galvania

And never EVER , under ANY circumstances look at gel tests under a microscope or a magnifying glass - unless you want to call absolutely everything positive and waste everybody's time

and you would - I hope - transfuse with group A, so if you wrongly called it a group A, rather than an AB, it would actually be better for the patient. I know of at least one case where an ABel was transfused with group AB blood and died as a result of a transfusion reaction. And if this is a donor, the amount of B antigen present MIGHT cause a minor reaction if transfused to a group A patient but would not do any serious harm. An what percentage of those weak reactions with B cells will actually be caused by this phenotype anyway? Probably less than patients having antibodies against LFAs that are not picked up in the antibody screen and who have a minor reaction due to the incredible bad luck of receiving a unit of blood that just happens to have the antigen

I would just like to add another comment to this discussion. CAT is NOT the best method for looking for weak ABO antigens or antibodies.

Yes, that's true, Malcolm. On the other hand, if you test with 2 different monoclonal anti-A reagents (and an anti-AB for good measure - a real one not an A+B) and they all come up 4+, I think it's fairly safe to say that the patient is a group A. I think that giving group O blood in this case is both wasteful of group O blood (unless you are swimming in it) and overkill

Well if this is a patient, the worst case scenario is that the patient is actually an AB with an atypical weak B antigen that is not being detected and a weak anti-B in his plasma. So what would happen if you gave group A blood - nothing. If it's a donor, then you surely have the possibility to do extra work ups - but you would be unlikely to cause any harm if you called the donor a group A. And there are SO many reasons for having weak ABO antibodies - not least because we have the tendency to be too clean around our kids!

HI sounds more like it.

I would just like to add one 'grain of salt' to this debate. You cannot detect all D variants - whether D weaks or partial Ds by serological methods alone. Neither D weaks or Partial Ds behave in a way that allow one to say that all D weaks or partial Ds react with such and such a strength. You will always miss some. You will miss some D+ donors because their D antigen is so weak that it is not detected by even the most sensitive of routine serological tests - or because despite using at least two different monoclonals the donor has an extremely unusual variant that is detected by neither. You will miss some 'D-neg' patients because they have sufficiently large numbers of D-antigen sites that they give a normal reaction with the anti-D reagents used. Follow the manufacturer's instructions for the reagent and method used and you will detect as many as you can hope to detect. And before you shoot the manufacturers because their reagent/instrument gave a 4+ reaction with a partial D known to have 10'000 D-antigen sites per red cell, and discovered because the lady made an anti-D and she is pregnant - please take a minute to think about the theory behind the serology. Maybe in years to come there will be a foolproof routine method for catching every single one……...

You make it sound as though the bio-Rad system does not pick up Kidd antibodies. I can assure you that just is not the case. If it were, then there would be a huge number of transfusion reactions in hospitals using the bio-rad system - and that is not the case. There may be all sorts of reasons as to why these particular antibodies are not showing up. First of all, can I suggest you try repeating manually. As this is a trial, there may be a problem in the way the instrument has been set up - although that is a long shot. Then, what I strongly suggest is that you contact your local bio-rad rep and let them send the samples to Switzerland. Switzerland will need to know exactly which lot of cells and cards you were using, they will want print outs of the results from the IH1500 - and as much information about the age of the sample, whether it was frozen etc as you can give them.

OK - to clear up some confusion (apologies for it being so late in the day!). Ortho in Europe has glass beads. Ortho in the States has gel which is similar to but not identical to the Bio-rad (ex-DiaMed) gel. As the original question was about Kidd antibodies, I will stick to that. You all know that Kidd antibodies love playing hide and seek. BOTH techniques will miss some Kidd antibodies - but not necessarily the same ones. So you might see some antibodies coming up in Ortho and not in Bio-Rad; and you will see others that do the opposite. And by the way, Immucor will pick up Kidd antibodies that no one else does but it's not sure that these are real but might be artefacts caused by Paraben. It is fair to say that some antibodies just 'prefer' one system over another. But across the board, it evens out.

