Skip to main content

Manufacturing and batch records

How do you adjust an API quantity for potency (as-is, anhydrous or dried basis) in a batch formula?

Put the lot’s assay on the as-is basis, then scale the label-claim quantity by it. As-is assay = anhydrous assay × (100 − water %) ÷ 100, or dried-basis assay × (100 − loss on drying %) ÷ 100. Quantity to weigh = label-claim quantity × 100 ÷ as-is assay. A lot assaying 98.7% anhydrous with 4.2% water is 94.55% as-is, so 50.00 kg of API needs 52.88 kg. State the factor and the compensating excipient in the master formula.

Illustration of a seal beside a production line of sealed vials.

1The formula for each basis.

A potency-adjusted master formula states the API at 100% content and scales it by each lot’s real content, because the batch must be formulated “with the intent to provide not less than 100 percent of the labeled or established amount of active ingredient”.¹ Convert the certificate’s assay to the as-is basis, the material as it sits in the container, then divide:

Quantity to weigh (Q) = Q₀ × 100 ÷ A, where Q₀ is the master formula quantity at 100% content and A is the as-is assay. The potency factor is 100 ÷ A.

Table 1. Converting a certificate’s assay to the as-is basis.
The certificate reportsAs-is assay A
Assay as isA = assay
Anhydrous, water W%A = assay × (100 − W) ÷ 100
Dried, loss on drying L%A = assay × (100 − L) ÷ 100
Solvent-free, W% and S%A = assay × (100 − W − S) ÷ 100, where S is residual solvents
Solution, C% w/wA = C, and Q is a mass of solution; for % w/v, divide C by the density in g/mL first

Pair each basis with its test: anhydrous with a water result, dried with loss on drying, which is not specific for water; ICH Q6A prefers “a detection procedure that is specific for water (e.g., Karl Fischer titration)”.² Karl Fischer water includes a hydrate’s water of crystallisation, so for an anhydrous label claim the water-corrected assay is already the as-is content; applying a hydrate-to-anhydrous molecular weight ratio as well removes the water twice. For a free-base claim and a salt assay, first multiply Q₀ by the salt’s molecular weight over the base’s. For potency in units, Q is the units required divided by the as-is units per milligram.

1.1For a reference or working standard.

The standard’s content enters every assay result, so it too must be on the basis of the standard as weighed. Use a primary standard as its supplier directs:³ one labelled to be dried before use is dried as the label says, and its stated value then applies as given. A Ph. Eur. CRS is used as received, and dried it “will no longer be suitable for its intended use”; its leaflet value already accounts for water and solvent.⁴⁵ Convert only a value stated on a different basis from the material as weighed. Assigning a working standard’s value is worked through in working standard qualification.

2Example: one lot, three readings of 98.7%.

Example, invented. Product A 100 mg tablets, Batch 26-041: 500,000 tablets of 250 mg. The label claim is 100 mg of anhydrous API A per tablet. The master formula holds 50.00 kg of API A at 100% content, 67.50 kg of microcrystalline cellulose as the compensating filler, 5.00 kg of croscarmellose sodium and 2.50 kg of magnesium stearate: 125.00 kg. QC released API A lot A-2611 with an assay of 98.7% on the anhydrous basis and 4.2% water by Karl Fischer.

Table 2. Example: the dispensing calculation for Batch 26-041.
StepCalculationResult
As-is assay98.7 × (100 − 4.2) ÷ 10094.5546%
Factor100 ÷ 94.55461.05759
API A to weigh50.00 kg × 100 ÷ 94.5546 = 52.8795 kg, rounded up to the balance’s 0.01 kg52.88 kg
Filler to weigh67.50 kg − (52.88 − 50.00) kg64.62 kg
Batch mass52.88 + 64.62 + 5.00 + 2.50 kg125.00 kg
Check52.88 kg × 94.5546% = 50.0005 kg of anhydrous API A100.0% of label claim

