writing method validation protocols

[olivia]

Hi could someone please tell me where to begin with writing a method validation protocol for validation of assay
Thanks
Liv

[tom jupille]

Here are two places to start:

http://www.fda.gov/CDER/GUIDANCE/cmc3.pdf

http://www.ich.org/cache/compo/363-272-1.html#Q2A

[Bruce Hamilton]

Obviously, for pharmaceutical work there is the freely available, downloadable ICH Q2 guide " Validation of Analytical Procedures - text and methodology ".

It defines the process to be used, but doesn't offer much in the way of document format, and detailed procedures.

In the review section of this site, I reviewed " Validating Chromatographic Methods - a practical guide ".

That book could be a good place to start for a novice, as I'm not sure there are many free detailed templates etc. floating around the Internet.
Their may be some around for environmental methods.

However:-

- because of the high levels of nausea associated with method validations , the organisation should have allocated a suitable budget and resources before starting the process.

- The systems and resources your organisation uses for validation should exist, and be checked off ( documented SOPs, qualified instruments, trained operators, appropriate chemicals, etc etc. )

- the validation protocols MUST be decided and documented in advance, and signed off by the relevant Quality people - both within your organisation, and the client's ( if applicable ).

Please keep having fun,

Bruce Hamilton

[Consumer Products Guy]

Not to hijack the thread, but I'd like to hear estimates of how many man-hours others estimate that a methods validation takes.

We formerly wrote a report after completing the validation, figuring that we had documented the project. However in recent years our QA arm has increased its staffing, and now requires a pre-validation protocol report, although they cannot show me where in the regulations that is required. And then a finished report as well. And they wanted that in their special font, and made a special template (probably from one's former employer), but we fought that, and told them we needed to have that report available to all our R&D, and won. I wish we were staffed half as well as they are !!!

As to Liv - you'll find there are NUMEROUS gray areas, and you'll just have to use your best judgement. I personally find forced degradation in the specificity part as the hardest, and we do obtain columns of three different lots of packings if available, so there's a hardware cost as well. We automate tests such as flow rate changes, mobile phase changes, temperature changes, and use few samples with those. we try to use a second system if we have one. We're hoping our QA understands new USP <1226> about applying compendial test methods to finished products, thus requiring only verification, not validation. They claimed for over a year that they hadn't heard of <1226>, but I had - apparently they never read journals like LC-GC Magazine, where I saw that. Don't get me started....

[juddc]

Honestly, I think it took longer to write up and approve the validation protocol for the last one I did than it did to do the work. I've been developing methods for quite a while, but this was my first attempt at a full-blown validation with a written protocol and it was something of a bear.

I think our approvals process needs some work, too, as I found mistakes in the protocol that should have been caught during the approvals process.

It was a learning process for me and a good one. My next protocol will be better.

All in all, the chromatography portion of the validation took about 3 weeks at a fairly deliberate pace.

[Dan]

As Tom and Bruce said: follow ICH Q2.

But, to add to that, you should have an SOP. You could have templates as part of (or associated with) the SOP. Some things to consider and to talk about in your SOP:

• You can describe the method within the protocol. You can write the final validated method as a separate document after the validation report is complete.
  Have an 'Objective' section in the protocol.
  List all the applicable tests (linearity, accuracy, etc.) with acceptance criteria
  Acceptance criteria should be practical. Avoid complicated statistics like: t-test, f-test, etc. as they are not applicable to such small data sets.
  Be careful with 'relative difference' and 'absolute difference'. If you use relative difference (like between mean values of two anlaysts), then define what it is relative to (it could be the mean of the data or it could be ((A-B)/A)X100%).
  The acceptance criteria for the system suitability should be "proposed" criteria as you may need to change the values once you complete the validation (e.g. you may not know the appropraite value for plates before the validation).
  Have a template for the validation report also. Try to fill in the report as you complete the validation tests. This can save time and help to spot problems before you get too far along.
  If (or when) problems do arise, consider using a 'protocol amendment' to describe the problem and the new way that you will proceed with the validation. Get QA approval on the amendment before proceeding with additional validation work.
  As to how long it takes: Well, that depends on the type of method (assay, purity, dissolution, etc.) and your available resources. It typically has taken 2-3 weeks for the bench work part of it and 1-2 weeks before and after for the protocol, report and method document writing, review and approval (of course, add more time for QA reviews and that final approval step). You would need 2 people full time and one person for 2-3 days as the second analyst. The 2 full-time people can split the workload: one for the documents and notebook checking and the other for the bench work.
  
  I have validated using only two different lots of column; each analyst in the precision experiment uses a different lot of column and, if possible, a different instrument.
  
  CPG, I don't know what a "pre-validation protocol report" is. It seems to me to be a useless thing, but I am only going by its name. If this is a "method development report", then it is useful. A method development report describes your experiments, conlcusions, etc. for the method development. You will find that this document is not only useful for your analysts for future validation work but also to help down the line If/when the method develops problems when it is put into routine use. Also, the FDA and EMEA like to have a method development report; I have seen those reports included as part of the submission documents.
  
  Regards,
  Dan