GMP Sampling Training: QC Sampling and Retention Samples

Opens with FDA's guidance revised in May 2023 after children died from cough syrups contaminated with DEG and EG through excipients, where contamination varied from drum to drum. Two FDA letters follow: Dynamic Blending Specialists in Utah, which sampled only some supplier lots, and Cohance Lifesciences in India, whose reserve samples of crumbling tablets were not tested. You then meet JKL Pharma, which makes metformin tablets and amoxicillin capsules in Mumbai and sterile ceftriaxone injections in Singapore, with locked sample stores and an EU importer in Dublin, plus a generic maker in London and a contract laboratory in Sydney. The module outlines the ten modules, practice questions and final exam, corrects four myths and ends with a recap.

What you will be able to do

• Explain why a test result depends on how the sample was taken
• Recognise the sampling failures behind three recent FDA cases
• Describe how long retained samples must be kept after expiry

Why it matters

Seeing how poor sampling let contaminated material through elsewhere shows why your own plans and sample stores matter.

Practice: Practice: Welcome: why sampling matters · 3 questions

Starts with a small tablet maker whose procedure says only take a representative sample, so one sampler takes three drums with a clean thief and another one drum with a kitchen scoop. You then see where the rules are written and what a written sampling procedure must set: method, tools, amount, containers and precautions, approved by quality. The module covers trained, qualified samplers with recorded training, documenting at the time, and labelling each sample with batch, containers sampled, date and sampler. A contract laboratory in Sydney that tests for several medicine makers shows the shared duties, and JKL Pharma's procedure uses a separate form for each sample type. It ends with retained sample amounts and periods, a two-sampler test and a recap.

What you will be able to do

• Check that a sampling procedure sets method, tools, amount and containers
• Label a sample with batch, containers sampled, date and sampler
• Explain how much retained sample to keep and for how long

Why it matters

A procedure two samplers would follow the same way is what makes a sample representative, not a word on the label.

Practice: Practice: The rules in brief · 3 questions

Opens with forty drums of one excipient batch in quarantine and three samplers giving three answers: three drums, seven, or all forty. You then learn that identity, full quality testing and building a retained sample are different jobs, and that a uniform powder and a liquid that separates need different thinking. The module explains the WHO n, p and r plans, including the square root formula, and the limits WHO itself places on them. It covers recording which containers were chosen, US statistical quality control criteria for product sampling, and printed cartons sampled under an ISO 2859-1 plan. A generic medicine maker in London shows a plan revised after a failure. It ends with the common mistake and a recap.

What you will be able to do

• Choose a sampling plan for the job it has to do
• Explain the n, p and r plans and where they do not apply
• Justify a plan by risk, supplier history and variability

Why it matters

A plan chosen for its purpose and written down with its reasons stands up when an inspector asks why you opened that number of drums.

Practice: Practice: Sampling plans · 3 questions

Starts with FDA's 2023 guidance on glycol contamination, citing a report from Indonesia of nine drum samples over the safety limit with wide variability between containers. You then learn that the user tests identity on every batch and normally every container, and that Annex 8 allows fewer only under a validated procedure, improbable for brokers and for injections. The module shows numbering each container and recording each identity result against its number, then combining samples for the full specification. JKL Pharma has validated no reduced plan, while a tablet maker validated one for an excipient bought directly from a maker in Frankfurt. It ends with high-risk components tested in every container for the glycol limit, checks and a recap.

What you will be able to do

• Explain why identity must be tested container by container
• Check whether a reduced identity plan meets the Annex 8 conditions
• Record identity results so that a failing container can be found

Why it matters

Testing identity in every container catches the one wrong drum that an average across containers would hide.

Practice: Practice: Identity of every container · 3 questions

Opens with a powder thief wiped with a cloth but not washed between two materials, so traces of Monday's powder enter Tuesday's sample and drum. You then learn what sampler training covers: plans, written procedures, techniques, risks and what to notice. The module explains sampling in a defined, recorded area, with special or dedicated environments for materials such as hormones and penicillins, and clean, inert tools that are cleaned, dried, stored and logged. It shows WHO's labelling at the time of sampling, the sampling record a questioned result starts from, and JKL Pharma's booth log. A sterile medicine site in Boston sampling a sterile active ingredient adds aseptic precautions. It ends with resealing and marking containers, checks and a recap.

What you will be able to do

• Apply clean sampling technique from opening to resealing
• Check that sampling tools are cleaned, dried, stored and logged
• Label each sample at the moment it is taken

Why it matters

Careful technique keeps the sample true and the batch clean, because every later result starts with the sampler.

