Skip to content

You are reading 1 of 2 free-access articles allowed for 30 days

Not cool: The trouble with freezing blood samples

Lunchtime on Saturday I heard my wife’s car pulling into the driveway. She had looked in on her elderly mother and then called to the supermarket on the way home. I gave her a hand to carry in the groceries. My priority was lunch, hers was that I sort out what food in the shopping bags needed to be refrigerated.

We all appreciate the importance of keeping certain foods cool to stop them going off. It’s easy to take that mind-set and apply it widely, but if you do that with blood samples, it can cause problems. In the majority of cases, blood samples do not like the cold, especially when for biochemistry tests.

The following case from a few years ago, reconstructed à la CrimeCall, is an example of how this problem often surfaces. It is a true case, though not typical, simply because in many cases when we discover the problem we have difficulty making contact with the GP surgery.

Picture the scene if you will. It’s late afternoon and, as part of the normal protocol, a scientist sees a plasma potassium result has been flagged. The potassium is 6.1 (mmol/L) and needs attention. A quick look at the lab system shows the sample to be from a 53-year-old woman; let’s call her ‘patient X’.

The sample was sent from a GP (‘Dr L’) and there are no clinical details and no previous samples. The other biochemistry results on this sample are all well within the appropriate reference intervals, and in particular there is no indication of renal impairment.

Though of course we could be dealing with a true hyperkalaemia, it is more likely an artefact caused by some pre-analytical error. The instrument automatically includes a check for haemolysis — a glance at the Haemolysis Index shows that’s not the cause. Next check is for sample age, but unfortunately there’s no date on the sample. So a phone call is needed to enquire when the sample was taken and to inform of the result.

A ringing tone is heard and then, oh joy, a real person answers.

‘Dr L, M, and N’s surgery. Can I help you?’  is the friendly greeting.

‘Good afternoon. Biochemistry in the Mater here. We have a high potassium on a patient of Dr L’s and we were wondering when the sample was taken. X is the patient’s name, address 5 Hospital Avenue, age 53.”

‘I remember X’s samples. That was taken this morning all right. I think she was in around 11. Hold on and I’ll just check.’

A minute later, the friendly voice returns.

‘Yes, that sample was taken at 11 o’clock this morning.’ [Okay. Not from yesterday, so the potassium is not elevated due to the sample being too old]. ‘And it was kept in the fridge until the courier collected it.’

Eureka.

So what happens when blood for potassium measurement is stored in the cold? It’s all about balance and control. Plasma levels of potassium average about 4mmol/L. But plasma usually constitutes just a bit more than half of the volume of a blood sample. The other half is composed primarily of red cells and some white cells. Those cells have over 20 times the concentration of potassium compared to the plasma. The result is that although in vivo the movement of potassium is tightly controlled, when the blood is removed from the body, that control is soon lost. Even a small leakage from red cell to plasma can result in an appreciable rise in measured potassium. For example, in a ‘typical’ blood sample a mere 3 per cent leak raises the plasma potassium level on average by 2.0mmol/L. That could change the result from a ‘no particular worry’ level of 4.4 to a ‘do I need to take urgent action’ level of 6.4 [our reference interval is 3.6 to 5.0mmol/L]. Potentially much more serious is the scenario where a ‘needs urgent action’ potassium of 2.3 appears as a ‘no worry’ result of 4.3.

The effect of haemolysis in falsely raising measured plasma and serum potassium levels is well known to doctors and nurses. Fortunately, modern instruments flag that on all results (or block results if severe haemolysis).

So how come we can’t detect the problem caused by refrigeration? This is because of the way cold affects the blood cell — it’s not a straightforward leakage, but is relatively specific. What keeps potassium in the cell is the energy-dependent sodium and potassium pump. In the cold, this Na/K ATPase doesn’t work properly so potassium leaks out, but as there is no haemolysis there is no visible warning. All we see is a potassium result higher than the true level (which may, or may not, be recognised). A similar situation applies when a sample is left overnight. Na/K-ATPase activity ceases, though in this case through exhaustion of glucose, and an artefactually-raised potassium is the result.

Does all this matter, or are we just being overly pedantic? If a patient doesn’t receive treatment just because cooling the sample masked a seriously low plasma potassium, then it matters. Even if no critical outcome results, if you as a GP have to recall a patient for re-testing when an elevated potassium results from refrigerating the sample, then it matters, to you and the patient.

So just remember that refrigeration is great for keeping the milk fresh, but if you want correct potassium results on your patient, then please keep their sample at room temperature.

Leave a Comment

You must be logged in to post a comment.

Scroll To Top