When Should You Consult a Protocol, Registry Entry, Supplement, or Previous Paper to Understand What Was Actually Done?

When is the main paper no longer enough?

You are reading a study and encounter a familiar sentence: “Further methodological details are provided elsewhere.” Or perhaps the paper never explains whether its primary outcome changed, how a complex intervention was delivered, why one analysis was designated primary, or what happened to a previously registered outcome.

At that point, should you keep appraising the paper as a self-contained document, or follow the documentary trail? The answer depends on what you are trying to judge. A protocol, registry entry, statistical analysis plan, supplement, or previous publication can provide information that the main article compresses or omits, but each source answers different questions and has different limitations.

Consult another study document when an unresolved detail could change your appraisal

Use the wider study record strategically rather than automatically

A journal article may not contain the entire study record

Published articles operate under editorial and space constraints. Complex studies can involve protocols, analysis plans, intervention manuals, registrations, data dictionaries, previous methodological publications, supplementary appendices, and other materials.

CONSORT 2025 explicitly requires randomized-trial reports to state where the trial protocol and statistical analysis plan can be accessed and promotes supplementary material for more detailed reporting of methods and results than may fit in the main article.

This means that consulting another document is not necessarily an attempt to rescue a poorly reported study. Sometimes the paper deliberately functions as one part of a larger, linked research record.

Start with the unresolved question, not with the document type

Do not download every supplementary file merely because it exists. First identify what information is missing that you need to judge the study.

Then choose the source most likely to answer that question.

Consult the protocol when you need to know what was planned

The protocol is especially useful when you need to understand the intended design before comparing it with what eventually happened.

CONSORT 2025 explains that a trial protocol provides important context for interpretation, replication, and risk-of-bias appraisal because it prespecifies methodological features such as the primary outcome. It can also help identify undeclared post hoc changes and selective outcome reporting.

Useful protocol questions include:

• What were the original primary and secondary objectives?
• What population and eligibility criteria were planned?
• What interventions and comparators were intended?
• What outcomes and time points were specified?
• What sample size and study design were planned?
• Were important methods changed after the study began?

The protocol is therefore particularly valuable when investigating what was prespecified and what appears to have been decided later.

Consult the registry when chronology matters

A registry can provide a dated public record of important study information. CONSORT 2025 requires trial reports to provide the registry name, identifying number, URL, and registration date.

The registration date matters. A record created before participant enrollment or before relevant results become available has a different evidential role from a registration completed retrospectively.

Registries can be useful for checking outcomes, target enrollment, study status, intervention arms, recruitment dates, and changes to registered information when version history is available.

However, do not assume the registry contains the complete protocol. CONSORT's explanation notes that registration provides a minimum set of protocol details but can leave substantial ambiguity about key protocol and statistical-analysis issues.

Consult the statistical analysis plan when the analytical details matter

The protocol may tell you that researchers planned a regression analysis. The statistical analysis plan may tell you which model, covariates, analysis population, interaction terms, transformations, missing-data procedures, sensitivity analyses, and decision rules were planned.

CONSORT and SPIRIT describe the statistical analysis plan as a document providing more detailed technical information about planned analyses than may appear in the main protocol.

It becomes especially useful when you are trying to identify the primary analysis, classify sensitivity analyses, understand missing-data procedures, or determine whether analytical choices preceded access to the results.

Consult the supplement when the main article deliberately compresses detail

Supplementary material often contains information that is important but too extensive for the main article: detailed statistical methods, intervention descriptions, additional participant characteristics, complete outcome tables, sensitivity analyses, subgroup results, questionnaires, technical appendices, and flow information.

CONSORT 2025 explicitly encourages online supplementary material to allow more detailed reporting than typical article length permits.

If the main text explicitly says “see Supplementary Appendix,” follow the reference when the omitted detail matters to your appraisal. Critiquing the paper for failing to explain something while declining to open the file where it says the explanation resides would be a rather avoidable methodological plot twist.

