Eleni Linos

You know, we always think about the whole concept of lab error, where, you know, the samples get mixed up or, you know, it's run through the wrong machine and all that.

And that certainly does happen.

But with the standards that are put in place, I mean, it's pretty unusual.

I mean, it's less than 1%.

And that's from data from 10 years ago.

I'm sure it's even better now.

There's two other types of variation or things that can go on to influence a number that you get back.

And one of them is the analytic variation, which is just how good that test is at

measuring something.

One of the best examples I try to think of as far as analytic variation is if you were to try and measure someone with a one foot ruler and, you know, you put it against the person and then moved it up and measured again, every time you measured the height of that person, their height would change.

But, you know, common sense would dictate that person's height hasn't changed.

So that's the analytic variation.

What the good news is, is that the analytic variation is actually very well controlled.

And in fact, it's probably small enough that we don't need to worry about it for the vast majority of tests.

Where the biggest issue around variation is that biologic variation, the natural fluctuation of things that go on in a body, the sort of noise that we can't eliminate due to the normal physiological processes.

I mean, for instance, things like, you know, your sodium level doesn't fluctuate a huge amount because there's sort of homeostatic mechanisms that deal with that.

But for instance, something like vitamin D, the variation in that, it can go up, you know, relatively day to day, month to month.

There's some seasonal variation.

And so it's that type of thing that goes on.

one of the problems with this variation, as you can imagine, is when we have these arbitrary thresholds for diagnosis.