ALCOA+ Data Integrity Principles
Use the nine attributes to assess whether manufacturing and laboratory records preserve trustworthy evidence from first capture through review and retention.

On this page · 11 sections
- 1What ALCOA+ means—and what it does not establish
- 2Map the framework to the right product and jurisdiction
- 3Verify the five ALCOA attributes in actual records
- 4Verify completeness, consistency, endurance and availability
- 5Decide what counts as the record, copy or printout
- 6Apply the attributes from capture through retention
- 7Three worked cases: dispensing, laboratory calculations and retransmission
- 8Reconcile interfaces and recover without rewriting history
- 9Find failure modes that polished reports can hide
- 10Test record integrity before accepting a system
- 11Where V5 fits in assembling and checking the evidence
How does ALCOA+ apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What ALCOA+ means—and what it does not establish
ALCOA+ is a guidance and industry framework for assessing record quality. PIC/S PI 041-1, section 7.4 names Attributable, Legible, Contemporaneous, Original and Accurate, then adds Complete, Consistent, Enduring and Available. The PIC/S data-integrity guidance connects these attributes to documenting events and supporting informed decisions.
ALCOA+ is not a statutory term or certification. Assess compliance against requirements applicable to the product, activity and jurisdiction; an “ALCOA+ compliant” label is not evidence that those requirements have been met.
The shorter acronym used in MHRA GxP guidance does not imply fewer expectations. MHRA section 3.10 explains that expectations do not differ with the acronym used: governance should ensure completeness, consistency, endurance and availability throughout the data lifecycle.
| Related term | Meaning and boundary |
|---|---|
| ALCOA++ | An extension whose additional terms should be checked against its author’s definition; do not assume a single official ALCOA++ list. |
| Data integrity | The completeness, consistency, accuracy, trustworthiness and reliability of data, maintained throughout their lifecycle. |
| Good documentation practice | Working controls for creating and maintaining records. These controls support the attributes rather than replace them. |
For example, an instruction requiring operators to identify their entries is a documentation control. Inspecting a completed weighing record to establish who made an entry tests attribution. Select an actual observation and check whether its author information identifies the individual rather than only a department or shared account. A written instruction alone does not demonstrate that the control operated.
02Map the framework to the right product and jurisdiction
Apply the boundaries above through a scope matrix covering product category, activity, market and record type. Start with applicable manufacturing practice and assess Part 211 applicability where relevant. For supplement, food, device and chemical operations, identify their own applicable duties before adopting a drug-sector example. This matrix is an implementation tool, not a prescribed regulatory form.
Keep separate fields for the binding requirement, guidance recommendation, locally chosen control and verification evidence. Assign an owner to check applicability and source version before approving the assessment. Describe the activity precisely: a packaging execution record, for example, presents different evidence questions from a laboratory processing history.
| Context | Source status | Specific point to map |
|---|---|---|
| EU medicinal-product GMP documentation | GMP requirement | Chapter 4, section 4.10 requires defined relationships between manufacturing activities and records, identified record locations, and secure integrity controls throughout retention. |
| MHRA GXP | Official guidance | MHRA section 2.4 excludes medical devices from this document’s scope. |
| PIC/S GMP/GDP remote inspection | Official guidance | PI 041-1 section 4.2 limits remote assessment to data governance systems; on-site assessment is normally required for data verification and evidence of operational compliance with procedures. |
For a supplement dispensing log, a local control might specify retaining the original weighing record. Identify the applicable authority separately rather than assigning a drug regulation by analogy. Name the evidence to inspect: the saved weighing result, material and lot reference, operator identity and correction history. If several systems contribute, identify which owner supplies each component and who resolves a missing record.
Record software reports and their locations in the evidence field. A useful matrix entry tells the reviewer what to open and compare; a generic entry such as “system supports ALCOA+” does not define a verification task.
