A few days ago I mentioned that I need to look at ways in which I can seek to improve the quality of the service we offer. Here's something inspired by a little episode in the lab this afternoon. It might come to nothing, but the intention is good...
- On the one hand having a blood sample taken isn't something that is done lightly. It hurts.
the other hand providing poor quality blood test results can adversely affect
patient treatment... and that hurts too.
It is possible to obtain blood count results from under-filled sample bottles. In fact the manufacturers of most anaylsers pride themselves on just how small a sample can be run. But should we run them? The EDTA anticoagulant in the bottles causes swelling of red cells. In a correctly filled blood sample bottle this swelling is to an expected uniform amount (as we have a precise amount of blood and a precise amount of anticoagulant) and therefore isn't a pre-analytical variable of importance. However in an underfilled sample...
many times have you looked at cumulative blood count results and seen that
someone's MCV which has been constant for over a week suddenly jump up by ten
per cent; only to find an under-filled blood bottle?
Does this sudden rise in MCV matter? Obviously it does. This issue could lead to a case of iron deficiency or a case of thalassaemia being missed.
what do we do? Reject all blood count samples that are underfilled? Of course not.
The haemoglobin level and cell counts in the sample are unaffected. It is just
the MCV which rises (and consequently the parameters calculated from that).
So given an under-filled sample we just report the haemoglobin and cell counts
on underfilled bottles and report the rest as "unable to provide a full
set of results due to an under-filled sample having been received"...
But following a difference of opinion this afternoon on whether or not the amount of blood in a sample was low enough to warrant being called under-filled, this begs the question of precisely how under-filled is "under-filled"?
Back in the day we would have got scientific at this point. We really would have got out a needle and syringe, exsanguinated the trainees, created a range of blood samples from "properly filled" to "barely a drop of blood in the bottle" and had a range of samples filled to half-millilitre intervals in between, and seen for ourselves how the MCV changes. With loads of sets of data (from loads of different exsanguinated trainees) we could have determined at which point the EDTA-induced swelling of red cells becomes significant.
But nowadays we (sadly) can't just stick needles into trainees. We need to get formal ethical approval from all sorts of committees before we can even consider this sort of experiment... (for which the trainees are probably grateful)
I've suggested to the boss that she might like to investigate how we get this approval. We'll not tell the trainees about this just yet... They tend to run and hide when this sort of thing happens. I know I used to...