How long for new gcms setup through method creation, etc

[Anonycms]

Hi all,

We've recently purchased a brand new agilent gcms setup.

We need to carry out in house testing of liquid to confirm nicotine content to GMP standards.

A consultant whom we may hire to come in and use the kit has quoted what we feel is a very long time to simply get "set up".

We already have the kit which has been installed by agilent and the relevant standards from sigma Aldrich and the correct column.

We don't have anything else such as method, validation and documentation - all of which needs to be GMP/GLP.

How long do you all think this should take 1 person working full time to do before they can issue/generate CoA's?

Not sure if it helps but a dual sim/scan will be run (specifically looking for nicotine then a generic scan looking at peaks for other major ingredients).

Many thanks!

[Peter Apps]

How long is a piece of string ? - there are too may variables that nobody here knows about. If you think that the consultant is padding his time, get another estimate from another consultant.

Peter

[Consumer Products Guy]

We need to carry out in house testing of liquid to confirm nicotine content to GMP standards.

We don't have anything else such as method, validation and documentation - all of which needs to be GMP/GLP.

I guess I'm curious - because with the structure of nicotine and the levels in a finished liquid (not smoke, environmental, trace, etc. - why HPLC assay using a UV detector was not chosen for this assay???

It sounds like you have to validate an assay procedure for your own finished product, assume you already have an established procedure for this. Either way, that will take some time, read weeks or more.

[James_Ball]

It is faster if you have a reference method to work from, but starting from scratch can take a while. You have to optimize the instrument then optimize any sample preparation steps, then calibrate and do a study to show that the method works and see how any deviation from that method affects the results.

[Don_Hilton]

Having developed many methods -- do not expect anything in less than 6 - 10 weeks. And that's without seeing what kind of range of samples you have, precision required, or any other details. Just the fact there are samples to be weighed or pipetted, internal standards to be added, solutions to be made, dilutions made, chromatography to be adjusted, integrations to be checked, data flow to be established for reporting, a validation study, and a formally written method of analysis - it takes time. And that 6 weeks assumes that everything is "easy" and you are starting with a method from someone else that you basically need to set up, check out, and write up. With the work requried to validate for GLP/GMP, this can easily stretch in time. I can't tell you how long the piece of string is - but I can tell you it is not short.

[lmh]

It's so good to see sanity about how long jobs take. In the environment where I work, I find it aggravating that when a lead scientist asks their own technician to set up a new method, it will probably be on the basis "let's have a progress meeting this time next week", whereas if they ask me to do the job, if I quote them for more than 4 hours of my time, it's almost certain they'll drop the job as too expensive. Even setting up a well-described literature method and checking it "works" is a tall order in half a day. People look at time and money quite differently.

If you bring in a consultant to set up a method, of course they've got to allow for things going wrong, and they can't cut corners on developing the method to a certain standard of specificity, repeatability, robustness and sensitivity - if they do, then when the method fails in the future, their reputation goes with the method. It won't come cheap. Peter Apps is right: if you think they're over-charging, get a couple more quotes, and ask what the quote includes.

To put consultant fees in context: I find that responding to a typical request for help on a new compound takes me getting on for an hour by the time I've done some literature searching to see how easy/difficult it will be, checked the appropriate technology etc.; and checked that standards are available, estimated how much time I need for a quick look-see, and composed a suitable reply describing what I can do, and asked future client what sort of number of samples/types of samples/levels of analyte they expect. If one in ten jobs actually proceeds to me doing some real work, then theoretically I have to add 10 hours to every single job just to cover the admin time of administering enquiries that don't proceed to real work...

[Anonycms]

Thank you for all the replies.

I'm not expecting it to happen in a few days, in my head I expected 6-12 weeks but have been quoted 6 months which seems excessive.

[adeputy]

There is one area which you have missed which is software validation. Regulatory agencies require that GMP/GLP results are generated on a qualified instrument using validated software. Between all the documentation and running of the test cases, a software validation can take a minimum of two months to complete.

[Consumer Products Guy]

I was asked this morning if we could identify which dye was used in a finished consumer product, may have legal implications.

I can't predict if we can do this, or how long it might take.

We have history here of quantitating exactly one dye by HPLC, the only one we've ever tried, about 15 years ago. Most of these dyes are sulfonic acid salts containing benzene rings, and we did that other one using a DAD detector with its visible lamp.

[Don_Hilton]

Depending on the complexity of the mixture and what you want to get out of the general profile, six months is not unreasonable. It brings to mind a project that I had several years ago to set up a profile of a number of flavorings as a QC method. We had several people on the project - and as I recall, that project took 6 months or so. While we had a number of flavorings, which added work, our validation was light, which saved work...