Technical guide for manufacturers, quality teams and laboratory purchasers. A reliable laboratory result depends on more than the analytical instrument. Sample selection, identity, custody, preservation and reporting limitations can determine whether a test is meaningful for the intended decision. This practical guide explains how to build an evidence trail aligned with the relevant concepts of ISO/IEC 17025:2017.
- Define the decision and test method before sampling.
- Keep sample identity unique from collection through reporting.
- Document custody, transport conditions and receipt exceptions.
- Verify laboratory scope and the meaning of the final report.
Why sample integrity matters before any test begins
A technically capable laboratory can produce a precise measurement on the wrong specimen. That is why procurement, quality and laboratory teams must consider the entire route from sampling plan to final report, not only instrument calibration. A sample may be collected from an unrepresentative location, mislabeled during transfer, exposed to unsuitable temperatures, or confused with a similar container. In each case, the final number may look credible while answering the wrong question. ISO/IEC 17025:2017 addresses laboratory competence and consistent operation, including sampling where applicable and handling of test or calibration items. For manufacturers, the practical objective is to establish which material was examined, under what conditions, by whom, and whether the reported result remains applicable to the original lot. The controls described below are examples of an evidence-based approach; the exact requirements depend on the method, matrix, contractual agreement and applicable regulation.
Start by defining the decision the result must support
Before collecting anything, write down the decision the result will inform: release a batch, investigate a complaint, qualify a supplier, verify a product specification or support a regulatory submission. Specify the product, lot, test parameter, acceptance criterion, applicable method and required turnaround. A vague request such as “test the material for compliance” leaves too much room for interpretation. For instance, a polymer manufacturer asking for tensile properties should identify the relevant specimen preparation, conditioning and test method rather than expecting a generic strength value. The laboratory should review whether it can perform the requested work and whether the customer understands any limitations. Where sampling is performed by the customer rather than the laboratory, the parties should identify that responsibility and the effect it may have on interpretation. A sound test request is the first piece of traceable evidence.
Verify laboratory accreditation and the actual scope
A logo or certificate number alone does not establish that a particular test is within a laboratory’s accredited scope. In India, the National Accreditation Board for Testing and Calibration Laboratories, or NABL, accredits testing and calibration laboratories against ISO/IEC 17025. Procurement should verify the laboratory identity, relevant location, validity, test parameter, method and scope through current official accreditation information. If a particular regulator requires additional recognition, accreditation alone may not satisfy that sector-specific rule. Do not confuse certification of a management system with accreditation of a laboratory for defined technical activities. Ask whether the report will identify results issued under accreditation, and clarify the treatment of any non-accredited parameters. These checks are especially important for export documentation, customer acceptance and high-consequence product decisions.
Design a sampling plan that represents the lot
A sample cannot represent an entire shipment simply because it came from the shipment. Define the population, sampling units, locations, frequency and selection approach based on the product, variability and purpose. A food ingredient held in several silos may need a different plan from machined components produced in a stable process. Where the laboratory undertakes sampling, ISO/IEC 17025 includes provisions for sampling plans, methods and associated records. When the customer samples, the laboratory should not imply it independently verified how representative those specimens were. Retain the sampling instruction and any justified deviations. For high-risk characteristics, consider whether stratification, multiple increments or special preservation is necessary. Sampling plans should be technically defensible, not chosen merely because they are convenient for the courier or shift supervisor.
Build an unambiguous sample identification system
Each primary sample should receive an identifier that remains linked to its source throughout receipt, storage, preparation, testing and reporting. A useful identifier is unique within the relevant system and does not depend on handwritten descriptions such as “blue bag” or “sample two.” Link the code to the lot, collection point, date, collector and requested tests. If a sample is split into subsamples, maintain the parent-child relationship. Barcodes can reduce transcription mistakes, but they do not eliminate the need for controlled data entry and verification. Labels must survive the actual handling conditions; ink that dissolves in a freezer or an adhesive that detaches during transport defeats the purpose. At handover, confirm that the physical item and the electronic or paper record agree before accepting custody.
Record custody transfers without creating paperwork theatre
Chain-of-custody records are especially valuable when sample ownership, location or condition may later be disputed. A practical log identifies the sample, the releasing and receiving parties, transfer time, purpose and observed condition. Use signatures or authenticated digital acknowledgments where appropriate. The point is not to collect signatures for their own sake: it is to reconstruct what happened and detect gaps. For routine in-house testing, a controlled laboratory information system may provide a sufficient audit trail. For external dispatch, document packaging, courier tracking and any special instructions. If a handover is missed, investigate rather than backfilling a fictional entry. A credible record includes anomalies and corrective actions, not just an uninterrupted sequence of green checkmarks.
