Posted by Sten Westgard, MS
- Do we have the right method?
- Do we have the right QC rules for that method?
- Do we have the right controls?
Labs often assume they've got all of the right stuff. It's humbling to examine studies or visit labs and determine that, alas, they have the wrong method, and they're doing the wrong QC on it, and they have the wrong controls, which only makes the problems worse.
The field frequently debates the relative importance of the phases of the total testing process: pre-analytical, analytical, and post-analytical. The pre-analytical phase is in its zenith right now, everyone concentrates on reducing errors in that phase, but on the whole, we still agree that each phase is part of a three-legged stool. If one of those legs is broken, the whole chair falls down. So labs cannot neglect any of the phases.
With methods, QC practices, and controls, it's another three-legged stool. Ideally, we need the right method, right rules, and right controls. If the method falls short, the right rules and controls can help a little bit - at least they can tell you you're tipping over. But one of the real problems we're beginning to see is control-induced outliers and lab-induced outliers. These are out-of-control results that aren't real; they're just ghosts generated by poor practices and/or poor controls.
I know, as if it wasn't hard enough working the laboratory already. Now I'm going to be warning you about the quality of your controls, too.
Sorry for the short notice about this workshop. But I hope to see you there.