All resourcesPractical guide

Audit completion reports: fields, exceptions and history

An audit report is useful when it can reconstruct the obligation, completion and exceptions without sending a leader back through messages and files. An overall completion percentage gives a quick view, but answering detailed questions still requires identifying people with overdue assignments and the relevant material versions. Make those visible so the next conversation can address the gap.

Author: Piotr Adamski · Updated:

A report needs to explain its scope

A completion report compares the people covered by a requirement with the status of assigned training on a specified date. Its starting point is an explicit population, rather than a percentage alone. Readers should be able to identify who was included, which version was checked and which gaps still need action.

A completion report should allow a reader to reconstruct who had a requirement, what they needed to complete and which exceptions remained at the selected time. A single percentage cannot do that. It can summarise the position, but the detail must explain the population, source versions and decisions behind the figure.

The ISO and IAF Auditing Practices Group's competence guidance distinguishes attending training from demonstrating competence [1]. The proposed report structure below is an operational design, not a universal format imposed by those publications. Confirm exact criteria against the applicable standard, contract or legal requirement.

Treat the export as an evidence view

Treat the report as a view of controlled records. The export should not create new training history or change the meaning of an assessment. It should present evidence that already exists and make any missing evidence visible. If the source records cannot explain a status, a more polished report will not fix the underlying gap.

Start by stating the purpose. A supervisor's action list should make the next conversation or practical check clear; a management review summary and an auditor's evidence package need different detail. Combining every available personal field into one export usually increases exposure and makes the actual question harder to answer.

Build the denominator from requirements

Count the people who were required to fulfil the selected requirement, rather than only the people with assignments. This prevents a common reporting error: unassigned workers disappear, leaving the report apparently complete. The population should reflect the applicable duties and scope at the relevant time. Specify the date logic. A report can cover requirements active on a particular date, assignments created during a period or completions achieved during a period. Those are different questions. An employee who completed yesterday may still appear in a snapshot of last week's unresolved requirements. Make the chosen meaning explicit.

An illustrative example shows the arithmetic. A procedure applies to fifty workers. Forty have assignments and all forty complete. Assignment completion is one hundred percent, but requirement coverage is eighty percent. Both figures are calculable; only the second exposes the ten missing assignments. Label metrics so a reader cannot mistake one for the other. Reconcile the population with a current or historical roster appropriate to the report. Include substitutions, agency staff and returning employees where the requirement applies. Record exclusions with a reason, rather than quietly removing people whose incompletion reduces the percentage. A transferred employee and an overlooked operator require different interpretations.

Identify the evidence represented by each row

At minimum, a row should identify the person, requirement, source release, assignment, deadline, completion and result where relevant. Add the training material and question-set version when those differ from the source procedure. These fields explain what was actually checked. Keep evidence types distinct. Acknowledgement, knowledge assessment, practical observation and authorisation are separate records. The report can display their combined status if the rule is clear, but a passed quiz must not appear as an observed skill merely because the final cell is green. State the acceptance criterion for the combined decision.

For repeated attempts, show either the sequence or a clear link to it. The latest result can support a current view, while the earlier unsuccessful attempt explains follow-up. A report that replaces history loses the evidence of how the organisation detected and addressed a gap. Avoid unexplained blank fields. No score can mean acknowledgement was required, assessment was not attempted, a practical decision was recorded elsewhere or data is missing. Use a stated status or reason. Readers should not have to infer which interpretation applies, especially when the report informs permission to perform a task.

Make unresolved items actionable

Separate missing assignments, incomplete activities, unsuccessful assessments, expiring evidence and expired authorisations. They need different responses. A reminder may help an incomplete course; it will not resolve a missing practical assessment or an unclear scope decision. Give each item an owner and next action so leaders can arrange the needed support without a fresh round of messages to discover who is handling it. The report should show who arranges access, who assesses the person, who approves an exception and who decides whether work can proceed. Where a deadline passes, the relevant operational consequence must come from the organisation's rules and risk assessment.

Record exceptions with their boundaries. An illustrative temporary arrangement could allow a worker to perform a task under named supervision until an assessment is completed. The record needs its scope, approval, end condition and review date. A free-text note saying "agreed with manager" gives little assurance once that manager is unavailable. Review unresolved items against planned work. A small number of gaps can matter more than a high overall completion rate if those people are about to perform the changed activity. Provide a view by task, location or shift where it helps the recipient make the next decision. The report should expose concentrated risk rather than average it away.

