Method Comparison / Method Transfer

[Sallybeetle]

Are there any rules on Method Transfer or Method Equivalency specifications when changing from a USP Microbiological quantitative assay to a HPLC quantitative assay? For example, set forth in USP / Compendial / Government / Industry?

Microbiological Assays tend to be extremely variable, which makes it hard to set a straight forward comparison specification in advance; for example, intermediate precision of HPLC replicates versus Micro replicates. The USP <1223> states that "Generally, a RSD in the 15% to 35% range would be acceptable" for the precision of a Micro assay alone. Therefore, the intermediate precision of melded HPLC and Micro results would pop an auditors eyes out (metaphorically, of course).

Could I have 1 - 3 samples analyzed by both methods and just report the results, citing the following:
1. USP <1223> quote on Micro Assay % RSD range; and,
2. "HPLC method is equivalent to, or better than, the results generated by the conventional method" (USP <1223>) as evidenced by the successful validation of the HPLC methodology.

Does anyone on Chromforum have an FDA tested procedure, or literature reference on the subject?


Your input will be GREATLY appreciated.

Thanks,
M.

[krickos]

Therefore, the intermediate precision of melded HPLC and Micro results would pop an auditors eyes out (metaphorically, of course).

Hi got some genereal input.

First I think you have misunderstood the ICH Q2 definition of intermediate precision. Intermediate precision adresses the variation within your laboratory due to different LC instrumnets, days, operator, columns of same brand, solutions etc Within one methodology NOT between methodologys.
So you typically do not meld data from two completly different methodologys, so hopefully no poping of eyes

One can of course use statistical tools like paired t-tests, F-tests etc to build ones case but seems almost superflous in this case

Second, in my experiance when it comes to method comparisions and submitting to authorities, a "head to head" comparsion of the results from 3 batchs/lots/samples seems to be common practice and enough together with a technical discussion.
If you submit changes to the Japanese market you likely need triplicate data per batch/lot/sample if data permits calculation of statistics (standard deviation etc).

Can not help you with the specifics of your analysis as it is outside my field.

[danko]

Hi Sallybeetle,

Logical thinking leads me to the following conclusion: Both the micro-bio-assay and the HPLC would convey “identicalâ€

[grzesiek]

but what is all of this worth?

if the methods are validated then you show that you can get better results in case of accuracy and precision by one method, showing the numbers, done.

[danko]

Hi grzesiek,

but what is all of this worth?

What are you referring to?

show that you can get better results in case of accuracy and precision by one method, showing the numbers, done.

Finally, if you look through the post prior to yours, do you find some things that remind you of accuracy and precision, or do you have some other conceptions of these terms?

Best Regards

[grzesiek]

"What are you referring to?" - all the work with comparing methods

As I understand the results of one method are considered to be superior (precision at lest) so pointing to the characteristics of the methods in validation protocols is in fact comparision of methods

"Finally, if you look through the post prior to yours, do you find some things that remind you of accuracy and precision, or do you have some other conceptions of these terms?" - I am aware of your post Danko, and I know nothing about the accuracy of the first method, so maybe you know more but I can say (knowing that I don't know how accurate the first method (microbiological) is) that results and so accuracies can be different, also mean says nothing about accuracy as far as I know

I'm not trying to rip off your post, please understand, I just say how I would approach the problem