In my lab we reloaded the millennium software as the administrator password was lost.
Does this calls for IQ/OQ/PQ of the software. If so, what should be its content.

Thanks,

Yes. But I am unclear how your lab could be cGMP if you did not have any controls on this. Sounds like you should do some other qualification to ensure data recovery.

There is a Auto Qualification for Millenium, when you run this program, you just have to check everything and you IQ/OQ is done.

I agree with anon 2 - sounds like your systems are out of control if this could happen.
If you are using Millenium32 with Oracle database and you lost the admin password does this mean you reinstalled the database as well? What happened to the data in it?

Are you a GxP lab? I truly expect the answer to be NO. If it is YES, then I suggest you need some serious training and probably support from an external company specialising in bringing labs up to standard. The FDA would have a field day in your lab.

In the previous posting I should have said I agree with Anon 1.

Anon2 is looking at things in a very simplistic manner if he thinks that running the auto qual is all that is needed. It is useful I agree, but have you validated the auto qual software? I suggest you take a long hard look at GAMP4 for guidance.

Anon 3: I guess that Waters has validated the auto qual software, they used it for our validation of our systems, so why not using it after they have used it?

From what I've seen from FDA, the Waters validation doesn't mean a thing. It's up to the user to completely validate every aspect of every function of every software/hardware that you use.

From Anon 3:-

I agree with Anon 4 - The level of validation that you perform is often down to judgement (scientific not opinion) based on your confidence in the vendor plus many other factors. Properly conducted vendor audits are also very valuable - again I refer you to GAMP4. You may be suprised how poorly some well known software and hardware vendors perform in an audit.

Validation is a strange beast - on a sliding scale you could do very little validation and be in production quickly (but at high risk) or you can do everything to the nth degree, experience little risk but never actually produce anything. The trick is to get the balance right - aim for acceptable risk at the lowest possible cost in terms of money, resource, time and lost production etc. If you feel that you can stand up in front of an FDA inspector and scientifically justify your position, then that is a good position to be in.

Make sure you document every decision you make. Don`t forget, if you generate faulty GMP data due to a software problem, from the FDA`s perspective you are responsible and not the software vendor.

This posting should not be considered as an exhaustive discussion of this subject - there are many factors to consider and I suggest you do some reading on the subject, as opinions do differ.

This posting is also not a comment on the quality of Waters validation.