The same figure gives three quantities, depending on the basis printed beside it:

Table 3. Example: one assay figure, three quantities. Weights are rounded up to 0.01 kg.
If 98.7% is on theAs-is assayAPI A to weighFiller to weigh
As-is basis98.70%50.66 kg66.84 kg
Anhydrous, water 4.2%94.55%52.88 kg64.62 kg
Dried, LOD 4.6%94.16%53.11 kg64.39 kg

Read the anhydrous result as if it were as-is and the operator weighs 50.66 kg, which holds 50.66 × 94.5546% = 47.90 kg of API A: 95.8% of the label claim, with every weighing in tolerance. Record the basis next to every assay value, not only in the specification.

Set the weighing tolerance around the adjusted 52.88 kg, the weight the second person checks as “correct as stated in the batch production records”.¹ A tolerance of ±0.10 kg accepts 52.78 to 52.98 kg, and its low limit delivers 99.8% of the claim; a tolerance above the target only keeps every accepted weight at 100% or more. The same ±0.10 kg around the nominal 50.00 kg would pass 50.10 kg: 94.7% of the claim. Whether the balance can weigh to that tolerance is a minimum weight question.

3Test cases for the spreadsheet or MES.

Input to and output from a computer system of formulas must be “checked for accuracy”, and Annex 11 asks validation test scenarios to consider “system (process) parameter limits, data limits and error handling”.⁶⁷ These cases come from the example, with a master record range of 90.0 to 102.0% as-is.

Table 4. Example: calculation test cases for Batch 26-041, with expected results.
CaseInputExpected resultTests
Basis conversion98.7%, anhydrous; water 4.2%52.88 kg API A; 64.62 kg fillerOutput accuracy, 211.68(b)
Assay above 100%100.8%, as is49.61 kg; or 50.00 kg if the factor is capped at 1The master record’s rule
Outside the range89.9%, as isBlocked, no weight shown, referred to QAParameter limit, Annex 11 4.7
Rounding98.7%, anhydrous; water 4.2%52.88 kg; 52.89 kg means the as-is assay was rounded before dividingOutput accuracy, 211.68(b)
Missing basis98.7%, basis blankRejected until a basis is enteredError handling, Annex 11 4.7 and 6

4What the master formula must say.

Part 211 does not mention potency adjustment by name. It permits “reasonable variations” in component quantities only if “they are justified in the master production and control records”, and wants “a statement concerning any calculated excess of component”.⁸ A factor that moves the API and filler from batch to batch is such a variation. For the dossier, EMA says that where the quantity of an active ingredient “is calculated from the actual assay value of the batch of that active ingredient”, “this should be stated and justified”, and an ingredient used to keep the batch mass constant “should also be indicated”.⁹ Because the factorisation sits in the registered batch formula (CTD 3.2.P.3.2), adding it to an authorised EU product changes that formula; EU variations covers the filing.¹⁰⁹

Copy these entries into the master record and fill the brackets.

Table 5. Master formula entries for a potency-adjusted API, each tied to its source.
EntryPaste and fillRequirement
Component[API name and grade]; label claim expressed as [anhydrous / as-is / free base]“Any special quality characteristic”, 211.186(b)(3);⁸ EU 4.17(c)¹¹
Quantity to weigh[Q₀] kg at 100% content × 100 ÷ as-is assay, where as-is assay = [formula for the basis], from QC-approved [assay method] and [water method] results211.186(b)(4);⁸ factorisation stated and justified⁹
Compensating excipient[a filler whose level does not affect performance, not the disintegrant or lubricant], reduced by the API excess so the batch stays [total] kgCompensating ingredient indicated⁹
RangeThe factor applies when the as-is assay is [x] to [y]%; outside it, stop and refer to QAA justified variation, 211.186(b)(4)⁸
RoundingFull precision until the weight; the weight rounded up to the balance’s readabilityNot less than 100%, 211.101(a)¹
Tolerance[± or +] [amount] kg around the adjusted weightWeight correct as stated in the batch record, 211.101(c)(2)¹
Overage[amount and reason], stated apart from the factor, or none211.186(b)(5);⁸ ICH Q8(R2) 2.2.2¹²
Calculation checkCalculated by [system or validated spreadsheet], tested against cases like those above; inputs checked by [second person or validated interface]211.68(b);⁶ Annex 11 sections 4.7 and 6⁷

5Who does what, and when.

