WHY ENGLISH?
Millions of routine records become research evidence only through a visible chain of coverage, code definitions, linkage, cleaning, validation and interpretation.
This guide reads the published 2015 RECORD statement as a 13-item STROBE extension and clearly dates the announced update that remains under development.
Begin with Identify Routinely Collected Data in the Title.ROUTE 2 · SECTIONS 9–16Codes and linkage
Begin with Report Linkage Quality.ROUTE 3 · SECTIONS 17–24Bias and analysis
Begin with Describe the Database Population.ROUTE 4 · SECTIONS 25–32Access and use
Begin with Date the Published and Developing Guidance.
Open the full article contents
- Identify Routinely Collected Data in the Title
- Define Routinely Collected Health Data
- State the Data Source and Coverage
- Describe the Study Population Selection
- Provide Complete Code Lists
- Explain Algorithms for Case Identification
- Report Algorithm Validation
- Describe Data Linkage
- Report Linkage Quality
- Describe Data Cleaning
- Explain Variable Derivation
- Handle Missing Data
- Explain Changes in Coding Systems
- Report Data Access and Cleaning Methods
- Name Database Linkage Across Levels
- Protect Privacy and Confidentiality
- Describe the Database Population
- Report Flow of Records and People
- Distinguish Data Availability From Clinical Absence
- Assess Misclassification
- Address Confounding and Bias
- Report Analytical Reproducibility
- Explain Data-Sharing Constraints
- Use STROBE Alongside RECORD
- Date the Published and Developing Guidance
- Help Students Read Big-Data Claims
- Help Parents Interpret Routine-Data Evidence
- Preserve a Versioned Data-Provenance Record
- Avoid Immortal-Time Bias
- Audit Time-Varying Definitions
- Use Negative Controls as Diagnostic Evidence
- Check Subgroup Capture and Fairness
People searching for the RECORD statement, routinely collected health data or electronic health-record research often assume that a very large database automatically produces a reliable answer. English matters because readers must distinguish the data’s original operational purpose from the research question and follow every code, linkage, cleaning, validation and access decision that creates the analytical dataset.
The official EQUATOR Network RECORD record, checked on 5 November 2026, identifies RECORD as reporting guidance for observational studies using routinely collected health data. The open-access 2015 RECORD statement describes a 13-item extension to STROBE. EQUATOR also lists an updated RECORD guideline as under development; that announcement is not treated here as a completed replacement.
RECORD supports transparent reporting of administrative data, electronic records, registries and other routinely collected sources. It does not make operational codes identical to clinical truth, remove confounding or authorise open release of sensitive data. The existing STROBE guide remains the owner for generic observational reporting; this article owns the routine-data provenance layer.
Identify Routinely Collected Data in the Title
Readers should recognise that the observational study uses routinely collected health data.
Data provenance check
Look for the data type, database, registry or electronic record source in the title or abstract. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A title names a national prescription database rather than presenting the study as bespoke prospective data collection.
What remains uncertain
A database label does not explain its contents. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Confirm the source and purpose in the methods. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Define Routinely Collected Health Data
These data were originally collected for administrative, clinical or operational purposes, not solely for the research question.
Data provenance check
Find the original purpose, contributing organisations and data flow. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Billing codes support reimbursement, while clinicians record diagnoses for care; both purposes shape completeness.
What remains uncertain
Large routine datasets are not neutral mirrors of health. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Ask why each field exists. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
State the Data Source and Coverage
Coverage determines who can appear in the analytical population.
Data provenance check
Look for jurisdictions, facilities, care settings, dates and populations captured. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A primary-care database does not capture every hospital event unless linked.
What remains uncertain
National scope can still exclude sectors or people. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Map represented and missing care pathways. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Describe the Study Population Selection
RECORD requires a clear route from source population to analysed cohort.
Data provenance check
Find codes, dates, enrolment requirements, exclusions and a selection diagram. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The cohort requires twelve months of prior registration so baseline conditions can be observed.
