Publication and dissemination of research
Ethics of dementia research
Research into dementia is, or should be, carried out for specific reasons. Usually, the aim is to answer a question, which, once known, may improve the lives of people with dementia and their carers or future generations of people with dementia and carers. Even research carried out for academic reasons (e.g. for the purposes of obtaining qualifications) should be worthwhile. Consequently, it is important to share the findings of such studies so that the potential gain to society can be as widespread as possible. Other reasons include respect for the people who participated in the studies, avoiding the unnecessary duplication of studies and enabling researchers to build on the existing body of research, thereby enabling them to constantly progress in their search for knowledge and understanding.
There are various methods of dissemination, which are partly determined by the type of research and those responsible for the research. Typically, researchers try to publish the results of their studies in peer-reviewed journals. The peer reviewers, who are usually respected scientists in the relevant field of research, do not check the researchers’ actual raw data and calculations but they verify that the research was carried out correctly and was accurately reported (based on the information provided by the researchers). There is therefore a certain element of trust involved but well conducted research should provide a trail of evidence, which others can follow in order to see how the researchers arrived at their conclusions (Krueger and Casey, 2009).
Other means of dissemination include speeches and poster presentations at conferences, information on websites, brochures and books, articles in magazines and newspapers, and governmental or non-governmental reports. The various means of dissemination may reflect different target audiences or the backgrounds of the researchers. The audiences of such disseminated information may be influenced in their assessment of its accuracy and credibility by how and by whom it is reported. On the whole, the general public does not have access to specialised peer reviewed journals as they are expensive and not readily available. Consequently, they often hear about the results of studies through secondary reports (e.g. on television, through the Internet or in newspapers and magazines).
Conflicts of interests based on financial issues
It is important that those reporting the results of their research are open about any information about themselves which might lead other people to suspect their motives or question their objectivity. Such information represents a potential conflict of interests but does not automatically imply that the researchers lack objectivity or integrity.
A potential conflict of interests would include researchers having received sums of money or presents from organisations which could be seen to benefit from the findings. An example might be a researcher who has received money or gifts from an organisation specialising in vitamin tablets reporting the findings of a study which suggests the benefits on cognition of certain vitamin tablets. Another example would be a researcher reporting on a clinical trial for a drug produced by a company in which s/he has shares.
Researchers may be under pressure from the institutions which employ them to have successful studies as those institutions may be dependent on the findings of those studies for subsequent funding (Seigel, 2003).
Other financial issues include the source of funding for a project. The findings of a study into the health benefits of milk, if financed by the milk industry, might, for example, be considered misleading were such information not openly declared. On the other hand, some people might find it perfectly logical that the milk industry would be interested in funding such a study (rather than one into the benefits of meat). The issue is not about which organisations should fund research but rather about being transparent about funding.
Whilst the pharmaceutical industry clearly has financial interests in discovering and developing drugs, those financial interests are completely transparent and understood by the general public. Moreover, the relationship between the companies and the general public is one of mutual benefit. Companies are clearly interested in proving that their product is effective and safe, and the general public is interested in having such products if and when they need them. The companies comply with stringent controls which help ensure that any conclusions they draw are accurate and based on the application of a rigorous scientific methodology. The good conduct of scientific research requires that measures are taken to avoid bias.
Academic interests and other outcome preferences
Seigel (2003) prefers to use the term “outcome preferences” due to the pejorative connotations of the term “bias” and as it has another meaning in the context of statistics. Sometimes, even in the absence of direct financial concerns, researchers may have outcome preferences. In the academic domain, for example, some researchers may be under overt or covert pressure to publish a certain number of articles per year and such publication may be linked to their career advancement (Maj, 2008). Zaki (2011) describes this as the “publish or perish” culture. As studies which do not produce results in support of the hypothesis or research questions are difficult to publish, this creates an undesirable pressure on researchers to produce certain results.
Prestige and peer recognition may add to this pressure to obtain significant and meaningful results. As noted by Seigel, “Fleming is not likely to have been celebrated for screening a host of potential antifungal agents, all found to be inert” (2003, p.3420).
The desire to improve health may also result in a particular outcome preference. Whilst such a desire is commendable, great care is required to ensure that researchers remain objective and that those involved in the secondary reporting of research results present such findings in a realistic manner.
Researchers’ commitment to a hypothesis or allegiance to a particular school of thought may also lead to outcome preferences (Coughlin, 2006; Maj, 2008; Seigel, 2003) which in turn may even lead to what Charles Babbage (1782-1871) described as “cooking”, namely retaining only those results that fit the theory and discarding others (Al Bareeq and Fedorowicz, 2008). In some cases, the firm belief in a particular theory could be based on years of research. Challenging such beliefs and having the courage to question one’s own previous work may be difficult for some researchers but a necessary requirement for good science.
Certain official bodies, which carry out or fund research, may have a preferred outcome (e.g. linked to policy, service provision or legislation) and want to approve documents and have the opportunity to comment on various drafts before they are published. This itself is not unethical. However, it would be unethical if such organisations censored the findings, only allowed the publication of those which supported their arguments, distorted arguments, rejected certain interpretations of the results by the researchers and/or delayed publication (Royal College of Psychiatrists, 2001).
Conflict of interests amongst publishers
Maj (2008) argues that just as researchers may have conflicts of interest, so too might those responsible for publishing such research or reporting on it. Writing in the context of medical and psychiatric research, Maj draws attention to the possible impact of the non-financial interests of publishing organisations as well as individuals or organisations acting as referees or evaluators of research for such journals. These might include strong political commitments, allegiance to particular schools of thought and even protection of their own research from perceived criticism. It is perfectly defensible to have strong political interests and to be heavily influenced by a particular school of thought. However, if such factors are not known or there is no balance in terms of publishers with alternative political views, there may be an imbalance in the publication of research, resulting in a biased portrayal of knowledge and the unnecessary duplication of studies.
Integrity of publishers and researchers
It would be unethical to publish inaccurate research findings based on studies involving the use of fraud, the fabrication of results, the falsification of data, plagiarism (Al Bareeq and Fedorowicz, 2008) or unethical practices such as failing to obtained informed consent. For studies involving participants with dementia, this would include failure to consider appropriate methods to obtain consent from people with dementia before resorting to proxy consent.
Publication of negative or inconclusive findings
The publication of negative or inconclusive findings may be hampered by a reluctance on the part of publishers to publish such findings as well as on the part of researchers to be publicly associated with such studies. Some researchers would gladly publish such results but find it difficult to find publishers. However, as mentioned above, failure to publish negative or non-conclusive findings (or much delayed publishing) may result in the duplication of research and exposing additional participants to risk and burden. Furthermore, it would be misleading and hence unethical to publish one study which was successful and to refuse a further nine which were inconclusive or negative involving the same chemical compound or intervention.
In addition to the misleading nature of such information, the failure of nine studies, if known, might affect the readiness of people to participate in a new study. This would clearly be problematic for the researchers but does not justify withholding such information.
Results which do not support the hypothesis of a particular study or which provide an inadequate or conflicting response to the research question are valuable in terms of the acquisition of knowledge. They might point to new hypotheses or research questions, or suggest the need for changes to future research protocols or choice of methods.
Negative findings are also of interest to lay people in their everyday lives. Many lay people react to the information they receive directly or indirectly about possible preventive measures and should be able to rely on researchers to give them the full picture with regard to research findings. For example, the publication of a study which suggests that certain food supplements or substances, used by lay people to prevent dementia, have no effect whatsoever, may help prevent those people from being misled, from wasting money in health shops and from having a false sense of security.
