Why do we xm?

[RR1]

Please could you tell me your views behind performing a crossmatch by IAT if your computer system didn't have an 'electronic issue' function?

Thanks!

[Malcolm Needs ☆]

Please could you tell me your views behind performing a crossmatch by IAT if your computer system didn't have an 'electronic issue' function?

Thanks!

Do you mean do an "immediate spin" cross-match instead, or do you mean just grab a unit off the shelf?

:confused::confused:

[RR1]

Yes- I suppose I mean would you do an instant-spin cross-match or an IAT crossmatch, and whichever one you choose to do ,what are the reasons behind the choice?

[Malcolm Needs ☆]

Yes- I suppose I mean would you do an instant-spin cross-match or an IAT crossmatch, and whichever one you choose to do ,what are the reasons behind the choice?

In that case then, as long as the present screen by IAT is negative, and there is no history of a clinically-significant antibody, BCSH Guidelines do not require an IAT cross-match to be performed, and I would be quite happy to cross-match by "immediate spin".

However, I do know of some places that will take units off the shelf in such a situation (as if they are performing electronic issue, but without the computer) and this is an absolute NO, NO, as far as I am concerned (and as far as BCSH Guidelines are concerned).

:eek::eek:

[TimOz]

Hi Rashmi,

I think the answer has to be considered for each facility depending on the "customer" base and the population as a whole. The question should not be is electronic release OK but how successful and safe is the antibody screen in use. The decision to NOT IAT crossmatch is predicated on the abilities of the antibody screen.

When training in Asian Hospitals (where electronic Blood Issue is not used) I usually ask them to consider their blood supply based on the antibody incidence and the statististics behind the success of the antibody screen that they use. To put it simply, what antibodies are in the population that the antibody screening cells used cannot detect? How common are they and how harmful are they? If the antibody screen has the right antigens in the right zygosities, then Electronic release is safe. If electronic release is not available, an IS XM (I HATE that name) may be used safety as this is only a "last chance" to detect an ABO error.

If the chances of undetectable, clinically harmful antibodies are significant, such as in Asian populations using Caucasian based screening cells, then this blood release algorythm is dangerous. The level of danger is proportional to the chance of an undetectable antibody. In many Asian countries, Mut, Mur and Mia antibodies (formerly Miltenberger) antibodies are harmful, common (maybe 20-30% of all alloantibodies) and will not be detected unless specific GP.Mur phenotype cells are used in the screen. If this is not available, an IAT crossmatch MUST be performed to detect ABO error and also any alloantibodies missed by the screen.

2 complications:

1. Many countries have no knowlege or data on antibodies missed by their screen. There are many publication that present the antibody incidence like it is actually correct - such as Ab Incidence is 0.8% and of these the specificies are: E 20%, c 15%, Lea 12% Jka 4%...... But look carefully. What method is used and what is the antigen profile of the screenign panel being used. What are they missing, In places like the Philippines, ~24% of antibodies are detected in IAT crossmatching and they cannot be detected on a screen or ID panel. Think of those stats for a minute as many alloantibodies will not be detected in a crossmatch, so the incidence of alloantibodies undetectable in their screening cells is higher than 24% and it may be double this. What are they? Are they harmful?

2. The second issue is that in the modern world, it is probably faulty logic to simply look at stats from a "population" like the Uk or Australia in isolation. The population mix is changing rapidly and many stats are history. If you group, screen and electronic release, what is the chance that you are testing a patient with antibodies that are "rare" in a predominantly Caucasian popualtion, but may be common in a patient of other ethnic origin? Can you detect it in your screen? I think this chance is increasing. Australia has a population that is 4% of Asian origin and rising and we have many East African immigrants. New Zealand has 9.2% Asian ethnicity and will be 14% in a few years. When do you change the screening cells to match the changing population? When does the Electronic release become unacceptable?

Finally, and this may be teaching many to suck eggs but I think it worth mentioning, back in the old days we repeated everything - patient group + patient check group, donor group + donor check group, antibody screen + IAT crossmatch. In this testing the IAt checks the ABO group compatibility and that there are no alloantibodies that will react with selected donor cells.

In electronic release, the algorythm is missing a significant check - patient group + patient check group, donor group + donor check group, antibody screen + electronic check on ABO group agreement and compatibility and that the antibody screen is negative. The antibody screen is not longer checked and low incidence antibodies are likely to be missed.

