Protocols, registrations and preprints: keep the research record clear

1. Name the record before describing its status

A protocol describes objectives and planned methods. A registration creates a structured record in a registry. A preprint shares a manuscript before journal peer review. A journal’s version of record is a further publication object. These may link to one another, but a draft protocol on a computer does not establish registration, and a registry identifier does not establish that results have been peer reviewed.

For a systematic review protocol, PRISMA-P helps organise reporting of the planned question, eligibility criteria, search, selection, extraction and synthesis. Choose the reporting guidance suited to the design. Completing a protocol checklist is preparation for scrutiny; it is not approval of the study or a substitute for methodological review.

Reference: PRISMA-P: systematic review protocol checklist

On small screens, scroll the table horizontally, or focus it and use the arrow keys.

2. Choose a registry appropriate to the design

ICMJE recommends that medical journal editors require clinical trial registration at or before the first participant’s consent for enrollment as a condition of consideration for publication. It accepts qualifying WHO primary registries and ClinicalTrials.gov; approval by an ethics review body does not satisfy its registration requirement. This is a publication-policy distinction. The responsible team must separately confirm the ethics, sponsor and local requirements that apply to its study.

WHO’s ICTRP provides a route to search records across registries; it is not a registry to which investigators submit a trial. PROSPERO is a prospective register of systematic reviews relating to health and social care. Check its current eligibility and stage requirements in the live registration guidance before starting a submission. Do not apply a clinical-trial timing rule to all review designs or assume that every review is eligible.

Reference: ICMJE: clinical trial registration · WHO: International Clinical Trials Registry Platform · University of York: the redesigned PROSPERO registry

3. Record the real sequence, including delays

Keep protocol version dates, registry submission dates, public posting dates and the dates of relevant research stages in separate fields. For trial publication policy, ICMJE uses the date registration materials were first submitted and may ask about a substantial delay before public posting. Save the actual receipt and check the posted record; do not make the dates agree by changing the history.

If research has started, describe its true stage and ask the appropriate registry or research office how to proceed. Later registration cannot turn completed decisions into prospective decisions. A clear explanation of what was known when a change was made is more useful to an editor than a label that hides the sequence.

Reference: ICMJE: clinical trial registration

4. Preserve the plan and explain amendments

An amendment should identify the original plan, the new decision, the date, reason and who approved it within the project. State whether relevant data or results had already been seen. Retain the earlier version and follow the registry’s update process. PRISMA-P explicitly calls for tracking and dating protocol amendments.

The example below is invented. It describes a review team’s planning record, not a completed registration or a real set of results. At the first two stages no formal screening or extraction has begun. The final row is a later hypothetical analysis-stage decision, included to show how its timing changes the description.

Reference: PRISMA-P: systematic review protocol checklist

On small screens, scroll the table horizontally, or focus it and use the arrow keys.

5. Distinguish a registration from a working folder

OSF separates editable projects from registrations. A submitted registration preserves a dated record; its uploaded files cannot simply be replaced. The service provides an update workflow and allows an embargo at registration. Use its current instructions for the particular action, and check what another reader can access before calling a link public.

As checked on 14 September 2026, OSF has announced that Projects will become read-only in February 2027, while registrations and preprints continue. Consult its transition instructions before choosing a working folder for an ongoing project.

Before submitting any public record, inspect filenames, acknowledgements, attachments and metadata for information that should remain private. A timestamp does not grant permission to disclose participant information, confidential contracts or material belonging to collaborators. Store necessary restricted material through the institution’s approved process and describe access conditions accurately.

Reference: OSF: registrations, updates and privacy · OSF: announced Projects transition

6. Make preprint status and later versions easy to see

Check the intended journal’s preprint policy before posting. ICMJE advises authors to use archives that clearly label non-peer-reviewed work, disclose prior posting to the journal and link preprints to later versions, including the published article. A preprint citation should identify its status. Follow the journal’s and archive’s rules about which files may be posted or updated.

Before sharing an Arabic summary, compare it with the same manuscript version used for the English summary. Preserve the study design, uncertainty and limitations. Label preliminary findings clearly and avoid describing a preprint as accepted, clinically established or independently validated. Keep a note of the version each summary describes so material changes can be corrected consistently.

Reference: ICMJE: preprints and overlapping publications

7. Reconcile the records before journal submission

Prepare a small crosswalk linking the protocol version, registry record, amendment log, preprint version and submitted manuscript. Compare objectives, outcomes, eligibility, analysis plans and dates. Explain departures in the methods or the location requested by the journal. Missing documentation should appear as an unresolved item with a responsible person, not as an invented identifier or completed task.

After publication, check the actual article link and update related records through their supported processes. Keep any results-reporting duty separate from journal publication: an accepted paper is not evidence that another registry task is complete. For unresolved requirements, consult the sponsor or research office using the facts of the project. Transparent records make review easier; they do not replace ethical approval, sound conduct or critical reading.

Questions for a practical final check

• Every record has a clear purpose, owner, version and date.
• The chosen registry fits the actual study design and stage.
• Submission and public posting dates remain distinct.
• Amendments explain what changed and what was already known.
• Preprint status and links are clear in both languages.
• Pending items and later publication updates are recorded honestly.