  1. Development and regulatory affairs choose the basis and the compensating excipient, and state the factorisation in the dossier’s batch formula.¹⁰⁹
  2. One person prepares the master record, a second checks it independently, and the quality unit approves it.⁸¹³
  3. QC approves the API lot and the assay and water values to use. A supplier’s certificate can replace your own assay in the US only with your own identity test and validated reliability of the supplier’s results; in the EU, only with an identity test on each batch, a justified reliance and a full analysis at intervals to check the certificate.¹⁴¹⁵
  4. At dispensing, a second person or validated electronic means checks the inputs before the weight is calculated.⁶⁷ After weighing, a second person, or one person checking automated equipment, confirms the weight against the batch record.¹¹⁵
  5. The batch record keeps the API lot and the quantities actually weighed; QA reviews it before release and investigates any unexplained discrepancy.¹⁶¹¹¹⁷

FDA’s drug inspection programme lists “Adequate procedure and practice for charge-in of components” and “Formulation/manufacturing at not less than 100 percent” among the production areas investigators cover.¹⁸ Be ready to trace any weighed quantity back to the certificate, and to show how the calculation was validated and who can change it.⁶

6Decisions at the balance.

Table 6. Decisions at dispensing, and the rule behind each.
SituationWhat to doBasis
Supplier and QC differUse the value QC approved, and investigate the gap211.84(d)(2);¹⁴ 211.192¹⁷
Hygroscopic APIIf the container has been open for weeks, retest water before using the factorRetesting “after exposure to air”, 211.87¹⁹
Batch assays lowInvestigate first; adding API to a failed batch is reprocessing, approved by the quality unit211.192;¹⁷ 211.115²⁰

7How FDA enforces it.

Your firm failed to revise your manufacturing formulation after replacing the (b)(4)% ethanol with a lower strength ethanol.

Aromas Para El Alma S.A., warning letter of 10 September 2024, citing 21 CFR 211.101. The firm switched a hand sanitiser and a spray to a lower-strength ethanol without changing the formula, and batches below the labelled ethanol content were released for distribution.²¹

You then adjusted the potency by adding additional APIs and re-mixing the batch. The batch was retested, and the potency assay was found within specification. However, the root cause of the low potency was not determined.

CTK OTC Laboratories LLC, formerly FMK Labs, warning letter of 19 May 2025, citing 21 CFR 211.192. A factor set in the master record before weighing is planned; topping up a batch that has already failed is not.²²

8EU GMP, PIC/S and the MHRA.

EU GMP sets the same duties in its own words: a manufacturing formula listing “all starting materials to be used, with the amount of each”, a batch record of “the quantities of each starting material actually weighed”, and an independent check of each weight, recorded.¹¹¹⁵ PIC/S, whose participating authorities include the MHRA, describes its GMP Guide and the EU’s as “practically identical”.²³²⁴ For what the batch record must hold, see what goes in a batch production record; for the certificate behind the factor, see what a certificate of analysis must contain.

9How it runs in Seal.

In Seal’s weigh and dispense workflow, potency or moisture corrections are kept with their source values and reviewed calculation, and the protocol defines rounding, tolerances and verification. The container scan, the gross, tare and net readings and the verification are recorded against the batch step, and the version used for the actual dispense is preserved rather than recalculated against a revised formulation.

