Writing a Research Protocol for Publication

A published protocol is a public record of what you planned before the data came in. Journals now enforce strict timing and checklist rules, so plan it early.

Guide · 11 min read · Directive Publications Editorial Team

What a protocol paper does that a registry record cannot

A trial registry record tells the world that your study exists and what it will measure. A published protocol goes further. It sets out, in peer-reviewed and citable form, how you will recruit, allocate, measure, analyse and monitor, before anyone knows the answer.

That record earns its keep when the results arrive. The SPIRIT 2025 authors note that comparing protocols with reports of completed trials helps identify selective reporting and undisclosed amendments, such as changes to primary outcomes or analyses. If every change was declared, that comparison works in your favour.

Three documents are easy to confuse, and each does a different job:

The protocol paper presents the full protocol for readers. It is not a chance to rethink the design.

The order that saves you a rewrite

A protocol can be declined for reasons unrelated to the science: late registration, pending ethics approval, or recruitment already over. Work in this order:

1. Choose the journal and read its protocol rules. Cut-offs, formats and required documents differ.
2. Get ethics approval and register the trial. Both must be done before you enrol, not before you submit.
3. Draft against the SPIRIT 2025 checklist. Use its section order as your outline.
4. Write outcomes, sample size and analysis in full. Your results paper will be compared against them.
5. Complete the checklist with page and line numbers. Explain every item that does not apply.
6. Submit with the supporting documents. Ethics approval, funding documentation, sometimes a questionnaire.
7. Keep protocol, registry and publication aligned. Record amendments as they happen.

Choose a journal that publishes protocols

Protocols are a distinct article type, and a journal in your field will not necessarily accept one, so start from the article types listed in the author instructions. For each candidate, answer these questions:

That fourth question decides how long review takes. BMJ Open will consider publishing without further peer review a protocol that has formal ethical approval and has undergone independent peer review to gain funding from a recognised, open access advocating research-funding body, such as those listed by the Jisc open policy finder. It asks for proof that both criteria are met when you upload the protocol, including peer-review reports showing the methodology was thoroughly evaluated by suitably qualified experts. Protocols that miss either criterion go to open external review, where the stated intention is not to alter the study design. Trials gives funder-reviewed protocols an editorial review by a Protocol Editor and sends the rest to full external review. Find your funder's reports before you submit, not after an editor asks.

A journal that offers to publish protocols is not automatically reputable, so run the usual checks from our guide to choosing the right journal without getting scammed. Treat the cover letter as part of the submission too; Trials makes a questionnaire in the cover letter mandatory, and our guide to writing a journal cover letter covers the structure.

Timing: submit well before recruitment ends

Protocol journals want planned or ongoing studies, and some set an explicit cut-off after which they will not consider a protocol at all. Two examples show how differently that line can be drawn:

• BMJ Open takes protocols for planned or ongoing studies of any design, including observational ones, but not once data collection is complete. Protocols nearing the end of data collection are handled case by case.
• Trials stopped considering protocols submitted after recruitment has ended from 1 January 2026, and from 1 July 2026 it no longer considers protocols submitted within 90 days of recruitment completion. The trial must already be registered when the protocol is submitted, and the journal no longer takes protocols for pilot or feasibility trials. The full requirements are in the Trials study protocol guidelines.

Both expect ethics approval before submission: Trials does not consider protocols without it, and BMJ Open says protocols for studies needing approval are unlikely to be considered without it.

Decision rule. Fix the journal's deadline on a calendar first, noting which event it hangs on: end of recruitment for one of the examples above, completion of data collection for the other. From that date, subtract the time you need for drafting and for sign-off by co-investigators and the sponsor. If the result has already passed, choose a journal with a later cut-off, or supply the protocol with the results paper instead; BMJ Open asks authors of research articles for the original protocol as a supplementary file.

Register before you enrol, and say so in the paper

The ICMJE recommends that editors require registration in a public trials registry at or before the time of first patient consent for enrolment, as a condition of considering a trial for publication. It accepts any primary registry of the WHO International Clinical Trials Registry Platform, or ClinicalTrials.gov, provided the record contains the minimum trial registration data set.

Three details catch experienced investigators:

• Ethics approval is not registration. ICMJE states that approval from an ethics committee or institutional review board does not fulfil the requirement.
• Registering at submission does not fix late registration. ICMJE says retrospective registration, for example at the time of manuscript submission, meets none of the purposes of registration, and that exceptions should be rare.
• The data sharing plan goes in the registry too. Clinical trials that began enrolling on or after 1 January 2019 must include a data sharing plan in their registration, and "undecided" is not an acceptable answer in an ICMJE data sharing statement.

In the manuscript, SPIRIT 2025 item 4 asks for the registry name, the identifying number with URL, and the date of registration. Copy those from the registry record, not from memory.