Practice: Practice: The sampler and technique · 3 questions

Starts with a tablet batch sampled only from the middle of the compression run, while the first drum held soft tablets from a press that took twenty minutes to settle. You then learn the in-process controls the rules list, such as weight variation, disintegration time, mixing, dissolution and pH, and the US rule for testing each batch against final specifications. The module covers WHO's list of bulk materials and how JKL Pharma checks tablets at the start, every thirty minutes, at the end and after every stop, with each sample time written into the batch record. A packaging site in Paris receiving patient leaflets shows packaging sampled one printed material at a time. It ends with retention samples from each packing operation and a recap.

What you will be able to do

• Plan in-process samples at the start, end and restarts of a run
• Sample finished product across the whole run in whole packs
• Apply safe sampling of printed packaging, one material at a time

Why it matters

Sampling where processes drift finds problems before customers do, instead of confirming only the steady middle of a run.

Practice: Practice: Packaging, in-process and finished · 3 questions

Opens with an assay result of 89 per cent against a lower limit of 90, retested and resampled until only a result within limits was reported, a pattern FDA's OOS guidance calls testing into compliance. You then follow a sample from container to laboratory: labelled and sealed at the time of sampling, transferred under its conditions and logged in a chain of custody. The module explains investigating an out-of-specification result starting in the laboratory, then production records and sampling procedures, using an active ingredient maker in Tokyo. It covers the maximum number of retests set in advance in a written procedure, when resampling is allowed, and why every result is reported without averaging. It ends with checks and a recap.

What you will be able to do

• Keep a chain of custody from container to result
• Explain why an OOS investigation starts with a laboratory check
• Apply retest and resample rules set in advance in a procedure

Why it matters

Reporting every result and following set rules for retests protects the site from the charge of testing into compliance.

Practice: Practice: Handling samples · 3 questions

Starts with a complaint about discoloured tablets, eighteen months after release, where the store holds a single bottle and the full release tests need three. You then learn the Annex 19 definition of reference samples as a record of the batch, for finished product, starting materials and packaging, with traceability records kept for inspectors. The module covers written agreements between sites, samples kept at the manufacturer where a mutual recognition agreement applies, and a contract packaging site in Frankfurt. You see JKL Pharma keep six bottles of metformin tablets per batch in the Mumbai store at 15 to 25 degrees, enough for two full runs. It ends with keeping periods past expiry, marketed packaging, a working laboratory nearby, and a recap.

What you will be able to do

• Explain what reference samples are and which materials need them
• Calculate the amount needed for two full runs of the release tests
• Check storage, keeping period and access to a working laboratory

Why it matters

Keeping enough reference sample means a complaint years later can be answered with full test data, not partial guesses.

Practice: Practice: Reference samples · 3 questions

Opens with a pharmacist's question about leaflet wording, answered within the hour from a retention sample after the artwork had been revised twice. You then learn that a retention sample is a fully packaged unit from each packing operation, kept a year past expiry, preferably where the certifying Qualified Person is located and open to the authority, under written agreements. The module covers the US yearly visual check of reserve samples, which Annex 19 does not require. A parallel distributor in Paris shows the 2026 rules for parallel trade, including photographic samples that are complete, high quality and traceable to the batch packaging record. You see JKL Pharma label each retention pack with its operation number. It ends with checks and a recap.

What you will be able to do

• Explain what a retention sample is and how long it is kept
• Check that retention samples can be found and reached quickly
• Describe the 2026 Annex 19 changes for parallel trade

Why it matters

A retention sample you can find within minutes shows a regulator exactly what a patient received.

Practice: Practice: Retention samples · 3 questions

Returns to Cohance Lifesciences, whose investigation of crumbling tablets named a raw material as the cause, but lacked chemical testing of reserve samples to confirm it. You then learn when retained samples are used: whenever a complaint, query or safety report concerns a batch, with a generic maker in London receiving a pharmacovigilance report. The module walks through examining the retention pack against the complaint sample, testing the reference sample, keeping the sample log up to date at the time, and recording approved disposal. It then recaps the course, from written plans and identity in every container to technique, handling and keeping samples, explains the scenario-based final exam, and ends with the course's closing message.

What you will be able to do

• Use retained samples to confirm the cause of a complaint
• Keep a sample log that records every removal at the time
• Summarise the sampling practices covered across the course

Why it matters

Testing retained samples during an investigation turns a stated cause into one you can show to the regulator.

Practice: Practice: Putting it together · 3 questions