Impurity Percent Recovery (API instead of known)

[jayroseville58]

We are performing validation method for related substance (known and unknown). Part of the method is the percent recovery studies. Can we just simply perform the percent recovery of known impurity but using the API (at 100% known impurity specs level)instead since we are dealing with unavailability of the former? Thanks.

[gtma]

Quite often I see the API is use as a surrogate for the unknown. However, I would consider using a risk-based approach where the know impurities with similar UV spectrum/RRF to the API be use as a surrogate. This minimize the validation work. Similar rationale can be used for other impurities. Has anyone used this approach?