What remains uncertain
That rule may exclude mobile or newly enrolled people. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Explain the trade-off. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Provide Complete Code Lists
Cohort, exposure, outcome and confounder definitions often rely on codes.
Data provenance check
Look for diagnostic, procedure, drug or laboratory code lists and where they are available. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The supplement lists ICD and medication codes rather than saying cases were identified algorithmically.
What remains uncertain
A code list without logic or version is incomplete. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Record vocabulary, version, combination rules and dates. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Explain Algorithms for Case Identification
Operational definitions may combine codes, timing, repeated entries and exclusions.
Data provenance check
Find the full algorithm and its rationale. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A case requires two outpatient codes thirty days apart or one hospital diagnosis, with a washout period.
What remains uncertain
Complexity can create hidden researcher degrees of freedom. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Translate the algorithm into plain-language steps. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Report Algorithm Validation
Readers need evidence that operational definitions identify the intended condition or event.
Data provenance check
Look for sensitivity, specificity, predictive values or comparison with chart review in a relevant setting. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A validated algorithm from another country is discussed cautiously because coding incentives differ.
What remains uncertain
Validation is context- and time-dependent. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Match evidence to source, population and period. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Describe Data Linkage
Linkage combines records across sources and can create missed or false matches.
Data provenance check
Find deterministic or probabilistic methods, identifiers, linkage variables and quality measures. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A trusted third party links hospital and mortality data using encrypted identifiers.
What remains uncertain
High match rate does not prove unbiased linkage. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Examine who is less likely to match. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Report Linkage Quality
Readers need evidence about linkage success and error.
Data provenance check
Look for match rates, validation, thresholds and handling of uncertain links. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The report compares characteristics of linked and unlinked records.
What remains uncertain
Aggregate rates can hide subgroup disparities. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Inspect differential linkage. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Describe Data Cleaning
Routine data contain duplicates, impossible dates and changing formats.
Data provenance check
Find deduplication, range checks, date rules and reconciliation across sources. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
An admission ending before it starts is corrected only under a prespecified rule; otherwise it is flagged missing.
What remains uncertain
Cleaning can silently change outcomes. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Preserve a reproducible change log. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Explain Variable Derivation
Analytical variables may be constructed from multiple operational fields.
Data provenance check
Look for windows, thresholds, aggregation and priority rules. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Medication exposure uses dispensing date, days supplied and a grace period that is reported explicitly.
What remains uncertain
Different derivations can produce different exposure time. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Tie every derived variable to code and rationale. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Handle Missing Data
Missingness in routine data may mean absent measurement, absent care or absent capture.
Data provenance check
Find amounts, patterns, assumptions and analytical handling. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
No laboratory value is not automatically interpreted as a normal result.
What remains uncertain
Imputation cannot recover information missing by design. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Distinguish not measured, not recorded and not linked. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Explain Changes in Coding Systems
Code vocabularies and policies change over time.
Data provenance check
Look for version transitions, mapping and sensitivity analyses. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
An ICD transition is marked in the time series and code mappings are validated.
What remains uncertain
A smooth mapped variable can hide a discontinuity. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Test before and after periods. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Report Data Access and Cleaning Methods
RECORD asks authors to describe how researchers accessed and cleaned data.
Data provenance check
Find extraction environment, approvals, query dates and whether the database owner performed preprocessing. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Researchers receive a frozen extract through a secure environment and record the release date.
What remains uncertain
The same named database can change between extracts. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Attach findings to a data version. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Name Database Linkage Across Levels
People, clinicians, facilities and regions may be linked in one analysis.
Data provenance check
Report identifiers, hierarchy and clustering. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Patient records are nested within practices and linked to area deprivation measures.
What remains uncertain
Ignoring structure can understate uncertainty. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Map levels before modelling. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Protect Privacy and Confidentiality
Routine health data can be sensitive even when direct identifiers are removed.
