Dear HIFA colleagues,
We have now completed our thematic discussion on Open access to medical research, sponsored by Oxford Pharmagenesis. https://www.hifa.org/projects/open-access
Thank you to everyone who contributed and those who followed the discussion.
There were 130 messages from 25 participants in 12 countries (Australia, Cameroon, Canada, Honduras, India, Jordan, Nigeria, Norway, Switzerland, Tanzania, UK, USA).
Our objectives were largely fulfilled:
- Develop a collective understanding of the challenges around open access
- Consider the open access needs of healthcare professionals (HCPs) in their capacity as authors and as readers
- Explore how to maximise the benefits/drivers of open access and how to minimize the disadvantages/barriers.
As we review the project, we invite you to say:
1. What did you like about the discussion?
2. What did you learn?
3. What could have been better?
A few thoughts from me (Neil) on these questions:
1. I enjoyed hearing diverse perspectives on OA from different countries and professional backgrounds
2. I learned about the value of a pluralistic, flexible, pragmatic approach that builds on the existing strengths of the global evidence ecosystem (especially journal publishing) as opposed to a rigid, ideological, revolutionary approach (but recognise that there is also value in exploring new models)
3. I would like to have seen more sharing of personal, practical experience in the form of anecdotes on the impact of open access.
The Open Access working group has reviewed the messages and notes the following 61 points emerging from the discussion. We have organised these under subheadings:
Q1 IMPACT OF OA
Direct impact
1. We were unable to identify a specific example of where OA to the full text of an article saved a life or improved quality of care, despite repeated calls to HIFA members, two independent approaches to ChatGPT, and a PubMed search
2. A HIFA member was moved to prepare and publish a preprint that concludes: 'the true life-saving value of open-access literature emerges not from singular dramatic cases, but from the aggregated probabilistic gains distributed across entire health systems' https://www.qeios.com/read/U2SYIR
3. OA papers are read and cited more than non-OA papers
4. Several participants from LMICs emphasised the general benefits of OA for their work and their organisations
Indirect impact
5. OA has several positive and negative impacts on the global evidence ecosystem, particularly on the components 'generate evidence', 'publish evidence', and 'synthesise evidence'
6. OA has a positive impact on post-publication communication, including communication on the HIFA forums
7. OA may have a negative epistemic impact if it overly represents English-speakers and research in HICs
8. OA facilitates evidence synthesis as a result of free access
9. OA makes evidence synthesis more time-consuming due to proliferation of low-quality publications
Q2. EXPERIENCE Of OA AS A HEALTHCARE PROFESSIONAL
10. HCPs information needs are met primarily by synthesised and repackaged evidence (eg drug formularies, systematic reviews, clinical guidelines, textbooks, decision tools) rather than by access to full text of research papers
11. High-quality evidence = evidence syntheses that are relevant to their questions, well conducted, timely and that deliver findings in forms that can be understood, packaged (for instance, into frameworks to inform guidelines and other decisions) and acted on by a range of stakeholders
12. There is wide agreement that free access to the full text of research is welcome
13. However, clinical decisions should rarely be based on a single research paper
14. We did not hear any specific examples of the use of OA articles by HCPs
Evidence literacy among HCPs:
15. 1 in 4 clinicians in LMICs believe that abstracts are adequate for clinical decisions
16. The cumulative element of evidence-informed practice is widely unrecognised
17. Single studies and preprints may be misinterpreted or misapplied
Access to subscription journals:
18. HCPs in HICs are often able to access subscription journals through their institutional subscription (with many exceptions)
19. HCPs in LMICs are often able to access subscription journals through Hinari (with many exceptions)
Q3. EXPERIENCE Of OA AS A RESEARCHER
Excessive APCs is the most important disadvantage of OA
20. Some journals are profiteering from excessive APCs
21. APCs that are inflated above costs can lead to business driven by quantity rather than quality
22. APCs discriminate against those with less money
23. APC waivers are confusing, haphazard and inconsistently applied
24. APC waivers are inequitable in that they label LMIC researchers as unable to pay
25. Some institutions have transaction agreements to cover APCs
26. Researchers can publish preprints or in APC-free journals if they cannot pay APCs (but the impact of their research and academic recognition may be consequently limited)
Q4. HOW TO DESIGN A NEW SYSTEM (1/5)
General considerations
27. Knowledge translation requires a coherent and functional global evidence ecosystem
28. The solution to translating evidence into practice requires a whole knowledge ecosystem approach in which different actors work together
29. We need to consider the impact of OA on the global evidence ecosystem and on each of the six components of the system (generate, publish, synthesise, repackage, avail, apply)
30. An ideal OA system is one that strengthens the global evidence ecosystem, not one that benefits any specific stakeholder group (researchers, publishers...)
