Validation of processes for production & service provision - Service industry

[Marc]

I dont mind emails from folks.
ken

I can understand that, but the point of this forum is to share publicly. That helps LOTS of people instead of just one or a few.

It also makes it look like you may be wanting to build up an email list.

 

[FDAINFO]

Re: Validation Times article on new FDA validation guidelines

Hello, all:

At the moderator's request, I have uploaded the January issue of Validation Times, which includes two articles (based on a recent ISPE Webinar) on the new guidelines.

So, let me know if I can be of further help.

Ken Reid
Editor

 

[Marc]

Re: Validation Times article on new FDA validation guidelines

We very much appreciate your sharing!

 

[QAnalyst1]

Re: Service industry - Validation of processes for production & service provision

Colin: I'm not saying what you've said is wrong, but the steps you've outlined here...isn't that just business as usual? I don't understand the reasoning behind the requirement to "validate a process" because when I think of what that entails, to me, it's what you SHOULD be doing in the first place, before you try to make a product anyway.

"What the clause requires is that you prevent problems as much as you can by ensuring that the process is appropriate and that the people are competent and that you provide any necessary instructions - all of which is designed to reduce the risk of the process going wrong."

Where am I going wrong here? Shouldn't the above be done as a matter of course, especially for an ISO certified company??? If you've done due diligence with planning in the first place, what need is there for validation of any of your processes, whether it's something you can't reasonably test or not? It's almost like it's saying an ISO certified company can get away with not controlling processes...otherwise, what need is there for "process validation" if the whole point of ISO, in large part, is to ensure a company has proper control and oversight of it's processes in the first place?

 

[Colin]

I couldn't disagree with what you say, every organisation should ensure that they do these things but in ISO 9001 they are not necessarily mandatory as long as the output can be verified. That is why 7.5.2 is there to deal with the situations where the output of the process cannot be verified.

 

[QAnalyst1]

Thanks so much for the reply. I read a lot of your explanations on here before posting and that's why I replied to you. I especially liked your replies and explanations above others I read.

I worked for an ISO certified company years ago and we documented everything...how to conduct training, qualification of operators, how to operate every piece of equipment, virtually every inspection and test, how to use inspection and measuring tools, etc. We literally had HUNDREDS of work instructions and forms, top to bottom. This was before ISO 2008...subsequently, we were awarded certification after our initial assessment audit with no follow up required because of the extent to which we documented our processes.

With 2008 then, this level of documentation is not necessary as long as the output can be "monitored and measured". If so, this represents a paradigm shift in thinking for me, and will take some getting used to, because I want to dive in and document everything, top to bottom, regardless. I remember a "special process" clause, but since we documented everything so thoroughly, I guess we had that covered because the only exemption we claimed was for Design.

So let me ask you this then:

1. If from one step of the process to the next you can't verify the output, but you do 100% inspection of the finished product pass/fail and then rework if possible, the process is still considered as needing validation, correct?

2. If by 100% inspection, you're only inspecting the product for flaws that occur during the manufacturing process, that's not necessarily testing it to customer specifications especially if there is no way for you to reasonably recreate the environment the product will be subjected to by the end user, or if there is any subsequent assembly it may be subjected to by the customer, such as software needed to control the component, or further processing by the customer, as in the customer adding your component to their assembly.

The process in the above scenario then needs validation, and that could involve a trip to the customers facility, or actively seeking feedback from the customer, (which you'll certainly get if something fails), correct?

Any guidance you can give me is greatly appreciated.

 

[Colin]

Thanks for the kind feedback and sorry for the delay in replying, I am really busy with work at the moment. The key point is in the wording of clause 7.5.2 ..."where the resulting output cannot be verified by subsequent monitoring or measurement"...

In other words, you are unable to tell if the process has produced conforming product or not - so you can't do inspection on it.

The usual ones are welding, painting, adhesives, etc. The one I usually use on courses is sticking 2 pieces of paper together, you can't be certain how well they have been stuck unless you tear them apart - destructive testing.

So in that case I would experiment with different adhesives, at different temperatures, humidity's, times for curing, amount of pressure etc until I found the right combination. I would then document the process and make sure that the person operating the process was completely competent to do the work.

There are still no guarantees that it will work correctly but at least I am giving myself a chance.

Hope this helps.

 

[QAnalyst1]

Yes it does help, thank you.