Reliable Context Article
ALCOA and ALCOA+ Explained: Data Integrity Attributes
The nine ALCOA+ data integrity attributes, as FDA, MHRA, WHO, PIC/S and EMA word them, with typical failures and the checks a record system must pass.
Summary
- 01ALCOA names Attributable, Legible, Contemporaneous, Original and Accurate. ALCOA+ adds Complete, Consistent, Enduring and Available; EMA's clinical-trials guideline adds Traceable as ALCOA++.
- 02The acronym is a shared checklist, not a regulation. The additions make explicit what inspectors already expected, and integrity must hold throughout the data life cycle.
- 03The attributes apply to paper, electronic and hybrid records. Dynamic records need their native form, and hybrid systems need a documented definition of the whole data set.
- 04Keeping the final result alone is insufficient. Failed, aborted and reprocessed attempts, relevant metadata and explanations for changes must remain reviewable.
- 05Turn each principle into evidence a system can demonstrate: named users, controlled timestamps, visible original values, validated transformations, tested restoration and timely retrieval.
Inside this article
The short answer
ALCOA is an acronym for five attributes that regulated records should have: Attributable, Legible, Contemporaneous, Original and Accurate. ALCOA+ adds four: Complete, Consistent, Enduring and Available. The European Medicines Agency's clinical-trials guideline goes one step further with ALCOA++, adding Traceable. The acronyms come from good documentation practice in pharmaceutical and other GxP (good practice) regulated work, and they now appear in the data integrity guidance of the US Food and Drug Administration (FDA), the UK Medicines and Healthcare products Regulatory Agency (MHRA), the World Health Organization (WHO), the Pharmaceutical Inspection Co-operation Scheme (PIC/S) and the EMA.
The acronym is not a regulation. In the United States, the binding requirements are in the current good manufacturing practice (CGMP) regulations, and FDA's guidance maps each ALCOA attribute to sections of 21 CFR parts 211 and 212 [1]. What the acronym gives quality and data teams is a shared checklist: for any record, can you show who made it, when, what it originally said, and that nothing has been lost or quietly changed since?
This article defines each attribute using the regulators' own wording, shows the failure that typically breaks it, and turns it into a property a record system must be able to demonstrate. It is educational, not regulatory or legal advice.
Where ALCOA sits in the guidance
Each regulator uses the acronym slightly differently.
| Source | Jurisdiction / scope | Acronym used | How it is framed |
|---|---|---|---|
| FDA Q&A guidance, Dec 2018 | US, drug CGMP | ALCOA | "Complete, consistent, and accurate data should be attributable, legible, contemporaneously recorded, original or a true copy, and accurate (ALCOA)" |
| MHRA GXP Data Integrity Guidance and Definitions, Rev. 1, March 2018 | UK, all GxP | ALCOA, with ALCOA+ explained | "There is no difference in expectations regardless of which acronym is used" (§3.10) |
| WHO TRS 1033, Annex 4, 2021 | WHO guidance, GxP | ALCOA+ | Data governance "should ensure the application of ALCOA+ principles" (§4.4) |
| PIC/S PI 041-1, 1 July 2021 | Participating inspectorates, GMP/GDP | ALCOA and ALCOA+ | Table of the nine attributes with a requirement for each (§7.5) |
| EMA guideline EMA/INS/GCP/112288/2023 | EU, clinical trials | ALCOA++ | Ten attributes "considered of universal importance to data" (§4.5) |
MHRA's note on history is useful: "ALCOA was historically regarded as defining the attributes of data quality that are suitable for regulatory purposes. The '+' has been subsequently added to emphasise the requirements" [2]. In other words, the four additions make explicit what inspectors already expected.
Two definitions frame everything else. FDA defines data integrity, for the purposes of its guidance, as "the completeness, consistency, and accuracy of data". MHRA defines it as "the degree to which data are complete, consistent, accurate, trustworthy, reliable and that these characteristics of the data are maintained throughout the data life cycle" [2]. The phrase that matters is "throughout the data life cycle": integrity is not checked once at entry; it must hold until the record is destroyed at the end of its retention period.
