Laboratory Method Verification

This record verifies one method, on one matrix, for use by named analysts. Using it to log routine results, or to record calibration of the instrument itself, produces a register nobody can query for either purpose.

ISO 17025 cl.7.2 separates the selection and verification of a method from its validation, and requires evidence the method achieves the performance the laboratory intends to rely on before it is used.

ISO 17025 is applied through national accreditation bodies and, in food testing, through regulation that requires accredited or officially recognised laboratories for certain results. What counts as adequate verification evidence differs by who is asking.

Filling in every field is mechanical. The judgement is in what counts as sufficient evidence before the method is approved for routine use.

These are the gaps an assessor, or an internal audit, finds most often in method verification records.

A protein method verified against lean beef mince passed every performance check: accuracy, precision, LOD, all clean. Applied to a cured, high-fat sausage product without a fresh matrix check, recovery drifted low enough to understate protein content on three consecutive batches before anyone noticed.

The cause, once traced, was fat interference the original verification never tested for because the original matrix had none. Reverifying against the actual product matrix, rather than assuming a beef result generalised to any meat product, was the fix, and matrix change is now written into the record as its own trigger.

The form only proves the method worked once. Keeping every authorised analyst, every reverification date and every marginal result actually followed up is the ongoing work.

KnowQuality

Holds the method verification register against every analyst authorisation and reverification date, and flags a method still marked routine past its due date.

KnowMaintain

Tracks the calibration and service history of equipment named in the verification, so an equipment change is caught before it silently invalidates a method.

KnowComply

Maps verified methods against accreditation scope and customer specifications that depend on them, surfacing where a method used routinely was never actually verified.

Ella

Coordinates the crew across method verification, calibration and proficiency testing, and holds every write for your approval before it touches a record.

Accuracy / recovery

How close a measured result comes to the true or spiked value, usually expressed as a recovery percentage. Low recovery understates what is actually present; high recovery overstates it.

Precision / repeatability

How closely repeat measurements of the same sample agree with each other under the same conditions. Poor repeatability means a single result cannot be trusted without replicates.

Limit of detection (LOD)

The lowest amount of a substance the method can reliably distinguish from background noise. Below the LOD, the method cannot say whether something is present at all.

Limit of quantification (LOQ)

The lowest amount the method can measure with acceptable precision, distinct from merely detecting it. A result between the LOD and LOQ is detected but not reliably quantified.

Matrix interference

The effect that fat, salt, protein or other components of the actual sample have on a method's performance, as distinct from its performance on a clean reference solution.