This journal requires authors to provide a statement of financial or other relationships that might lead to conflict of interests. Is this appropriate? Is this enough? What should we be doing in the future?
What is conflict of interest?
Financial interests are relatively easy to define and include employment, consultancy, sponsorship of research, lecture fees and support for participation in scientific meetings. Although financial interests are easy to quantify, Krimsky1 usefully reminded us that the existence of a financial interest does not imply a conflict and the potential financial gain is only one of many factors that can generate such conflicts.
Relationships between clinicians and pharmaceutical companies are widespread. As such, we cannot ignore them; rather, we have to understand and then come to terms with them. In the past 15 years, my own relationships with industry have included: paid employee; retained consultant; investigator in regulatory studies; collaborator in mutually interesting projects; receipt of lecture fees, travelling expenses and accommodation; recipient of research grants and free drug for local research projects, and, finally, a few book tokens for filling in questionnaires. Should all these interests be declared? If so, how often and to whom?
Research is an intellectual activity and few of us engage in a project without some prior view on the hypothesis we are testing. Commitment to a line of research, possibly extending over some years, may generate real conflicts when designing experiments or reviewing the work of others.2 Are these conflicts less important simply because they are unpaid? Possibly they are more worrying because they are harder to spot and easier to hide. Pressure groups are an everyday feature of modern medicine. The anti-smoking, anti-alcohol and even anti-capitalist brigades are well organized and effective. Do we need to be as cautious of them as we already are of the cigarette, alcohol and pharmaceutical industries?
What about our clinical practice? Some clinicians earn the bulk of their income from a particular procedure, such as coronary angioplasty or gastroscopy. Are they likely to be objective when approving, reviewing or funding research that challenges this? Anaesthetists may scoff at this as irrelevant to our speciality, but areas of our practice, including lumbar endoscopy for low back pain3 and accelerated detoxification from opioids,4 are procedures which are potentially lucrative in private practice but not yet evidence based. People who practice these are best placed to research them but are they free from conflict of interest?
Why should we bother? Is there a problem?
Conflicts of interest are common and there is nothing wrong with having a conflict of interest. What is more of a problem is not to declare a conflict of interest.5 Major discussions on conflict of interest have followed research that attempted to investigate correlations between interests of authors and the outcomes of their research.
Calcium-channel antagonists
Calcium-channel antagonists are widely used in hospital and primary care and represent a valuable sector of the pharmaceutical market. Their role and safety in a range of indications have been questioned in a large number of publications. Stelfox6 and colleagues reviewed articles examining the controversy about the safety of calcium-channel antagonists and found that authors who supported the use of calcium-channel antagonists were significantly more likely than neutral or critical authors to have financial relationships with manufacturers of these drugs (96% compared with 60% and 37%, respectively; P < 0.001). The paper was criticized on technical grounds and because the authors of the critical papers were primarily epidemiologists who do not require research funding from pharmaceutical companies but may have other conflicts of interest of their own.7 Non-financial conflicts of interest may include fanaticism about a single issue, political commitment, philosophical bias and commitment to a particular theoretical framework for solving a problem.2
Passive smoking
The potentially harmful effects of passive smoking have driven widespread changes in attitudes towards smoking and smokers. Barnes and Bero8 reviewed 106 review articles on the health effects of passive smoking and reported that the only factor associated with concluding that passive smoking is not harmful was whether an author was affiliated with the tobacco industry. Their findings support allegations that most published reviews are unsystematic and that their quality is low.
Sevoflurane and Compound A
The olefin, Compound A, is generated by interaction of sevoflurane with soda lime. In 1994, Anesthesiology published two high-quality papers demonstrating the toxicity of Compound A to the rat kidney and exploring the effects of increasing concentration and duration of exposure.9 10 Sevoflurane is manufactured by Abbott Laboratories. The senior author of the toxicology papers was Professor Eger, who is a long-time consultant to Ohmeda, the manufacturer of desflurane, a rival inhalational agent. The apparent conflict of interest concerned reviewers and members of the Anesthesiology Editorial Board and an accompanying editorial was published.11 Eger subsequently explained the detailed process he had followed to allow external scrutiny of the work and to comply with editorial requirements.12 A letter from the Editor gently reminded readers that this work could have been done at another laboratory where the conflict of interest could have been minimized.13
Looking back on this episode with several years of hindsight, we can reasonably conclude that the science was good and the process transparent. Professor Eger has continued his crusade against sevoflurane and by May 2000, Medline listed 42 publications from him about sevoflurane and Compound A, including: descriptions of the breakdown of sevoflurane by soda lime;14 sevofluranes slower recovery than desflurane;15 the effects of sevoflurane and Compound A on renal function;16 the kinetics and effects of Compound A;15 its toxicological affects on hamster ovary cells,17 and its interaction with acetaminophen (paracetamol).18 Professor Eger is a clinician scientist of considerable stature; however, it might have been preferable if an independent research worker with no affiliations to the manufacturer of a rival drug had undertaken such an exhaustive series of investigations into the potential toxicity and disadvantages of sevoflurane.
