When Should a Calibration Procedure Be Revised?

A revision is warranted when the technical basis of the work changes

A calibration procedure should be revised when it no longer describes the method actually being performed, or when it no longer provides adequate control of the measurement result. The trigger is not the age of the document. It is a change that could affect how the laboratory obtains, evaluates, reports or supports a result.

ISO/IEC 17025:2017 requires laboratories to use appropriate methods and procedures for their activities, keep them current, and control the data and information associated with laboratory activities. Its requirements for document control also require documents to be reviewed and approved before issue, identified with their revision status, and protected against unintended use of obsolete versions.

Not every change requires a full technical reassessment. Correcting a typographical error, updating a job title, or clarifying an instruction without changing its meaning may be handled as a minor editorial revision. The laboratory should nevertheless assess and record why the change is non-technical. A changed acceptance criterion, revised uncertainty calculation, replacement reference standard, or altered measurement sequence is substantive even if the procedural edit appears small.

Typical triggers for technical revision

A procedure should be reviewed when any of the following occurs:

• A method changes. This includes a new measurement technique, changed test point sequence, different stabilisation time, altered mounting arrangement, or a revised correction model.
• Reference standards or supporting equipment change. A replacement standard may have different uncertainty, resolution, range, loading behaviour or environmental sensitivity. Its suitability must be demonstrated rather than assumed from its calibration certificate.
• The uncertainty budget changes materially. New evidence may identify a previously omitted component, a revised sensitivity coefficient, better characterisation data, or a changed contribution from the reference standard. The procedure may need amended controls, calculation steps or reported uncertainty.
• Environmental controls are revised. A procedure should state conditions that materially affect the result, along with the required monitoring and response to excursions. This is particularly important where thermal equilibrium, humidity, vibration or pressure affects the measurand.
• A nonconformity or technical investigation identifies a weakness. Repeated transcription errors, inconsistent analyst practice, an out-of-tolerance reference standard, or an invalid calculation formula can show that the written method is insufficient or unclear.
• The customer or intended use changes. A laboratory may introduce a new measurement range, tighter acceptance limits, a customer-specified decision rule, or an additional reporting requirement. These may require a separate procedure rather than an overloaded revision to an existing one.
• External requirements change. Updates to a normative method, regulatory requirement, manufacturer instruction, or accreditation scope should trigger review. The laboratory must determine applicability; not every external publication automatically changes its controlled method.

Decide whether the change needs verification or validation

The level of technical evidence should match the significance of the change. Under ISO/IEC 17025 clause 7.2, a laboratory verifies that it can properly perform a method before introducing it. Where it modifies a method outside its intended scope, or develops a non-standard method, validation is required to the extent necessary to demonstrate fitness for intended use.

For example, changing a digital pressure controller for an equivalent unit may require documented verification of range, stability, connection configuration and uncertainty impact. Introducing a new pressure generation principle, extending the range from 70 MPa to 100 MPa, or changing the correction model for head-height effects is more likely to need a defined validation or method-extension study.

The procedure revision should not be approved merely because the equipment is technically capable. The laboratory needs evidence that the complete method, including personnel instructions, calculations, quality controls and uncertainty statement, produces results suitable for the stated application.

Approval must precede routine use

The technical author is not normally the only person who should approve a substantive revision. Approval should involve personnel with authority to judge the method's technical validity and its alignment with the laboratory's management system. Depending on the laboratory's structure, this may include the technical manager, discipline lead, quality manager, or authorised signatory.

Before approval, confirm that the revision has addressed the practical controls needed for implementation:

• applicable equipment, reference standards and software are available and suitable;
• affected staff have been trained and, where necessary, authorised;
• forms, worksheets, calculation templates and certificate wording are aligned;
• the uncertainty budget and decision rule remain appropriate;
• required intermediate checks or quality-control activities have been updated; and
• superseded documents can be removed from points of use.

Document control should make the revision identifiable, including its number, issue date, approver and change history. A change summary is valuable because it tells reviewers what was deliberately altered, rather than requiring them to compare documents line by line.

Set an effective date that can be operated in practice

An issue date and an effective date need not be the same. The issue date records when the revision was approved. The effective date identifies when routine work must follow it.

A future effective date is appropriate where training, software configuration, procurement or controlled distribution must occur first. During that period, the previous approved procedure remains in force. The laboratory should avoid ambiguous arrangements in which some analysts begin using the new method informally while others continue with the old one.

For urgent corrections affecting result validity, the effective date may be immediate. The laboratory should then stop affected work, withdraw the obsolete procedure, brief staff promptly and assess any work already in progress. Urgency does not remove the need for documented technical review and approval.

Preserve the evidence for work performed under earlier versions

Work completed under a previous approved procedure is not automatically invalid because a newer version is issued. Its validity depends on why the procedure changed and whether the earlier method remained technically suitable at the time of use.

Technical records should identify the procedure version used, directly or through an unambiguous controlled linkage. This is essential when reviewing historical certificates, investigating a complaint, or assessing the impact of a changed uncertainty model. A certificate issued after the effective date of a new procedure may still properly relate to work performed under the prior version, provided the record makes that clear and the result has been reviewed appropriately.

Where the revision corrects a substantive defect, perform a documented retrospective impact assessment. Define the affected date range, equipment types, measurement ranges and reported results. Consider whether recalculation, customer notification, certificate amendment, recall, or no further action is justified. The decision should be based on evidence, not on the inconvenience of the investigation.

A controlled calibration procedure is therefore a maintained technical basis for measurement, not a static instruction. Revising it at the right point, approving it with appropriate evidence, and retaining clear links to earlier work allows the laboratory to adopt improvements without losing confidence in the results it has already issued.

References

• ISO/IEC 17025:2017, *General requirements for the competence of testing and calibration laboratories*, particularly clauses 7.2, 7.5, 7.6, 7.8 and 8.3. International Organization for Standardization.
• JCGM 100:2008, *Evaluation of measurement data - Guide to the expression of uncertainty in measurement* (GUM). Joint Committee for Guides in Metrology.