Part of the fundamentals and standardization pillar of the TAAG Learning Center. It complements the guide on AOAC-PTM certification and the one on traditional microbiological culture.
Table of contents
Two different questions about the same method
Before using a microbiological method in routine work, a laboratory has to answer two questions. The first is whether the method works: whether it detects or quantifies the target microorganism with acceptable performance against a recognized reference. The second is whether it works in its hands, with its staff, its equipment and the foods it analyzes every day. The first question is answered by validation; the second, by verification.
The distinction is not a technicality. ISO/IEC 17025:2017, which governs the competence of testing laboratories, requires the laboratory to verify that it can properly apply a method before introducing it, demonstrating that it achieves the required performance and keeping the records. A validated method that nobody verified in the laboratory using it leaves a gap that often shows up in audits. Accreditation requirements are covered in the article on ISO/IEC 17025 in food laboratories.
Validation: establishing the method's performance
Validation consists of determining a method's performance characteristics and providing objective evidence that it meets the requirements for an intended use. In food microbiology this almost always means comparing an alternative method (for example, a PCR kit or a chromogenic medium) with a reference method, which is usually a standardized traditional culture protocol such as ISO 6579-1 for Salmonella.
The ISO 16140 series
Technical committee ISO/TC 34/SC 9 replaced the old ISO 16140:2003 with a series of standards, each designed for a different scenario:
- ISO 16140-1:2016: common vocabulary for validation.
- ISO 16140-2:2016: protocol for validating alternative methods, mostly commercial, against a reference method. It has an amendment published in 2024 that revised some of the calculations.
- ISO 16140-3:2021: protocol for verifying, in a single laboratory, reference methods and alternative methods that have already been validated.
- ISO 16140-4: single-laboratory validation. Its results are valid only for the laboratory that ran the study.
- ISO 16140-5: factorial interlaboratory validation for non-proprietary methods, with fewer participating laboratories than in part 2.
- ISO 16140-6:2019: validation of alternative confirmation and typing methods, which start from a suspect colony rather than from a food sample.
The series is completed by ISO 17468, which sets the technical requirements for establishing or revising a standardized reference method.
What a validation includes under ISO 16140-2
The protocol has two phases. The first is a method comparison study carried out in an organizing laboratory: contaminated samples are analyzed with both methods to measure sensitivity, the relative level of detection (RLOD) is determined, and inclusivity and exclusivity are checked with 50 strains of the target microorganism and 30 non-target strains. In quantitative methods, the comparison study evaluates relative trueness and the accuracy profile. The second phase is an interlaboratory study, in which several laboratories analyze the same samples to estimate the method's reproducibility.
The scope of the validation depends on the food categories studied. To declare a method valid for a "broad range of foods", at least five categories must be covered. That scope statement is the first thing worth checking on a certificate.
Who certifies
In Europe, independent bodies such as MicroVal, NF Validation (AFNOR) and NordVal (NMKL) issue certificates based on ISO 16140-2 and publish the reports on their websites. In the United States the most widespread scheme is that of AOAC International, which we explain in the guide on AOAC-PTM certification.
What European regulation says
Regulation (EC) 2073/2005 establishes in its Article 5 that the analytical methods in Annex I are the reference ones, and that using alternative methods is acceptable when they are validated against that reference and, if proprietary, certified by a third party according to the EN ISO 16140 protocol or other similar internationally accepted protocols. Regulation (EU) 2019/229 added key definitions: what is meant by a "broad range of foods" and what a certification body independent of the method's manufacturer is.
Verification: demonstrating that the method works in your laboratory
ISO 16140-3 defines verification as the demonstration that a validated method performs in the user's hands according to the specifications of its validation. It divides the process into two stages.
Stage 1: implementation verification
The laboratory demonstrates that it knows how to run the method using a single food, chosen from those analyzed in the validation study. If the results resemble those of the validation, the underlying issue (technique) is resolved.
Stage 2: verification of the laboratory's foods
The laboratory then demonstrates that the method works with the foods it claims to analyze. To do so it picks a "challenging" food from each category included in its scope: products with low water activity, with natural antimicrobials such as spices, with lecithin, with a lot of fat. The training guide from the ISO committee itself shows examples such as ice cream, powdered egg, seasoned meat and cocoa powder.
