%Recovery Validation Criteria

[Pear]

Hi,

I am working on a few new methods to quantify processing residuals in final products. Things like solvents etc.

Normally, we work within the FDA specification (ORA Lab Manual Vol. II - Methods, Method Verification and Validation (ORA-
LAB.5.4.5)) of 97-103% for APIs and 95-105% for final dosage forms. But a residual solvent is neither an API or final dosage form.

I'm seeing a general acceptance criteria of 80-120% for accuracy %recovery during method validation, but I can't seem to find a concrete source for this. I'd like to not have to develop new methods to a stricter requirement than necessary, but without a reliable source I don't see how I can justify the 80-120% range.

Would really appreciate if anybody with more experience/knowledge in this area could provide some input please.

[Consumer Products Guy]

Normally, we work within the FDA specification (ORA Lab Manual Vol. II - Methods, Method Verification and Validation (ORA-
LAB.5.4.5)) of 97-103% for APIs and 95-105% for final dosage forms. But a residual solvent is neither an API or final dosage form.

The company I worked for did not make APIs; we bought APIs and put into a final dosage product. Your post reminds me of the time I had to ARGUE from same ORA source with our QA Director to get him to accept the 95-105% range (he wanted 98-102% which we almost always met, but I felt for our topical products it was ridiculous to have tighter standards than the FDA).
********************

But a residual solvent is neither an API or final dosage form.

I'm seeing a general acceptance criteria of 80-120% for accuracy %recovery during method validation, but I can't seem to find a concrete source for this. I'd like to not have to develop new methods to a stricter requirement than necessary, but without a reliable source I don't see how I can justify the 80-120% range.

Would really appreciate if anybody with more experience/knowledge in this area could provide some input please.

Another time I had a method for an "impurity" that I wanted to do a validation for; the best I could find was from Australian Veterinary publication, title page and page 3 attached below. This showed different %Recovery guidelines depending on the level of such impurity. There may be newer information out there. Our QA Director looked down on this publication because it was Veterinary; there were other, and multiple reasons I believed this QA Director was not competent.


****************

[DR]

I do believe that the specs for residuals that are generally in use come from the USP (or whatever compendium your neck of the woods subscribes to). I'd check the appropriate monograph in the appropriate compendium. Failing that, go for the appropriate general chapter for the test(s) you're running.

[Consumer Products Guy]

I do believe that the specs for residuals that are generally in use come from the USP (or whatever compendium your neck of the woods subscribes to).

At least back then: we didn't find USP to be helpful for this.

[DR]

I can see where that would be the case.
Sometimes, any guidance provided, even from sources like the USP is better considered to be a starting point as any auditor worth their salt will require justification backed by in-house data if they see something they don't have experience with and consider to be a little fishy.
No matter what spec is landed on, you need to know what your accuracy, reproducibility, sensitivity (LOD, LOQ) and intermediate precision look like for residuals, cleanouts and impurity methods.