The Center for Biomedical Research Transparency (CBMRT) was established in 2017 with a mission to increase transparency in biomedical research reporting. CBMRT aims to ensure that all high-quality biomedical research results – including negative and inconclusive results – are accessible to researchers and clinicians in the interests of patient safety and research efficiency. In pursuing its mission, CBMRT runs two main initiatives: Null Hypothesis (H0) and the Biomedical Transparency Summit series.
The goal of CBMRT’s H0 initiative is to help address the systematic issues associated with publishing negative, inconclusive, and replicative results. Despite the importance of reporting research outcomes, many experimental results remain unpublished, sitting in lab books around the world (“dark data”). H0 is a special edition journal for well-designed yet negative or inconclusive studies, and replication work that is presently difficult to publish through traditional scholarly channels. It is published in partnership with major scholarly societies and journal publishers, using their existing infrastructure to source and peer review papers. As emphasized by Dr Sandra Petty, CEO of CBMRT:
“There is significant value in publishing these under-reported research outcomes in terms of informing future research design, reducing funding wastage, and promoting patient safety.”
In partnership with Neurology® and the American Academy of Neurology, CBMRT launched Neurology® H0 in 2018. To date, over 30 papers have been submitted, with accepted papers published ahead of print online and freely available at Neurology Today. The first print edition of Neurology H0 was circulated to over 34,000 subscribers as a special edition in April 2019. The H0 concept is now being expanded into other therapeutic areas during 2019, with the goal being to build momentum and positive exposure for the publication of negative, inconclusive, and replicative research studies.
CBMRT’s Biomedical Transparency Summit (BMTS) initiative is a five-year series of free convenings held in the US and Europe. BMTS aims to promote discussion and collaboration among diverse stakeholders towards greater transparency at a global level. The first European BMTS – held May 15, 2019, in Paris, France – was very well received, with participants from across Europe and the US. Medical publishers, academics, policy makers, industry, research funders, patient organizations, and technologists were all well represented. Wellcome Trust funding enabled a large number of early career researchers to attend; CBMRT is keen to expand attendance of early career researchers and representatives of patient-centered research organizations at future Summits. See the section below for some highlights from the European BMTS.
CBMRT is building its network of ambassadors; this network promotes Null Hypothesis and extends the call for high-quality negative/inconclusive/replicative studies for publication. If you are interested in joining CBMRT’s ambassador network (receive a quarterly newsletter and invitations to BMTS 2020, and have opportunities to meet/collaborate with stakeholders in biomedical research transparency), please email CBMRT at email@example.com.
Highlights from first European BMTS, May 15, 2019, in Paris, France
Plan S and Wellcome Open Access
An early stakeholder in the open access movement, Wellcome was the first Plan S funder to provide technical details on how a funder could implement an open access policy to support Plan S. Dr Diego Baptista (Wellcome Open Research Coordinator) highlighted some key elements of Wellcome’s open access policy, which was adapted to align with the revised Plan S guidance released in May 2019, that will take effect on January 1, 2020:
- All Wellcome-funded articles must be made freely available immediately upon publication under a CC-BY license.
- Wellcome will no longer fund open access publication costs in subscription journals (unless journal has transformative open access agreement – support until 31.12.21).
- Publications regarding public health emergencies must be published immediately before peer review on an approved platform (new requirement).
- Wellcome-funded organizations must sign or publicly commit to DORA (San Francisco Declaration on Research Assessment) – also new.
Dr Baptista emphasized how Wellcome sees DORA as a key enabler of open access and is jointly funding a DORA community manager to support its implementation. Widespread adoption of DORA should ultimately see researchers being assessed based on their research and its societal impact, with the actual venue of publication becoming less of a driver. The Wellcome Open Research platform is similarly enabling; papers are published within 28 days of submission (with average article processing charges [APCs] of £825 [free for Wellcome grantees]), and the open peer review process takes an average of 72 days.
Beyond open access: Open peer review
Dr Lin Lee (Chief Editor of BMC Medicine) presented a delightfully data-rich case for open peer review. BMC Medicine has had an open peer review system since its launch in 2003 and has been keenly tracking progress in a changing environment:
“Forty-four of the BMC series medical journals use only open peer review, with no opt out for authors or reviewers.
These journals publish over 8,000 papers a year, demonstrating that at least in the medical disciplines,
there is clearly a community willingness to undertake open peer review.”
However, there are ancillary issues, specifically: why are reviewers more reluctant to sign up for open peer review? Are quality, accountability, and accept/reject drivers the same for open peer review? And what does it all mean for editors? Dr Lee cited results from a recent meta-analysis study that suggest the quality of the peer review report is higher and the recommendation to reject manuscripts for publication is lower under an open peer review process.
“It’s naïve to think that open peer review does not have any effect on editorial processes, so editors need to be mindful.
Although, generally, we do not see differences in recommendations/report quality in open peer review,
this does happen in specific instances, and there must be editorial mechanisms to deal with this.”
Pre-prints for clinical medicine
Dr Theodora Bloom (Executive Editor, The BMJ) described the near exponential growth in preprints over the last five years, driven by an increasing preference for research funders to see “interim research outputs,” along with a greater openness to preprints as part of grant applications. Some funders are also providing guidance on how to choose a preprint repository and are launching their own open research platforms, which provide a streamlined pipeline from pre-print through open peer review to submission for publication (Wellcome, Gates Foundation). There is also a generational aspect, with younger researchers more likely to preprint. It is noteworthy that an estimated two-thirds of preprints end up being published.
Dr Bloom also noted the balance to be struck between the benefits of more rapidly and openly sharing results and the potential risk to patient safety of sharing results that have not been fully peer reviewed. As BMJ’s representative in discussions towards launching the eagerly awaited medRxiv clinical preprint server, Dr Bloom is involved in the design of medRxiv’s measures for balancing speed against risk mitigation. These include upfront screening and author declarations (authorship, permissions); a series of automated, Cold Spring Harbor Laboratories Press and affiliate checks; and escalation of “high risk” preprints to medRxiv management.
Cochrane and transparency
Founded in 1993, Cochrane Reviews continues to be driven – in the face of changing evidence and pressure to provide more open access – by the reflection of Archie Cochrane 40 years ago:
“It is surely a great criticism of our profession that we have not organised a critical summary, by specialty or subspecialty, adapted periodically, of all relevant randomised controlled trials.”
Dr. David Tovey (outgoing Editor-in-Chief of Cochrane Library) described the move away from a reliance on data published in scientific journals towards fuller, more comprehensive, granular, and “real world” data. There is also widespread demand for results/data to be made available much more quickly, and for the inherent knowledge to be better translated for use by those who need it to make health decisions.
For its part, Cochrane is beginning to implement “living systematic reviews,” which aim to maintain reviews with data updates almost in “real time.” It has increased content linking and multi-lingual content and added a PICO (Population, Intervention / Comparator, Outcome) search annotator to its library. Dr Tovey observed that machines have increasing capability to undertake some of the “heavy lifting” of review production, and other tools, such as RobotReviewer, can help to assess risk of bias. He also noted the opportunity to better involve journalists in the process of accurately translating science to the public.