What is measured and with what acceptance criterion
Qualitative methods: eLOD50
In detection methods, the 50% level of detection (eLOD50) is estimated: the contamination at which the method gives a positive in half of the portions analyzed. It is called "estimated" because the number of samples is much smaller than in a full validation. The acceptance criterion is that the eLOD50 obtained must not exceed four times the LOD50 of the validation. In the ISO training guide's example for Salmonella according to ISO 6579-1, the reference LOD50 was 5.7 CFU, the limit stood at 22.8 CFU and the laboratory obtained 3.0 CFU, so it complied. The LOD50 concept is developed in the article on limit of detection and measurement uncertainty.
Quantitative methods: reproducibility and bias
In counts, implementation verification measures the intralaboratory reproducibility standard deviation (SIR), with a design aligned with ISO 19036. The criterion is that it must not exceed twice the lowest mean reproducibility value obtained in the validation. Food verification measures the estimated bias (eBias): portions are contaminated at three levels, in duplicate, and the difference from the expected level must not exceed 0.5 log10 at any level.
Confirmation and typing methods
Here only implementation verification is required: target and non-target strains are tested (five of each type in the guide's example) and 100% agreement with the expected result is required.
The 2028 deadline for non-validated reference methods
Many ISO reference standards were published without performance data. For them, ISO 16140-3 offers a transitional protocol in its Annex F. According to the ISO committee's transition document, that annex can be used until 31 December 2027; from 1 January 2028, only validated reference methods may be verified under ISO 16140-3. ISO standards are voluntary, but accreditation bodies often take them as a criterion.
The same document clarifies when re-verification is needed. A method that is already accredited does not need a new verification unless it changes. A new method or a new food category within the scope does need it. If a validated method undergoes a major technical change with a relevant impact, it does too.
If the verification does not meet the criteria
The standard calls for a root cause analysis before repeating. Common causes are inoculum preparation and level, sample homogenization, equipment calibration, lots of media or reagents and the analyst's application of the protocol. If the problem persists, the next step is to consult the method provider (if proprietary), the standardization body (if a reference method) or the certification body.
Common mistakes
- Assuming that a validation certificate replaces verification in the laboratory.
- Applying the method to foods of a category that is not in the scope of its validation.
- Verifying only with an "easy" food and not with one representative of the laboratory's real difficulties.
- Not keeping verification records, which are required under ISO/IEC 17025.
- Changing a step of the protocol, such as the enrichment time, and continuing to treat the method as validated.
Conclusion
Validating and verifying answer different questions: whether a method works and whether it works in your laboratory. The ISO 16140 series organizes both processes, from the validation of alternative methods under ISO 16140-2 to verification in a single laboratory under ISO 16140-3, with concrete acceptance criteria such as eLOD50 or eBias. Checking the certificate's scope, verifying with representative foods and keeping records prevents a well-validated method from becoming a weak point of the laboratory.
About TAAG
Discover how TAAG documents the validated scope of its molecular detection kits to make them easier to verify in each customer's laboratory.
Frequently asked questions
What is the difference between validating and verifying a microbiological method?
Validation establishes the method's performance and demonstrates that it is fit for an intended use, normally by comparing it with a reference method. Verification demonstrates that a specific laboratory obtains that performance when applying it with its own equipment, staff and foods.
Does a method with a MicroVal, NF Validation or AOAC certificate need to be verified?
Yes. The certificate backs the method, but ISO/IEC 17025 requires each laboratory to verify that it can apply it properly before using it, and ISO 16140-3 describes how to do so.
What is the eLOD50?
It is the 50% level of detection estimated during the verification of a qualitative method: the contamination at which the method detects the microorganism in half of the portions analyzed. ISO 16140-3 accepts the result if it does not exceed four times the LOD50 obtained in the validation.
What happens with ISO reference methods that have no validation data?
Until 31 December 2027 they can be verified with the transitional protocol in Annex F of ISO 16140-3. From 1 January 2028, only validated reference methods may be verified under that standard.