Posted by Sten Westgard, MS
We recently got the following question by email:"1. Do we need to perform method validation when lab replace a current analyzer by the new one with the same model that being used? 2. Could we just do correlation study for this situation?"
In theory, if you replace your Instrument A with another Instrument A, same brand, same model, they should be identical in performance, right? However, in the real world we know that not even two instruments of the same model are exactly alike in performance.
Dr. Westgard answers, without quoting the chapter-and-verse of the various regulations:Just to reinforce that general advice, we also asked Nancy Yeransian, MT(ASCP)SC, Manager of Accreditation Services for CAP's Laboratory Accreditation Program. Here is her response:"Good Laboratory Practice would be to perform an initial, short, method validation study to assure comparability of results. This should include a quick comparison of methods study and verification of control results. Then you could transfer control ranges for a short period before updating to new control limits from data on new instrument."
"A replacement instrument is treated like a new instrument from a method performance view. The user has to ensure that the new instrument performs as expected in their laboratory. The Method performance Specification section of the Laboratory General Checklist would apply.
This means, accuracy, precision, reportable range, sensitivity, etc all need to be verified. Typically patient reference intervals/normal ranges would not need to be redone since the same methodology is in use. Just like the same make and model of a car. You can get a good one or a lemon and you need to verify performance."
It would be nice to assume everything is the same, but unfortunately, the reality is that there are differences in performance. We need to find out how big those differences are - and make sure they aren't clinically significant. As the saying goes, we we assume, we make an.... out of you and me. Except when the laboratory makes assumptions, the patient is the one who lives with the consequences.
Update: we also got some advice from Megan Sawchuk, Associate Director of the Standards Interpretation Group from the Joint Commission:
"For a replacement instrument, it is a combination of meeting the requirements and using reason. We usually recommend an abbreviated approach rather performing than the full verification the laboratory may have performed when the instrument was first installed. For example, performing several runs of QC on both the new and the old method could readily provide the data needed to verify the specifications (accuracy, precision, reportable range), establish or verify the QC ranges, and provide correlation information to verify the previously established normal ranges are still applicable."
This shows that the requirements will vary depending on your regulation and accreditation environment.