This may be OK but you have to do the math and after all, if you only get the groups right, you are most of the way to safe transfusion anyway.

Edited February 17, 2010 by TimOz

[adiescast]

I almost hesitate to speak after the wizard of Oz...

It is possible for us to overthink some situations. I would definitely agree that in a situation where your screening cells do not represent your population and your population has a high incidence of dangerous antibodies that might not otheriwse be detected, you really would be best off doing IAT crossmatch. In a stable population which is pretty well represented by the screening cells, the electronic and immediate spin crossmatch scenarios have proven to be safe and cost effective.

I absolutely agree with Malcolm about the faux electronic crossmatch (sort of like "I'll have a cheeseburger, hold the cheese."). It is a travesty and it is all too common.

[RR1]

Thanks everyone, so it would be better to perform immediate spin for the purpose of detecting ABO incompatibilities if no other antibodies / historical antibodies were present.

How useful is the IAT xm in the detection of ABO incompatibility; I gather up to 25% of incompatibilities of group AB blood to B recipients cannot be detected in this test? Additionally, some patients may only have IgM anti-A and anti-B and therefore may not react in this phase anyway with other grouped blood...... does anyone have concerns that there are lab folk that think the IAT xm detects all ABO incompatibilities as well ?

From the info supplied by Tim, it looks as though over the years we may all need to consider returning to both full IAT xm as our populations change. I suppose these decisions would be based on haemovigilance data .

[Malcolm Needs ☆]

Thanks everyone, so it would be better to perform immediate spin for the purpose of detecting ABO incompatibilities if no other antibodies / historical antibodies were present.

How useful is the IAT xm in the detection of ABO incompatibility; I gather up to 25% of incompatibilities of group AB blood to B recipients cannot be detected in this test? Additionally, some patients may only have IgM anti-A and anti-B and therefore may not react in this phase anyway with other grouped blood...... does anyone have concerns that there are lab folk that think the IAT xm detects all ABO incompatibilities as well ?

From the info supplied by Tim, it looks as though over the years we may all need to consider returning to both full IAT xm as our populations change. I suppose these decisions would be based on haemovigilance data .

If you are performing an "immediate spin" cross-match, then, presumably, you would be performing this in a tube, rather than by column agglutination technology (CAT). This would detect ABO IgM antibodies, in almost all cases, as long as there is a slight incubation period of a couple of minutes or so (which is why I always put "immediate spin" in quotation marks [whenever I remember so to do - it's an age thing]). If you look back at various NEQAS (for those not in the UK, this is the National External QUality Assurance Scheme) reports, they do suggest that a short incubation time is essential to detecting such antibodies.

In addition, addition of EDTA to the process helps, but, of course, these days almost everyone uses EDTA samples anyway.

CAT will detect IgM antibodies actually, as these will sensitise the red cells during the incubation period, thus causing agglutination, and the agglutinated red cells will stay at the top of the column (in most cases). Whether this is true of other "new-fangled" techniques is another thing entirely!

I agree with you entirely that we have to be ready to adapt the antigenic make-up of our screening cells, with our changing ethnic mix of patients, but in the UK (for the moment anyway) we are lucky that most of the ethnic minorities lack a high-frequency antigen more often than they stimulate an antibody against what is a low-frequency antigen in our native population (whatever our native population may be; Saxon? Viking?, Jute? Norman?). I also agree that haemovigilance schemes, such as SHOT and/or SABRE are probably our best allies in this.

:):)

[RR1]

Thanks Malcolm, I was also asked recently why we didn't use our automation for performing full crossmatching (be this IS or IAT). I suppose this depends on the technology used and would generally be achievable using the card analysers but not using the Galileo/ ECHO as there is no functionaliy to perform IS Xm on this equipment.

However performing IS on any analyser would be very expensive compared to tube techniques, but this would be a safer option and help reduce transcription errors- with a full audit trail of resting.

[Malcolm Needs ☆]

Thanks Malcolm, I was also asked recently why we didn't use our automation for performing full crossmatching (be this IS or IAT). I suppose this depends on the technology used and would generally be achievable using the card analysers but not using the Galileo/ ECHO as there is no functionaliy to perform IS Xm on this equipment.

However performing IS on any analyser would be very expensive compared to tube techniques, but this would be a safer option and help reduce transcription errors- with a full audit trail of resting.