In electronic batch records, calculations such as potency adjustment are versioned process logic: a changed formula is a controlled revision tested against representative cases before it takes effect, and each calculated value keeps its formula version, inputs, rounding and result. Neil can draft dispensing checks from your dispensing instructions; the manufacturing owner verifies the devices, units and exception behaviour and releases the workflow under change control.

References

  1. 121 CFR 211.101, Charge-in of components, paragraphs (a) and (c), eCFR, current to 5 November 2026. https://www.ecfr.gov/current/title-21/section-211.101 (read 8 November 2026)
  2. 2ICH Q6A Specifications, section 3.3.1(e), water content, ICH, 6 October 1999. https://database.ich.org/sites/default/files/Q6A%20Guideline.pdf (read 8 November 2026)
  3. 3ICH Q7 Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients, section 11.17, ICH, 10 November 2000. https://database.ich.org/sites/default/files/Q7%20Guideline.pdf (read 8 November 2026)
  4. 4Should Ph. Eur. CRSs be dried before use?, EDQM FAQs, European Directorate for the Quality of Medicines & HealthCare. https://faq.edqm.eu/pages/viewpage.action?pageId=1377209 (read 8 November 2026)
  5. 5What is the water/solvent content of a Ph. Eur. CRS?, EDQM FAQs, European Directorate for the Quality of Medicines & HealthCare. https://faq.edqm.eu/x/twMV (read 8 November 2026)
  6. 621 CFR 211.68(b), Automatic, mechanical, and electronic equipment, eCFR. https://www.ecfr.gov/current/title-21/section-211.68 (read 8 November 2026)
  7. 7EudraLex Volume 4, Annex 11 Computerised Systems, sections 4.7 and 6, European Commission, in operation 30 June 2011. https://health.ec.europa.eu/system/files/2016-11/annex11_01-2011_en_0.pdf (read 8 November 2026)
  8. 821 CFR 211.186, Master production and control records, eCFR. https://www.ecfr.gov/current/title-21/section-211.186 (read 8 November 2026)
  9. 9Guideline on manufacture of the finished dosage form, EMA/CHMP/QWP/245074/2015, section 4.2, European Medicines Agency, adopted 20 July 2017. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-manufacture-finished-dosage-form-revision-1_en.pdf (read 8 November 2026)
  10. 10ICH M4Q(R1), section 3.2.P.3.2 Batch Formula, ICH, 12 September 2002. https://database.ich.org/sites/default/files/M4Q_R1_Guideline.pdf (read 8 November 2026)
  11. 11EudraLex Volume 4, Chapter 4 Documentation, sections 4.17 and 4.20, European Commission, in operation 30 June 2011. https://health.ec.europa.eu/system/files/2016-11/chapter4_01-2011_en_0.pdf (read 8 November 2026)
  12. 12ICH Q8(R2) Pharmaceutical Development, section 2.2.2 Overages, ICH, August 2009. https://database.ich.org/sites/default/files/Q8%28R2%29%20Guideline.pdf (read 8 November 2026)
  13. 1321 CFR 211.22, Responsibilities of quality control unit, eCFR. https://www.ecfr.gov/current/title-21/section-211.22 (read 8 November 2026)
  14. 1421 CFR 211.84(d)(2), Testing and approval or rejection of components, eCFR. https://www.ecfr.gov/current/title-21/section-211.84 (read 8 November 2026)
  15. 15EudraLex Volume 4, Chapter 5 Production, sections 5.35 to 5.38, European Commission, in operation 1 March 2015. https://health.ec.europa.eu/document/download/4a1fdb4f-6f6f-49c4-b264-8056e5bbe078_en?filename=chapter_5.pdf (read 8 November 2026)
  16. 1621 CFR 211.188, Batch production and control records, eCFR. https://www.ecfr.gov/current/title-21/section-211.188 (read 8 November 2026)
  17. 1721 CFR 211.192, Production record review, eCFR. https://www.ecfr.gov/current/title-21/section-211.192 (read 8 November 2026)
  18. 18Compliance Program 7356.002, Drug Manufacturing Inspections, Part III, D. Production System, FDA, issued 16 September 2022, implemented 17 October 2022. https://www.fda.gov/media/75167/download (read 8 November 2026)
  19. 1921 CFR 211.87, Retesting of approved components, drug product containers, and closures, eCFR. https://www.ecfr.gov/current/title-21/section-211.87 (read 8 November 2026)
  20. 2021 CFR 211.115, Reprocessing, eCFR. https://www.ecfr.gov/current/title-21/section-211.115 (read 8 November 2026)
  21. 21Warning letter to Aromas Para El Alma S.A., MARCS-CMS 672728, FDA, 10 September 2024. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/aromas-para-el-alma-sa-672728-09102024 (read 8 November 2026)
  22. 22Warning letter to CTK OTC Laboratories LLC, formerly FMK Labs, Inc., MARCS-CMS 705479, FDA, 19 May 2025. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/ctk-otc-laboratories-llc-formerly-fmk-labs-inc-705479-05192025 (read 8 November 2026)
  23. 23Publications: GMP Guide, PIC/S, on the EU and PIC/S GMP Guides. https://picscheme.org/en/publications?tri=gmp (read 8 November 2026)
  24. 24Members, PIC/S, United Kingdom, MHRA. https://picscheme.org/en/members (read 8 November 2026)