Preserve the context of an export

Include the generation time, reporting period and applied filters. Identify the source system or register and the person or role generating the report where appropriate. These details explain why a later report may differ without implying that one is wrong. Use a stable report identifier where useful. A checksum can help detect whether a particular file changed, but it does not prove that its source data was true, that the participant identity was correct or that training was effective. EDPB guidance discusses measures supporting confidentiality, integrity and availability; it does not turn a hash into a guarantee about every event inside a file [3].

Decide whether the export can be edited. If someone adds follow-up notes, label the resulting file as a working copy and preserve the original snapshot where necessary. Otherwise a reviewer may mistake manually modified statuses for a direct system output. The workflow should explain how authorised corrections return to the underlying records. Retain accessible references to the source evidence. A report containing broken links or identifiers with no retrieval method cannot support reconstruction. Test one sample before sending the package, including whether the intended recipient can view the relevant supporting record without receiving unnecessary access to the entire employee register.

Limit personal data to the report's purpose

Employee names, identifiers and individual results can be personal data. EDPB guidance emphasises defined purposes, minimisation, accuracy and storage limitation [2]. A paper register and a spreadsheet are not exempt merely because they are simple tools. Design the report around the recipient's legitimate need. A shift supervisor may need the person's identity, task status and required next action. An auditor sample may need the source version and assessment trace. A management review can often use aggregate trends with controlled access to detail. Do not add a home address, national identifier or medical information to a training export simply because those fields exist elsewhere.

Restrict distribution and review access. Shared folders, mail attachments and local downloads all create copies. Define who may receive the report, how they access it and what happens when their role changes. EDPB security guidance recommends appropriate access and protection measures [3]; the exact implementation should match the organisation's environment. Set retention according to the purpose and applicable obligations. There is no universal duration for every training report across all sectors and countries. The underlying evidence may need to remain available for one period, while an extra working export can become unnecessary earlier. Keep that distinction in the retention decision.

Test calculations and historical consistency

Recalculate a small sample manually. Verify the denominator, status rules, date boundaries and treatment of exclusions. Use cases that exercise the logic: a person with no assignment, an unsuccessful attempt followed by success, a transfer, an expiry and a procedure revision. A formula working for ordinary completions can still misclassify exceptions. Compare the report with the source records. Select one fulfilled requirement and one unresolved item. Confirm that versions, dates and identities match. Where a practical assessment is held separately, ensure the report does not claim it exists merely because online training is complete.

Check how the system represents history. If an employee changes name or department, a historical report may use a current display value while the event retains its original scope. Decide which behaviour is appropriate and document it. The essential requirement is that the record remains interpretable and does not imply a different event. Ask a recipient to explain one row using the report alone. If they cannot identify what the status means or what to do next, revise the labels and definitions before asking a supervisor to use the report in a discussion with an employee. User comprehension is part of verification. A report can be technically accurate while inviting an operational misunderstanding.

Use a focused evidence package

Before an audit or review, agree the question being answered and assemble the relevant scope. Provide the report, its definitions and a small set of supporting records. Keep additional evidence available rather than filling the package with unrelated certificates. Volume is not a substitute for traceability.

CrewShift can support assignments, completion results, history and report exports. Create an account and evaluate the output using the same sample cases you use to verify your current register. Ask the team leader to identify who needs help and the next action, and note any manual searching still required. Confirm whether you can identify the source release, required population and exceptions in the actual workflow. Do not claim capabilities such as cryptographic immutability or guaranteed audit acceptance without specific evidence.

Improve the reporting process

After the review, record gaps in the reporting process as well as gaps in training. Missing assignment rules, inconsistent source identities and uncontrolled copies can each need a different correction. Repeatedly repairing the export by hand leaves the same issue waiting for the next review.

A report is ready when another person can understand its scope, follow a selected result to evidence and identify the owner of remaining work. Completion percentage remains useful as a summary. The supporting trace makes it a figure that can be checked and gives the leader a basis for spending review time on unresolved learning needs with the team.

Questions worth asking

Is an overall percentage enough?

It supports management review, but an audit usually needs drill-down to people, versions, results and exceptions.

How long should reports be retained?

Base retention on legal, contractual and internal requirements. Do not keep a report longer without a defined purpose.

Why count people without assignments?

They may still have an applicable requirement. Excluding them can produce a high assignment completion figure while hiding missing coverage.

Does an export checksum prove training happened?

No. It can help detect a change to the file, but it does not establish participant identity, truth of source events or effectiveness of training.

Sources and scope

  1. ISO / IAF Auditing Practices Group — guidance on auditing competence (2015 edition context)
  2. European Data Protection Board — data protection basics
  3. European Data Protection Board — secure personal data
  4. ISO 10015:2019 — competence management and people development

This informational material does not replace licensed standards or individual legal advice.