31. Systems should be designed collectively and in cooperation regionally and globally.
32. We need to support a multitude of publishing models. A universal model is neither achievable nor desirable.
Model: Cap APCs
33. Rationale: Journals (whether OA or non-OA) are a vital part of the global evidence ecosystem. We need an evolution.
34. Research funders should cap APCs to prevent inflated APC costs
35. Journals should be transparent about real costs of article processing
36. Aligned with Plan S, which advocates that funders or research institutions should take responsibility for APCs, which should reflect costs
37. Plan S already has wide engagement from all relevant stakeholder groups
Model: Gates Foundation OA policy/Stop paying APCs
38. Rationale: Journals are profiteering from APCs, we need a revolution
39. Pros: supports exploration of non-APC models
40. Pros: reduces journal profiteering
41. Pros: aligns with Plan U (below)
42. Cons: reduces the role of (APC-funded) journals, which are a key part of the global evidence ecosystem
43. Cons: breakaway from Plan S consensus
44. Cons: reduces choice from researchers
45. We are unable to identify any funder that explicitly agrees with or replicates Gates OA Policy
Model: Mandatory preprints
46. Plan U advocates that researchers should post a prepint on an OA repository before publication in a journals
47. Plan U support is much less than Plan S
48. The Gates OA Policy requires that researchers post a preprint
Limitations of preprints
49. Preprints are liable to overinterpretation or misinterpretation.
50. Preprints can offer early insights but should not be the basis for clinical decision-making.
51. Preprints have a limited role (eg rapid communication among researchers in a fast-moving field)
52. Preprints are less robust than final papers. They are susceptible to be misinterpreted and misused, adding to misinformation and disinformation
53. Preprints are even less useful to the majority of users than the final research paper
54. All parties involved in medical research should take a more cautious and responsible approach to the use of preprints
Diamond OA:
55. Diamond OA journals (no cost to access; no APC) are commendable, but how are they funded?
56. Diamond OA journals are often funded by institutions and research funders
Researcher incentives:
57. Some University Appointment and Promotion panels continue to stigmatise OA papers
58. New academia could incentivise OA
Artificial Intelligence:
59. AI could increase efficiency of publishing processes, and reduce costs and APCs
60. A major benefit of OA is that it is visible to AI and therefore contributes to AI processes
61. AI makes it easier for unscrupulous authors to submit fake papers
NEXT STEPS
1. Further discussion on any of the above points is welcome. Please send your contributions to hifa@hifaforums.org
2. We shall shortly email three outputs of the discussion: full compilation; edited version; and selected extracts
3. We shall draft and publish a short blog on the project.
Best wishes, Neil
On behalf of the HIFA Open Access working group
https://www.hifa.org/projects/open-access
HIFA profile: Neil Pakenham-Walsh is coordinator of HIFA (Healthcare Information For All), a global health community that brings all stakeholders together around the shared goal of universal access to reliable healthcare information. HIFA has 20,000 members in 180 countries, interacting in four languages and representing all parts of the global evidence ecosystem. HIFA is administered by Global Healthcare Information Network, a UK-based nonprofit in official relations with the World Health Organization. Email: neil@hifa.org