Chromatography Forum

Is anyone out there in a GxP environment NOT using validated software to calculate system suitability parameters? We don't have the costly Waters Empower Syst. Suit. add-on, so we are doing all the calculations with Excel.

Which, is, of course, an audit NC waiting to happen since we haven't validated the spreadsheet.

Any other (affordable) options besides good old pen and paper?

What if you validate, lock, and password protect the spreadsheet?

Depends on what calculations you are dong but if you design a spreadsheet and lock the cells containing the calculation formulae leaving only data entry cells accessible . Validation can be done by comparing a series of data sets calculated manually Vs spreadsheet values.

in a GxP environment, you may also need a SOP for the spreadsheet operation, change control, version control and performance qualification.

Once the validation is completed the spreadsheet can be used routinely and you only need to QC the data entry. Annual PQ can be performed by repeating the validation test data sets and comparing them against the original manual values.

before you start, consult your QA and get their input on what they feel is necessary and try to keep them happy. After all to need them to sign off on the approval.

I've worked in GLP for over 20 years and have used excel for the bulk of that time to calculate system suitability. I've not had an issue with and QA units I've worked with.

I've used excel for calculating results in a GMP environment. Although we did validate the spreadsheets and never heard any negative feedback during FDA audits. I think the key is validating and locking the calculations. The data can be easily exported into excel and copy/pasted into the spreadsheets.

If you some additional other details, Ludwig Huber over at
www.labcompliance.com offers a range of training options for using Excel in regulated laboratory environments at quite reasonable prices.

The biggest hurdle may be obtaining your organisation's QA/Regulatory Affairs approval. However, careful planning can bring them onside.

The following has worked for me previously...

Firstly, email QA/RA , cc your senior manager, with details of the existing non-compliance - along with the price of the expensive modules. Note that's it's now their problem, and you're happy to discuss options, but you want it fixed by month's end.

Secondly, advise your manager that there are cheaper options, but then need investigating to ensure they don't compromise efficiency, and that you'll probably require a more powerful computer - or some other toy you want to upgrade.

Wait, then at month's end, provide the purchase order for items to your manager for signing, noting that QA/RA haven't responded.

Please keep having fun,

Bruce Hamilton.

The major concern w/ use of spreadsheets is that you lack source code (and can't get it from MS). This is why you need to check, lock, control your Excel templates. As long as you have adequately documented and followed procedures in place to govern this, OK.

Now - as to what you feed your XL sheets - area slices from exported raw data files, or numbers that reflect hand measured peak dimensions?

(The answer to this will lead to more questions.)

If you want the source code, youd could use OpenOffice Calc instead of Excel. But that doesn't solve the problem of validation.

A picky auditor might even find that password-protecting Excel sheets is not enough - there are several tools to open a password protected sheet without knowing the password.

What kind of system suitability parameters are you talking about?
RSDs of retention times and areas can be calculated in normal tables. Resolution, S/N, number of theoretical plates should be computed by any chromatography software. As far as I understood, the SST option allows to predefine criteria, stops the batch if these criteria are violated and sends you an SMS.

Alex

hi Alex,

Surprisingly enough, the Empower version that we have cannot calculate resolution, tailing factors, S/N, etc. We literally have to print out the chromatogram and measure the peaks with a ruler ..

The add-in which provides this functionality costs the equivalent of 6 months' salary for a lab tech in my country

We may be missing something, though... please let me know if there is an easier way (i.e. by predefining a formula) ?

Thanks everyone for your suggestions, will certainly give validating the spreadsheet a shot!

More accurate than a ruler... if you can see x-y coordinates (time, mV or µV) while reviewing suitability chromatograms, you can make notes as to the Y values at the top and baseline, then calculate your percentage needed for your peak tailing calculation (typically 5%) and determine the X values at that height on the same review screen.

If you can't get running X-Y values within your software, you can do a screen dump, pull it into MS-Paint and use the X-Y coordinates within that program's status bar to get your coordinates. This was covered in a really old "Data File" column in LC-GC (Feb. "91 or '92?).

{or invite the boss to pay for the software upgrade/option - nobody ever told him chromatography is cheap}