Build the manuscript on the current SPIRIT checklist

SPIRIT 2025 is the current guideline for randomised trial protocols: 34 minimum items plus a diagram of the schedule of enrolment, interventions and assessments. Its authors state that it supersedes SPIRIT 2013, which should no longer be used or cited, and Trials now accepts only the newer guidance and checklists. The editable checklist, the timeline template and the explanation and elaboration document are on the SPIRIT-CONSORT website.

If your group's template was built on the 2013 checklist, rebuild the outline. The notable changes:

• a new open science section covering registration, access to the protocol and statistical analysis plan, data sharing, funding and conflicts of interest, and dissemination (items 4 to 8)
• a new item on patient and public involvement in design, conduct and reporting (item 11)
• more emphasis on harms, and on describing interventions and comparators in enough detail to allow replication
• a new item on trial monitoring (item 29), replacing the old item on auditing

SPIRIT is written around the two-group parallel randomised trial, though most items apply to any trial. For special cases such as factorial designs, patient-reported outcomes or AI interventions, its authors recommend the existing SPIRIT extensions until those are aligned. For other designs, check the EQUATOR Network library; PRISMA-P, for instance, covers systematic review protocols.

One caution: a complete checklist does not make a good trial. The authors point out that a poorly designed trial can address every item by fully describing its weaknesses.

Write the details your results will be judged against

Title and abstract

Follow the journal's format exactly. Trials asks for titles in the form "[descriptive title]: study protocol for a [design] trial" and an abstract of no more than 350 words under Background, Methods, Discussion and Trial registration. BMJ Open wants the study type in the title and abstract headings of Introduction, Methods and analysis, and Ethics and dissemination, with registration details last.

Outcomes

SPIRIT item 16 asks you to define each primary and secondary outcome by four elements: the measurement variable, the analysis metric, the method of aggregation, and the time point.

• Incomplete: "Blood pressure will be assessed during follow-up."
• Complete: "Systolic blood pressure, analysed as change from baseline, summarised as the mean in each group, at 12 weeks."

Use the same wording in the abstract, the methods and the registry record.

Sample size

Item 19 asks how the sample size was determined, including all assumptions supporting the calculation. State the target difference and why it is worth detecting, the assumed variability or control-group event rate and its source, the significance level, power, allocation ratio, and any allowance for loss to follow-up. A statistician should be able to reproduce your number from the text alone.

Analysis

Item 27 covers the methods for comparing groups, who is included in each analysis, how missing data will be handled, and any subgroup or sensitivity analyses. Item 5 asks where the full statistical analysis plan can be accessed. Trials lets you build the plan into the protocol or attach it as an additional file.

Status and version

Give the protocol version date and identifier (item 2). Trials also asks for the date recruitment began and the approximate date it will be completed. Leave results out: Trials tells authors not to include baseline or pilot data and will ask for them to be removed.

Complete the checklist so the paper is not returned

SPIRIT 2025 recommends that protocols come with a completed checklist showing where each item is reported, and journals enforce it. Trials returns manuscripts whose checklist lacks page and line numbers, or lacks a short explanation beside items that do not apply, and warns that simply writing "N/A" will lead to the manuscript being returned.

Write not-applicable entries as reasons: "Item 28b: Not applicable. No interim analyses are planned, so there are no stopping guidelines."

Before you upload:

1. Every SPIRIT item has a page and line number, or a stated reason it does not apply
2. The schedule of enrolment, interventions and assessments is where the journal wants it
3. Registry details and primary outcome wording match the registry record
4. Protocol version date and identifier are stated
5. No baseline, pilot or interim results appear anywhere
6. Ethics approval and funding documents are ready, with translations if required

After publication: amendments and the results paper

Studies change after the protocol is published; what matters is that every change is visible. SPIRIT 2025 item 31 asks for plans to communicate important protocol modifications, and its authors say those amendments should reach ethics committees and trial registries as they occur, then be described in the report of the completed trial.

In practice:

• keep a dated amendment log that maps each change to a protocol version
• update the registry record whenever an amendment touches registered details
• if you finalise the statistical analysis plan after the protocol is out, publish it before data lock; Trials, for example, takes standalone plans as an Update
• in the results paper, cite the protocol, list every deviation, and give the reason for each

Then publish the results whatever they show. A published protocol makes a missing results paper visible, and a null result reported against a pre-specified plan is still a contribution, as our piece on negative and null results explains. Report the completed trial against CONSORT 2025, whose wording was harmonised with SPIRIT 2025.

Where to spend the extra time

If you have one more day before submission, spend it checking that the document agrees with itself, because the paper will later be judged against its own text.

Hand it to someone who was not in the design meetings. Can they state the primary outcome and its time point without reading twice, and reproduce the sample size from the text alone? Any hesitation marks a sentence to fix now, while it costs a revision rather than a published deviation.

When the trial is complete and you are preparing the results paper, Directive Publications links to its article types, author guidelines and data availability policy from the For Authors page, so you can check requirements against your protocol before you submit your manuscript.