Papers versus reviews
Review articles and textbooks allow the authors to express their own opinions; the degree of such expression is limited by the style and content of the piece. A modern literature review describes the preliminary search process, uses explicit criteria for including or excluding its sources and is explicit about the quality of the supporting evidence available for each of the questions it addresses.19 At the other extreme, an undergraduate textbook may express the personal opinions of the author unsupported by any facts whatsoever.
Scientific papers usually adopt the Abstract, Introduction, Methods, Results and Discussion structure. Opportunities for inaccuracy exist in every section of the paper. The abstract may selectively quote a subset of the results. The introduction may exaggerate the importance of the clinical problem to which a new compound offers a solution. Methods may be structured so that a particular outcome is highlighted or underplayed. Results cannot be altered except by fraud, but analysis is a subjective business unless exhaustively specified before a study commences. Finally, the discussion allows the authors to set their findings in the context of other work and broader clinical issues; again, this is subjective.
Editors and assessors
Submission of a manuscript to a journal is an act of trust. Editors must act impartially within the declared editorial policy of the journal. Likewise, external assessors who are sent a manuscript for review must perform their duty without prejudice. The British Journal of Anaesthesia now requires assessors to reveal whether they have a conflict of interest when reviewing a manuscript. Such conflicts are defined as financial (e.g. shares in the sponsoring company or its competitor, or receipt of funding from the sponsor) or being an academic competitor of the authors.
Responsibilities of the reader
Review articles and original scientific reports contain subjective elements and therefore offer a distorted view of reality; the only variable is the degree of distortion. Readers must adopt a critical approach to published material and be prepared to make judgements about the quality of the methodology and analysis that is presented to them.
What are the consequences for journals?
Conflicts of interest damage journals as well as authors. In 1997, the New England Journal of Medicine published a review20 by Dr Gerry Berke of the book Living Downstream: an Ecologist Looks at Cancer and the Environment, by Sandra Steingraber. The review described the book as an environmental polemic, not. .. supported by serious scientific analysis, obsessed. .. with environmental pollution as the cause of cancer. The reviewer did not reveal that he was the medical director of a chemical company that was currently under criticism for polluting drinking water and contaminating soil. The publication of this review greatly embarrassed the New England Journal of Medicine.
In the future, journals that are less than rigorous in identifying and addressing conflicts of interest may find themselves in litigation with aggrieved authors, companies or publishers.
Interactions with pharmaceutical companies
Commercial drug development is driven by the limited patent protection, usually 20 years, given to new molecules. Pharmaceutical companies need to refresh their portfolios with patent-protected compounds to replace the revenues lost when older drugs are lost to generic manufacturers.
For a new, and usually more expensive, drug to replace an older one, it must claim novel benefits. These claims must be supported by data. Companies therefore seek data that support their ambitions for their products. Clinical investigators advise pharmaceutical companies and test their compounds. Inevitably some of the commercial enthusiasm for a project transfers itself to the investigator.
The pharmaceutical industry is currently in rapid retreat from anaesthesia, with few compounds in development for a market which is relatively small compared with those for drugs to treat cardiovascular disease, cancer or chronic pain, and unlikely to expand. Current commercial battles (Table 1) will run on for a few years and new ones may replace them. Rather than hide these conflicts or stigmatize those who participate in them, we should campaign for greater transparency so that readers may decide for themselves.
|
The British Medical Journal has provided constant editorial innovation in biomedical publishing. In 1998, it revised its policy on conflicts of interest and now requires authors to declare competing interests, suggesting that authors may be concerned that admission of a conflict of interest may imply wickedness or that they may be so confident that they are not influenced by a conflict of interest that they do not declare it.21
Asai and Shingu22 examined the instructions to authors of English language anaesthesia journals. Seven of the 11 examined during 1998 required some statement about conflict of interest; however, the extent of disclosure required is variable. Anesthesiology provides a comprehensive draft covering letter and lengthy conflict of interest statement which must be completed for all submitted manuscripts 23. The British Journal of Anaesthesia at present manages with a single sentence (see above).
