GMP data integrity means that manufacturing, laboratory and quality records stay complete, consistent and accurate from the moment they are created until the end of their retention period. The yardstick is ALCOA+: every record must be attributable, legible, contemporaneous, original and accurate, and must also remain complete, consistent, enduring and available.
Key takeaways
- ALCOA+ comprises nine attributes: attributable, legible, contemporaneous, original, accurate, complete, consistent, enduring and available.
- The US FDA's December 2018 guidance defines ALCOA; the UK MHRA's March 2018 guidance defines the four extra attributes in ALCOA+.
- According to the FDA, audit trails are reviewed as often as the data itself, for example after each significant manufacturing step and before batch release.
- The weakest links are hybrid records, where paper and electronic records coexist, and shared user accounts.
- A risk-based six-step programme starts with a data flow map and scales up from a single-line pilot.
Why does GMP data integrity start with the record?
In a plant making medicines or active pharmaceutical ingredients, product quality and record quality cannot be separated; suppliers such as pharmaceutical packaging makers meet similar expectations in customer audits. The decision to release a batch rests on the assumption that weighing, blending, filling, in-process checks and lab results are trustworthy. If the record is questionable, the batch becomes hard to defend even when the product itself tests fine.
What we typically find on site looks like this. The batch record is on paper, instrument printouts are stapled to it and some values are corrected afterwards in biro. Everyone logs in to the lab instrument PC as the same "admin". Nobody can say for certain which version of the stability spreadsheet is the current one.
As of 2026, in Türkiye, manufacturers of human medicines are inspected against Good Manufacturing Practice (GMP) by the national regulator, the Turkish Medicines and Medical Devices Agency (TİTCK); veterinary medicines and cosmetics fall under different rules and oversight, and this article focuses on human medicines GMP. Sites that supply the EU, UK or US must also meet the data integrity expectations of those markets' regulators. This guide treats ALCOA+ not as an abstract checklist but as concrete requirements that drive system and process decisions.
What do ALCOA and ALCOA+ mean in GMP data integrity?
ALCOA is an acronym for five attributes data must have. The US Food and Drug Administration's Data Integrity and Compliance With Drug CGMP guidance (FDA, December 2018) states plainly that complete, consistent and accurate data should be attributable, legible, contemporaneously recorded, original or a true copy, and accurate. CGMP stands for Current Good Manufacturing Practice, the term used in US regulations.
The UK regulator's MHRA 'GXP' Data Integrity Guidance and Definitions (March 2018) defines the "+" as four further attributes: complete, consistent, enduring and available. MHRA stresses that expectations are the same whichever acronym you use: ALCOA+ simply spells out the same requirement more fully.
In practice, the nine ALCOA+ attributes boil down to one question: when an inspector asks about a batch from two years ago, can you show what was done, who did it and when, and prove that nothing was quietly changed since?
What do data integrity gaps cost a pharmaceutical manufacturer?
Data integrity weaknesses hurt most during an inspection. A finding rarely stays confined to one record; the regulator starts to question every other record produced by the same system. That means corrective action, retrospective review and lost confidence among export customers.
Trust in data is low across industries, not just pharma:
A survey of 565 data and analytics professionals by Drexel University's LeBow College of Business and data software company Precisely found that 67% of respondents don't fully trust the data their organisation uses, and 64% named data quality as the top challenge to data integrity. (Drexel LeBow & Precisely, 2025 Outlook: Data Integrity Trends and Insights, September 2024)
Poor data is expensive, too. Data quality expert Thomas C. Redman, writing in MIT Sloan Management Review in 2017, estimated the cost of bad data at 15% to 25% of revenue for most companies.
In pharmaceuticals, rejected batches, longer release times and inspection findings come on top of that cost. The everyday data quality and duplicate record problems every business knows become compliance problems in a regulated plant.
How are ALCOA+ principles met in systems and processes?
Knowing the ALCOA+ principles is not enough; each one has to become something a system or procedure actually does. The table below summarises this and doubles as a gap assessment checklist.
| Principle | What it means on site | System or process control | Typical gap |
|---|---|---|---|
| Attributable | Each entry tied to one person | Personal user accounts, role-based permissions | Shared "operator" or "admin" login |
| Legible | Permanent and readable | Digital record, meaningful field names | Scribbles, overwritten figures |
| Contemporaneous | Recorded as the work happens | Server-time timestamps | Batch record filled in at end of shift |
| Original | First capture or certified copy | Automatic capture from the instrument | Values retyped from an instrument display |
| Accurate | Reflects what really happened | Tolerance checks, calibration lock | Measuring with an out-of-calibration gauge |
| Complete | Nothing lost | Justified correction instead of deletion | First result of a repeated test discarded |
| Consistent | Sequence and timing make sense | Enforced step order | Record dated before the preceding step |
| Enduring | Survives the retention period | Backups and restore tests | Instrument data left on a local drive |
| Available | Readable when requested | Searchable archive | File that only opens in an old software version |
The enduring and available principles, the last two rows of the table, are easily overlooked. A record can be correct today yet fail if it cannot be read throughout its retention period. That is why regular backup and restore testing belongs inside the data integrity programme rather than being a separate IT chore. The same logic applies to warehouse and cold room records, as our pharmaceutical warehouse temperature mapping guide explains.
