Many dispensing errors start before anyone touches the weighing balance in the dispensing hood. These errors start with a certificate of analysis (CoA) that was read quickly, with the assay taken from the wrong line or the water content ignored. The CoA is the one document that connects a particular lot of raw material to the numbers in the batch record, so it is worth knowing which lines matter & how it may impact the weight of any ingredient.
Watch: Dispensing process
API Requirement Calculator, which gives the gross API to weigh after overage and assay correction.
What a CoA is, and what it is not
A CoA is a lot-specific document from the manufacturer or the supplier. It lists the tests performed on that lot, the specification for each test and the result. It is not the specification itself, and it is not general product information. Two lots of the same material from the same supplier can have different assay and water values, and each lot has its own certificate.
Most GMP frameworks also expect more than the certificate alone. A supplier CoA is normally relied on only when the supplier has been qualified, and at least an identity test is performed on the material received.
The lines that change dispensing
| Line on the CoA | What to look for | Effect on dispensing |
|---|---|---|
| Assay | The percentage, and the basis it is stated on (as-is, dried or anhydrous) | Changes how much material gives the required active |
| Water content or loss on drying | The percentage, and the method used (LOD or Karl Fischer) | Needed when assay is on a dried or anhydrous basis |
| Potency | The unit, such as µg/mg or IU/mg | Converts the label claim into the weight of material |
| Salt, base or hydrate form | Which form the label claim refers to | A conversion factor may be needed |
| Lot number | Must match the container label exactly | Ties the numbers to the material that is weighed |
| Retest or expiry date | Must be in date at the time of use | Material outside the date cannot be dispensed |
| Particle size and other physical tests | Compared with the specification | Does not change the weight, but can change processing |
Assay and the basis it is stated on
The assay says how much of the material is active ingredient. The basis matters as much as the number. An assay on the as-is basis already includes the water and other volatile matter in the material, so it can be used directly. An assay on the dried or anhydrous basis is calculated as if the water were absent, so the water content must be taken into account to know how much active is in the material you actually weigh.
Effective content as received = assay (dried basis) × (100 − water %) ÷ 100.
Practical example 1: Assay on an anhydrous basis
A batch of Metformin needs 100.0 kg of API i.e Metformin.The CoA gives an assay of 99.2% on the anhydrous basis and a water content of 1.5%.
- Effective content as received = 99.2% × 0.985 = 97.71%.
- Weight of Metformin to dispense = 100.0 ÷ 0.9771 = 102.34 kg.
If the water had been ignored and only the assay used, the weight of Metformin would have been 100.0 ÷ 0.992 = 100.81 kg, which is 1.5 kg short of active. The API Requirement Calculator divides by an assay percentage, so enter the effective content as received (97.71) when the CoA is on a dried basis.
Practical example 2: Two lots of the same material
Now let's discuss a case where lots are changed, meaning 2 lots are used and even these are form the same supplier.
| Lot | Assay (anhydrous basis) | Water | Effective content as received | Weight for 100.0 kg of active |
|---|---|---|---|---|
| Lot A | 99.2% | 1.5% | 97.71% | 102.34 kg |
| Lot B | 98.5% | 3.0% | 95.55% | 104.66 kg |
The second lot needs 2.32 kg more material for the same amount of active. If the batch record still showed the weight from the first lot, the batch would be under-dosed due to less quantity of API. This is why each lot is calculated from its own certificate.
Potency units and salt forms
Some materials, especially antibiotics, state potency in micrograms of active per milligram of material or in international units, and not as a percentage. Convert the label claim into a weight with the potency.
- For example, if a unit must contain 250 mg of activity and the potency is 950 µg/mg, the material needed per unit is 250 ÷ 0.950 = 263.16 mg. The Unit Converter helps with the unit changes.
- When the label claim refers to the free base but the material is a salt, a conversion factor is needed. With a supposed factor of 1.39, a 5 mg claim as base needs 5 × 1.39 = 6.95 mg of the salt. Hydrates need the same kind of correction. Check which form the label claim and the CoA each refer to, because they are not always the same.
A checklist before you rely on a CoA
- The lot number on the CoA is the same as on the container label and in the receiving record.
- The material name, grade and pharmacopoeia or specification are the ones in your approved specification.
- All results are within the specification, with units stated clearly.
- The assay basis is stated, and the water content is given if the basis is dried or anhydrous.
- The retest or expiry date has not been passed.
- The certificate is signed or authorised, and it is dated.
- Your own release test, at least identity, is complete and consistent with the CoA.
- If the lot differs from the previous one, the dispensing quantity has been recalculated.
Common mistakes
- Using the assay from a previous lot or from a typical value in the supplier’s brochure.
- Ignoring the water content when the assay is on a dried or anhydrous basis.
- Mixing the assay on one basis with the water content from a different method.
- Forgetting the salt or hydrate factor.
- Reading a typical or a range value in place of the actual result for the lot.
- Not recalculating when a new lot arrives, even though the numbers differ.
An unnoticed assay error is also one of the reasons a batch yield looks wrong at the end. See Yield Above 100% for how a wrong theoretical quantity shows up.
Frequently asked questions
Which number on a CoA matters most for dispensing?
The assay, together with the basis it is stated on. If the basis is dried or anhydrous, the water content matters just as much.
Can I use the average assay from several lots?
No. Use the result from the CoA of the specific lot you are weighing, because assay and water content can differ between lots.
Is a supplier’s CoA enough on its own?
Usually not. Most GMP frameworks expect a qualified supplier and at least an identity test on the material received, in addition to the certificate.
Does overage change how I read the CoA?
No. The CoA tells you the actual content of the lot. Overage is a separate, approved addition. See API Overage: How Much Is Too Much?.