Minimum handover fields
- Unique sample identifier and associated lot or source
- Sender, recipient, transfer time and transfer purpose
- Container and seal condition, when relevant
- Temperature or preservation record where required
- Exception, decision and responsible reviewer
Control packaging, transport and environmental exposure
Transport conditions should protect the characteristics being measured. Some samples require temperature control, light protection, tamper evidence, upright orientation, moisture barriers or specific time limits. The appropriate controls depend on the method and product; there is no universal temperature or maximum transit duration. Where necessary, use validated packaging and suitable monitoring devices, and review recorded excursions on receipt. A photograph of a sealed box is not proof that the contents stayed within specification. Record the packaging configuration, dispatch and arrival times, and the acceptance criteria for transport conditions. For an unstable analyte, a temperature excursion may invalidate a result even when the container remains physically intact. The laboratory and customer should agree beforehand on how such deviations will be assessed and reported.
Treat receipt inspection as a technical decision point
Receiving staff should verify identification, quantity, container integrity, requested testing, relevant preservation conditions and discrepancies. ISO/IEC 17025 addresses the recording of deviations from specified conditions and consultation with the customer when suitability is in doubt. An item that arrives leaking, unlabeled or outside a critical transport limit should not automatically enter routine testing. Quarantine or clearly flag the item while competent personnel decide whether to reject it, request a replacement or proceed with a documented limitation. Keep the original observation, decision maker, customer communication and any report qualification together. This protects both the laboratory and the purchaser: an exception that is visible at receipt can be assessed; one discovered only after release becomes much harder to explain.
Separate storage control from test preparation
Sample storage and sample preparation are distinct stages with different risks. Storage needs defined location, environmental controls, access arrangements and retention periods where relevant. Preparation may introduce dilution, cutting, homogenization, extraction, drying or other changes that must be traceable to the original item. Use suitable equipment, trained personnel and controlled methods. If several samples are prepared in parallel, arrange work to prevent cross-contamination and mix-ups. Record key preparation details that could affect interpretation, including relevant weights, volumes, conditions and deviations. Where a sample is consumed or altered during analysis, note that fact in the retention plan. A controlled storage room does not compensate for an undocumented preparation step, and an excellent preparation worksheet cannot repair a lost sample identity.
Maintain technical records that allow reconstruction
A defensible technical record lets another competent person understand how the result was produced. Depending on the method, this may include instrument identity, calibration status, standards or reference materials, environmental conditions, sample preparation, raw observations, calculations, analyst identity, review and any departures from the procedure. Retain original data or controlled electronic equivalents with appropriate access and change history. A spreadsheet showing only the final answer is often inadequate for investigating an unexpected result. For automated systems, ensure that data transfer, software configuration and audit trails are controlled according to risk. Records should be sufficiently complete to support technical review without becoming a dumping ground for irrelevant screenshots. Quality is demonstrated by meaningful traceability, not document volume.
Distinguish calibration, method suitability and result validity
Calibration provides evidence about measurement equipment; it does not prove that every sample was suitable or every method was correctly applied. Laboratories must also consider method selection, verification or validation as applicable, competence, measurement uncertainty where relevant, and ongoing validity checks. A calibrated balance can still produce a misleading result if the wrong specimen was weighed or the method’s range was exceeded. Ask how the laboratory controls blanks, reference materials, replicates, proficiency testing or other appropriate quality-control activities. The exact mix depends on the test and risk. When reviewing a result, examine whether the quality-control evidence relates to the actual batch of work, rather than assuming that a calibration certificate from months earlier proves the entire analytical process.
Handle nonconforming work and suspect results transparently
A mislabeled sample, failed quality-control check, equipment malfunction or missing preservation record may trigger assessment of nonconforming work. The laboratory should determine the significance, stop or repeat work where necessary, evaluate previous results that could be affected, and communicate appropriately with customers. Not every discrepancy requires discarding all data, but decisions must be justified and recorded. A corrective action should address the cause rather than merely instructing staff to “be careful.” For example, repeated transposed sample numbers may call for redesigning labels, scanning at transfer points or restricting manual re-entry. When the problem affects a released report, controlled amendment and notification processes matter. The objective is trustworthy decisions, not protecting an error-free appearance.