How do you build a data integrity programme? A six-step plan
In a data integrity programme, working by risk rather than digitising everything at once protects both compliance and budget. We recommend this order:
- Map the data flow. From goods receipt to batch release, record where each critical entry is created, where it is copied and who approves it. Mark paper, spreadsheets and instrument software separately.
- Rank by criticality. Records that directly affect product quality and patient safety (weighing, filling, sterilisation, results reporting) come first.
- Reduce hybrid records. Wherever paper and electronic records coexist is the weakest link. Capturing values straight from the instrument, and guiding operators through a terminal instead of a paper form, fixes both contemporaneity and originality. We cover the shop-floor side in our paperless manufacturing guide.
- Fix identity and access. Remove shared accounts, define roles and revoke leavers' access on the day they go. Who may approve what should be written down.
- Define and review the audit trail. Who changed which value, when, from what to what and why must be held in a form that cannot be deleted. FDA guidance says that where CGMP regulations set a review frequency for the data, the audit trail should be reviewed at the same frequency, for example after each significant manufacturing step and before batch release. Where no frequency is set, decide one using process knowledge and risk assessment.
- Validate and maintain. Document that each computerised system is fit for its intended use, route changes through change control and put periodic reviews on the calendar.
Audit trail review, step five, is the one most often skipped. Switching the audit trail on is not the same as reading it; a trail nobody reviews only makes the finding easier for the inspector to spot. Turning raw logs into meaningful review mirrors SIEM and log management in information security.
Which system should hold which record for data integrity?
Data integrity is not solved by a single piece of software; each record needs to be born in the right system. Production steps, weighings, machine parameters and operator sign-offs belong in a manufacturing execution system (MES) as an electronic batch record. See our MES vs ERP comparison for how MES and enterprise resource planning (ERP) divide the work.
Non-conformances, deviations, corrective and preventive actions (CAPA) and calibration records live in quality management system software. Blocking any entry from an out-of-calibration instrument is the most concrete expression of "accurate". We cover module selection in our quality management software guide.
The link from raw material lot to finished pack comes from barcode-based production traceability. Without it, completeness and consistency are hard to claim; the basics of forward and backward tracing are in our production traceability guide. In food, the same discipline appears as food traceability and digital HACCP.
How does Digital Bridge run a data integrity project?
Digital Bridge starts a data integrity project with discovery and requirements analysis, not a product pitch. Together with your quality and production teams we map the flow of critical records and mark, using the ALCOA+ table, where each one falls short of ALCOA+. The output is a prioritised gap list and a written proposal setting out scope, phases and cost. This work usually runs as part of our digital transformation consultancy.
Next comes a pilot on a single line or process. A typical pilot captures weighing and in-process check values automatically from the instruments on one filling line. Operators sign off with their personal account on an industrial data terminal, and every correction lands in the audit trail with its reason. Hardware and software come from the same team, so device connection and screen flow have one owner.
Then comes integration. The Digital Bridge MES works with Logo, SAP, Mikro and custom ERP systems, so materials and work orders are never keyed in twice. We also fold data governance and quality work into the programme to settle record ownership, definitions and retention rules. Our data governance framework guide explains how ownership is set up at company level.
Validation documentation remains your quality team's responsibility; we design the system to produce the requirements, test and change records that documentation depends on, and organise those records together with you.
What are the most common data integrity mistakes?
- Scanning paper and calling it digital. A scanned record is legible, but it does not become contemporaneous or attributable.
- Only switching the audit trail on. Without a defined review frequency and owner, the trail is a control on paper only.
- Leaving spreadsheets in critical calculations. A workbook whose formulas can be edited and whose version is unclear is a consistency and originality risk.
- Changing every system at once. Without a criticality order, resources are spread thin and the team burns out.
These data integrity mistakes usually reflect overall digital maturity; for the bigger picture, see our digital transformation roadmap and digital transformation topic page.
Next step: a gap assessment of critical records
If an inspection is coming up, or an export customer has started asking about data integrity in supplier audits, the quickest win is a data integrity gap assessment of your critical records. Let's work out together which records fall short on which ALCOA+ principle and pick the right line and process for a pilot. Get in touch and in the first conversation we will listen to how your records flow today and agree where to start.