The joint position statement on preprints and peer review

The 2021 AMWA–EMWA–ISMPP Joint Position Statement on Medical Publications, Preprints, and Peer Review recommends safeguards for rapid publication. It addresses preprints, post-publication and traditional peer review, quality control, training and medical writers’ responsibilities.

Scope and purpose

The statement recommends practical standards for data integrity, publication quality and transparency across all stakeholders. ^{1, 2, 3} Its recommendations address both urgent public-health emergencies and the continuing need for publication quality. ⁴ It uses “medical writer” to include professionals preparing documents for regulatory submission or publication in peer-reviewed journals. ^{2, 3, 5}

The statement argues that rapid publication’s value should be weighed against inadequate validation, because lowering oversight may establish a difficult-to-reverse precedent and erode standards and public trust. ^{2, 3} Its accompanying guidance calls for full disclosure of peer-review status and transparency about quality-control processes. ⁴

Preprint safeguards

The statement calls for preprint standards developed jointly by a convened body of all stakeholders. ^{6, 7} Its suggested safeguards are these. ⁸

• Cite a preprint only in the manner of a personal communication: in the text using its link, DOI or both, rather than in the reference list, and clearly identify it as a preprint.
• Distinguish preprints from peer-reviewed articles through both a watermark stating that review has not occurred and a clear disclosure in the article’s body that the findings have not been formally peer reviewed; this might reduce readers’ tendency to treat the work as fully vetted.
• Strengthen pre-publication vetting through more extensive, consistent checks by server hosts using a comprehensive checklist, and encourage authors to mark subsequently published preprints and link them by DOI to the full publication.

^{6, 9, 8}

Post-publication peer review

For articles undergoing post-publication peer review, the statement recommends the preprint safeguards and mainstream indexing, where applicable, only after full peer review. ^{10, 11} It says such articles should not be used as references before review is complete and publication approved, and adds that, if cited, the in-text approach recommended for preprints should be used. ^{9, 12} It also recognises that the problems of traditional peer review apply to this model. ^{9, 12}

Traditional peer review

The statement divides its proposed improvements among authors, editors and publishers. ^{13, 14}

Responsible party	Suggested action
Authors	Submit previous rejection comments to a second-choice journal with itemised rebuttals and an updated manuscript: portable peer review.
Authors	Be more accepting of editorial referrals to cascade journals.
Journal editors	Accept, request or require portable peer review to reduce further review cycles.
Journal editors	Consider commercial back-end services that expedite peer review, such as ResearchSquare.
Journal editors	Form rapid-response reviewer teams with appropriate expertise and a quick turnaround.
Publishers	Standardise formatting requirements to expedite resubmission.
Publishers	Offer fast-track options for potentially practice-changing work.
Publishers	Consider incentives for reviewers.

^{13, 14}

Quality control and training across formats

The statement recommends four quality-control measures across publication formats. ^{15, 16}

• Use existing publication guidelines and checklists for high-quality development.
• Include clinical trial protocols and statistical analysis plans as supplementary material.
• Ask every author to sign a form confirming full access to the relevant reported data and accepting responsibility for submission; the contributor statement should identify at least two authors who accessed and verified the underlying data, as suggested in the revised Lancet guidance.
• Have journals explain the initial quality review editors perform on new manuscripts.

^{15, 16}

Its training recommendations cover these activities. ^{17, 18}

• Train authors, peer reviewers and editors adequately in the nature and technical aspects of peer review.
• Use guidelines, such as COPE’s guidance, alongside the statement’s reviewers’ checklist.
• Educate medical journalists and the public about the differences between preprints or pre-publications and peer-reviewed literature.

^{17, 18}

Medical writers’ role

The associations’ statement cites evidence suggesting that professional medical writing improves publication quality and speed, and an association with reduced misconduct-related retraction risk. ^{17, 19} It recommends these roles for medical writers. ^{17, 19}

• If a qualified medical writer is part of the team, involve them as early as possible, with access to the clinical study report if available, source data and related documents, including statistical outputs and patient narratives, as far as data-protection regulations permit.
• Give writers an active role in publication development, editing, referencing and appropriate checklist use; involve writers and statisticians in peer review, with writers applying common critical-appraisal criteria to complement subject-matter review.
• Consider writers for pre-publication vetting, training or both.

^{17, 19}

Each party, including readers, must take responsibility for exercising judgement and selecting rigorous, transparent publishing sources, according to the statement. ^{17, 19}

The reviewers’ checklist

The appendix is intended for journals and can also guide authors and medical writers reviewing manuscripts before submission. ^{20, 21} A suitably qualified team should comprise at least two reviewers, preferably including editors, subject experts, medical writers, statisticians and trained researchers; subject-matter peer reviewers are not required for preprints. ^{20, 21} Every item must be checked by at least one accountable reviewer, although each reviewer need not complete every field. ^{20, 21}

The checklist covers source documents, attestations, preprint editorial review and additional peer-review checks, in that order. ^{20, 22, 23}

Item	Medical writer reviewer	Clinical reviewer	Biostatistical reviewer	Peer reviewer A	Peer reviewer B
Source documents, if available and required by the journal
Clinical study protocol
Redacted clinical study report synopsis
Statistical analysis plan
Tables, listings and graphs
Attestations
Data-sharing statement using the ICMJE template
Author contribution form
Signed author forms confirming full access to reported data and responsibility for submission
Author confirmation that verified data were used to develop the manuscript
Conflict-of-interest statement, with the ICMJE template recommended
Funding sources for the study and supporting activities
Suggested preprint editorial review
No offensive or nonscientific content
No plagiarism
Sound basic statistical methods, including the analysis population and missing-data handling
Where public registration is required, endpoints and eligibility criteria align with it; all endpoints appear in results for primary clinical-trial reports
Clear, consistent sections: abstract matches results and the discussion addresses the introduction’s hypothesis
Discussion points and conclusions supported by reported data
Adherence to relevant guidelines, such as CONSORT, STROBE, PRISMA, SPIRIT or CARE; specify the guidelines
No ethical concerns
Additional peer review
Statistical considerations: adequate sample-size calculation, including the comparator; adequate statistical methods; checks for random errors; sources of bias addressed
Methodological quality: confounding, including concomitant treatments; inadequate disclosure; misinterpretation
Study design: adequate and relevant endpoints, adequate eligibility criteria, blinding, adequate follow-up, adequate reporting of complications and adequate data presentation

^{20, 22, 24, 23}

The author contribution form should ask whether medical writing support was used, and the signed data-access and submission-responsibility forms may be merged with it. ^{24, 23} The methodological-quality criteria are adapted from MEDDEV 2.7/1; other appraisal criteria may be used, and additional reviewer columns and signature lines can be added as needed. ^{24, 23}

Publication and translations

The statement first appeared in Current Medical Research and Opinion under a Creative Commons Attribution 4.0 licence. ² A 2021 AMWA announcement gives the publication date as 1 April, whereas EMWA’s announcement of the Chinese and Japanese translations gives March 2021. ^{25, 26} Spanish was announced as the first translation in September 2021, available through EMWA; Chinese and Japanese versions were announced by December 2021 and made available through ISMPP. ^{27, 26}