The original five
Attributable
PIC/S: "It should be possible to identify the individual or computerised system that performed a recorded task and when the task was performed. This also applies to any changes made to records, such as corrections, deletions, and changes where it is important to know who made a change, when, and why" [3].
Typical failure: shared logins. FDA answers the question directly: "When login credentials are shared, a unique individual cannot be identified through the login and the system would not conform to the CGMP requirements in parts 211 and 212." Shared read-only accounts are acceptable for viewing data, but not for actions such as second-person review [1]. MHRA is equally direct: "shared logins or generic user access should not be used", and where a system supports individual access, "this function must be used. This may require the purchase of additional licences" [2].
Checkable property: every create, change and delete event names one individual or one identified system, with a timestamp.
Legible
PIC/S: records should be "readable and unambiguous in order for it to be understandable and of use". EMA adds a requirement that matters for data pipelines: "changes to data, such as compression, encryption and coding should be completely reversible" [4].
Typical failure: a correction that overwrites the original entry, or a proprietary file format that can no longer be opened.
Checkable property: the original and every corrected value can be displayed in human-readable form for the full retention period.
Contemporaneous
PIC/S: "The evidence of actions, events or decisions should be recorded as they take place." EMA specifies that "the time point of the observation and the time point of the storage should be kept as part of the metadata", and that date and time should be captured automatically and "set by an external standard" [4]. FDA ties the attribute to §§ 211.100(b) and 211.160(a) and states: "You must document, or save, the data at the time of performance to create a record in compliance with CGMP requirements" [1].
Typical failure: results written on scrap paper and transcribed at the end of the shift; a backdated entry; a workstation clock that users can change.
Checkable property: each value carries two system-generated times, when it was observed and when it was recorded, from a controlled clock.
Original
PIC/S: "the first-capture of information, whether recorded on paper (static) or electronically (usually dynamic...). Information that is originally captured in a dynamic state should remain available in that state." A verified "true copy" can stand in for the original [3].
Typical failure: printing dynamic data. FDA explains that a paper printout may be the original record for a pH meter or balance, but electronic records from instruments such as FT-IR are dynamic and "a printout or a static record does not preserve the dynamic record format that is part of the complete original record" [1].
Checkable property: the first-captured file or entry is retained in its native form, and any copy is linked to it with a verification of sameness.
Accurate
PIC/S: "Records need to be a truthful representation of facts to be accurate", supported by qualification and calibration, procedures, deviation management and trained staff. EMA adds that "the process of data transfer between systems should be validated to ensure the data remain accurate" [4].
Typical failure: a validated instrument feeding an unvalidated spreadsheet whose formula silently rounds; a manual transcription with no second check.
Checkable property: every transformation between capture and report is validated or independently verified, and the chain can be shown.
What the acronym gives quality and data teams is a shared checklist: for any record, can you show who made it, when, what it originally said, and that nothing has been lost or quietly changed since?
The additions in ALCOA+
Complete
PIC/S: "All information that would be critical to recreating an event is important when trying to understand the event. It is important that information is not lost or deleted." For electronic data, completeness "includes relevant metadata" [3]. The US regulation is explicit for laboratories: records "shall include complete data derived from all tests necessary to assure compliance with established specifications and standards" [5].
Typical failure: keeping only the final result. FDA's answer to whether a laboratory may save only the final results of reprocessed chromatography is "No": "each result" must be "retained for review" [1]. Trial injections, aborted runs and deleted files are the classic inspection findings.
Checkable property: the system can list every attempt, including failed and superseded ones, and no record can be deleted without leaving a trace.
Consistent
PIC/S: information should be "created, processed, and stored in a logical manner that has a defined consistency", including standard approaches to "chronological sequencing, date formats, units of measurement, approaches to rounding, significant digits". EMA asks for processes "to detect and/or avoid contradictions" [4].
Typical failure: a timestamp sequence where a review precedes the result it reviews; two sites recording the same unit differently.
Checkable property: event sequences are monotonic and units, formats and rounding rules are defined and enforced.
Enduring
PIC/S: records should "remain intact and accessible as an indelible/durable record throughout the record retention period" [3].
