Gamp & Fda

Hi Folks,

Can anybody give me any pointers on where I can get a general overview as to what GAMP is all about in relation to PLC control...ditto for FDA? Any links/pointers would be much appreciated.

Thanks a million,

John

The only GAMP I know of is related to the pharmacetical industry:
http://www.gamp.org/

FDA to me is Food and Drug Administration (US)

This is actually a pretty basic question when dealing with pharaceutical companies. Is this for homework, or a job interview?

GAMP is Good Automation Manufacturing Practice, and is 'just' a methodology for

• Know what you want the system to do
• Write the code to do what you said, and
• Prove that it does it.

It's pretty common sense stuff, but requires a whole lot of documentation.

The FDA wants (insists, actually) that pharmaceuticals follow GAMP to ensure that the process that makes the drugs we take produce the same drug from batch to batch, with 100.000% repeatability and reliability.

The process of documentation and testing is called Validation. It is both a royal pain in the southern end of the spine, and a neccessity ("But it's lousy logic - can't I just change that one line of code?" "Not without first writing up Why it's wrong, proposing the solution, describing how the change will impact other parts of the code, writing the test for the change (how will the change make things "better"), and getting all that paperwork signed by at least 5 people (all of whom are always in meetings)).

Check GAMP.ORG

Allen,

You must have a huge amount of patients, all that to change one rung of code?

I have patience, not patients

Once a system has been validated - YES, all that, just to change one rung of code. Even on a packaging line (which doesn't 'necessarily' change the quality of the product.

And the FDA doesn't kid around about it, either. They have this way of imposing big fines (or just shutting a line down) if the rules aren't followed.

If you wondered sometimes about the high cost of drugs, this is part of the answer (but only part. Add in the delay between the granting of the patent (which lasts for 14 years) and the time it takes for the testing (first on cells, then animals, then people), and only then can you actually build the line (and if that's the paperwork for a change, imagine what it's like for a whole batch system - I'm working on the documentation now for one "simple" system). Plus it may take 10 researchers to find one new drug (but they all have to be paid).

And so on. It's not "just a big rip-off". Although they do make some pretty good money on it.

Whoops, long night last night, you knew what I meant.

Sounds pretty interesting none the less, bet there is a lot of legal issues as far as who is responsible for what. The lawyers make out pretty good too I would imagine.