Write reports that make limitations visible
A useful test report identifies the tested item, relevant method, results, units, report identity and other information required by the applicable reporting provisions. Where sampling was not performed by the laboratory, do not imply that the laboratory established the representativeness of the original lot. If the sample arrived outside specified conditions and testing proceeded by agreement, relevant qualifications should be visible to the reader. Statements of conformity need particular care: the applicable specification and decision rule should be understood and agreed as required. A simple pass/fail label can conceal uncertainty or a mismatch between the buyer’s criterion and the method. Procurement should read the notes and scope, not only the bold conclusion. A report is evidence about defined work, not a blanket guarantee for every item in a shipment.
Illustrative case: a disputed packaging material batch
Consider a fictional manufacturer receiving a customer complaint about a batch of polymer packaging. Three bags are sampled from different pallets, but two containers are labeled with the same handwritten code. One is shipped in a damaged outer carton; the laboratory receives all three without a clear sampling record. The eventual test results differ. The correct response is not to average the numbers and declare the batch compliant. The manufacturer and laboratory should reconstruct the sampling and transport history, identify which results can be confidently linked to which source, and determine whether new sampling is necessary. A revised plan could use unique labels, pallet references, sealed containers, transfer acknowledgments and a receipt checklist. This is an invented training example, not a description of a real Shark client, laboratory or investigation.
Audit-ready evidence checklist for manufacturers and laboratories
An effective audit sample follows one item from request to report. Select a completed job and retrieve the purchase or test request, sampling responsibility, plan where applicable, identifier, transfer log, transport requirements, receipt acceptance, preparation record, instrument and method information, quality-control evidence, technical review and final report. Verify that dates and sample codes agree across systems. Then select an exception case to see how deviations were handled in practice. Auditors should ask personnel to demonstrate the actual workflow, including the electronic audit trail, rather than accepting a policy document alone. The checklist is a practical organizing aid and not a substitute for the full requirements of ISO/IEC 17025 or the applicable accreditation scheme.
Documents to retrieve for one completed job
- Approved request and applicable method.
- Sampling plan or recorded customer sampling responsibility.
- Sample identification and custody history.
- Transport and receipt acceptance evidence.
- Preparation and original technical records.
- Quality-control, review and issued report.
- Any deviation assessment and customer communication.
Create a 30-day improvement plan with clear owners
In the first week, map the current sample journey and identify where identity or condition could be lost. During the second week, pilot unique identification and receipt acceptance on one product family. In the third week, test the process with an intentionally simulated discrepancy and check whether staff know who can authorize the next step. In the fourth week, review evidence from completed jobs and revise instructions where actual practice differs. Assign owners from quality, laboratory operations, logistics and IT as appropriate. Useful indicators may include unmatched sample identifiers, missing transfer records, temperature excursions, receipt holds and report corrections; avoid turning these into invented benchmark targets. Management should review whether controls improve confidence in decisions, not merely whether a training attendance sheet was signed.
Frequently asked questions
Is ISO/IEC 17025 a product certification?
No. It addresses competence and consistent operation of testing and calibration laboratories. It does not by itself certify every product tested.
Does NABL accreditation cover every test a laboratory offers?
No. Verify the current accredited scope, specific parameter, method, location and validity through official NABL information.
Who is responsible when the customer collects the sample?
Responsibilities should be agreed before testing. The laboratory must not imply it performed sampling or verified representativeness when it did not.
Does a chain-of-custody form guarantee valid results?
No. It supports traceability; method suitability, preservation, competence and quality controls also matter.
What should happen if a sample arrives damaged?
Record the deviation, assess suitability and consult the customer as appropriate before deciding whether testing can proceed and how limitations will be reported.
Primary references and next steps
For authoritative requirements, consult ISO/IEC 17025:2017 from ISO, NABL testing laboratory accreditation guidance and NABL’s accreditation overview. This article is educational and does not reproduce the standard or substitute for its controlled text.
Related reading: Why calibrated equipment can still produce wrong results and Why supplier declarations are not enough.
Discuss documentation and audit-readiness support
For organizations reviewing laboratory-related evidence within broader quality and compliance programmes, explore Shark Certification’s consultancy services or contact the team to discuss your requirements. This does not imply that Shark operates an ISO/IEC 17025-accredited testing laboratory or holds accreditation for a particular test.
Editorial image credits: featured image by Trnava University; inline images by Talha Hassan, Julia Koblitz and Louis Reed, all via Unsplash. These are contextual stock photographs, not photographs of Shark facilities, customer work or certified processes.
