We have had three occurrences of air being spotted down steam of the pump (Graseby 500) towards the end of blood infusions recently. The pumps are being used by experienced, trained staff who have had competency based training on the product.
The staff ensured that the air was not injected however their confidence has been knocked.

The bubbles have been between 3 and 12mm long in a standard blood set (not I believe enough to be dangerous but still worrying) with one infusion having 4 separate bubbles.

The giving sets on close inspection show many small bubbles above the pump typically smaller than 0.5mm diameter and the same below the pump but up to 2mm diameter; these bubbles are not apparent when the infusion starts. Watching the small bubbles during an infusion showed them slowly getting larger but not moving down the set. (sizes of bubbles at end of infusion)

The infusions are typically of around 250ml infused over two hours twenty minutes, so infusion rates around 100ml an hour.

The incidents have occurred with different pumps, bloods and batches of giving set.

I have tested these pumps to Graseby’s Checklist and found no problems; they have also been tested by a third party; I could not get air to go past the first blister and its air detection system, Graseby are also investigating this.

The blood is coming down immediately after matching so will still be relatively cold.

Does anyone know what is causing this problem?
What is a safe level of air to be injected into a vein? (Yes I have read the Infusion Pump Technical Article).

Lee

Hi Lee

I'm making the obvious assumption, that you're using the appropriate blood giving set.

Is this just a recent occurance or have you been troubled with it for some time?

I know blood outgasses when it is warmed, hence the need for an air trap on blood warming sets, but I'd be surprised it could happen as the blood is warming up to room temperature.

I've just spoken to our ICU staff and they have had no problems with blood and air in line, unlike some other fluids they put through.

Is this noticed at the beginning or latter part of the infusion?

I have a feeling that your pumps are OK but you've got problems outside your control.

Sorry not to be much help but I would be looking at the blood / admin lines myself as a next stage.

Good luck and keep us informed

Brian

Hi Brian

Yes they are Graseby Blood sets.

The first occurrence was in February then two this month. We have two other departments running bloods who have not noticed any problems.

The problem is noticed at the end of the infusion.

Lee

Must be out gassing in the line between the pump and patient Lee. I can't see any other reason.

Does the tubing become warm for any reason, radiators, sunlight ??

The fact that you get a few small bubbles above the pump in the short section of tubing whilst the larger quantity collects in the longer section would tend to support this view, also that the effect is cumulative and most noticable and the end of infusion.

You've got yourself a good one there mate!

Brian

Hi All,
Could it be cavitation causing air to be drawn into the blood due to too low a resevoir level in the chamber?
Did come acros this many, many years ago..........
Regards

Ed Bennett
Southwest Medical Ltd

Two factors that I would look at:- 1) Bag height 2) It's getting hot (as Brian mentions)!

Brian
You are right it has certainly got me thinking. The patients normally sit near some large windows I will sort out some monitoring to see if problems occur with higher temperatures or strong sunlight.

Ed
If I understand you correctly I would expect to see air above the pump.

Geoff
The patients sit in comfortable chairs, whilst the pumps are mounted on pole stands at a comfortable height for the nurses to operate(so at or just above the patients head height)with the bags hanging from loops that are 6 foot from the ground.

Lee

Hi Lee,
In your original post you said "The giving sets on close inspection show many small bubbles above the pump typically smaller than 0.5mm diameter and the same below the pump but up to 2mm diameter". The bubbles above the pump is what put me in mind of cavitation. If there really is no air bubbles above the pump then it can only be getting in at or around the pump head or being generated therin?
Will be watching with interest................

Ed Bennett
Southwest Medical Ltd

Hi, we dont use pumps at all primarily because the peristaltic pump part damages the blood cells as it forces them through.(unless transfusing paeds of course)
we use gravity sets and we dont warm the blood unless giving it to a hypothermic patient.
we dont have problems with air bubbles.
why dont you contact one of the transfusin specialist nurses, there is one at almost every hospital.

Hi Quinny

Small point, Graseby 500's don't use peristaltic action, they have a couple of small pistons which pump the fluid through compressable chambers in the giving set. Hence the cells don't get crushed - bit like using a series of sequenced syringe drivers!

Interestingly, Alaris have use peristaltic acion with their pumps and sell dedicated sets for blood delivery. Anyone able to provide a reasoning behind that? Is it the fact that the percentage of cells damaged is insignificant?? I know our BT service insist that all blood is delivered via pumps unless in an emergency.

Brian

sorry meant "transfusion specialist nurses"

Surely it depends more upon the thickness (of the walls) of the giving set where it gets acted upon by the pumping mechanism, and the "compliance" (stiffness) of the tubing, rather than the type of mechanism? After all, most IV pumps "squeeze", do they not? I don't recall clinicians complaining about those old fashioned roller type pumps! And aren't they the type used in Heart-Lung Machines? What is needed is what I can only describe as "a massaging" of the tubing (ie, a gentle pumping action).

Ed
What I was trying to say was that I would expect to see the same size bubbles and quantities above and below the pump if they were being created in that way.

