When the Protocol, Registry, and Manuscript Tell Different Stories

Differences among a protocol, trial registry, and manuscript are not automatically misconduct. Studies change: recruitment may be slower than expected, an instrument may become unavailable, a safety concern may require modification, or a statistical method may fail. The integrity problem is an unexplained difference that changes what readers believe was planned.

The protocol records the approved design. The registry creates a public, time-stamped summary. The manuscript tells the final scientific story. Research governance should make those three artifacts consistent where they should be and transparently reconcile them where legitimate changes occurred.

01Compare the elements that can change a conclusion

A useful crosswalk includes:

• study title, design, phase or model, and setting;
• eligibility criteria and target population;
• intervention and comparator;
• primary and secondary outcomes, measurement methods, and time points;
• sample-size target and rationale;
• recruitment dates and study status;
• analysis populations and primary model;
• subgroup, sensitivity, and multiplicity plans;
• important protocol amendments; and
• registration, protocol, analysis-plan, and data-sharing links.

Outcome comparison deserves exact language. “Depression at 12 weeks” is not necessarily the same as “change in depression score from baseline to week 12.” “Any cardiovascular event” is not the same as a prespecified composite. The WHO Trial Registration Data Set requires the outcome name, measurement method, and time point; use those components when matching documents.

02Preserve chronology

For every difference, record what changed, why, when, who approved it, and which data were available at the time. A change made before recruitment is evaluated differently from one made after unblinded outcome patterns were known.

Link amendments to dated protocol and statistical analysis plan versions. If the registry was updated late, preserve both the original and revised registration dates. Do not treat an updated current page as evidence that the new outcome was always planned.

SPIRIT 2025 encourages consistency among the protocol, related documents, and registry record. CONSORT 2025 asks final reports to distinguish prespecified from non-prespecified analyses and makes the protocol, analysis plan, and registration part of the open-science record.

03Classify the discrepancy

Not every mismatch has the same consequence. Use categories such as:

• Editorial: wording differs without changing meaning.
• Administrative: contact, site, or status information changed.
• Operational: recruitment, timing, or implementation changed.
• Scientific: outcome, population, intervention, or estimand changed.
• Analytical: model, covariates, missingness, multiplicity, or subgroup strategy changed.
• Reporting: a planned element is omitted, relabeled, or selectively emphasized.

Scientific, analytical, and reporting differences usually require explicit manuscript disclosure and may need ethics, sponsor, regulatory, or journal attention depending on context. An editorial difference can still be corrected, but it need not be presented as a major amendment.

04Assign a reviewer who did not write the manuscript

Authors are close to the evolving study and may read two similar phrases as equivalent because they know the intended history. An independent research-office or methods reviewer is more likely to detect that a time point, threshold, or population changed.

The reviewer should work from source documents, not only an author-prepared summary. Each manuscript claim should point to the governing protocol or plan version, or to a documented amendment. Unresolved differences should be escalated before submission.

05Correct the record, not just the prose

When a registry permits updating, bring it current while preserving transparent dates and change history. Add a manuscript explanation of material changes and their timing. If an analysis is exploratory, label it. If the revised outcome cannot support the original confirmatory claim, narrow the conclusion rather than using wording to disguise the change.

The same discipline applies after publication. A discovered inconsistency may require a registry update, correction notice, supplementary clarification, or—in severe cases—reassessment of the conclusion. The response should match the consequence.

06Make the crosswalk a release gate

Before any manuscript leaves the institution, require a completed protocol–registry–manuscript table signed by the principal investigator and reviewed by a methods or integrity function. Store it with the publication record.

Consistency is not achieved by forcing three documents to look identical after the fact. It is achieved when every material difference has an honest, time-stamped explanation that a reader can use to understand how the study changed.

Publish the registration number and direct protocol or analysis-plan locations whenever available. Making the crosswalk verifiable outside the institution turns an internal quality exercise into durable research transparency.

This article is educational and intended for research purposes. It does not provide individual medical advice, diagnosis, or treatment. No patient data were used.