Consult a previous paper when the current study is built on an established cohort or project

Large cohorts, registries, longitudinal projects, and research programs often publish dedicated design or methods papers. A later analysis may legitimately refer readers to those publications for recruitment procedures, measurement protocols, sampling frames, cohort construction, or data infrastructure.

CONSORT guidance for trials using cohorts or routinely collected data similarly notes that access to documentation about the underlying cohort or database can help readers understand participant characteristics and data-collection methods.

A previous paper is therefore useful when the current article depends on infrastructure established elsewhere. But verify that the methods still apply to the current analysis. A cohort may have changed instruments, recruitment procedures, waves, sites, or eligibility criteria over time.

Consult another source when the intervention description is incomplete

Complex interventions can be difficult to describe fully within the article. Manuals, protocols, supplements, and dedicated intervention papers may contain the information needed to determine what intervention was actually examined.

CONSORT 2025 asks trial reports to describe intervention and comparator conditions sufficiently for replication and, when relevant, indicate where additional intervention materials can be accessed.

This is particularly important when a broad intervention label conceals content, dose, provider training, tailoring, implementation, or fidelity.

Consult another source when participant flow is unclear

If the paper reports that 1,000 people enrolled but only 720 entered the primary model, supplementary material may provide the missing participant-flow diagram or analysis-specific denominators.

CONSORT 2025 requires numbers randomized, receiving intended intervention, and analyzed for the primary outcome, together with losses and exclusions after randomization and their reasons.

If these details are absent from the main text but explicitly supplied elsewhere, use the linked material before concluding that the fate of recruited participants cannot be determined.

Consult earlier documentation when an outcome or analysis looks unexpectedly prominent

Suppose the paper emphasizes an outcome that seems only loosely connected to the stated objective, or a subgroup finding dominates the conclusion despite an unremarkable overall result.

That is a good reason to check earlier documentation. Was the outcome originally primary? Was the subgroup analysis prespecified? Did the analysis plan identify the model before results were known?

CONSORT 2025 requires reporting important changes after trial commencement, including outcomes or analyses that were not prespecified, and asks authors to distinguish prespecified from post hoc additional analyses.

Do not assume the earliest document is automatically correct

A protocol tells you what researchers planned, not necessarily what they ultimately did. A registry may contain outdated information. A statistical analysis plan can be amended. A previous paper may describe an earlier wave of data collection.

Therefore, compare documents chronologically rather than treating any single one as the unquestionable truth.

CONSORT's explanation recommends documenting changes to protocol and analysis-plan versions along with their rationale and timing.

Version history can matter more than the latest version

If your question is whether an outcome was prespecified, downloading only the final amended protocol can defeat the purpose. The final version may already contain changes made after the study began.

Look for dates, version numbers, amendment histories, archived registry versions, and explanations for changes. The relevant document is the version that existed before the information capable of influencing the decision became available.

Documents can disagree

You may find one outcome in the registry, another hierarchy in the protocol, and slightly different wording in the publication. Do not automatically assume misconduct or decide which document you prefer.

First establish dates and versions. Then look for documented amendments, clarifications, and differences in terminology. Some apparent discrepancies are legitimate changes or differences in level of detail. Others remain unexplained.

If the conflict cannot be resolved, preserve it as an uncertainty in your appraisal.

Do not let supplementary searching become an excuse never to finish reading

Not every appraisal requires a documentary investigation. If the main paper contains enough information to answer the question you are evaluating, additional documents may add little.

The threshold should be consequentiality: would resolving this missing or ambiguous detail materially change how you understand the design, analysis, risk of bias, applicability, or strength of the conclusion?

Using different documents to answer different appraisal questions

Common mistakes when consulting other study documents

Follow the documentary trail only when it can answer a consequential question

Before opening another source, write down what you are trying to establish. This keeps supplementary reading targeted and prevents different documents from becoming a pile of methodological trivia.

Keep notes on both source and date. “Protocol version 1.2, before recruitment” is more informative than “protocol says...” when later amendments exist.

Before searching beyond the paper, know what document you need and why

Questions about protocols, registries, supplements, and related papers

The paper is sometimes only one layer of the study record

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.