03Verify the five ALCOA attributes in actual records
The core definitions below follow PIC/S PI 041-1 section 7.5. The shop-floor examples, laboratory examples and verification tests are practical applications. Select an identifiable activity and follow its evidence from observation to reported result, rather than assessing only the appearance of a final report.
| Attribute | Operational meaning | Shop-floor example | Laboratory example | Evidence to inspect |
|---|---|---|---|---|
| Attributable | Identify the individual or system performing a task and when it occurred. Connect changes to who made them, when and why. | An identified operator records a material addition against the batch and step. Distinguish the performer, recorder and reviewer. | An acquisition identifies the analyst, instrument and sample, rather than showing only the person approving the report. | Trace the event to its actor and role. For a correction, inspect author, time and reason. Check whether an automated event identifies its originating system. |
| Legible | Keep information readable and unambiguous, including notation, units and contextual fields. | A weighing entry identifies whether the recorded mass is gross, tare or net and states its unit. | A result presents units, method identification and interpretable qualifiers; an unexplained symbol is not sufficient context. | Open the rendered record and referenced instructions. Check decimal notation, labels and annotations. For handwritten entries, ask a reviewer unfamiliar with the activity to interpret them. |
| Contemporaneous | Record actions, events and decisions as they take place; distinguish occurrence time from later entry or correction. | Record the weighing during execution instead of reconstructing the sequence from memory after the shift. | Distinguish acquisition time from subsequent processing time. A later calculation does not change when measurement occurred. | Compare event, entry and correction timestamps. Investigate late entries using available supporting evidence; one timestamp should not be assumed to describe every action in the workflow. |
| Original | Identify the first capture. FDA’s CGMP guidance, Q1(a) describes this attribute as “original or a true copy”. | Determine whether the weighing was first recorded as an instrument output, controlled paper entry or electronic event. | Identify the acquisition record and its sample and method context, rather than starting and ending with the approved result summary. | Trace a reported value back to first capture. Identify each intermediate transcription, calculation or transformation so the reviewer can follow how the reported value was produced. |
| Accurate | Represent the facts faithfully, including correct recording and processing. | Compare balance output with the batch entry, checking instrument status at use, units and transcription. | Check calculation inputs, method version, processing parameters and rounding against the reported result. | Independently check the processing and compare recorded values with source evidence. Assess measurement validity separately: correct arithmetic does not establish that the underlying observation or measurement was valid. |
For electronic identity, test how an operator badge scan connects the active user to an event, including shift handovers. Observe whether the next operator’s work remains assigned to the previous user. This tests the implemented workflow rather than merely confirming that a badge reader exists.
For measurement control, inspect as-found and as-left calibration evidence alongside maintenance and use history. A current status label does not explain the instrument’s condition when an older measurement was made. Record the sampled activity, evidence inspected and unresolved issue so another reviewer can reproduce the assessment.
04Verify completeness, consistency, endurance and availability
The four additional attributes introduced above address the record set and its continued usability. Their meanings follow PIC/S PI 041-1 section 7.5; the examples and tests below translate them into practical review activities.
| Attribute | Practical meaning | Shop-floor example | Laboratory example | Verification test |
|---|---|---|---|---|
| Complete | Preserve the information needed to reconstruct the activity, not just its final summary. | Account for completed, interrupted and superseded steps, with their status. | Account for valid, failed, aborted and reprocessed runs, with status and rationale. | Compare batch execution histories and instrument acquisition lists with retained records. Investigate unexplained omissions. |
| Consistent | Create, process and store information coherently, including chronology, identifiers, units and rounding conventions. | Match material additions to the correct batch and execution sequence. | Connect acquisition, processing and review versions to the reported result. | Trace identifiers across records and compare date formats, units and version references. Investigate contradictions rather than silently normalising them. |
| Enduring | Keep records intact, durable and accessible throughout the applicable retention period. | Preserve the meaning of equipment references after equipment retirement. | Keep older analytical records interpretable after methods or software change. | Open representative older records and reconstruct their meaning. Include records from retired equipment or superseded systems in the sample. |
| Available | Make records accessible in readable form to the personnel responsible for review throughout retention. | Retrieve a historical batch record together with its linked deviations. | Retrieve source data with processing and review context. | Test historical record requests using the intended reviewer’s permissions. Record missing links, access restrictions and retrieval obstacles. |
Retaining an attempt does not make its result reportable or suitable for averaging. Separate evidence retention from result selection; use laboratory-result investigation to document why results are accepted, rejected or investigated. The retained history should let a reviewer examine that reasoning, not conceal it behind the final value.
