GMP & medical devices

Data integrity training: who needs ALCOA+ and 21 CFR Part 11 competence

In brief
  • FDA guidance expects data to be attributable, legible, contemporaneous, original or a true copy, and accurate, and expects staff to be trained to prevent and detect data integrity issues.
  • 21 CFR Part 11 makes competence a control: people who develop, maintain or use electronic record systems need the education, training and experience for their tasks.
  • Operators, QC analysts, reviewers, IT system owners and QA each need different depth; reviewers must be able to review audit trails, not only results.
  • Investigators examine the data and systems first, then test whether people understand them, so pair training records with exam results and real effectiveness checks.

Data integrity is often treated as an IT problem or a laboratory problem. It is neither, or rather it is both and much more. Every batch record an operator completes, every chromatogram an analyst processes, every review a supervisor signs and every user account a system owner creates either protects the reliability of your records or puts it at risk. When the US Food and Drug Administration examines a site, it is testing whether the people behind those records understand what trustworthy data means and behave accordingly.

This guide is for QA heads, training managers, laboratory managers and IT system owners in pharmaceutical and biologics manufacturing who need to decide who should be trained in ALCOA+ principles and 21 CFR Part 11, how deep that training should go for each role, and what evidence will hold up during an FDA inspection. It is built on the FDA's own published expectations, which we summarise in plain language below.

What the FDA expects, in brief

The FDA's guidance Data Integrity and Compliance With Drug CGMP: Questions and Answers, issued in December 2018, sets out the agency's current thinking. Like all FDA guidance it is not binding in itself, but it explains how the agency reads the CGMP regulations. Four points from it shape any training plan.

  • Data integrity means complete, consistent and accurate data. The guidance says such data should be attributable, legible, contemporaneously recorded, original or a true copy, and accurate: the principles known as ALCOA.
  • Audit trails are part of the record. The people who review a CGMP record should review its audit trail too, in the same way a reviewer of a paper record looks at crossings-out. Where the regulations set a review frequency for the data, the audit trail follows it; otherwise the frequency should be set by risk.
  • Access must be controlled and attributable. Only authorised people should be able to change records, system administrator rights should sit with people independent of those responsible for the content of records, and shared login accounts that make it impossible to identify who did what are a concern.
  • Training is expected. The guidance says personnel should be trained to prevent and detect data integrity issues, as part of routine CGMP training.

Alongside the guidance sits the regulation itself. 21 CFR Part 11 sets the conditions under which the FDA treats electronic records and electronic signatures as trustworthy, reliable and generally equivalent to paper records and handwritten signatures. It applies to electronic records created, modified, maintained, archived, retrieved or transmitted under any FDA record-keeping requirement. Among its controls for closed systems are validation, secure computer-generated time-stamped audit trails, authority checks, written policies holding individuals accountable for actions taken under their electronic signatures, and a requirement that people who develop, maintain or use these systems have the education, training and experience to perform their assigned tasks. In other words, Part 11 itself makes competence a control.

Finally, the CGMP regulations for finished pharmaceuticals in 21 CFR Parts 210 and 211 require everyone engaged in manufacturing, processing, packing or holding a drug product to have the education, training and experience to do their job, with CGMP training conducted by qualified individuals on a continuing basis and often enough that employees remain familiar with the requirements that apply to them.

ALCOA+ in practice: who makes each principle real

ALCOA is the FDA's own formulation. The plus extends it with further attributes now in wide use across the industry: complete, consistent, enduring and available. The European Medicines Agency's guideline on computerised systems and electronic data in clinical trials, for example, sets out an extended list that also adds traceable. Whichever version your procedures use, each attribute depends on particular people more than others, and that is the starting point for training by role.