Gift and ghost authorship
According to the International Committee of Medical Journal Editors (ICMJE), also known as the Vancouver group, an “author” is generally considered to be someone who has made substantive intellectual contributions to a published study. More specifically,
“An author must take responsibility for at least one component of the work, should be able to identify who is responsible for each other component, and should ideally be confident in their co-authors’ ability and integrity.“ (ICMJE, 2010, p.2)
Authors should also be involved in the drafting of the article, its revision and final approval before publication. The prevalence of gift authorship (based on failure to comply with ICJME criteria) was found to range from 0.5% to 60% of research papers as measured in a selection of peer-reviewed journals and 60% of senior researchers in one university are believed to have benefited from gift authorship (Pignatelli et al., 2005 and Mowatt et al., 2002 in Hodgkinson, 2007; Zaki, 2011).
Gift authorship is linked to issues of honesty, integrity, trustworthiness, possible manipulation and the inappropriate use of power. If the list of authors is not in accordance with the recommendations made by the publishers, it could also be construed as scientific misconduct (Albert and Wager, 2003). In some organisations, junior researchers may find it difficult, if not impossible, not to include the names of senior researchers as authors of their research.
Gift authorship may result in people who are not experts in the topic being considered as such and even being asked to review the work of other researchers. They may also be asked to accomplish other tasks, which are beyond their level of expertise, and be perceived as more skilled than their colleagues which may result in unfair professional advantages (Zaki, 2011). Moreover, the issue is not just about giving credit to the people who actually carried out the research but also of responsibility and accountability for published work (Hodgkinson, 2007). Accepting such gifts may therefore represent a professional risk.
Research ethics committees increasingly recommend involving patients and patient representative groups in research which focuses on issues of relevance to them. The inclusion of such groups as authors on articles may be interpreted as a validation of the related studies. In cases where this is gift authorship, researchers may be benefiting from a perceived sign of approval from those organisations. In other cases, patient groups may be consulted during the study and such consultation may be simply reported in the article or in the acknowledgements section of an article. This can sometimes by misleading as, whilst it may be true that organisations or people with dementia were consulted, their opinions and comments might not necessarily be reported or taken into consideration.
Ghost authorship is perhaps not as widespread as gift authorship. It may involve a professional writer who is paid by a commercial organisation to write an article about research in which the former played no active role. Alternatively, it may involve a person who was heavily involved in the research not being included in the list of authors of a published article about that research. The use of professional authors may conceal a possible conflict of interests or serve to present data in the best possible light but if openly acknowledged may simply improve the quality of articles (Hodgkinson, 2007).
A useful article by Albert and Wagner (2003) provides guidance on how to handle authorship disputes. Some of their suggestions are included in our recommendations below.
- Researchers should question their findings, consider competing explanations for their findings and take the necessary measures to check the validity or trustworthiness of their conclusions.
- All published work should be accompanied by a statement of possible conflicts of interest.
- All researchers should be willing and authorized by funders to submit their findings to expert scrutiny if justifiable reasons for such scrutiny are provided.
- In order to avoid confusion, researchers should clearly distinguish in published articles and reports between research criteria for Alzheimer’s disease and clinical criteria for Alzheimer’s disease.
- Organisations responsible for the secondary reporting of research findings should strive to be objective and clear in their reporting, providing additional explanations if necessary in order to facilitate understanding of the possible implications of the findings.
- Unrealistic or sensationalist reporting should be avoided.
- Patient groups and related organisations should challenge sensationalist reporting of dementia research and try to present findings in a more balanced way.
- Publishers should be encouraged to publish the results of research which was inconclusive or negative.
- Governments should set up an obligatory national registry for clinical trials combined with a database with details of the results of all clinical trials, which should be freely accessible to the general public.
- Researchers should follow the requirements of the International Committee of Medical Journal Editors regarding authorship (please see the references section for details).
- Researchers should discuss authorship issues right at the start of a project, preferably at a face-to-face meeting and make a written record of any decisions taken.
- Researchers should acknowledge contributions from people who do not qualify for authorship but did contribute towards the research project.
- People or organisations who were consulted in the course of a study should not be individually named in a published article without their consent.
- Local customs and practices which condone gift or ghost authorship should be challenged by researchers.
- All listed authors should be given the opportunity to see the final draft of an article before it is submitted for publication and to withdraw their names if they so desire.
Last Updated: jeudi 29 mars 2012