Classify a failed test precisely. A record that exists but cannot be opened is a different problem from an acquisition missing from the retained dataset. Likewise, a broken batch link differs from a disagreement between two recorded values. Assign an owner and evidence-based resolution to each finding. Repeat the affected test after correction; restoring access does not, by itself, resolve a content discrepancy discovered during the review.
05Decide what counts as the record, copy or printout
Identify the record set before deciding whether a scan, PDF or printout can replace it. The PIC/S Metadata and Raw Data definitions treat metadata as integral to the original record and say information captured dynamically should remain available in that state. Assess dynamic versus static records by retained content and interaction, not file extension.
For US drug CGMP records, FDA guidance, question 9 permits electronic true copies provided they preserve the original record’s content and meaning, including metadata. The practical assessment therefore asks what the proposed copy preserves, not merely whether it looks like the source when opened.
Practical copy-assessment steps
- Locate first capture: identify where each observation first became recorded information.
- Map context: connect results to units, identities, timestamps, method versions and relevant change history.
- Test interaction: identify queries, detailed views or reprocessing needed for review that a fixed page cannot provide.
- Define the record set: list source data, metadata, paper components and electronic assets, with stable links to their production context.
- Compare the proposed copy with that set. Check content, relationships and necessary functionality. Resolve any gap that affects required content, meaning, metadata or functionality before approving a replacement; if a gap cannot be resolved, restrict what the copy may be used for and retain the source record.
MHRA’s raw-data guidance identifies the printout as raw data for basic equipment that does not store electronic data or provides only printed output. The PIC/S Hybrid Systems definition identifies electronic and paper data together as the complete hybrid dataset. In a hybrid record system, assess both components rather than selecting whichever is easier to retrieve.
| Assumed setup | Practical verification |
|---|---|
| Balance with print-only output | Inspect the original printout’s identity, units and connection to the dispensing entry. |
| Chromatography system retaining interactive data | Test access to native data and processing context, not only the report PDF. |
| Paper execution sheet with electronic measurements | Retrieve both components and match measurements to the correct execution. |
Record which functions and relationships were compared, who performed the assessment and any limitation on the copy’s intended use. Stable identifiers assist this work but do not establish completeness by themselves.
Record
Connects manufacturing or laboratory context.
- record_id
- batch_id
- sample_id
- source_record_id
Capture
Preserves the observation and its origin.
- capture_id
- record_id
- actor_id
- event_at
- value
- unit
Change Event
Explains changes without hiding previous values.
- change_id
- record_id
- changed_by
- changed_at
- old_value
- new_value
- reason
Review
Links a person's decision to a version.
- review_id
- record_id
- record_version
- reviewer_id
- reviewed_at
- decision
Retained Asset
Locates evidence and its retention rule.
- asset_id
- record_id
- storage_uri
- format_version
- retention_rule_id
06Apply the attributes from capture through retention
Use this practical sequence to assign evidence handoffs. Local procedures identify the sender, recipient and decision authority; the sequence is not a required organisational chart.
- Define scope: the process owner and quality representative identify the intended decision and expected evidence, such as dispensing records or laboratory acquisitions.
- Capture: the operator or analyst creates the record using the controls and attribute checks described above.
- Preserve context: the custodian maintains relationships between results, sample or batch identifiers, method versions and supporting records.
- Transfer and reconcile: sending and receiving owners compare expected and received records, then assign discrepancies for investigation.