AttributeWhat it means in daily workWho most affects it
AttributableEvery entry and change can be traced to the person or system that made itOperators and analysts, who must never share logins; system owners, who set up unique accounts
LegibleRecords can be read and understood throughout their lifeEveryone who writes on paper; system owners, who preserve readability of electronic data
ContemporaneousRecorded at the time the work is done, not later from memoryOperators and analysts; supervisors, who must not allow back-filling
Original or true copyThe first capture of the data, or a verified copy that keeps its content and meaningAnalysts and system owners, who decide what is kept and how
AccurateCorrect, with changes documented and justifiedEveryone who records data; reviewers, who check it
CompleteAll data, including repeats, failed runs and metadata, is retainedAnalysts and reviewers, who must see the whole picture, not only the passing result
ConsistentData is recorded and processed the same way every time, in the expected sequenceQA and process owners, who write the procedures; system owners, who configure workflows
Enduring and availableData survives the full retention period and can be retrieved for review or inspectionIT and system owners, responsible for backup, archiving and retrieval

Five roles, five different training needs

A single data integrity module for the whole site will tick a box but leave real gaps. The roles below need different depth and different emphasis.

Production operators and technicians

Operators generate a large share of CGMP records, on paper batch records and increasingly on electronic systems on the shop floor. They must be able to:

  • record entries at the time of the activity, in the correct place, in their own name;
  • correct an entry properly, without obscuring the original, and give a reason;
  • protect their passwords and never use another person's account;
  • recognise and report pressure or shortcuts that threaten data integrity.

Good Documentation Practice (GDocP) Training builds exactly these habits, with US FDA GMP Awareness for Pharmaceuticals (21 CFR Parts 210 and 211) as the CGMP foundation for new starters.

QC analysts and laboratory staff

Laboratories deserve particular attention, because instrument systems hold dynamic electronic data, processing methods can be changed, and tests can be repeated. Analysts need everything operators need, plus a working knowledge of audit trails, metadata, the difference between original electronic data and a static printout, and the rules for handling invalid or repeated results. They should understand why unofficial trial runs and selective reporting are serious problems. Data Integrity (ALCOA+) and 21 CFR Part 11 is designed for this group.

Reviewers and approvers

Supervisors, second-person reviewers and laboratory managers carry a responsibility that is easy to underestimate. The FDA expects them to review the audit trail along with the data, not only the final report. They must be able to read an audit trail, judge which entries matter, spot patterns such as repeated aborted runs or changes to integration parameters, and know when to escalate. They also need to understand what their electronic signature means and the accountability that goes with it. Reviewers should take the data integrity and Part 11 course in full, and practise audit trail review on their own systems.

IT staff and system owners

The people who configure, administer and validate computerised systems decide whether the technical controls in Part 11 actually work. They need to understand validation for intended use, user access management, separation of administrator rights from people who generate data, audit trail configuration, time synchronisation, backup, archiving and the generation of accurate copies for inspection. Computer System Validation (GAMP 5) gives this group the risk-based validation method, and the data integrity and Part 11 course explains the regulatory reasoning behind each control.

Quality assurance

The quality unit owns the data governance system: procedures, periodic review of systems, investigations into data integrity failures, and the culture that makes people comfortable reporting problems. QA staff need the deepest understanding of all, from the FDA guidance and Part 11 to the wider CGMP framework. US FDA Current Good Manufacturing Practice (cGMP) Implementation for Pharmaceuticals (21 CFR Parts 210 and 211) covers that framework, and FDA Inspection Readiness Training prepares QA to host an inspection and present data integrity evidence with confidence.

What an FDA investigator is likely to examine

An investigator testing data integrity is not looking for a training certificate first. They are looking at the data and the systems, and then asking whether the people involved knew what they were doing. Based on the FDA's published expectations, prepare for questions such as these.

  1. Who can do what in this system? Expect to show user lists, role permissions and how administrator rights are separated from those of analysts and operators. Shared accounts that can modify data will draw attention.
  2. Who reviews the audit trail, and how often? Have the procedure, the risk basis for the frequency, and records showing that reviews actually happened.
  3. Are electronic records complete? The FDA states that all records required under CGMP are subject to inspection, including electronic records. Be ready to show original data, metadata and audit trails in the system itself, not only as printouts.
  4. How are blank forms controlled? The guidance expects blank forms, including electronic worksheets, to be controlled by the quality unit or another document control method.
  5. Were people trained, and does it show? An analyst asked how they would correct an entry or what they do with an aborted run reveals far more than a training matrix.
  6. Have electronic signatures been set up correctly? Part 11 requires each electronic signature to be unique to one individual, the identity of the person to be verified before the signature is assigned, and the people using electronic signatures to certify to the FDA that they are intended to be the legally binding equivalent of handwritten signatures.