What should the British Journal of Anaesthesia do next?
The editorial board is reviewing our policy on conflict of interest. From the beginning of 2001, the Editors of the British Journal of Anaesthesia plan to increase the degree of disclosure required from authors and the extent to which that information is revealed to the reader. Ideally, the editors of all leading anaesthesia journals would agree a uniform approach to this matter. This may take some time!
J. Robert Sneyd
Department of Anaesthesia,
Derriford Hospital,
Plymouth, PL6 8DH
References
1 Krimsky S. Conflict of interest and cost-effectiveness analysis. J Am Med Assoc 1999; 282: 14745
2 Horrobin DF. Beyond conflict of interest. Non-financial conflicts of interest are more serious than financial conflicts. Br Med J 1999; 318: 466
3 Igarashi T, Hirabayashi Y, Shimizu R, Saitoh K, Fukuda H. Thoracic and lumbar extradural structure examined by extraduroscope. Br J Anaesth 1998; 81: 1215
4 Hensel M, Kox WJ. Safety, efficacy, and long-term results of a modified version of rapid opiate detoxification under general anaesthesia: a prospective study in methadone, heroin, codeine and morphine addicts. Acta Anaesthesiol Scand 2000; 44: 32633[ISI][Medline]
5 Whitaker L. Conflict of interest. Does the BMJ attempt independent ascertainment?. Br Med J 1996; 313: 1555
6 Stelfox HT, Chua G, ORourke K, Detsky AS. Conflict of interest in the debate over calcium-channel antagonists. New Engl J Med 1998; 338: 1016
7 Meltzer JI. Conflict of interest in the debate over calcium-channel antagonists. New Engl J Med 1998; 338: 1696
8 Barnes DE, Bero LA. Why review articles on the health effects of passive smoking reach different conclusions. J Am Med Assoc 1998; 279: 156670
9 Gonsowski CT, Laster MJ, Eger EI, Ferrell LD, Kerschmann RL. Toxicity of compound A in rats. Effect of a 3-hour administration. Anesthesiology 1994; 80: 55665[ISI][Medline]
10 Gonsowski CT, Laster MJ, Eger EI, Ferrell LD, Kerschmann RL. Toxicity of compound A in rats. Effect of increasing duration of administration. Anesthesiology 1994; 80: 56673[ISI][Medline]
11 Saidman LJ. Unresolved issues relating to peer review, industry support of research, and conflict of interest. Anesthesiology 1994; 80: 4912[ISI][Medline]
12 Eger EI. Motivation, bias and scientific integrity. Anesthesiology 1994; 81: 2701
13 Saidman LJ. Motivation, bias and scientific integrity. Anesthesiology 1994; 81: 271
14 Eger EI, Ionescu P, Laster MJ, Weiskopf RB. Baralyme dehydration increases and soda lime dehydration decreases the concentration of compound A resulting from sevoflurane degradation in a standard anesthetic circuit. Anesth Analg 1997; 85: 8928[Abstract]
15 Eger EI, Gong D, Koblin DD et al. The effect of anesthetic duration on kinetic and recovery characteristics of desflurane versus sevoflurane, and on the kinetic characteristics of compound A, in volunteers. Anesth Analg 1998; 86: 41421[Abstract]
16 Eger EI, Koblin DD, Bowland T et al. Nephrotoxicity of sevoflurane versus desflurane anesthesia in volunteers. Anesth Analg 1997; 84: 1608[Abstract]
17 Eger EI, Laster MJ, Winegar R, Han C, Gong D. Compound A induces sister chromatid exchanges in Chinese hamster ovary cells. Anesthesiology 1997; 86: 91822[ISI][Medline]
18 Laster MJ, Gong D, Kerschmann RL, Eger EI, Martin JL. Acetaminophen predisposes to renal and hepatic injury from compound A in the fasting rat. Anesth Analg 1997; 84: 16972[Abstract]
19 Greenhalgh T. Papers that summarise other papers (systematic reviews and meta-analyses). Br Med J 1997; 315: 6725
20 Berke JH. Book review: Living Downstream: an Ecologist Looks at Cancer and the Environment. New Engl J Med 1997; 337: 1562
21 Smith R. Beyond conflict of interest. Transparency is the key. Br Med J 1998; 317: 2912
22 Asai T, Shingu K. Ethical considerations in anaesthesia journals. Anaesthesia 1999; 54: 1927[ISI][Medline]
23 Anesthesiology home page: http://www.anesthesiology.org/