You CAN trust the results in gel. But as with any technique you have to understand the theory before you can interpret them correctly. Same goes for all methods

They may not be contaminated. The question is, are they working correctly? In other words, if you put up your controls, especially anti-Fya, are they working correctly?

Anti Jra is quite common in Korea and they sometimes send some samples to our reference lab in Switzerland for testing. The reference lab have previously reported difficulty in adsorbing out allo-anti-Jra too

It looks to me as though you may have some spontaneously agglutinating IgM in the sample. Do you have some neutral cards? If so, put 50ul of the cell suspension on the neutral card and centrifuge directly. That won't necessarily differentiate between an agglutinating IgG and an IgM but it would tend towards the latter if it were positive. Also, in case it is LISS-enhanced, you could try suspending your cells in Cell Stab or Alsevers or PBS and repeating. Just beware of cells 'training' through the gel. And make absolutely sure that whatever you use has well come to room temperature before you use it! Anyway, from those results, even though the control is positive, you clearly have a major component of IgG present.

In my experience good monoclonal anti-D reagents will pick up all but the weakest of D antigen variants (by which I mean all normal D+ and all D variants with a reasonable number of D antigens on their surface). The positive DAT will not interfere with them unless it's caused by clinically significant cold haemagglutinin disease. So only the very very weak D variants will come up anyway with the weak D test in an IAT but not in direct testing. So if you did not do the weak D testing you might miss a few cases. Those few would have to nonetheless be able to stimulate mum into making an anti-D. I would be very interested to know exactly how many cases you see where the routine testing (using a sensitive method and good quality monoclonal anti-D reagents) is negative and the Weak D testing was positive (with a negative DAT). I wonder if it's more or less than the number of D variant mums you miss because they group as normal D+ ........... Also, for the anti-D to give you a false negative result because all the antigen binding sites are already loaded, your DAT would have to be VERY strong - and mum would have a known (hopefully), very high titre, antibody. So you would have to carry out an eluate anyway on baby's cells and find for example that mum has a high-titre anti-D and you can elute anti-D from the baby's apparently D-negative red cells. That's when you know that baby really is D+ and all the D sites are saturated. I've seen this twice in 'real life'.

the trouble with titres is that the result is method dependent - so unless the methods are identical, comparing across labs is not much use

You are talking about patients? The biggest problem as I see it is that you will get a false positive result due to the positive DAT and then transfuse with D+ blood. Either you go to a lot of time and effort and strip off the antibodies from the red cells as Malcolm said - or, probably a lot easier - and safer - treat them as D-negative

You will not know just on the basis of this single test. Looking under the microscope can help (not at the card!) to see if rouleaux is present. The diagnosis can also help. Also a cold auto-antibody will disappear if the test is carried out strictly at 37°C; the rouleaux will not.

Cold auto antibodies can give you false positives in the forward group - but will also give you a false positive in the control well, so this will be picked up. The reverse group will also be positive. I have yet to see a case of 'normal' rouleaux that is so heavy that it will do this, but it is possible if bloods are taken directly after administration of plasma expanders.

The manual result is no more accurate than the result obtained on the analyser. Personally I would report as unable to interpret due to mixed field post transfusion

Would you call that a limitation of the automat? Personally I would not - any more than saying that sampling from the top of the tube is a limitation of manual sampling. It implies that there is a problem and it lies with the automat - when in fact it's a simple manifestation of a biological property of red cells

Much deserved. I allocated you a winners cup. The icon says'thanks' but read the icon more as much deserved congratulations One of my colleagues will be in Brighton. I'll get him to come up and shake your hand (and maybe buy you a beer....) anna

Likewise. This is very common and can occur on any automats

And going back to the first case. Were those results 'made up' or did you actually have a patient that gave those results?

sounds a bit too much like alchemy to me........but I don't pretend to be a chemist

Tabbie - what exactly is behind this question? I get the impression there is something you are not telling us??????