Publication-ethics guidance and professional codes

Publication-ethics guidance supports decisions about the integrity of research and its publication. Guidance bodies advise on difficult cases, while ethical codes identify duties shared by researchers, reviewers, editors and readers.

Guidance and case advice

The Committee on Publication Ethics (COPE) advises editors and publishers on publication ethics and handling suspected misconduct, and provides a forum for discussing cases. ^{1, 2} It does not investigate individual cases: editors should ensure that the appropriate institution or employer investigates. ¹ The World Association of Medical Editors (WAME) is another international publication-ethics body. ³

COPE's searchable case database addresses plagiarism, authorship disputes, informed consent and conflicts of interest through anonymised cases, summaries of advice and, sometimes, accounts of the eventual resolution. ⁴ Examples include an author sending an editor a gift during review, consent forms signed after experimental treatment, and an author asking for removal from a paper apparently after another author rejected romantic advances. ⁴ COPE also provides guidance for new editors and flowcharts for suspected plagiarism or data fabrication, with translations into languages including Chinese, Arabic, Farsi, Spanish and Croatian. ⁵

When data fabrication or falsification is suspected but evidence is incomplete, questions include what action is appropriate, who should act, how quickly to respond, whether further investigation is needed and what sanctions, if any, should follow. ²

The Center for Biomedical Research Transparency aims to increase transparency in biomedical research reporting. ⁶

An ethical framework for research

Researchers have ethical duties to participants, institutions, disciplines, professions and publication outlets; reviewers, editors and readers share responsibility for interpreting and applying research with integrity. ⁷ Ethical conflicts need a framework of values and often have no black-and-white answer. ⁸ The six tenets of the Society for Technical Communication's ethical guidelines have the following applications to research. ^{8, 9}

1. Legality. With human subjects, especially in medical research, privacy laws apply; once results are published, copyright governs their later academic or commercial use by others; researchers sponsored by or partnering with a corporation must not compromise its intellectual property or proprietary information in reporting their findings; and the same holds for data a government treats as sensitive or classified.

2. Honesty. It applies to researchers, reviewers, editors and readers, although it is often simply assumed. It governs the gathering of data, which can be falsified, or obtained from a survey slanted to give the desired result; the presentation of data, where lying with statistics and misleading visuals are what reviewers and editors should referee; and the interpretation and spreading of results by readers. A cautionary scene shows the failure: four laboratory partners, under pressure to get an A, heat 10 grams of aluminium to Y°C, immerse it in 100 millilitres of water at Z°C, measure the resulting temperature change and apply the formula for specific heat; one asks what temperature another has, the answer is "What do you need?", a third says 71°C would be about right, and the fourth, squinting at the thermometer, announces 70.9°C.

3. Confidentiality. It overlaps legality on participants' privacy: researchers are bound to protect participants' anonymity in recording, archiving and reporting data, and to make them aware of confidentiality issues through disclosure and release forms. It also governs the blind peer review that decides which articles a journal accepts.

4. Quality. For reporting research, quality has four attributes.

   1. Objectivity means having no bias or self-interest in interpreting and presenting results — the example is a pharmaceutical or medical project whose primary goal is market share rather than patient wellness — so that data are judged on their merits rather than slanted to serve the researcher or the sponsor. The advertising claim that four out of five dentists surveyed recommend sugarless gum for their patients who chew gum shows how: it omits how many dentists were approached and how many declined, and its restrictive clause and the word recommend turn a narrow caution aimed at patients already chewing gum into what reads as a broad endorsement.
   2. Thoroughness means appropriate and sufficient methods and populations, with statistical analysis or an equally rigorous check of validity and reliability, because a single study is unlikely to be incontrovertible and the more a writer knows of related studies and their limits, the more credible and ethical the use of the findings. If 60 of 100 US technical communicators judged Helvetica more legible than Times New Roman in a side-by-side comparison of two one-page business letters, it would be wrong to say that a majority of technical communicators prefer Helvetica, or to switch an organisation's documents to it because research proves it more legible; the only logical and ethical claim is that the finding is potentially significant and merits research with other populations and other documents.
   3. Accuracy runs through collection, measurement, analysis and reporting, and means avoiding exaggeration and gross simplification as well as measuring correctly. A television station's report that 60% of callers support a proposition and 40% oppose it cannot be judged without knowing when and for how long calls were taken, whether they were solicited, how many there were, whether any came from the same caller, and whether 60% is exact or rounded and in which direction; the more accurate report is that the station asked viewers to call, received 36 calls from different numbers on Monday evening between 6 and 9 PM, and counted 21 in favour and 15 against.
   4. Clarity means presenting information as simply as possible without oversimplifying, and the question it raises, clear for whom, is answered by the audience: a research journal article can be written at a scholarly level without lapsing into the undefined jargon of its specialty or needlessly elaborate academic language that seems intended to impress rather than inform.

5. Fairness. Rooted in honesty and respect for others, it asks a researcher on a controversial subject to present all sides of the debate even when disagreeing with some of them, and it matters most in constructing surveys, gathering data, and administering tests and interpreting their results; culturally biased tests that purport to show one race genetically inferior in native intelligence are an example of what its absence produces.

6. Professionalism. Communication products and services should be evaluated constructively and tactfully, and communicators should seek definitive assessments of their own performance; this applies directly to decisions to publish research with minor editing, with revisions, or not at all.

^{8, 9}

Harm from publishing findings

A proposed ethical concern is that research can be permissible to conduct yet harmful to publish, as when dual-use findings could help terrorists reproduce dangerous work. ¹⁰

Two medical examples illustrate the concern about publication harm. ^{10, 11, 12} In a Danish study, antibiotics reduced symptoms of a widespread chronic disease; a reviewer warned that changes in practice could increase antibiotic resistance and harm health. ¹¹ In a British statin debate, an article suggested adverse effects outweighed benefits for people at low or moderate cardiovascular risk, and an estimate suggested about 200,000 people might have stopped treatment, with roughly 2,000 subsequent cardiovascular events expected. ^{10, 12} The journal and authors withdrew claims that adverse events occurred in 18–20% of patients. ¹⁰

The ethical approach to these publication risks has four elements. ^{11, 12}

1. Publication of the Danish and British studies can cause significant harm to individuals.
2. Editors of medical journals have a moral responsibility for potential adverse effects of publishing research.
3. Refusing publication is not an adequate way to fulfil that responsibility.
4. Internationally recognised ethical codes should provide a basis for assessing and mitigating potential effects of publishing medical research.

^{11, 12}