Typical failure: data on an instrument's local drive with no backup; a vendor system decommissioned without migrating its audit trail.
Checkable property: backups and archives are tested by restoration, and system retirement includes migration of records and their metadata.
Available
PIC/S: records "should be available for review at any time during the required retention period, accessible in a readable format to all applicable personnel who are responsible for their review" [3]. EMA phrases it as "available when needed".
Typical failure: an archive that can be restored only by a vendor, in weeks, while an inspector is on site.
Checkable property: a defined record can be retrieved, with its metadata, within a stated time.
Traceable (EMA's ALCOA++)
EMA: "Any changes to the data, to the context/metadata should be traceable, should not obscure the original information and should be explained, if necessary. Changes should be documented as part of the metadata (e.g. audit trail)" [4].
Summary table
| Attribute | Short definition | Common failure | Property to demonstrate |
|---|---|---|---|
| Attributable | Who did it, and when | Shared or generic logins | Every event names one person or system |
| Legible | Readable, unambiguous, reversible | Overwritten corrections | Original and corrected values both visible |
| Contemporaneous | Recorded when it happened | Transcription later; backdating | Observation and record times, controlled clock |
| Original | First capture, or verified true copy | Printing dynamic data | Native first capture retained |
| Accurate | Truthful representation of fact | Unvalidated transfer or formula | Validated transformations end to end |
| Complete | Nothing critical lost or deleted | Only final results kept | Every attempt listed, deletions traced |
| Consistent | Defined, logical, contradiction-free | Out-of-order timestamps | Monotonic sequences, enforced formats |
| Enduring | Intact for the retention period | No tested backup | Restorable archives with metadata |
| Available | Retrievable for review when needed | Vendor-only restore | Retrieval within a stated time |
| Traceable (ALCOA++) | Changes explained, original not obscured | Silent edits | Audit trail with who, what, when, why |
- Attributable: every event identifies a person or system and a time.
- Legible: original and corrected values remain readable throughout retention.
- Contemporaneous: observation and recording times come from a controlled clock.
- Original: retain the native first capture and verify any copy.
- Accurate: validate or independently verify every transformation.
- Complete: retain every attempt and leave a trace of deletions.
- Consistent: enforce event order, units, formats and rounding rules.
- Enduring: test restoration and migrate records with their metadata.
- Available: retrieve a defined record and its metadata within a stated time.
EMA's clinical-trials guideline adds Traceable as ALCOA++: changes should be explained and documented without obscuring the original information.
Paper, electronic and hybrid records
The attributes apply to every medium. WHO calls for regular review of data for consistency with ALCOA+ principles that "includes paper records and electronic records" (§4.11), and where a system "cannot support ALCOA+ principles by design (e.g. legacy systems with no audit trail), mitigation measures should be taken for defined temporary periods", such as add-on software or paper-based controls [6]. MHRA requires that for hybrid systems, where paper and electronic records coexist, "it should be clearly documented what constitutes the whole data set" [2].
The audit trail is the mechanism that makes most attributes demonstrable for electronic records. FDA defines it as "a secure, computer-generated, time-stamped electronic record that allows for reconstruction of the course of events relating to the creation, modification, or deletion of an electronic record", with an HPLC example that includes user name, date and time, integration parameters and reprocessing details [1]. WHO describes it as capturing "the 'who, what, when and why' of the action" [6]. The site's audit trail glossary entry gives a shorter definition.
A worked example: one assay result
The example is fictional. Halden Therapeutics AG is a fictional company; the analyst, instrument and numbers are invented.
