ISO/IEC 17025 accreditation in food laboratories: what it assesses and how to read it

Certifying a method is not accrediting a laboratory

When a quality team receives a microbiological analysis report, it is actually trusting two things at once: that the method is suitable and that the laboratory that ran it is competent. The first is backed by validation schemes such as AOAC-PTM or the ISO 16140 series. The second is backed by accreditation under ISO/IEC 17025, the international standard of general requirements for the competence of testing and calibration laboratories.

It is also worth distinguishing it from ISO 9001 certification. Certification assesses whether an organization has a quality management system; ISO/IEC 17025 accreditation is granted by a national accreditation body and assesses, in addition to the management system, the technical competence to produce valid results in specific tests.

What the standard contains

The current edition is the third, published in November 2017. Its structure is organized into five blocks of requirements:

• General requirements (clause 4): impartiality and confidentiality.
• Structural requirements (clause 5): legal identity, responsibilities and organization.
• Resources (clause 6): personnel, facilities and environmental conditions, equipment, metrological traceability and externally provided products or services, such as culture media or strains.
• Processes (clause 7): review of requests, selection, verification and validation of methods, sampling, handling of test items, technical records, measurement uncertainty, ensuring the validity of results, reports, complaints, nonconforming work and data control.
• Management system (clause 8): with the option of meeting it directly or through an ISO 9001 system already in place.

The scope: the part almost nobody reads

A laboratory is not "accredited" in the abstract. It is accredited for a scope: a list of specific tests, each with its method and the type of sample or matrix. A laboratory may have accreditation for the detection of Salmonella by ISO 6579-1 in food and not have accreditation for the yeast and mold count, or for PCR detection in environmental samples. If the test you need is not in the scope, the result may be technically correct, but it is not covered by the accreditation.

That is why, when evaluating a laboratory, the first step is to ask for the certificate and the current scope annex, and check that the method, microorganism and matrix match what you are going to send.

How it applies in a microbiology laboratory

Verified and validated methods

The standard requires verifying that the laboratory can properly apply a method before introducing it, and validating non-standard or modified methods. In food microbiology, the internationally accepted way to do this is the ISO 16140 series, explained in the article on validation and verification of methods.

Discipline-specific requirements

ISO/IEC 17025 is a general standard, valid for any type of testing. For food chain microbiology, ISO 7218:2024 sets out the general requirements and guidance on facilities, personnel, equipment, sample preparation and culture techniques. ISO 11133 governs the preparation and performance control of culture media, and indicates which control strains to use to assess productivity, selectivity and specificity, identified by their World Data Centre for Microorganisms (WDCM) number.

Measurement uncertainty

Clause 7.6 requires evaluating measurement uncertainty. In microbiological counts it is estimated with ISO 19036:2019, as explained in the article on limit of detection and uncertainty. When the report includes a statement of conformity (meets or does not meet a limit), clause 7.8.6 further requires documenting the decision rule used and the risk it entails.

Ensuring the validity of results

Clause 7.7 asks for a procedure to monitor the validity of results. In practice it combines internal controls (positive and negative controls, reference strains, duplicate analyses, media control, trend monitoring) with external controls. Among the latter, participation in proficiency testing or interlaboratory comparisons is the requirement that carries the most weight: the laboratory analyzes samples sent by an external provider, without knowing the expected result, and its performance is compared with that of other laboratories. Providers of these tests can be accredited under ISO/IEC 17043, whose second edition was published in 2023.

In a microbiology laboratory, those internal controls are usually seen in very specific tasks: checking the sterility and productivity of each lot of medium before use, including in each run a positive control with a reference strain and a negative control, periodically repeating counts between analysts to compare results, and recording incubator and refrigerator temperatures. When a control fails, the laboratory must treat it as nonconforming work, assess which results may have been affected and decide whether analyses must be repeated or the customer notified. That traceability is what later makes it possible to defend a result in an audit or a complaint.

When accreditation is mandatory

European Union

Regulation (EU) 2017/625 on official controls requires, in its Article 37, that designated official laboratories operate in accordance with EN ISO/IEC 17025 and be accredited under that standard by a national accreditation body. The regulation provides for limited exceptions in some cases.

United States: the LAAF program

On 3 December 2021, the FDA published the final rule of the Laboratory Accreditation for Analyses of Foods (LAAF) program, under the FSMA law. When it applies, certain analyses must be carried out in laboratories accredited by an FDA-recognized body: analyses to be removed from an import alert or to admit detained imported foods, those required by specific rules on shell eggs, sprouts and bottled water, direct FDA testing orders and those for administrative appeals. Laboratories in the program must be accredited under ISO/IEC 17025:2017. The FDA applies the program in phases and publishes on its LAAF dashboard the analytes whose deadlines are already in force.

International recognition

Accreditation bodies that sign the ILAC Mutual Recognition Arrangement recognize each other's accreditations, under the principle of "accredited once, accepted everywhere". For an exporter, this means that a report issued by a laboratory accredited in its own country is more likely to be accepted in the destination market.

What to ask a laboratory before sending it samples

• Are the test, method and matrix I need within the current accredited scope?
• Which reference or alternative method does it use, and with what validation?
• Which proficiency tests does it take part in for that microorganism, and with what recent results?
• Can it report the measurement uncertainty of counts, and what decision rule does it apply if it declares conformity?
• What are its sample acceptance criteria at reception? Shipping conditions are detailed in the article on transport and storage of samples.

What accreditation does not guarantee

Accreditation backs the laboratory's competence to run the test. It does not guarantee that the sample is representative of the lot, or that it arrived in good condition, or that the chosen sampling plan is appropriate. Those decisions remain with the plant's quality team and are designed within the HACCP system. An impeccable result on a badly taken sample is still bad data; sampling design is explained in the article on attribute sampling plans.