Process validation - sample size

**AchkoAl** · 07-04-2012, 07:45 AM

Hello,

According the guidance - "Powder Blends and Finished Dosage Units..." we have some tendency about the sample size determination. But I'm sure that the approach has to be individual. In my opinion we have to take different sample size for different cases. If we have to evaluate powder blend of 100 kg and of 500 kg, is the number of samples taken the same? Or, if we have API less than 5% in one product and more than 50% in another one, can we evaluate both with the same approach. Please, share some experience about these issues.

Regards.

**ramyaraj725** · 07-05-2012, 03:46 AM

Dear Friend,

You query is about sample size (quantity of a single sample) or number of samples? If your requirement is sample quantity, then calculate the "quantity required for single analysis" (according to your test method) and collect the triple quantity (for duplicate analysis in case of OOS). But, if your concern is number of samples, as its a process validation study, better to take at least 3 individual samples from blender and prove the blend uniformity in different locations. Batch size and % API cannot change the above rules.

Regards
Veerraju

**AchkoAl** · 07-05-2012, 06:59 AM

Dear Veerraju,

Thank you for your answer.
My query was about number of samples that we have to take in order to make appropriate validation of blend uniformity. I'm looking for a formula, which can give me the exact number of samples depending on batch size, quantity of API in a single dosage unit and %API in a single dosage unit. Anyway, may be there isn't such a formula.

Regards