Training evidence then supports the picture. The strongest records show the role's requirement, the course completed, and a final exam that tested understanding. A numbered certificate that can be verified by scanning its QR code lets QA answer a request for proof in seconds, and every Sterling Academy course provides one.

Three misconceptions to clear up

"Part 11 is only about software." The regulation explicitly requires that people who develop, maintain or use systems have the necessary education, training and experience, and that written policies hold individuals accountable for actions taken under their electronic signatures. A validated system used by untrained people is not enough.

"Our reviewers check the results, so we are covered." Reviewing a final report without its audit trail can miss deleted runs, reprocessed data or changed parameters. The FDA expects audit trail review to be part of record review.

"Data integrity training is a one-time event." The CGMP regulations expect training on a continuing basis. Systems are upgraded, procedures change and new people arrive. Retrain after significant system changes and include data integrity in routine CGMP refresher training.

A practical 90-day plan

Days 1 to 30: map the roles and systems. List your GxP computerised systems and paper record types, then the people who create, review, approve and administer records in each. Assign every person to one of the five roles above. This becomes your training needs analysis, and also a useful check on who holds administrator rights.

Days 31 to 60: train the people with the most influence first. Start with QA, reviewers and system owners, because they set and enforce the rules for everyone else. Assign the data integrity and Part 11 course, the validation course and the CGMP implementation course according to role.

Days 61 to 90: reach the whole workforce and test it. Roll out good documentation practice training to operators and analysts, then check effectiveness: a sample of records reviewed for correct corrections and contemporaneous entries, an audit trail review exercise for reviewers, and short interviews mirroring the questions an investigator would ask.

The Sterling Academy company dashboard supports each stage: assign courses by role, follow completion across departments and sites, and download training records ready for an inspection or an internal audit.

Bringing it together

Data integrity rests on people more than on technology. The FDA's expectations, and Part 11 itself, assume that operators record honestly and at the time, analysts keep complete data, reviewers look at audit trails, system owners configure controls that work, and QA builds a culture where problems are reported rather than hidden. Training each role for its own part, and keeping evidence that the training worked, is how a site shows an investigator that its records can be trusted. Build the competence. Prove it with confidence.

Training a team? Assign each person only the courses their role needs, track progress on one dashboard and download the evidence before the audit. Ask us for a bundle price for two or more courses.

Frequently asked questions

What does ALCOA+ stand for?

ALCOA stands for attributable, legible, contemporaneous, original or a true copy, and accurate, as set out in FDA guidance. The plus adds attributes widely used in the industry, such as complete, consistent, enduring and available.

Does 21 CFR Part 11 require training?

Yes. Among its controls for closed systems, Part 11 requires that people who develop, maintain or use electronic record and signature systems have the education, training and experience to perform their assigned tasks.

Who should review audit trails?

FDA guidance says the people responsible for reviewing a CGMP record should also review its audit trail, and the quality unit reviews production and control records. Reviewers therefore need specific training in reading and judging audit trails.

Are shared login accounts acceptable?

Shared accounts that can create or change data are a concern to the FDA because actions cannot be traced to an individual. The guidance treats shared read-only accounts that cannot change data as acceptable for viewing.

How often should data integrity training be repeated?

The CGMP regulations expect training on a continuing basis and often enough for people to stay familiar with the requirements. Most sites include data integrity in routine refresher training and retrain after significant system or procedure changes.

Which course should a QC reviewer take first?

Data Integrity (ALCOA+) and 21 CFR Part 11 covers audit trails, metadata, electronic signatures and the FDA's expectations, which are the core of a reviewer's role. Good Documentation Practice training is a useful companion.

Make your next audit a two-minute answer

Current-edition courses built by experts who take organisations through certification, with a final exam, a verifiable certificate and a dashboard for companies.