Letters to the Editor
Conflict of interest in medical journals
- Chris Johnstone, Peter Grant, Liljana Sokolova, Rajko Igić, Peter McLaren, John Dowden
- Aust Prescr 2015;38:79-81
- 1 June 2015
- DOI: 10.18773/austprescr.2015.036
I would like to thank John Dowden for summarising this important issue so well in his editorial (Aust Prescr 2015;38:2-3). Simply reporting a conflict of interest is pointless because it does not seem to change the interpretation or impact of the information presented. Clinical practice is largely determined by opinion leaders. With respect to publishing restrictions, this is the group we need to target, particularly because of their financial conflicts which are the most concerning.
These people write the most influential literature such as reviews, editorials, meta-analyses, guidelines and position statements. A contemporary example is the literature around the efficacy of stroke thrombolysis, which has polarised clinicians worldwide. This controversy is complicated by the fact that virtually all the interpretative literature supporting the therapy has been written by people with financial conflicts.
What we need is a policy that prohibits the involvement of authors with financial conflicts in these interpretative types of articles. This is achievable, and in the long run will discourage clinicians from accepting money from industry for fear of limiting their academic careers.
Director, Emergency Medicine
Caboolture Hospital, Qld
Yes, readers are informed when a conflict of interest is declared. But no reader knows how to adjust the take-home messages from such articles to compensate for possible bias.1 Declaring a conflict of interest shifts the responsibility of dealing with potential bias from the writer, through the editor, to the reader. Is this fair?
The author asks how hard will it be to find non-conflicted authors in Australia. Do we need someone who has done the primary research when any cardiologist with information-literacy skills could evaluate new antihypertensive drugs? In the age of evidence-based medicine, writing these articles is mostly literature reviewing. For opinion-based parts of an article, why is the opinion of someone less deeply involved less valuable than one likely to be influenced inappropriately?
You don’t need to refuse to deal with people with a conflict of interest. Could you get a non-conflicted author to write the article, then ask your conflicted reviewer to edit it and give the final say to the writer? Or you could get the conflicted author to write the complicated physiology part, and another author to write the diagnostic or therapeutic sections?
No, we shouldn’t be concerned about authors funded by the National Health and Medical Research Council (NHMRC) because the aim of NHMRC aligns with doctors’ duty of care to their patients. In contrast, profit-seeking drug makers’ primary obligations are to their shareholders. If those don’t align with patients’ interests we have a problem.
Lastly, is Australian Prescriber complicit if doctors are flouting the medical board requirement to not accept gifts of more than trivial value,2 when authors report funding from drug makers for conference attendances?
Retired health professional
We read a very good editorial on conflict of interest by John Dowden, but we find that it deserves comment. We think that all authors who submit a manuscript of any type should disclose their financial and personal relationships that might bias or be perceived to bias an article’s content. Therefore, when the editor of a journal writes an editorial, he also should report the conflict of interest, like some used to do.2 The editorial board should vote to approve the editorial without participation of the editor.
However, if an editorial focuses on therapeutics, other treatments or potential advice for medical practice, the editorial should be reviewed by external referees who report no conflict of interest. This is in addition to the author declaring any conflict of interests.3 Thus, when the editor is the author, they should also disclose their conflict of interest, as all other authors,4 for editorials that will be published in the journal.
Editors should not publish other types of articles such as research or review papers in their own journal. Perhaps the editor may publish information related to the journal or historic articles in the journal, provided it is peer-reviewed. However, the situation of editing and publishing in small and developing countries often makes the peer-review process difficult due to various restraints. Fortunately, Australia is not this type of country.
Institute of Public Health Sombor, Serbia
Pharmacologist and physiologist
Hospital of Cook County
Rajko Igić accepted travel and local living expenses while Editor-in-Chief of the Scripta Medica, Banja Luka (2010 to 2013).
While I believe that we should be on guard against the promotional activities of the drug companies, I often wonder if we might be throwing the baby out with the bathwater. I do believe that many of the past developments in health care have been assisted by close association with the providers of our therapeutic armamentarium. Examples include the early association with a good apothecary, the development of the syringe, right up to Ian Frazer and the human papillomavirus vaccine, and Fiona Woods and spray-on skin.
The practice of denying all access to industry is rather akin to the army introducing a new assault rifle without talking to the soldiers about their requirements or gaining feedback on its problems after its introduction. It has been suggested that generic manufacturers have added fuel to this discussion as they do not develop new drugs or technologies and therefore have no need to educate the users or hear their feedback.
What John Dowden and others, who may occasionally suckle at the government teat, consistently fail to mention is how devastating it would be to their career or their position on a government panel if they were to spout a line of doctrine that was diametrically opposed to the government or their academic institution. I have never seen mention of this in any article. Surely the loss of income from not towing the line is just as persuasive as a plate of soggy salad sandwiches?
For example in the February 2015 issue of Medical Journal of Australia, an article concluded that cancer patients achieved better outcomes if they were treated in a major cancer institute.5 The contact details for that author showed that his employment was at a major cancer institute. Are academics so beyond reproach that they don’t feel the need to at least state the obvious?
I welcome continued discussion on the issue of the pervasive influence of company marketing on medical decision making, but I feel that a little perspective on this issue is warranted.
Peter McLaren hosts clinical meetings that may have industry-sponsored catering.
John Dowden, the Editor of Australian Prescriber and the author of the editorial, comments:
All of the new drug comments and most of the articles published in Australian Prescriber are written by authors without financial conflicts of interest. While some authors may declare such conflicts, the Editorial Executive Committee believes that the journal’s editorial processes reduce the risk of bias. We have confidence in those processes and believe it is not unreasonable for readers to be able to make their own assessments. This is why it is important to publish any relevant conflicts of interest that have been declared.
The response of readers to conflicts of interest has been studied. A randomised trial involving BMJ readers found that their views of an article were more sceptical if conflicts of interest were declared.6 As Dr Grant suggests, readers are more concerned if the author of the article is an employee of a company, rather than the recipient of a research grant.7
Professor Igić, Dr Sokolova and Dr McLaren raise the question of whether an author’s employment should be declared as a conflict of interest. Our Editorial Executive Committee will consider this, but it would mean duplicating the author’s details given at the start of every article.
The authors of articles in Australian Prescriber do not have to be involved in primary research, but sometimes those who have participated in clinical trials know more about the problems of a new drug than the published literature might reveal. As Dr Johnstone points out, opinion leaders are influential and they can be used by the pharmaceutical industry for promotional purposes.8
The industry is moving towards greater transparency regarding the payments made to health professionals.9 Once this information is available, perhaps it will help the Medical Board of Australia to define what constitutes ‘trivial value’.
Some of the views expressed in the following notes on newly approved products should be regarded as preliminary, as there may have been limited published data at the time of publication, and little experience in Australia of their safety or efficacy. However, the Editorial Executive Committee believes that comments made in good faith at an early stage may still be of value. Before new drugs are prescribed, the Committee believes it is important that more detailed information is obtained from the manufacturer's approved product information, a drug information centre or some other appropriate source.
Director, Emergency Medicine, Caboolture Hospital, Qld
Pharmacologist and physiologist