ISO17025 General requirements for

the competence of testing and

calibration laboratories
Selected sections from ISO 17025
 What does it mean if a laboratory
is Accredited to ISO 17025?
• An assessment of the competence of the
laboratory (staff, equipment, policies,
procedures, methods etc) to perform certain
specified tests, using specified methods in
specified matrices.
 General comments about the
content of this presentation
• Some procedures refer to a method or a
group of methods e.g. Measurement
uncertainty, trend analysis of CRMs and QCs.
• Some procedures refer to a specific batch of
samples processed e.g. Use of CRM’s or QC
solutions
– LIMS system used to automate and store data and
metadata for samples and batches
 5.9 Assuring the Quality of Test Results
• 5.9 Assuring the Quality of Test & Calibration Results
• 5.9.1
• Note Std
• There shall be QC procedures to monitor validity of results, data recorded,
trends, statistics. These activities shall be planned & reviewed:
• a) regular use of certified reference materials and/or secondary
reference materials.
• b) participation in inter-laboratory comparison or proficiency-testing
programmes.
• c) replicate tests or calibrations using the same or different methods.
• d) re-testing or re-calibration of retained items.
• e) correlation of results for different characteristics of an item.

• 5.9.2 QC data analysis. Action plan for incorrect results. Preventing the
reporting of incorrect results.
5.4 Test & calibration methods, method
validation
• 5.4.3 Laboratory developed methods
• Introduction of method planned, assigned to qualified persons equipped with adequate resources.
• Plans up-dated as development proceeds, effective communication amongst all persons involved.
• 5.4.4 Non-standard methods
• Note in Std: Agreement with the customer, include clear specification of the customer’s requirements and
the purpose of the test and/or calibration. Validated appropriately before use.
• 5.4.5 Validation of methods:
• 5.4.5.1
• Confirmation by examination and provision of objective evidence that particular requirements for specific
intended uses are fulfilled. Eg nutrient precision provided with results
• 5.4.5.2
• Notes Std
• Validate non-standard methods; laboratory-developed methods; standard methods used outside intended
scope; amplifications/modifications of standard methods. As extensive as is necessary for application/field
of application. Record results, procedure, statement of fitness for intended use.
• 5.4.5.3
• Notes Std
• Range and accuracy as assessed for the intended use relevant to the customers needs. (e.g. The
uncertainty, detection limit, selectivity, linearity, repeatability, reproducibility, robustness )
 5.4 estimation of uncertainty
• 5.4.6 Estimation of Uncertainty of Measurement
• 5.4.6.1 Test & calibration labs
• Uncertainty procedure for all calibrations and types of calibrations.
• 5.4.6.2
• NB Notes in Std
• identification of all the components of uncertainty
• reasonable estimation is attempted.
• Reporting of results do not give wrong impression of uncertainty.
• Estimation based on knowledge of performance of method,
measurement scope, and validation data.
• 5.4.6.3
• Notes Std
• All uncertainty components of importance in the given situation
taken into account using appropriate methods of analysis.
 5.6 Measurement uncertainty
• 5.6.3 Reference standards and reference materials
• 5.6.3.1 Reference standards.
• Programme & procedure for calibration, traceable. Used for calibration
only unless not invalidated. Calibrated before and after any adjustment
• 5.6.3.2 Reference materials.
• Traceable to SI units/to certified reference materials. Internal checked
• 5.6.3.3 Intermediate cheeks.
• Checks to maintain confidence in calibration status of reference, primary,
transfer/working standards, reference materials according to defined
procedures and schedules
• 5.6.3.4
• Note Std
• Transport and storage. Procedures for safe handling, transport, storage
and use in order to prevent contamination or deterioration and to protect
their integrity
 5.7 Sampling
• 5.7 Sampling
• 5.7.1 Sampling plan & procedures available at the location where sampling is undertaken.
• Notes Std Based on appropriate statistical methods, address factors to control to ensure validity of
results. E.g. Temperature record for frozen samples
• 5.7.2 Customer required deviations, recorded in detail with sampling data, results, communicated
to all staff.
• 5.7.3 Records of data & operations, procedure, sampler, environmental conditions, location,
statistics

