How to Create an Effective Sampling Plan for Packaging Material Testing

When a shipment of packaging materials arrives at a facility, testing every single unit in the batch is neither practical nor economical. Instead, quality assurance teams rely on a carefully designed sampling plan – a systematic, standardized procedure for drawing a representative set of samples from a batch and using those samples to draw conclusions about the entire lot. Done correctly, a sampling plan gives you reliable quality data without the cost and time burden of 100% inspection. Done poorly, it can let a substandard batch slip through undetected. This post walks through exactly how an effective sampling plan for packaging material testing is built – from defining your batch to labeling your specimens and making an acceptance decision.

What is a sampling plan? 🔗

A sampling plan is a documented procedure that specifies how many units to draw from a batch, where and how to draw them, and what criteria to use when accepting or rejecting the batch. As the EU GMP Guidelines, Annex 8 on sampling of packaging materials states, valid quality conclusions cannot be based on tests carried out on non-representative samples – making correct sampling an essential part of any quality assurance system. The plan does not exist in isolation; it must align with the material type, the quantity received, the production methods involved, and what is known about the supplier’s own quality system.

Key terms you need to know 🔗

Before executing a sampling plan, everyone involved must work from the same definitions. Misunderstanding even one of these terms can compromise the entire process.

Batch 🔗

According to WHO Technical Report Series No. 929, a batch is a quantity of material produced during a given cycle of manufacture under conditions that are stable and have not been modified. If the process is continuous, the batch is defined by a set time period during which conditions remain constant. This definition matters because the integrity of a sampling plan depends on the assumption that the batch is internally consistent – that is, all units within it were made under the same conditions.

Set of samples 🔗

A set of samples is the collection of individual units drawn from the batch in accordance with the sampling plan. These units represent the batch during testing. The number of units in the set is not arbitrary – it is calculated based on the batch size, the required confidence level, and the nature of the tests to be performed.

Specimen 🔗

A specimen is a portion prepared from a sample and used directly in a specific test. For example, if a roll of plastic film is sampled, rectangular strips cut from that roll for a tensile strength test are the specimens. One sample may yield multiple specimens for different tests, each prepared according to the relevant test method’s requirements.

Original sample vs. final sample 🔗

The WHO guidelines draw a clear distinction: an original sample is collected directly from the material as it is received. A final sample is what is actually submitted for the test procedure – it may be a portion of the original sample, prepared or conditioned as required by the test method. Keeping this distinction clear avoids confusion during documentation and analysis.

Why batch homogeneity is the foundation of sampling 🔗

The entire logic of sampling – the idea that a small subset can represent the whole – rests on one critical assumption: the batch is homogeneous. According to WHO guidelines, a material is considered homogeneous when all units within it share the same origin and have been produced under the same conditions. If a batch contains materials from different production runs, different suppliers, or different storage environments, it is non-homogeneous, and a standard sampling plan may not give you reliable results.

In practice, several factors affect homogeneity: raw material source, machine settings, ambient conditions during production, and operator practices. When these variables are held constant throughout a production run, the resulting packaging materials are more likely to perform consistently. When they shift – say, due to an equipment reset midway through a run – the batch should ideally be split into sub-lots, each treated as a separate batch for sampling purposes. The Codex Alimentarius General Guidelines on Sampling (CAC/GL 50-2004) recommend stratified sampling in such cases, where the batch is divided into more homogeneous subgroups and samples are drawn from each.

When non-homogeneity is suspected, quality assurance teams should not simply increase sample size and proceed. The appropriate response is to investigate the source of variation, document it, and adjust the sampling strategy accordingly – for example, by sampling from different zones or layers of the batch rather than drawing all samples from a single location.

Designing the sampling plan: step by step 🔗

Step 1: Define the batch clearly 🔗

Before any sample is drawn, the batch must be formally identified and defined. This includes confirming the batch number, production date, supplier, quantity received, and the relevant material specification. The standard procedure for packaging material sampling requires verifying on arrival that the supplier is on the approved vendor list and that the goods receipt details match what is physically present. Only after this verification should the sampling process begin.

Step 2: Determine sample size using a statistical method 🔗

Sample size must be determined statistically – not based on convenience or habit. The most widely used international framework for this is ISO 2859-1, which provides acceptance sampling procedures indexed by the Acceptable Quality Level (AQL). The AQL defines the maximum percentage of defective units that is still considered acceptable for a batch.

Under this framework, the batch size determines a sample size code letter, which then links to the required number of units to be inspected and the acceptance/rejection numbers at the chosen AQL. As QualityInspection.org explains, more than 95% of quality inspectors performing random sampling follow ISO 2859-1 as their guiding standard. Common AQL values in use include 1.0% for critical or medical-grade materials and 2.5% for general consumer or industrial packaging. A lower AQL means tighter inspection – fewer defects are tolerated before a batch is rejected.

Step 3: Select the inspection level 🔗

ISO 2859-1 provides three general inspection levels (I, II, and III) and four special levels (S-1 to S-4). General Inspection Level II is the standard default for most packaging material testing scenarios. Level I is used when less discrimination is acceptable or when the risk is lower. Level III applies when tightened scrutiny is required – for example, when a supplier has had previous quality failures. Inspection level affects sample size, not the AQL itself – a higher level simply means more units are drawn from the same batch at the same AQL threshold.

