Data Integrity per ALCOA+: From Principle to Practice

Data integrity is the foundation of every GxP decision. If the data is not trustworthy, neither is the product. The acronym ALCOA+ sums up what regulators expect, and it now appears in guidance from the FDA, the MHRA, the WHO and PIC/S. The requirement itself is not in dispute. Practice is what decides it: are the systems built so that the principles hold without relying on individual discipline?

What ALCOA+ actually requires

• Attributable: every record can be assigned to a person and a point in time.
• Legible: readable and durable, across the entire retention period.
• Contemporaneous: recorded at the time of the activity, not afterwards.
• Original: the original record or a verified, documented copy.
• Accurate: correct, complete and free of unjustified corrections.

The plus is the part that usually goes missing

Complete, consistent, enduring and available. This is precisely where the uncomfortable findings arise. Complete means that discarded runs and repeated measurements are documented too, not only the result that happened to fit. Consistent means that the sequence of events is evident from the records. Enduring and available mean that the data is still readable at the end of the retention period, even when the system that produced it was replaced long ago.

An audit trail is only as good as its review

Annex 11 requires an audit trail for GxP-relevant data. That is the easier half, because most systems can do it. The harder half is the review: who checks the entries, at what interval, against which criteria, and what happens to an entry that looks wrong? Without a documented answer, the audit trail is a file nobody reads.

Data integrity is not created at the end through inspection, but at the beginning through design.

Hybrid workflows are the real risk

Purely electronic systems are usually manageable, and so are purely paper-based processes. It is the mixture that becomes critical: a measurement is produced electronically, transcribed onto a form and signed there. Two records now exist, and it has to be defined which of them is the original record and how the transcription was verified. Leave that open and you create exactly the gap the audit will find.

A walkthrough shows more than a checklist

The pragmatic starting point is a data integrity walkthrough along the data flow, from capture through processing, release and transfer to archiving. At every transition the questions are the same: who is allowed to change things here, what is logged, where does the original sit, and how would anyone get back to it in five years? That reveals the gaps before the auditor does.