- Review: the designated reviewer assesses the evidence and exceptions, recording unresolved issues for the decision owner.
- Approve: the authorised human decision-maker determines whether the evidence supports disposition and records the decision. People approve and release; successful data transfer is not approval.
- Retain and retrieve: the records owner assigns retention periods from applicable requirements, documents the record locations and schedules the availability checks described above.
For US drug CGMP records, FDA guidance, question 8 says to follow the data-review frequency specified in applicable CGMP regulations for audit-trail review. Where no frequency is specified, it recommends process knowledge and risk assessment, including data criticality, control mechanisms and impact on product quality.
Implement that decision in the local audit-trail review workflow: identify relevant trails, review scope, reviewer, timing and escalation route. Retain the review outcome and track unresolved exceptions to the person responsible for the resulting decision.
Destruction is a separate decision from the copy assessment above. MHRA’s retention and destruction guidance recommends considering data criticality and applicable legislative retention requirements in destruction procedures. As a local control, confirm the approved record set, retention obligations and disposal authorisation before destroying source material. A successful scan or migration is not, by itself, permission to delete the source.
- Define scopeIdentify decisions, source records, owners and retention rules.
- Capture eventsRecord actor, value, unit and event time together.
- Preserve contextLink source data, metadata and related attempts.
- Reconcile transfersMatch event identities, units and sequence across interfaces.
- Review exceptionsAssess changes, gaps and anomalies against source evidence.
- Approve decisionsAn authorised person documents disposition and review scope.
- Retain and retrieveTest readable retrieval; control eventual authorised disposal.
07Three worked cases: dispensing, laboratory calculations and retransmission
These hypothetical training cases apply the attributes covered above. The checks are industry-practice examples using assumed procedures, not additional regulatory requirements or product acceptance criteria.
Case 1: Preserve an out-of-tolerance dispense
Assume an approved instruction specifies 5.000 kg net with a tolerance of ±0.5%. The allowance is 5.000 × 0.005 = 0.025 kg, giving a permitted interval of 4.975–5.025 kg. An observed net mass of 5.040 kg is 0.040 kg above target: (0.040 ÷ 5.000) × 100 = 0.8%. It exceeds the upper limit of the assumed weighing tolerance band by 0.015 kg.
To test original, attributable and accurate recording, inspect the balance capture and executed dispensing record. Verify material, lot, work order, instruction version, operator, timestamp, balance identity and tare/net basis. Preserve 5.040 kg; replacing it with 5.000 kg would conceal the exception. Record the subsequent action and its authorization. The calculation alone does not decide material disposition.
Case 2: Include every method-defined valid replicate
Assume an approved laboratory method specifies the unweighted mean of three valid reportable replicates, expressed as percent label claim and rounded to one decimal place. Results of 98.4%, 100.1% and 99.5% sum to 298.0%; 298.0 ÷ 3 rounds to 99.3%. Selecting only the two higher results gives 199.6 ÷ 2 = 99.8%, contrary to the assumed method.
For completeness and accuracy, inspect instrument records, replicate identifiers and the calculation worksheet. Check sample identity, method version, validity decisions, reporting basis, formula and rounding. Investigate the omission; preserve the selected summary and its correction history while verifying the method-defined result. This is neither permission to average invalid runs nor a rule for every analytical method.
Case 3: Distinguish retransmissions from measurements
Assume an interface procedure permits one business update per measurement event. A sender creates 120 unique events; the receiver logs all 120 original messages plus four unchanged retransmissions: 120 + 4 = 124 deliveries, but only 120 measurements.
For completeness and consistency checks, compare the sender’s event register with receiver and processing logs. Reconcile the actual event-ID sets: equal counts can hide one missing and one unexpected event. Compare payloads, units, timestamps and source identities; flag reused IDs with changed content. Retain retransmission evidence and verify one business update per expected event. This establishes delivery and processing coverage, not measurement accuracy.
08Reconcile interfaces and recover without rewriting history
For GMP/GDP interfaces, PIC/S PI 041-1 §9.4, expectation 1 recommends assessing interfaces during validation for correct, complete data transfer. The following are engineering practices, not prescribed fields or claims about software capabilities.
For an instrument-to-laboratory-to-batch-record path, define an interface contract: stable event ID, source-system ID, batch/sample ID, value, unit, event time, receipt time and record status. See MES-to-LIMS integration.
Define unit conversions, timestamp offsets, numerical precision, schema versions and status mappings. Keep source and transformed values distinguishable; record the conversion, rounding or mapping rule and version. Never map an unknown status silently to an accepted result.
For recovery: detect gaps; preserve source evidence; contain affected decisions; identify missing and duplicate events; reconcile against the designated authoritative event set; replay through controlled handling that prevents duplicate application; verify results; and document recovery and human review. Check event IDs, values, units and statuses—not just delivery receipts.
During downtime, use an approved fallback record capturing operator identity, context and actual event time. Record later transcription time separately; never backdate entries to conceal interruption. After restoration, interface repair or replay, verify readability and correct sample, batch and source-record relationships before affected decisions resume. A designated reviewer should confirm the checks and record unresolved exceptions.
09Find failure modes that polished reports can hide
A polished summary does not establish record integrity. In GMP/GDP guidance, PIC/S PI 041-1 section 6.1.9 recommends controls to prevent, detect, assess and correct breaches, and checking their performance. These illustrative practice tests are not evidence of incident frequency.
| Failure mode | Detection test | Control or recovery |
|---|---|---|
| Administrator bypass | Compare privileged activity for a sampled period with authorised tasks. | Prevent: restrict elevated access. Recover: investigate unmatched actions and identify affected records. |
| Review dashboard excludes invalidated events | Reconcile dashboard coverage against source event IDs and statuses, including invalidated events. | Prevent: test status filters. Recover: correct coverage and repeat affected reviews. |
| Export loses required precision | Compare exports with source values and specified rounding near decision limits. | Prevent: specify required precision. Recover: regenerate from preserved source data and assess affected decisions. |
| Unowned paper/electronic reconciliation | Sample cross-medium record links in both directions; identify who resolved mismatches. | Prevent: assign reconciliation ownership. Recover: investigate gaps and document supported links. |
| Archive lacks a working reader | Retrieve an older record using retained interpretation tools; inspect values and metadata. | Prevent: maintain tested readers. Recover: restore the environment or verify migration before use. |
If a test fails, preserve evidence and contain further risk. Assign an investigation owner, assess affected records and decisions, and use MES–QMS investigation routing where appropriate. Repeat failed challenges after remediation to verify effectiveness rather than treating action closure as proof.
10Test record integrity before accepting a system
Use this as a practical acceptance checklist, not a compliance certification. Agree pass/fail criteria before supplier demonstrations; inspect records, metadata and histories rather than compliance labels.
- Scope: Select test records for a shop-floor weighing and a laboratory result. Map their source data, interfaces, metadata, retention needs and review decisions into testable validation requirements. Identify operator, administrator, reviewer and data-owner roles.
- Demonstrate: Capture a value, correct it and invalidate a laboratory run. Check attribution to the actual user, recording time, original and revised values, reason for change and retention of invalidated data. Export and retrieve an older record; verify readability and its connection to supporting metadata.
- Challenge: Interrupt a transfer and reconcile source and destination records for missing or duplicate entries. Test known changes and deletions. PIC/S PI 041-1, section 9.6 recommends verifying audit-trail functionality during validation to capture all changes and deletions of critical data associated with manual activities.
- Assign: Name owners for gaps, training and acceptance. For IT suppliers and service providers, including cloud services, MHRA section 6.20 recommends documenting contract-giver and contract-acceptor responsibilities in a technical agreement or contract, ensuring timely access to data, including metadata and audit trails, for the data owner and national competent authorities upon request.