Quinny
I think I've spoken to everyone in the trust about this who might be able to help. I thought I would ask the EBME community and I will try to speak to our local Diabetic Centre next week.
As Brian says the Grasebys do not use a peristaltic action but the Alaris and the Baxters do (or did).

As a Trust we have standardised on the Graseby 500, I can obtain a Baxter Colleague for the Unit to try but this problem has happened three times in three months with the unit doing many infusions every day.

In the past we had Baxter 6200, 6201 then Colleagues and we have not seen this problem however we were not doing anywhere near this amount of infusions and other areas in thr Trust have not had these problems.

I am beginning to think that it may be due to the patient area which has large windows over looking a court yard as Brian suggested, tests will be on going both of temperature and weather conditions.

Lee

Yes, and you mentioned that "the blood is coming down immediately after matching so will still be relatively cold". Perhaps the bubbles are due to some sort of "rapid heating effect". Temperature is the key here, I reckon, not anything to do with the pumps.

Hi, been watching this thread for a while (fascinating!) I think Brian and Geoff are both correct! I have seen this before (rarely) when either a large pressure or temperature gradient was present! After all, and as you point out, It’s a carefully prepared closed system!

Ed

Lee also mentions that the phenomena only occurs towards the end of infusions lasting over two hours twenty minutes. From cold ... to hot? Could we be talking about hemolysis due to heat? Blood, the very stuff of life!

venous air embolism (VAE)...<50 mL causing symptoms in adults, and as little as 200 mL being potentially fatal (2). A PID (Pressure Infusion Device) pressurized to 300 mm Hg (standard operating pressure in the PIDs tested) can cause delivery of this fatal dose in as little as 4 seconds.

Anesth Analg 2003;97:1074-1077
© 2003 International Anesthesia Research Society

Regards

Ed

ah yes, we use alaris...thatll be why

Maybe this will shed some light!

Regards

Ed

Gentlemen
The bubbles observed are undoubtedly as a result of out gassing due to temperature change. It is not good practice to infuse blood that has not been allowed to warm up to a temperature approaching ambient. It is also good practice to use a blood warming device in-line.
The bubbles are definitely not due to cavitation as the negative pressures needed to cause this phenomenon will not be seen in a pumped infusion system.
Also, there will be minimal hemolysis in such a system as again the shear stresses, pressure difference, temperature change etc. are too low.
With regard to fatal doses of air, as Eddie points out, it is normally regarded as >50ml into the venous system and it depends on the infusion rate.

I got called out this morning to investigate this problem.

My findings are that the inside wall of the upper chamber of the cassette becomes opaque with deposits / staining of the blood solution and the IR detection fails as the device sees an obstruction of the IR path.At this point the 500 becomes an air + fluid pump allowing anything to pass as there is still header pressure from above maintaining the contacts of the upper piston.

The same scenario can be replicated by a small amount of felt pen on the cassette in the area of the IR beam.

Hi Rich
Best thing is to report this to the manufacturer and point them to this thread! There is obviously something seriously amiss with the IR sensitivity threshold? Can you replicate this with other 500's?

Eddie

Hi Rich

Don't follow this one!

The theory as I understand is that the IR beam requires an intact fluid path to maintain its integrity. Any absence of the transmitted IR pulses (due to air in fluid or contamintaed optical path) will be detected and immediatley throw up an air detection alarm and stop the pump. Valves within the pump engine will be in the closed position since the motor is no longer running, therefore fluid can no longer flow to the patient - even if additional pressure was to be applied upstream of the pump.

If you're able to demonstrate anything different then you should certainly have grave cause for concern.

Brian

Hi Rich,

In the tests I ran with saline any air I tried to get through the pump built up in the top blister.

Even if the air sensor was fooled I would expect the top blister to fill with air followed by the second and third before an air appeared out of the bottom of the pump.

I haven't measured the volume of air required to fill the cassette when it is in the pump but it is about 7ml to fully inflate the blisters. Even half of this would be a very large amount of air to be entering the pump and much more than I would expect from a properly primed set.

However although it does not help you much it is nice not to be the only one with this problem!

Lee

Hi Guys,

Sorry if this seems alarming ( excuse the pun ) but try this..

Get a marker pen and shade in a small area of the cassette in an area of the IR path.

Now run the pump and inject some air - make sure you keep a normal header pressure otherwise you will get the "no flow above pump" alarm. Don't forget that down stream air will only compress until a certain point where it equals the header pressure so it will pump just the same as fluid - as long as there is header pressure, the top piston will still activate the bag contacts.

Now you have a G500 which pumps air and fluid until the cows come home.

Brian said...

"The theory as I understand is that the IR beam requires an intact fluid path to maintain its integrity. Any absence of the transmitted IR pulses (due to air in fluid or contamintaed optical path) will be detected and immediatley throw up an air detection alarm and stop the pump"....... this is how I understood it to be for a long time but I was amazed to find that the fluid distracts the beam and it is the presence of air in the top of the cassette that allows the tansmision of IR light from the emitter to detector - this is when it alarms!!!

