Helen MacDonald
👤 SpeakerAppearances Over Time
Podcast Appearances
Then you have some analytical variation in lab techniques, which they say in our kind of modern era causes very minimal variation in our tests.
But the big issue is ourselves, the test subjects and the biological variation that humans display.
And there's not very much we can do to reduce our biological variation, except essentially what they say is that we need to understand it.
And in this article,
the authors have done a really great job, I think, of trying to put together some helpful information for jobbing doctors on the test results that they see most commonly.
And it's a bit techie, but if you stick with it, there's lots of useful information in there.
And in particular, they have a kind of master table and it exists as an
interactive infographic online.
So they look at tests like bone mineral density, haemoglobin, cholesterol measures, vitamin D, TSH, HbA1c and they try and do a couple of things in this paper.
One is to answer
What is the analytical and biological variation around a single measure that you might take, like any single HbA1c result?
And then they try and understand what kind of variability do you get if you do serial measures?
If you take two measures, is the difference that you're seeing between those two tests a real difference or is that just biological variation?
And this might help you work out things like, does a person really have diabetes?
If you're measuring the response to starting a drug such as a statin and you're looking at someone's
LDL, is it accurate enough to be helpful?
And then if you alter their dose in response to having not reduced their LDL sufficiently, is that helpful measure to go on and measure again?
So I think it's just very interesting and I wanted us to hear from the authors of this paper the answers to some of these questions in their own words.
I was surprised that with these serial measures, some tests seem to vary not very much, like bone mineral density only varied about 2-5%, but some of them varied hugely.
They said TSH varied 41-50% in serial measures.