- Verify: Record each test’s expected and actual results and unresolved gaps with their operational consequences. Have trained reviewers repeat selected tests without supplier coaching. Reassess affected tests after relevant configuration, interface, workflow or service changes; retain evidence and the accountable acceptance decision.
11Where V5 fits in assembling and checking the evidence
Worked review: an in-tolerance value entered incorrectly
This hypothetical practice case concerns a manually transcribed gravimetric dispense. Assume a target net mass of 10.000 kg and an approved weighing tolerance band of ±0.020 kg, with inclusive endpoints. The acceptable interval is therefore 9.980–10.020 kg. Assume the displayed values are directly comparable without further correction, and the site retains the original balance printout, operator identification and source-system entry history. These are example conditions, not recommended tolerances.
The reviewer selects the manually entered net mass as the critical field to test. The assembled record shows 9.998 kg and a variance of −0.002 kg. The reviewer retrieves the associated printout and matches its batch, material lot, bag and balance identifiers to the dispense step. The printout shows 9.980 kg. Its timestamp and operator identification also match the recorded operation.
For manual entry of critical data in GMP/GDP computerised systems, PIC/S PI 041-1, section 9.7 recommends verification by a second operator or validated computerised means. Here, the reviewer performs a source comparison rather than accepting the displayed variance as the check.
| Check | Observed evidence | Initial finding |
|---|---|---|
| Record identity | Batch, lot, bag and balance identifiers match. | Pass for the sampled dispense. |
| Attribution and timing | The identified operator and event timestamp agree across the records. | No discrepancy found in these fields. |
| Transcription accuracy | Source: 9.980 kg. Entered value: 9.998 kg. | Fail: entered mass is 0.018 kg too high. |
| Variance calculation | 9.980 − 10.000 = −0.020 kg. | Recorded −0.002 kg uses the wrong input. |
| Process tolerance | 9.980 kg equals the inclusive lower limit. | Pass under the stated example conditions. |
Both the source value and the incorrect entry fall inside the acceptance interval. A tolerance-only check would therefore miss this error. The recorded variance is mathematically correct for 9.998 kg, but that does not make its input reliable. The reviewer records the transcription failure separately from the process-tolerance finding.
Escalation and documented reviewer outcome
The reviewer opens a deviation under the assumed site procedure and checks the originating system’s entry and change history. In this example, 9.998 kg was present at first entry; no later edit explains the mismatch. The audit-trail review workflow records that finding. PIC/S PI 041-1, section 9.8 recommends investigating and recording significant variations from expected outcomes found during audit-trail review.
An authorised correction changes the entered mass to 9.980 kg and the variance to −0.020 kg, retaining the prior entry and identifying the correction’s author, actual time and reason. The correction references the printout and deviation. A second reviewer checks both corrected fields against the source and calculation. This independent double check tests the corrected evidence; it is not simply another signature beside the original error.
To establish the immediate scope, the reviewer compares the remaining manually entered dispense weights for this work order with their retained sources. Assume no further mismatches are found. The review names the records examined rather than concluding that all site data are reliable. A separate follow-up action assesses whether field layout, transcription practices or the checking procedure contributed to the error.
The completed review records the initial failure, evidence locations, correction history, recalculation, expanded comparison and secondary verification. Its outcome is specific: the sampled transcription discrepancy is resolved; the source dispense met the assumed tolerance; the follow-up action remains tracked. Retain this reasoning with the batch execution history for the batch-release review, applying the approval responsibilities covered above.
Frequently asked questions
Q.Does a small paper-based operation need electronic records to apply ALCOA+?+
No. MHRA section 4.3(i) recognises paper recording of manual observations or activities. For a small site, start with one completed operation and test whether a reviewer can reconstruct it without relying on staff memory. Account for issued forms, check that entries identify their authors, and verify that corrections remain visible. If instruments also retain electronic data, include those files in the hybrid-record assessment. Use the findings to improve the existing process before deciding whether digitisation would help.
