Knowledge into action
Six short cases with source-based explanations and two step-by-step educational calculators. Choose a case or change example inputs and inspect the effect.
For learning only; fictional inputs are not limits for a real product. A qualified specialist must derive the PDE. Approving cleaning limits, sampling and release requires a documented product, equipment and market assessment.
A is the residue from the previous product. Batch size and dose belong to the next product B. The example assumes carryover into that batch and allocates its limit across the entire shared product-contact surface.
Concentration and LOQ must refer to the same extraction solution after any analytical dilution is accounted for. This is not directly a rinse-sampling calculation. The surface limit is an independently established example input.
All scenarios below are fictional. Discuss the question, then reveal the explanation; this page does not issue a competence certificate or a real batch decision.
Fictional teaching case
Fictional case: the initial assay is 93.8% against a 95–105% specification. Two retests give 99.1% and 99.4%. No causative laboratory error has been identified. The analyst proposes deleting the first result.
Are the retests sufficient for that action? What evidence would you request?
No. Retain all results and complete the documented laboratory and, where no clear laboratory cause is found, manufacturing investigation. A passing retest alone does not justify invalidating the OOS. Batch disposition requires quality-unit assessment of the complete evidence.
Common pitfall: Selective averaging or repeating tests until they pass.
Working output: Original results, hypothesis plan, investigation evidence and disposition rationale.
FDA · OOS investigations
Section IV.B.1, Retesting; Section V, Concluding the investigation.
Fictional teaching case
Fictional case: a sterile batch passes its sterility test, but the production record shows an unqualified filling intervention and an environmental monitoring excursion.
Does the final-product test alone resolve the deviation?
No. Assess the intervention and excursion alongside process records, monitoring and the contamination control strategy. A passing sterility test does not compensate for inadequate sterility assurance. A documented investigation and qualified assessment must inform disposition.
Common pitfall: Treating a sterility test as an independent guarantee of process control.
Working output: An intervention and excursion timeline, CCS review and documented impact assessment.
EU GMP · Annex 1
Sections 2.6 and 2.7: CCS review and limits of reliance on final testing.
Fictional teaching case
Fictional case: solution LOQ is 0.2 µg/mL, extraction volume 10 mL, recovery 50% and sampled area 25 cm². The example’s approved surface limit is 0.1 µg/cm².
Can the method’s sensitivity establish a result below that limit?
Surface-equivalent LOQ = 0.2 × 10 ÷ 0.5 ÷ 25 = 0.16 µg/cm². It exceeds the example limit, so “below LOQ” does not establish compliance with that limit. Review analytical suitability and the sampling plan.
Common pitfall: Comparing µg/mL directly with µg/cm², or applying recovery correction twice.
Working output: A unit-consistent calculation, documented recovery and sensitivity linked to the limit.
EU GMP · Annex 15
Section 10, especially 10.11–10.12. The calculation is an illustrative mass balance, not an equation quoted from the annex.
Fictional teaching case
Fictional case: three analysts share one account with editing rights. Only the final result was printed; it differs from an earlier report, without clear attribution of reprocessing.
What is missing before this record can be relied upon?
Retrieve original data, metadata, audit trails and the reprocessing rationale. Review permissions and attribution to individuals. The final number alone cannot reconstruct the activity. A shared read-only account differs from one permitting edits or an attributable review.
Common pitfall: Relying only on the printout or assuming a passing result makes the record complete.
Working output: A reconstructable data package, access review and corrective actions.
FDA · Data integrity Q&A
Questions 1(b–c), 4, 5 and 7.
Fictional teaching case
Fictional case: a team uses the EU implementation date of M13A to conclude that every dosage form and study design is covered in the same way.
What is wrong with this conclusion?
First check dosage form, release, systemic action and study design. M13A concerns immediate-release solid oral dosage forms for systemic delivery. EMA describes replacement of applicable parts for non-replicate designs, not cancellation of every requirement or adoption by every country.
Common pitfall: Automatically applying the EU implementation date to another market or dosage form.
Working output: A scope matrix: product, design, market, applicable source and version.
ICH M13A · EMA
Guideline scope and EU implementation notes on the EMA page.
Fictional teaching case
Fictional case: an author cites a ClinicalTrials.gov identifier as proof that the US government approved the intervention’s safety and effectiveness.
What does registration establish, and what still needs appraisal?
Registration makes study information available; it does not mean the US government approved its safety or science. Compare registry and protocol with reported results, and appraise design, bias, effect size and uncertainty before drawing conclusions.
Common pitfall: Treating a registration number or complete reporting as a quality certificate.
Working output: A registry-to-publication comparison, documented appraisal and identified missing data.
ClinicalTrials.gov
ClinicalTrials.gov registry-use disclaimer.