• 5.8 Handling of Test and Calibration Items

• 5.8.1 Transport, receipt, handling, protection, storage, retention, disposal, protect integrity/
interests
• 5.8.2 Identify, for life of item, not confuse, accommodate sub-division and transfer of items.
• 5.8.3 Departures recorded on receipt. Where doubt consult customer before proceeding and record
• 5.8.4
• Notes Std
• Avoid deterioration, loss/damage storage, handling, preparation. Follow instruction.
Environmental conditions maintained, monitored, recorded. Storage secure, protect condition and
integrity.
 5.10 Reporting the Results
• 5.10 Reporting the Results
• 5.10.1 General
• Notes Std Accurate, clear, unambiguous,
objective, in accordance with specific instructions
in methods.
• Information requested by client & for
interpretation of results Refer 5.10.2, and 5.10.3
or 5.10.4.
• Internal customers/written agreement, report
simplified. Information 5.10.2 to 5.10.4 readily
available
CSIR In-house additional requirements
• ISO5.9 Assuring the Quality of Results
• 5.9.1 Quality Control (QC) Procedures. The laboratory shall comply with the quality
control (QC) procedures required by applicable agencies when testing for specific
analytes. The laboratory shall have QC procedures (SOPs) specific to each test
technology addressing, as appropriate, the use of:
• 5.9.1.1 Reagent/method blank analyses;
• 5.9.1.2 Trip blanks and field blanks;
• 5.9.1.3 Replicate/duplicate analyses;
• 5.9.1.4 Spiked sample analysis;
• 5.9.1.5 Blind samples;
• 5.9.1.6 Surrogate standards;
• 5.9.1.7 Laboratory control samples (LCSs);
• 5.9.1.8 Control charts or the equivalent (e.g., quality control database);
• 5.9.1.9 Calibration standards, blanks, and calibration devices (e.g., electronic
conductivity meter, NIST-traceable thermometer);
• 5.9.1.10 Reference material samples; and
• 5.9.1.11 Internal standards.
CSIR In-house additional requirements
• 5.9.2 Quality Control Practices.
• 5.9.2.1 The laboratory shall continually evaluate its performance
(system process control) for each method and matrix which
includes the determination of accuracy and precision.
• 5.9.2.2 Supervisory personnel shall conduct a documented review
of the data calculations and QC results.
• 5.9.2.3 Deviations or deficiencies in QC shall be reported to
management, and such reports shall be documented. QC data shall
be retrievable for all analytical results.
• 5.9.2.4 Method detection limits (MDLs) shall be determined and
documented.
• Note: Confirmation of MDLs shall be done as appropriate or as
required by the method.
CSIR In-house additional requirements
• 5E.9.3 Acceptance Limits. Acceptable performance
limits for analytical instrumentation as well as each
method shall be documented based upon the
continuing statistical evaluation of data generated by
the analysis of quality control samples, unless specific
minimum acceptance limits are established by the
method or the client.
•
• 5E.9.4 Where applicable, the following minimum QC
shall be practiced in the laboratory:
•
CSIR In-house additional requirements
• For Inorganics/ Classical Chemistry:
• 5E.9.4.1 One calibration check standard and associated blank in 10
samples tested; the lab should repeat analysis of all affected samples if
the calibration check standard is outside + 10% of expected value unless
the method specifies otherwise (Broader acceptance ranges must be fully
justified and documented).
• 5E.9.4.2 One reagent, method, or digestion blank (carried through
preparation) in 10 (or per batch).
• 5E.9.4.3 One matrix spike in 10 (or per batch).
• 5E.9.4.4 One duplicate or matrix spiked duplicate in 10 (or per batch).
• 5E.9.4.5 One laboratory control sample in 10 (or per batch).
• Note 1: Analysis of a low concentration or near MDL (2 to 3 times MDL)
standard with each batch of 10 samples or less is recommended to assess
analytical system performance near the MDL (required in CLP SOWs for
metals when client needs results near MDL).
CSIR In-house additional requirements
•
• 5.9.5 The laboratory shall establish control limits
for all the above types of QC samples and be able
to explain and document the basis for such
established limits
•
• 5.9.6 Control Charts. Control charts or control
data shall be used to track laboratory
performance with the associated acceptance
limits for each matrix and to evaluate instrument
performance
Current problems for nutrient and
oxygen
• Nutrient CRM availability and integrity
• Oxygen CRM?
• Southern Africa – interlaboratory study
availability (cost).
 Way forward??
• Sample preservation/handling/transport protocol
(after sampling but before analysis)
• Methods specified(sampling,preservation & analysis)
• Minimum reproducibility requirements
• CRM availability and use
• Estimation of measurement uncertainty
• Interlaboratory comparison study availability
• Lab analysis duplicate limits
• Use of internal QC with specified limits
– Eg +2 and -2 sd from mean <5%?? spread
What can laboratory say about the
result based on the above tests
• Assign one of 3 flags to an analysis
– Good – passes all tests
– Questionable – failed interlab but passed others
– Bad – Failed more than 2 tests or Failed
reproducibility

– The only problem with this approach is that in

terms of the requirements of ISO17025, the lab
should not release these results unless the flag
assignment is good.