Q.What should an operator do when an entry was not made at the time of the activity?+
Use the site's late-entry procedure rather than making the record appear contemporaneous. As a practical approach, identify the entry as late, record the actual entry date and time, explain the omission, and distinguish the activity time from the recording time. State what supports the reconstruction, such as an instrument timestamp; do not substitute an unlabelled estimate for missing evidence. Escalate uncertainty through the deviation process.
Q.When can an instrument printout stand in for its electronic source record?+
Only after the assessment in “Decide what counts as the record, copy or printout” supports that use, including its safeguards for dynamic records. Record that decision for the instrument and method, identify the approved record location, and tell reviewers whether the printout is a review copy or the retained record. Keep electronic sources available pending a separately authorised decision on replacement or destruction that addresses applicable retention obligations.
Q.Does every system event need the same audit-trail review frequency?+
No; use the frequency approach in “Apply the attributes from capture through retention.” For implementation, build an event catalogue for the system: identify entries concerning results, settings, permissions or communication failures, and assign each category an owner and review route. In the audit-trail review workflow, record the scope searched, time interval covered, findings and follow-up. Test filters with known events so an empty report is not mistaken for evidence that nothing relevant happened.
Q.How should failed runs, aborted tests and reprocessed laboratory results appear in the record?+
As a practical completeness check, make each attempt traceable to its sample, sequence or run identifier, operator, timing and disposition. Preserve the initial result and connect subsequent processing or testing to its reason and review decision. Distinguish an instrument failure from a valid measurement that fails acceptance criteria; neither should disappear merely because a later result passes. For out-of-specification handling, reviewers should be able to follow the investigation from the first observation to the reported conclusion.
Q.What data-access arrangements should a manufacturer establish with a contract laboratory or cloud provider?+
For cloud and IT services covered by MHRA section 6.20, the guidance recommends contractual responsibilities covering timely access to data, metadata and audit trails, archiving, continued readability and tested business continuity. For a contract laboratory, a practical extension is to specify access to the underlying analytical files behind each report. Agree retrieval times, usable export formats and exit arrangements, then test them using a representative dataset. Link these expectations to acceptance evidence through requirements traceability.
Q.What does ALCOA++ mean when it appears in a supplier specification?+
Start with the terminology boundaries in “What ALCOA+ means—and what it does not establish.” Then ask the supplier to turn its wording into acceptance criteria: for each claimed addition, name the record affected, the action being controlled and the evidence a tester should see. For a traceability claim, demonstrate navigation from a reported result to its source and subsequent changes. Capture the agreed tests in the specification using data integrity by design.
Q.What evidence should support a supplier's claim that a system is ALCOA+ compliant?+
Apply the claim boundaries in the definition section and the tests in “Test record integrity before accepting a system.” For EU GMP computerised systems, Annex 11 section 3.3 recommends reviewing commercial off-the-shelf documentation against user requirements. Ask for demonstrations in the proposed configuration, retaining test outputs, configuration details and unresolved gaps. Include attributable recording tests with actual operator and administrator roles. Record who assessed each gap and what must be resolved before acceptance.
Primary sources
- EudraLex Vol 4 Part I Chapter 4 — Documentation
- EudraLex Vol 4 Annex 11 — Computerised Systems
- FDA guidance: Data Integrity and Compliance With Drug CGMP — Questions and Answers (2018)
- MHRA: GXP Data Integrity Guidance and Definitions (2018)
- PIC/S PI 041-1: Good Practices for Data Management and Integrity in Regulated GMP/GDP Environments (2021)
Explore this topic
ALCOA+ sits inside 2 overlapping topic clusters in our glossary. Every neighbour is one click away.
Electronic records, signatures, audit trail and ALCOA+ data-integrity principles.
HCT/Ps, blood cGMP, donor eligibility, ISBT 128 labeling and forward-tracing lookback.
Want to see how ALCOA+ could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
