Here’s our summary of key points from selected presentations at the 2018 European meeting of ISMPP.
Panel discussion: medical publications and data disclosure – ensuring compliance, efficiency and effectiveness for publication and data disclosure teams
Full disclosure of data from clinical trials is important for many reasons. It facilitates collaboration and the critical evaluation of scientific research, thus improving the quality of the conclusions drawn from the data; it reduces duplication of effort; and it builds patient trust.
In a survey of 1018 US adults conducted in June 2016, only 9% said that they thought pharmaceutical and biotechnology companies put patients before profits. Earning trust is an important priority for the pharmaceutical industry and many people, including 47% of survey respondents, have suggested that greater transparency could be a way to achieve this goal.
Research published at the end of 2015 as part of the AllTrials campaign showed that a median of 57% of clinical trial data on new drugs were publicly disclosed. Levels of disclosure varied dramatically by company, with some meeting the required baseline disclosure levels for every clinical trial they had funded and others falling behind.
In recent years, new legislation mandating greater data transparency has come into force. Although many pharmaceutical industry spokespeople support the drive towards more transparent research, concerns about patient privacy and commercially sensitive data have been raised, and these remain a legal grey area.
The key message from the panel at the discussion was the importance of collaboration between teams: there are many different groups involved in the research and publication process for pharmaceutical companies, and all of these need to work together to achieve appropriate disclosure of research. Training initiatives are an important part of the process but achieving inter-team collaboration on the scale of a large international company conducting thousands of trials per year still represents a significant challenge.
When the AllTrials initiative was discussed, it was clear that the public nature of the campaign has brought the issue to the attention of the executive boards of pharmaceutical companies and prompted change, although the methodology of the underlying research has not always accurately captured disclosure levels. One of the main problems is that many pharmaceutical companies host their own repositories and disclose their research in those rather than in independent registries such as ClinicalTrials.gov.
It was agreed that the overall aim of posting all research in the same repository is a good one but that bringing about this type of collaboration between companies would present a fresh set of challenges, on top of the already difficult task of achieving within-company consistency.
Looking to the future, it was anticipated that integrated disclosure planning may replace publication planning; that a single process of development and disclosure of all studies, whether mandated or not, should – and would – ultimately become the norm.
“…the biggest challenge in our company is the expertise of the researchers and the publication professionals – there are some teams that get it, and some that don’t”
Speed research: does ‘open access’ affect the amount of attention an article receives online?
An increasing proportion of medical research is being published with open access but the discussion about how open access affects a publication’s impact is ongoing. Although open access journals tend to have lower impact factors than the most prestigious journals, many people argue that their increased accessibility could benefit the papers they publish by enhancing their ‘performance’ against non-traditional metrics such as those gathered by Altmetric; there have been some reports suggesting that this might be the case.
This session began with some background information on the landscape of altmetrics and social media in the context of medical research. Online engagement has become an important part of medical education, and social media is increasingly seen by physicians as a good way to share and discuss new medical research, with as many as 85% reporting that they scan social media at least once a week.
For their research, the presenters had selected general and internal medicine journals, all of which were hybrid open access. Each journal was categorized according to its impact factor: high, medium or low. Journals that did not support altmetrics were excluded. Articles that could be viewed without prior registration or payment were defined as open access.
The study looked at the 10 most recent articles from each journal, a total of 94 articles, of which 34 were open access. The mean Altmetric scores of open access articles were much higher than those of restricted access papers, at 113.7 and 31.6, respectively. This result was observed in all impact factor categories but the effect was more pronounced for papers in high- and medium-impact journals.
A suggested area of future research was whether altmetrics would be found to correlate with changes in patient or healthcare provider behaviour.
Round-table session: open access
What are the barriers to open access?
- Authors have mixed feelings about open access. They often perceive open access journals (or, more specifically, online-only journals) to be of inferior quality. Impact factor is still the most important determinant of target journal for most researchers. A way to overcome this would be for pharmaceutical companies to introduce open access policies but, at present, although many are supportive of open access, they rarely mandate it in the way that, for example, the Wellcome Trust and the Bill & Melinda Gates Foundation do.
- Authors often do not know about the breadth of open access options available.
- Cost is frequently cited as an issue with expanding open access, although article processing charges rarely exceed 1% as a proportion of total research cost. ‘Dissemination budgets’, as distinct from publication budgets, were mentioned; these may be one way to allow for growing open access costs
- The value added by the services paid for by the article processing charge is not always considered to be very high. For example, the fully formatted and typeset article only appears when the print article is published, which can be months after online publication.
Licensing: what are the benefits of an open licence?
- If an article is not published with a CC BY licence, publisher permission for translating papers into other languages must be sought; there are frequently other barriers, too.
- Publishers are reluctant to allow all authors a CC BY licence as it cuts into their reprint income and undermines their business model. Online-only journals frequently struggle with their revenue streams for these reasons.
Speed research: should we consider patients in communication plans?
Today’s patient is very likely to be ‘Patient 2.0’: well-informed, and able to play a role in drug research, development and approval processes. With this in mind, the research discussed in this session was initiated following a report that fewer than 20% of journals offer patient-level information.
Taking dermatology as the therapy area to be explored, the researchers sent a survey to 1200 patients and caregivers from a non-medical background; of these, 100 patients and 50 caregivers responded. The sample was skewed towards educated patients: about 30% were educated at college level or higher, which is a similar proportion to that observed among those seeking health information online.
Among the findings highlighted in this session were the following.
- A total of 50% of survey respondents said that they had sought information from medical journals; 72% of these felt that the information in such publications was relevant and empowering, and helped them to optimize their treatment and stay up-to-date with treatment strategies.
- The ‘patient-friendliness’ of information published in medical journals was thought to be somewhat lacking: 53% of respondents felt relatively confident with publication-level information but expressed the desire for more widespread use of plain language.
- Around 40–50% of respondents said that they spent too much time looking for information and suggested that there was too much jargon and too much extraneous information: they wanted results to be easy to find and clear to understand.
- The idea of patient lay summaries was popular: 90% of respondents said that patient lay summaries would be useful to them in discussions with their healthcare providers and would enable better engagement with their diseases or conditions.
Parallel session: facing the challenges of publishing unfavourable, negative, equivalence or non-confirmatory data
The challenges of publishing what is typically considered ‘less exciting’ research are particularly pressing for the pharmaceutical industry. Good Publication Practice version 3 guidelines require the publication of all scientific and clinical research in a “complete, accurate, balanced, and timely manner”;[1] however, the bulk of industry-sponsored research is not exciting phase 3 trials but safety trials and epidemiological studies, and publishing the results of such research in traditional journals is a time-consuming and costly process that adds no appreciable value to the quality of the work. In recent years, so-called ‘mega journals’ following the model developed by PLoS ONE have been proposed as a way to address these difficulties.
Posters
Do journals restrict access to commercially funded research?
This poster explores the different open access variants made available by high impact factor medical journals, and seeks to clarify the availability of these variants to authors whose work was funded by the pharmaceutical industry.
Read the preprint of this research here
Ongoing ORCID implementation within the pharmaceutical industry: an Open Pharma and GSK Vaccines initiative
This poster details the progress of an initiative implemented by GSK Vaccines in collaboration with Open Pharma, in which a requirement for GSK internal authors to publish using an ORCID identifier was piloted.
[1]Battisti WP et al. Good publication practice for communicating company-sponsored medical research: GPP3. Ann Intern Med 2015;163:461–4. Available from: http://annals.org/aim/fullarticle/2424869/good-publication-practice-communicating-company-sponsored-medical-research-gpp3 (Accessed 23 January 2018).
