papers - regardless of which journal you have in mind.

Conflict of interest
The author heads the medical and scientific editorial department of Deutsches Ärzteblatt.

thanksgiving
The author thanks Professor Wolf-Dieter Ludwig (Berlin) for critically reading the manuscript.

Author's address
PD Dr. med. Christopher Baethge
Head of the medical and scientific editorial team
Email: [email protected]

Transparent text

Dtsch Arztebl 2008; 105 (40): 675-9
DOI: 10.3238 / arztebl.2008.0675

The German version of this article is available online:
www.aerzteblatt.de/english