A journal published an article about clinical recommendations for a condition that supposedly was the result of a consensus between two scientific societies of different medical specialties. The article underwent peer review and no problems were identified at that stage. However, about one month after publication the journal was contacted by one of the scientific societies raising concerns that it had not been consulted or involved in drafting the article, and added that none of its members was an author on the paper. In the paper, there is only mention of the consensus between the two scientific societies in the title, and the body of the paper only includes the actual clinical recommendations. Moreover, the authors of the published article only included their clinical affiliations and did not include affiliations with one of the two scientific societies.
The journal contacted the authors of the paper and relayed to them the concerns raised by the aforementioned scientific society. The authors replied and reported that they had investigated the matter. They concluded that one of the authors had indeed acted inappropriately in giving false information on behalf of one of the scientific societies.
Questions for COPE Council
- What actions should the journal take?
- Is it appropriate to issue a correction which takes out the reference to the consensus between the scientific societies?
- Are there any additional measures the journal should take?
Advice on this case is from a small number of COPE Council Members. Most cases on the COPE website are presented to the COPE Forum where advice is offered by a wider group of COPE Members and COPE Council Members. Advice on individual cases is not formal COPE guidance.
Based on the additional information provided by the authors, publishing an expression of concern right away would be appropriate.
COPE views editors as stewards of the literature, so if the literature is now misleading then in needs to be corrected. If the paper purports to be from representatives of two learned societies, the leadership of both societies should be brought into the picture, as the paper may misrepresent one if not both society's positions. It would be good practice to contact the presidents of both societies to confirm these are their guidelines. If, having consulted the societies there is no consensus, and the paper reflects just one viewpoint, then the readers need to know that. Depending on the reply from one or both societies, the paper could be corrected (if there was some misunderstanding in the consultation process and some joint process did take place, but only one society agreed on the guidelines). At worst, both societies deny the guidelines and the paper should be retracted and the institutions also informed. If the complainant’s society confirms they had no knowledge of the guidelines and did not go through any consensus process, then that would be grounds for retraction. In that case, the statement that there was a consensus in the title and the implication that there was a complete systematic consensus process was a falsification and the 'consensus guideline' itself was a fabrication.
If this is a journal that regularly publishes clinical guidelines or recommendations, then there should be robust editorial office processes to assure greater transparency and validation. Clinicians depend upon such guidelines to make decisions for treatments and care plans, so complete disclosure of the process used to develop the guidelines is warranted. Societies are an important part of the process, so it would be good to incorporate a few additional steps in the journal review processes. A suggestion is to look for good reviewers from each of the societies to conduct the review and ask the specific question of the reviewers (and the authors) about the level of involvement of each society. Requiring review by someone from each relevant society would trigger the need to contact all the societies involved to look for appropriate reviewers. That action could have prevented this problem.
Second, the journal should require sign-off of the relevant authorities at each society prior to submission of the manuscript. If it is not already required, a complete disclosure statement about authorship and affiliations should be in place. There are always differences in interpretation between the "positions" that societies take on clinical issues and a good reviewer who is also a practicing clinician would likely identify those differences. So, rounding out the review panel for something that has such long-term impact with guideline statements is critically important. For clinical practice guidelines, there should be a comprehensive, systematic review of the literature on which to base recommendations. There are guidelines for the conduct of these reviews such as PRISMA and Cochrane, with which any society dealing with evidence-based clinical practice should be familiar. For consensus clinical recommendation papers, there is usually a set pattern of reporting. The consensus process is at least outlined, if not given in great detail to show how the decision makers were selected, evidence appraised, bias reduced, and consensus reached. For clinical practice guidelines in general, the Equator Network has the AGREE guidelines.The author guidelines for the journal could require use of such a structured approach for guidelines and practice recommendations.
Third, competing interest declarations by all authors must be on file. The journal should also ask for a competing interest statement from the society that seems to be promoting this guideline. Many drug and device manufacturers fund projects for societies. They should be funded as unrestricted educational grants, but that is not always the case.
Whatever the outcome, the editorial board might like to consider conducting a post-mortem on this incident and look at best practices and their internal editorial processes.