Method validation problem, help needed

[along]

Dear all,
Do you think I need to redo all my validation studies for the case described below:-

Our laboratory developed a set of clean up procedures for the analysis of pesticide residues in vegetable oil matrix.
According to original method, the final sample is analysed with GC-PFPD, and few parameters were validated i.e. linearility, recovery, RSD%.

Recently, the detector was down and I switched the analysis to another GC-FPD (Caution: detector was different), I built a new calibration curve and found that the linearity was still there within the analytical range.

My question is:
1. Do I need to redo the recovery study?

I understand that the LOD and LOQ may be different in that two system, however the samples I am working on were roughly 4X higher than the lowest calibration concentration.

In my opinion, I think the changing of detector has no direct effect on the clean up procedures performance (correct me if I am wrong).

Can anyone submit your suggestions and views, thank you.

If any one of you agreed to my point of view, can you tell me where can I find the proof (journat, discussion paper...etc).

Thank you very much

[Bruce Hamilton]

I think this is the wrong group to ask.

You need to ask the quality section of your organisation, the clients, and/or the relevant regulatory agency.

If I were in your situation, I would revalidate, just for my own peace of mind. Changing to a different detector is a significant modification that should be clearly documented, including a simple risk analysis to justify any decisions.

Please keep having fun,

Bruce Hamilton