Chromatography Forum

Dear fellow chromatographers,

We are currently using a software package coupled to two hardware set- ups ( X and Y) to analyse our products. The method and equipment are validated in their current set-up.

We want to change to another software package and one hardware set-up( Y). We will still be using the same analytical method and column.
What should we validate to comply ( GMP)?

regards

Philippe

You will need to define a set of requirements for what you need the software to do and what you need the hardware to do. You then need to test against each of those requirements to ensure they are met.

If you have already validated the operational ranges of the hardware, you could just get away with testing the software to ensure it can control the hardware, e.g if you change a parameter in the software, does the hardware respond. However, you still need to validate the other functions of the hardware - data acquisition, storage, processing and reporting.

The actual methods you run on the software/hardware combination is a relatively minor component, more to do with "configuration" of the software to control the hardware, then acquire, store, process and report data.

However, if you know you will only ever run ONE method on this hardware/software, you could just write your requirements and perform testing to cover that situation. This is not very likely though and means you have to go through potentially a more lengthy process if you decide to put more methods on later.

Tim,
Thanks for your reply.
Your items you mentioned are indeed incorparated in our validation plan.

But the item which I am dealing with is to confince the management that the results are comparable if I use the new software ( datahandling: integration settings).
What woud you suggest as being comparable ?
Mean value and the %CV ? LOD, LOQ comparison ?

Philippe

I've just checked and the way we compared between the CDSs was to run the exact same samples, with same mobile phase/column, etc. on both systems. If you can also use the same HPLC, that is even better, because then only the CDS will have changed. We were fortunate when we changed CDS in that the old system just used an A/D convertor on the analogue output of the detector, while the new one used a digital output (it was ChemStation), so we got EXACTLY the same chromatography going into both systems.

Using the response factors from CDS1 for the two samples generated, we calculated +/-2% limits and verified that the response factors (Amount) generated by CDS2 were within these limits for the respective samples, for the peaks of interest.

For the Efficiency, Tailing Factor and Resolution for the first standard and first sample generated by CDS1, we calculated +/-5% limits for each peak of interest. We then verified that the values generated by CDS2 lay within these limits.

We also compared the output from both systems against the standard acceptance criteria for the method, e.g. Efficiency greater than 2000, RSD <=2%, etc.