An analyst at Halden's Basel laboratory runs an assay on batch HT-2026-0412. The first injection gives 97.1% of label claim; the analyst reprocesses with a different integration and gets 99.4%. The audit trail should allow an inspector to reconstruct the following:
| Time (UTC) | User | Action | Value | Reason | Attributes evidenced |
|---|---|---|---|---|---|
| 2026-04-14 09:12:03 | a.keller | Injection 1 acquired (raw file inj1.raw) | — | — | Attributable, Contemporaneous, Original |
| 09:14:40 | a.keller | Integration v1, result | 97.1% | — | Accurate, Complete |
| 09:31:12 | a.keller | Reprocessing, integration v2 | 99.4% | "Baseline drawn through impurity peak; SOP-LAB-022 §5.3" | Traceable, Legible (v1 still visible) |
| 10:05:55 | m.roth | Second-person review | both results viewed | "Reprocessing justified" | Attributable (reviewer), Complete |
The failure version of the same morning has the same final number and no first row, no second row and no reason. That is exactly what FDA's Q14 and PIC/S's "Complete" requirement prohibit: the 97.1% result is part of the record whether or not it was reported.
a.keller acquires the first injection in the raw file inj1.raw.
a.keller records the first integration result, which remains part of the record.
a.keller reprocesses with a different integration and documents the reason; the earlier result remains visible.
m.roth views both results and records that the reprocessing is justified.
From principle to system property
Inspectors test ALCOA+ by asking for evidence. A practical way to prepare is to translate each attribute into a query the system must be able to answer without manual reconstruction:
- For record X, who created it and who changed it, with times? (Attributable)
- Show the value as first captured and every later value. (Original, Legible, Traceable)
- What was the gap between observation time and record time? (Contemporaneous)
- List every run, including failed and reprocessed ones, for batch Y. (Complete)
- Are any event sequences out of order? (Consistent)
- Restore this record from archive, with metadata, now. (Enduring, Available)
- Which transformations sit between the instrument and the certificate of analysis, and how was each validated? (Accurate)
The same questions apply outside the laboratory, to any fact a regulated organization relies on: a supplier's registered address, a product's approval status, a clinical trial's phase. Each value should be attributable to a source and a person, recorded when read, retained in its original form, and changeable only with a trace. The site's method note on retaining source copies and its pharma sector page apply that reasoning to external facts.
Self-inspection checklist
- Individual user accounts are enforced for every system that creates or changes GxP data; shared accounts are view-only.
- System clocks are synchronized to an external standard and cannot be changed by users.
- Audit trails are on, cannot be disabled by users, and record who, what, when and why.
- Dynamic electronic records are retained in native form; printouts are not treated as originals.
- Every test attempt, including aborted and reprocessed runs, is retained and reviewable.
- Corrections never obscure the original entry.
- Formats, units and rounding are defined in procedures and enforced in systems.
- Backups and archives are restored on a schedule, including metadata.
- Retrieval time for an archived record is known and acceptable for an inspection.
- Hybrid systems have a documented definition of the complete data set.
Reliable Context publishes a related guarantee for organizational facts: a cited, verbatim, dated source behind every value, with any past state reproducible.
Sources
- FDA, Data Integrity and Compliance With Drug CGMP: Questions and Answers, Guidance for Industry, December 2018 - ALCOA definition and CFR mapping (Q1, fn 5); audit trail; shared logins (Q5); static vs dynamic (Q10); time of performance (Q12); reprocessing (Q14).
- MHRA, GXP Data Integrity Guidance and Definitions, Revision 1, March 2018 [2] - ALCOA vs ALCOA+ (§3.10); data integrity definition (§6.4); shared logins (§6.16); hybrid systems.
- WHO, Guideline on data integrity, TRS 1033 Annex 4, 2021 (PDF) - ALCOA+ glossary definition; data governance; audit trail definition.
- PIC/S, PI 041-1, Good Practices for Data Management and Integrity in Regulated GMP/GDP Environments, 1 July 2021 - attribute-by-attribute requirements (§7.4-7.5); true copies (§7.7).
- EMA, Guideline on computerised systems and electronic data in clinical trials, EMA/INS/GCP/112288/2023 (PDF) - ALCOA++ including Traceable (§4.5).
- 21 CFR 211.194, Laboratory records (eCFR) - "complete data derived from all tests".
External Sources (6)
Published by
Reliable Context Research
Reliable Context is a system of record for organizations that fuse private and public information and must defend every value they act on: one canonical record per real thing, ambiguity shown rather than hidden, and provenance behind every value.
Editorial note
Researched and written by Reliable Context Research with AI-assisted research tools; sources are cited inline and checked before publication. Report an error: https://reliablecontext.com/corrections.