Step 4: Apply random sampling 🔗

Samples must be drawn randomly to avoid bias. Every unit in the batch must have an equal probability of being selected – this is the principle behind a random sample as defined by WHO guidelines. In practice, this often means using random number tables or software to select container or unit positions, rather than picking from the most accessible area of the pallet or shipment. Sampling only from the top layer, front row, or visually appealing units is a common error that compromises statistical validity.

For large shipments where physical access is difficult, systematic or stratified sampling can serve as a practical alternative, provided the approach is documented and justified. The Codex sampling guidelines note that each stratum should be more homogeneous than the original lot, and samples drawn from each stratum should follow random principles within that subdivision.

Step 5: Prepare and label specimens 🔗

Once samples are drawn, they are sealed, labeled, and transferred for testing. Each sample container must carry a clear label showing the batch number, date of sampling, sampling location, sampler’s identity, and a unique sample identification number. An “UNDER TEST” label is affixed to all sampled units in the batch that remain in the store pending results. After specimens are prepared for destructive tests, any remaining material from those samples is destroyed after analysis and the destruction is documented. For sensitive materials like aluminium foil or PVC/PVDC film, sampling is carried out under a Laminar Air Flow (LAF) bench to prevent contamination.

Step 6: Make the acceptance decision 🔗

After testing, results are compared against the approved material specification. The standard operating procedure for packing material testing and release requires that quality control record all test results – physical parameters, dimensions, print accuracy, appearance, and any other specified attributes – and report non-conformances to QA immediately. Based on these results, the batch is either released to the packaging store with an approved status label or rejected and transferred to a quarantine area pending investigation or return to the supplier.

Types of packaging materials and how sampling differs 🔗

Packaging materials are typically classified into three categories, and the sampling approach can differ between them. Primary packaging materials – those in direct contact with the product, such as bottles, blisters, and foils – generally require 100% sampling from each consignment given the higher risk of contamination or defect. Secondary packaging materials, such as cartons and labels, are sampled according to AQL-based plans determined by batch size. Tertiary packaging materials, like shipper boxes and stretch film, are typically tested using a smaller fixed sample – often five units from the consignment – with additional testing carried out in the QC lab on a single specimen per test parameter.

Documentation and traceability 🔗

A sampling plan is only as strong as its documentation. Every action in the sampling process must be recorded: when the batch arrived, who drew the samples, which units were selected, when they were submitted for testing, and what the results showed. This creates a full audit trail from raw material receipt to release or rejection. GMP standard operating procedures require that the sampling inspector also record the inspection history for each supplier – tracking defect levels across consecutive deliveries so that switching rules can be applied. If five consecutive lots from a supplier pass under tightened inspection, the plan can revert to normal inspection for the next delivery. Conversely, repeated failures trigger escalation to tightened inspection or supplier review.

Modern facilities increasingly use barcode or RFID-based tracking systems to automate sample identification and link results to batch records in ERP or quality management systems. This reduces transcription errors and speeds up the release process while maintaining full compliance with documentation requirements.

Common mistakes that undermine sampling plans 🔗

Even well-designed sampling plans can fail in execution. The most frequent errors include: drawing samples only from accessible or visually uniform areas of the batch (selection bias); failing to verify batch homogeneity before applying a standard sampling plan; using an outdated or incorrectly applied AQL table; and not maintaining adequate documentation of who sampled, when, and from where. A key misconception is treating AQL as a quality target – it is a statistical acceptance threshold, not permission to produce defective material. The actual manufacturing goal should always be zero defects; the AQL simply defines the point at which a batch statistically warrants rejection.

What do you think? If a batch of primary packaging material is found to be non-homogeneous – say, materials from two different production shifts were combined – how should the sampling plan be adjusted before testing begins? And given that a lower AQL means tighter inspection and higher costs, how should a quality team decide what AQL threshold is appropriate for a specific packaging material?

References

1. https://health.ec.europa.eu/document/download/22d03c04-8512-4336-89dd-556771fca388_en
2. https://www.who.int/docs/default-source/medicines/norms-and-standards/guidelines/quality-control/trs929-annex4-guidelinessamplingpharmproducts.pdf
3. https://www.fao.org/uploads/media/Codex_2004_sampling_CAC_GL_50.pdf
4. https://www.pharmaguideline.com/2011/07/sop-for-procedure-for-sampling-of.html
5. https://www.iso.org/standard/1141.html
6. https://qualityinspection.org/inspection-level/
7. https://www.inspectionmanaging.com/blogs/testing-and-standards/how-do-the-aql-inspection-levels-in-iso-2859-1-affect-sampling-size
8. https://pharmatimesofficial.com/project/sop-for-testing-and-release-rejection-of-packing-materials/
9. https://www.gmpsop.com/what-is-acceptable-quality-limit-aql-in-samping/
10. https://www.alekvs.com/what-is-iso-2859-a-practical-guide-to-sampling-inspection/