Well, no not really, because you can't perform an IS in a card. Even if you only incubate for a couple of minutes, you still need to centrifuge for about 10 minutes, whereas with tubes, you would incubate for a couple of minutes, and then centrifuge (low) for about a minute (or less).

This is why I can never understand Hospital Blood Banks who sy they cannot perform IS because they haven't got tubes, centrifuges, SOPs, etc, and yet are major trauma receiving hospitals; it doesn't add up to me - sounds like an excuse!

:confused:

Edited February 21, 2010 by Malcolm Needs

[rravkin@aol.com]

Thank you TimOz for an eye opening post. But one question, the imediate spin xm will for the most part catch ABO discepancies, should each facility do some level of demographic analyis of the local population that they serve and compare to the demographics of the reagent screening cells in order to determine the need for routine IgG cross matches which would catch any other incompatibilities?

[TimOz]

Rashmi,

I think the ISMX is a good test BUT Malcolm is absolutely correct in that it has a failure rate if not done correctly. The problem with it is if it is done immediately, it will fail 4-5% of the time. It really needs a 2 minute incubation. It is also temperature sensitive as you are trying to detect IgM antibodies (and only IgM) it will work better at lower temperatures. RT is OK but higher temperatures will increase the failure rate. If you do the testing you will see that the failures are in children, elderly or lowered immune status individuals who have low levels of ABO antibodies. As I have said before, the ISXM is NO SUBSTITUTE for reliable grouping but is a good test to use as a check if barcodes and electronic checks are not available. So the ISXM should NOT be done immediately and is not a crossmatch. Can we change the name to "Saline ABO check" or "2 minute Saline ABO check"? This is a far better name as it is descriptive.

And RRAVKIN,

I beleive the answer is YES!

Countries like Australia and the UK has detailed guidelines for the antigens and that must be used on screening cells and defines antigens that should be expressed as a homozygous phenotype. These guidelines are decided by august committees from population studies, incidence studies in their population and risk from published clincial cases. I understand that ISBT (but I cannot find tis referrence) state that antibodies should be screened for for any red cell antigen that has an incidence greater than 5% in a population. I presume this also says that the antibodies have to be proven harmful. I will keep looking for this referrence.

The hard bits are this:

1. How are we keeping up with population changes and

2. How do you find alloantibodies to antigens not on the screening cells you buy? The answer has been to look and antibodies found in crossmatches that do not react on your screen or make your own screening cells. It also used to be a general view that you should use screening cells that are "representative of" your population, but now it is probably better to actually look at the specificities of antibodies found in some detail and make sure the screening cells are appropriate. As examples, the Dia antigen is fairly rare in Caucasians but 5-10% in Japanese and Korean populations. They therefore require screening cells to have Dia. Dia is 1% in Malay populations. Do you need it therefore in Malaysia and Indonesia? How about San Francisco and Sydney? Maybe not now but maybe in the future.

As I said earlier this is part of incremental blood safety. I have a slide I use for training that I have totally made up but I think useful to consider when thinking about this issue. It goes like this. Imagine a barchart. The numbers are rubbery and vary with testing methods etc but it looks like this:

Relative safety Testing protocol

60% Random blood

97% Group compatible

99.4% XM

99.4% Group and screen release

99.8% Group screen and XM

99.98% Group, screen, Crossmatch and phenotype matched

The idea is that increasing testing gives incremental safety. The biggest safety difference is in getting the groups right and we should always remember this. Improving the alloantibody detection gives increased safety. Ab screening can fail to detect lower incidence antibodies. Crossmatching can fail to detect weak antibodies when crossmatching heterozygous units. It is all a cost / time / safety benefit analysis. On a barchart like this it looks like alloantibody detection makes very little difference but if you say it like this:

Using antibody screening in Asia with Caucasian cells may give you 99.4% safety and this may seem high but it means 6 in 1000 transfusions are not safe. Adding crossmatching now takes it to 2 in 1000 so this is quite an improvement. My next slide says WIIWYD? (What If It Was Your Daughter?) This usually makes people think a little bit differently about the stats.

These numbers are totally different in the UK and the US and the decisions may have different outcomes when choosing testing protocols but the point is the same. Get the blood group right. Consider time and cost and choose testing regimes that are the safest you can with the resources you have - in your population.