Saturday, September 26, 2026

Computer System Validation & Data Integrity: Proving Your Records Deserve Trust

Computer System Validation & Data Integrity: Proving Your Records Deserve Trust
Quality & Compliance / Pharmaceutical Manufacturing

Computer System Validation & Data Integrity: Proving Your Records Deserve Trust

Every batch record, chromatogram and validation report in this series now lives on a computer system. If that system isn't validated and the data can't be trusted, none of the paper trail behind it means anything.

⏱ 11 min read 📋 21 CFR Part 11 / Annex 11 💾 CSV & Data Integrity

01Why this sits underneath everything else

Process validation, cleaning validation and method validation all produce one thing in common: data. If the system that captured, stored or calculated that data isn't itself validated — and if the data can't be shown to be complete, consistent and accurate — every other validation exercise in this series rests on an unproven foundation.

Regulators have been explicit that data integrity failures aren't a new category of risk so much as a long-standing CGMP expectation finally getting dedicated attention, driven by an increase in inspection findings involving disabled audit trails, unreported failing results, and records signed by people who didn't actually perform the review.1,5

Computer system validation (CSV) is the discipline that proves a system does what it's supposed to do, reliably, before it's trusted with GMP data. Data integrity is the ongoing assurance that once data exists, nothing — deliberately or accidentally — has quietly changed its meaning.

GAMP
Recommended reading

GAMP 5: A Risk-Based Approach to Compliant GxP Computerized Systems

The industry-standard framework this article's risk categories are drawn from — essential if you're building or auditing a CSV program rather than just reading about one.

Find it on Amazon →

02The regulatory foundations

FrameworkIssuing bodyCore contribution
21 CFR Part 11 — Electronic Records; Electronic SignaturesU.S. FDARequirements for electronic records and signatures to carry the same weight as paper and handwritten equivalents
Data Integrity and Compliance With Drug CGMP: Q&AU.S. FDA (CDER/CBER/CVM, 2018)Clarifies how existing CGMP expectations apply to electronic data, audit trails and system access1,5,6
EudraLex Vol. 4, Annex 11 — Computerised SystemsEuropean Commission / EMAEU GMP requirements for validation, access control and audit trails on computerised systems
GXP Data Integrity Guidance and DefinitionsUK MHRA (Revision 1, March 2018)Defines data governance, data lifecycle, and the risk-based control expectations behind ALCOA+2,3,4

03The CSV lifecycle

Like process and cleaning validation, computer system validation follows a lifecycle model rather than a one-time sign-off. Click each stage to expand it.

Define user requirements, assess the system's GAMP risk category, and write a validation plan proportionate to that risk. A high-risk custom application warrants far more rigor than a low-risk, unmodified commercial tool.

  • User Requirements Specification (URS) agreed before build or configuration
  • Supplier assessment for any vendor-provided system

Install and configure the system, then execute installation, operational and performance qualification testing — the same IQ/OQ/PQ logic used for equipment, applied to software.

  • Test scripts traced back to each requirement
  • Access controls, audit trail function and data backup verified as part of testing, not assumed

Once live, the system is maintained under change control, with periodic review confirming it remains in its validated state — and audit trails are actually being reviewed, not just switched on.

  • Scheduled periodic review, risk-based on system criticality
  • Change control assessment for patches, upgrades and configuration changes
CSV
Recommended reading

Computer Systems Validation: Quality Assurance, Risk Management, and Regulatory Compliance for Pharmaceutical and Healthcare Companies — Guy Wingate

A detailed, applied companion covering how the lifecycle above translates into an actual URS, test protocol and periodic review program.

Find it on Amazon →

04Risk-based categories (GAMP 5)

Not every computer system needs the same level of validation rigor. GAMP 5 groups software into categories that scale effort to risk — switch tabs to compare them.

Category 1 — Infrastructure software. Operating systems, database engines, network components. Validated as part of the applications that run on them, generally through IT infrastructure qualification rather than a standalone GxP validation exercise.

Category 3 — Non-configured (or default-configured) products. Off-the-shelf software used as-is. Requires verification that it works as intended for its actual use, leaning on supplier documentation and risk-based testing.

Category 4 — Configured products. Commercial software configured to a specific business process (a LIMS or MES, for example). Requires validation of both the underlying product and the configuration applied to it — this is where most CGMP systems live.

Category 5 — Custom applications. Bespoke code written for a specific purpose. Warrants the highest rigor: full software development lifecycle documentation, source code review where relevant, and the most extensive testing.

05ALCOA+ in practice

Data integrity expectations are commonly summarized with the ALCOA(+) acronym — a shorthand for the attributes regulators expect of GxP data.3

A

Attributable

Traceable to the person or system that generated it.

L

Legible

Readable and permanent, not obscured or erased.

C

Contemporaneous

Recorded at the time the activity happened.

O

Original

The first record, or a verified true copy.

A

Accurate

Correct, complete, and reflecting what actually occurred.

+

Complete

Includes all data, including repeats and reprocessing.

+

Consistent

Timestamped and sequenced in the expected order.

+

Enduring

Retained for as long as required, not just convenient.

+

Available

Accessible for review or inspection throughout its retention period.

DI
Recommended reading

Data Integrity and Data Governance: Practical Implementation in Regulated Environments — R.D. McDowall

Written by a widely respected voice on lab data integrity; particularly strong on audit trail review practices and electronic record governance beyond the acronym.

Find it on Amazon →

06Data integrity self-check

Readiness checklist

0 of 7 complete

07Where programs fail inspection

  • Audit trails technically on, functionally unreviewed. Enabling the feature isn't the requirement — someone needs to actually look at it, on a defined schedule.6
  • Shared logins. If multiple people can sign in as "QC1," no individual action is genuinely attributable, which undercuts the entire audit trail.
  • Treating validation as a go-live event. A system validated once at installation and never periodically reviewed drifts out of its validated state the same way an unmonitored manufacturing process does.
  • Printouts as the "original" record. A paper printout of electronic data is a copy, not the original — and can strip away metadata (timestamps, audit trail entries) an inspector will expect to see.
Worth remembering: data integrity failures don't usually start as fraud. They start as a workaround for a system that's inconvenient to use correctly — which is exactly why validation, access design and audit trail review need to make the compliant path the easy path.

08References

  1. U.S. Food and Drug Administration. Data Integrity and Compliance With Drug CGMP: Questions and Answers — Guidance for Industry. December 2018. fda.gov
  2. UK MHRA. GXP Data Integrity Guidance and Definitions. Revision 1, March 2018. assets.publishing.service.gov.uk
  3. Lachman Consultants. "MHRA March 2018 GXP Data Integrity Guidance and Definitions." lachmanconsultants.com
  4. gmp-compliance.org. "MHRA GxP Data Integrity Definitions and Guidance for Industry: New Draft Version for Consultation." gmp-compliance.org
  5. Federal Register, Vol. 83, No. 239 (December 13, 2018). Notice of guidance availability. govinfo.gov
  6. Alston & Bird. "New Guidance Refines FDA's Thinking on Data Integrity." alston.com

Disclosure: This article contains Amazon affiliate links. As an Amazon Associate, this site may earn from qualifying purchases at no extra cost to you. Recommendations reflect genuine, independent picks for readers building or auditing a CSV and data integrity program — they are not a substitute for your organization's own quality and regulatory guidance.

This content is for general professional education and does not constitute regulatory or legal advice. Always consult current guidance from your applicable regulatory authority.

Page source (HTML)

No comments: