Chromatography Forum

Hi!

What kind of data would be necessary to declare equivalency between two methods? For instance, to declare a given analytical HPLC method is equivalent to a USP monograph, what sort of information would need to be generated? What sort of statistical work-up would be required?

Many thanks,

Putnam

Putnam,
Typically we do a n=6 assay using both methods. The criteria are not more than 3.0% difference between the means, and a pooled SD of not more than 3% for an assay. The criteria are a little looser for a dissolution.

This could be infinitely more difficult if you were trying to declare equivalency for a related substances / impurity method.

We do assay/rel subs. Our criteria for assay is n=9 with an RSD of 2%; related subs is RSD of 10%.

Regarding assays in finished product, wouldn't a validated house method be considered equivalent to the USP method by the very fact that it is validated? A validated method, whether it is a USP or a house method, is shown to be accurate and precise.

For impurities, though, the situation is different. I do not think that they need to be equivalent. If the house method quantitates a greater # and amount of impurites than that OK too. So some kind of correlation study is needed.

Equivalency between two assay methods is quite straight forward by looking at overall RSD between two methods.

For Impurity estimation , one has to prove that in-house method is equivalent or better than the offical method in terms of resolution , no. of impurities , sensitivity etc. It is done by preparing a study report of comparative data of one representative sample using both methods and prove that in-house method is better.if method is equivalent , it is safe to use official methods for regulated markets.

JM

Even if you want to bring a USP method in house, you need to do a technology transfer. The USP strongly suggests this.

How about replacing a USP TLC method with a validated house method (for impurities)? Does a comparison study even have to be done considering that the FDA has begun disapproving impurity limit tests altogether? - they want to see hard numbers!