validation of different dosage strength

[archie]

Hi all...

If I have one dosage form with 3 different dosage strength, did I need to validate my analytical method one by one?
I mean, is it needed to do the validation three times because I have 3 different dosage strengh?

Or can I just do it once?

Do you know where I can find the literature about that?

Thanks a lot....

[danko]

All 3 strengths should be validated in the same validation process – anyway if your plan is to analyse these in the same analytical setups. Validation is the documentation of a method’s performance in real life.

Best Regards

[Consumer Products Guy]

We recently did a re-validation for a product for which the only difference for the new product was a reduced active level. Our "change" was to double the sample size so that we could use the same method with original sample size for samples still on OTC stability.

But QA said we needed to validate...

[cody84]

Well as part of validation you have to show precision, linearity and accuracy across a range. If the other dosage is within the range, logic would say it doesn't need to be validated. HOWEVER if you work at a place like CPG or I, your QA will tell you to validate anyways.

[cody84]

Actually to give a better suggestion for your situation, where you're starting with 3 dosages before ANY validation,
This is what I would do

For accuracy and range have your low spiked value at 80% of the lowest dose, and your 120% value would be from the highest dose. This means (if results pass your recovery and %RSD requirements) that your method can determine all three doses accurately in presence of matrix.

Next for linearity do the same thing; 50% of the low dose for lowest sample and 150% of highes dose sample, then whatever evenly in between. This is all done with standard to check the response without matrix.

Have sample prep the SAME for all doses.
I would then only do precision for the mid level dose.

[jinzhaowang]

Actually to give a better suggestion for your situation, where you're starting with 3 dosages before ANY validation,
This is what I would do

For accuracy and range have your low spiked value at 80% of the lowest dose, and your 120% value would be from the highest dose. This means (if results pass your recovery and %RSD requirements) that your method can determine all three doses accurately in presence of matrix.

Next for linearity do the same thing; 50% of the low dose for lowest sample and 150% of highes dose sample, then whatever evenly in between. This is all done with standard to check the response without matrix.

Have sample prep the SAME for all doses.
I would then only do precision for the mid level dose.

Full ACK. Another question is, in the case of specificity, should only the lowest dose be testd? What if the composition of drug products varied significantily?

[paulw]

I would make the target concentration such that all three dosage forms had to be diluted down to that concentration. That way you get by with just one linearity and just one accuracy of 80/100/120 (or 70/100/130 if you have to do content uniformity). You would then do a precision on each dosage form.

You mention the drug products composition vary significantly. If this is the case and each dosage form presents a separate matrix, each will need to have specificity done separately, and I have to amend my statement about one accuracy, you may need to repeat accuracy with each matrix to insure that there is no matrix effect on your assay.

If you have to look for related substances or quantitate impurities by % area, then you just need to make sure you can detect at those levels on your target concentration.

[prepcolumn]

If you perform assay and related substances validation, you may perform a single validation (with arranging the concentration) if your strenghts are dose proportional (i mean the strenth is proportional with formulation and tablet weight). If they are different you have to validate one by one.

If you perform dissolution validation you have to validate them seperately