AQuestions and answers

Is a potency adjustment an overage?

No. An overage is a deliberate excess above the label claim, for example to cover a documented manufacturing loss, and ICH Q8(R2) discourages overages that compensate for degradation. A potency adjustment adds only enough material to deliver 100% of the label claim from a lot with less than 100% content. Declare the two separately in the batch formula.

Does the excipient have to be reduced to keep the tablet weight?

Neither Part 211 nor the EMA guideline requires it. Without compensation, the batch mass and each tablet’s weight rise with the factor, so the target weight and theoretical yield change from batch to batch. Taking the excess out of a filler keeps them fixed. Either way, EMA expects the compensating ingredient, if there is one, to be named in the batch formula.

What is the difference between assay, purity and potency?

Assay is the measured content of the active substance, as a percentage on a stated basis. Purity describes what else is present: related substances, water and residual solvents. In a dispensing calculation, potency is the as-is content used to work out the weight; for antibiotics and many biologics it is activity in units per milligram, measured by bioassay. A lot can show 99.8% chromatographic purity and still be only about 95% active substance as is, because water and solvent do not appear on the chromatogram.

What happens if the API assay is above 100%?

The factor falls below 1 and the formula asks for less than the nominal quantity: a lot at 100.8% as-is needs 49.61 kg for a 50.00 kg claim. A master record can instead cap the factor at 1, which still meets the intent to provide not less than 100% of the label claim. Whichever rule you choose, write it into the master record, justify it, and test it.

How is the quantity calculated when potency is in units per milligram?

Divide the units the batch needs by the lot’s potency in units per milligram. Use the potency on the basis the certificate states: if it is on the dried basis, convert it to as-is with the loss on drying first, exactly as for a percentage assay.

Is potency adjustment required by GMP guidelines?

Part 211 does not name it, but it requires each batch to be formulated with the intent to provide not less than 100% of the labelled amount of active ingredient. FDA enforces that intent: in a 2024 warning letter it cited a firm under that rule for keeping its formula unchanged after switching to a lower-strength ethanol, which left batches below their labelled content. Weighing the nominal quantity of a lot below 100% as-is content is the same failure on a smaller scale, so adjust for potency or justify the alternative in the master record. For EU dossiers, EMA asks that factorisation be stated and justified in the batch formula.

See your process in Seal.

Bring a procedure or a recurring problem. See how your team configures it with Neil and improves the next run.

Book a demo