The cause of the problem in our case was a small pocket of blood which seemed to adhere itself to the inside of the top chamber of the cassette..... Frightening!!

Rich

DO FOLLOW THIS ONE! (Standard procedure I would have thought Brian!)



I for one would like to hear their professional opinion!

Ed

Hi Rich

Yes we can reproduce this too! We are not making the system opaque but just enough to slightly attenuate the IR source! My point about sensitivity was that at some point in the design process the ‘sensitivity’ of the air detector must have been decided! It is common for a small amount of blood to ‘stain’ the surface due to ‘surface adhesion’. (Apparently blood likes to do that!). But this phenomenon should be designed in to the system!

Ed

This does not seem like a fail safe situation.

I thought a general principal of any safety system was: no or unexpected signal indicates a problem of some sort hence an alarm. Only a good (expected) signal indicates a safe condition.
A false negative is annoying but safe, a false positive is dangerous.

What happens if the IR sender unit fails? A failure of the Air in Cassette alarm?

I think this needs to be brought to the attention of the MHRA.
I will await the safety notice - in about a year or twos time.

Robert

I totally agree ROJO ! The whole logic here doesn’t seem right???
Undoubtedly there is probably some current detection circuit in the IR Send. Also (only in comparison to other system) I always thought that air detector circuits operated within a pre defined hysteresis (Tri-logic)and not a simple digital yes/no principal! This meant that an IR presence always existed at the reciever! Which again brings me back to the word “sensitivity” or in English “Sensitivity”!

Ed

Yes... I'm afraid there is no indeterminate state of some air, or a bit of air or not quite as much air etc.... it's 0 and 1 and that's all you've got!!

The initialisation process cannot be checking the presence of a beam from the emiter to the detector as the pump may have had the cassette in for some time and so no opportunity to test the beam.

The only time it could test the beam is when there is no cassette after initalisation.

I often wonder how it is possible that when investigating FC19 that usually more than 1 of the IR wires have perished to the point where they have both completely parted from the connections or completely broken??. Is it feasible that the unit will run for some time with TX IR device disconnected? ( red+orange ).

Rich

Can I put this one up again?

We have now had 6 cases of air bubbles found in blood infusions, the biggest bubble was 12mm long which is less than 0.2mL, they have occurred in two different departments, with different giving set batches and different ages of Graseby 500 pump. The rest of the details are the same as in the first post.

The second department assured me that they had never had this problem until I pointed out an infusion they were running had a 10mm bubble in it.

I still think that it’s a build up of the gas that is being produced as the blood warms up (champagne bubbles). But I need some kind of evidence to present to our risk management people to assure them that it is safe to use, still waiting for Graseby or the MRHA to come back with something.

A friend of this site has recommended that the staff give an extra squeeze to the drip chamber but the staff say that “that is not what it says in the instructions” and that they will not be able to use the chamber to check for drips.

Has any one else got this happening? Any more suggestions?

Help……………

Lee

Hi Lee

I can see this is going to go on for a while. I recon that since you have made people aware, they are looking for and found instances where this is occurring. Most staff tend to set up a pump and if it runs for a few minutes or so with out alarming they walk off to do something else and don't tend to notice anthing out of the ordinary unless the "bing bing bong" goes off.

I'm not sure about the gas build up bit.... I wouldn't have thought that much gas could be suspended by fluid in normal atmospheric pressure.

Hi Rich

If I am right then most of the gas would appear after the pump (as this is a much longer pipe run) and some of this gas is collecting at a high spot (normally between the cannula and the area where the tube is stuck to the patients arm). This gas could be the accumulation of all that available from 250mL of blood, I don’t know how much gas could be expected as the blood warms up from about 5 degrees centigrade up towards room temperature, but I am looking at between 0.1ml to 0.3ml if you add up the larger bubbles.

Does anyone know of any devices to remove or trap air in the market that could be put between the Graseby giving set and the cannula?

Lee

Hi Lee

I can see where you're comming from as the length of the tubing could be basically an air fluid seperator. The only thing that comes to mind here is that the air should be seperating in the tubing from the cassette to the bag and in the cassett itself. By this time most of the air should have dissapeared up line.

It's the sheer volume of air which I can't get my head around and the scenario we had when the pump just carried on pumping air and fluid because the cassett became stained.

Two things to point out, blood tends to have more dissolved gas than a normal fluid such as saline because that is one of its functions. Therefore there will be more outgassing occuring due to warming up to room temperature.
Secondly, I am not aware of any end point filter that can be used with blood to trap air, neither is there a bubble trap available.

Maybe there are two seperate scenarios here

1. The longest part of the tubing which extracts heat from the environment, warms up the blood which subsequently off gassses.

2. Staining of the casette which presents an IR block subsequently disabling the AIR IN CASSETTE alarm.

I don't think the problem we have is the same as the one you've seen with the staining Rich. We are not seeing the volumme of gas that would be required for some of it to be pumped through the set.

Lee