Data provenance check
Find governance, minimisation, disclosure control and secure access. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Small-cell results are suppressed under a stated rule and analysis occurs in a controlled environment.
What remains uncertain
De-identification does not eliminate re-identification risk. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Report protections without exposing sensitive details. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Describe the Database Population
Readers need to know how the database compares with the target population.
Data provenance check
Find age, sex, geography, service use and other relevant distributions. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The database underrepresents people receiving care outside participating providers.
What remains uncertain
Weighting may address some differences but not unmeasured absence. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Keep target and captured populations distinct. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Report Flow of Records and People
Routine-data studies may involve records, episodes and unique people.
Data provenance check
Show counts through linkage, cleaning, eligibility and analysis. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Five million claims become 800,000 episodes and 420,000 unique people before cohort criteria.
What remains uncertain
Record counts are not participant counts. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Label every unit. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Distinguish Data Availability From Clinical Absence
A missing code may reflect no condition, no visit, no test or incomplete capture.
Data provenance check
Look for the authors’ assumptions and supporting validation. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
Absence of a prescription claim is not treated as proof that no medicine was used outside the system.
What remains uncertain
Operational silence can be misread as clinical truth. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Use cautious verbs. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Assess Misclassification
Coded exposures and outcomes may be wrongly classified.
Data provenance check
Find validation, sensitivity analyses and discussion of direction. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The analysis repeats under stricter and broader code definitions.
What remains uncertain
Similar results do not rule out shared error. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Explain plausible differential and nondifferential effects. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Address Confounding and Bias
Routine datasets can contain many variables while missing key reasons for treatment or care.
Data provenance check
Look for causal reasoning, measured confounders and residual limitations. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The study adjusts for recorded severity but acknowledges unrecorded clinician judgement.
What remains uncertain
Big data do not remove confounding by indication. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Separate predictive association from causal effect. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Report Analytical Reproducibility
Code lists, protocols and analysis code allow others to understand the pipeline.
Data provenance check
Find repository links, access conditions and versioned scripts where possible. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The cohort-building code is shared without patient data, along with a synthetic test dataset.
What remains uncertain
Code sharing cannot overcome inaccessible source data. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Share what is legally and ethically possible. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Explain Data-Sharing Constraints
Legal, ethical and contractual limits may prevent open release.
Data provenance check
Look for a clear access process rather than a vague claim that data are unavailable. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The report links to the data custodian’s application process and eligibility criteria.
What remains uncertain
Controlled access is not the same as public access. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Describe how a qualified researcher can seek review. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Use STROBE Alongside RECORD
RECORD is an extension, not a replacement, for observational-study reporting.
Data provenance check
Read the relevant STROBE items together with the 13 RECORD items. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A cohort report uses STROBE for design and participants and RECORD for codes, linkage and data provenance.
What remains uncertain
Using only the extension leaves generic reporting gaps. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Assign each checklist its job. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Date the Published and Developing Guidance
Reporting standards can change and should be dated.
Data provenance check
As checked on 5 November 2026, EQUATOR lists the 2015 RECORD statement and separately marks an updated RECORD guideline as under development. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
The article uses the published 2015 checklist while alerting readers to verify whether the update has since appeared.
What remains uncertain
An announced update is not yet a replacement guideline. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Record the version and check official sources. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Help Students Read Big-Data Claims
Students should not equate a huge record count with a strong causal study.
Data provenance check
Ask how people entered the database, how variables were coded and what could be missing. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A student notices that millions of claims come from fewer unique patients and that an outcome algorithm has limited validation.
What remains uncertain
Sample size cannot repair systematic error. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Lead with provenance before precision. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Help Parents Interpret Routine-Data Evidence
Families may encounter routine-data studies in health, education and policy discussions.
Data provenance check
Translate the source, population and outcome definition before discussing a headline association. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A parent asks whether the database includes children of the same age and care setting.
