Posted by Sten Westgard, MS
Over at Advance for Medical Laboratory Personnel, they've posted an online article on Real-Time Quality. It's a quick overview of quality control techniques that give you Real-Time information, including "Westgard Rules" (really what they mean is statistical QC), and "delta checks" like anion gap, BUN/creatinine ratio, and MCHC.
One interesting excerpt from the article:
That's a perfectly valid statement. For labs that feed garbage into control chart calculations, they will get garbage out. Statistical QC is not a magic wand. Also correct is the idea that statistical QC can only make interpretations on the data that's observed, not on data that isn't observed. Errors could occur between measurements that might not be detected. However, we should note that most labs are not considering getting more real-time with more frequent controls. The trends are in the other direction. Statistical QC once a day is about as real-time as most labs get.
What's even more interesting about this article is this new distinction of "real time" quality. Maybe I'm reading too much into the label. Statistical QC used to be basically the only kind of QC that was done. It was just called QC. Now we've added "real time" to the front end, perhaps to distinguish this type from other emerging types of QC. I suppose "Equivalent QC" is not real-time because the QC is reduced to weekly or monthly. So in contrast to real-time quality with statistical QC, "Equivalent QC" could now be called "too-late-time" quality.
Comments