What remains uncertain
Population evidence informs questions but does not decide individual care. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Use it to support a conversation with qualified professionals. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Preserve a Versioned Data-Provenance Record
Routine databases change as providers, code systems and cleaning pipelines evolve.
Data provenance check
Look for extract dates, release identifiers and a chain from source fields to analysis variables. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A result is tied to the March 2026 frozen extract rather than the database name alone.
What remains uncertain
Two extracts from the same source may not reproduce the same cohort. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Record versions and transformation hashes where permitted. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Avoid Immortal-Time Bias
Exposure definitions can accidentally require a period during which an outcome could not be counted.
Data provenance check
Find the time origin, exposure assignment and person-time classification. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
People are classified as exposed only from the date they meet the prescription rule, not from cohort entry.
What remains uncertain
A large dataset does not protect against a misaligned clock. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Draw the timeline before reading the hazard ratio. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Audit Time-Varying Definitions
Clinical practice and coding behaviour can change during long study periods.
Data provenance check
Look for policy changes, new tests, reimbursement rules and revised algorithms. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A screening programme increases diagnosis coding midway through follow-up, so calendar period enters the analysis.
What remains uncertain
Adjustment may not capture every behaviour change. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Interpret trends with institutional history. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Use Negative Controls as Diagnostic Evidence
A negative-control outcome or exposure can reveal residual bias under carefully justified assumptions.
Data provenance check
Find why the control should not share the proposed causal pathway and what bias it can detect. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
An implausible association with a negative-control outcome raises concern about confounding or data capture.
What remains uncertain
A null control result does not prove the main estimate is unbiased. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Treat controls as diagnostics, not certificates. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Check Subgroup Capture and Fairness
Routine data systems may record some populations or identities less completely than others.
Data provenance check
Look for missing demographic fields, linkage differences and subgroup-specific validation. RECORD extends STROBE for observational research using routinely collected health data, so the reader should be able to trace the source, codes, linkage, cleaning, access and validation decisions that produced the analytical dataset.
A cohort algorithm has lower sensitivity in a language-minority subgroup because care occurs outside the linked network.
What remains uncertain
Aggregate validity can conceal unequal error. Routinely collected data can be large without being complete, unbiased or fit for a new research purpose. Report subgroup capture before making equity claims. Keep the operational record separate from the clinical or causal interpretation, and state the validation evidence that supports each variable or cohort definition.
Frequently Asked Questions
What is RECORD?
RECORD is reporting guidance for observational studies conducted using routinely collected health data.
How is RECORD related to STROBE?
RECORD is a 13-item extension used alongside the relevant STROBE items.
What counts as routinely collected health data?
Data originally collected for clinical, administrative or operational purposes rather than solely for the research question.
Why are code lists important?
They show how diagnoses, procedures, medicines or events were operationally identified.
What is data linkage?
It is the process of connecting records across sources or levels, using identifiers or probabilistic matching.
Does a high linkage rate prove unbiased linkage?
No. Errors may differ across subgroups even when the overall rate is high.
Does a missing code mean a condition is absent?
Not necessarily. It may mean no visit, no test, incomplete capture or recording differences.
Can very large datasets remove confounding?
No. Size improves some aspects of precision but does not create missing causal variables.
Why may routine data not be openly shared?
Privacy, law, ethics and data-custodian agreements may require controlled access.
Is there a new RECORD guideline?
As checked on 5 November 2026, an update is under development; the published 2015 statement remains the completed source used here.
Useful Next Reading
- Read the official EQUATOR Network RECORD record.
- Read the open-access 2015 RECORD statement.
- Check the official status of the developing RECORD update.
- Use STROBE for the underlying observational-study report.
Closing thought: RECORD becomes practical when a reader can reconstruct how operational records became research variables and can see where codes, linkage, missingness, validation and access limits bound the final claim.