How many times do you perform an antibody identification?

[javvcr]

this is the reason for my question!

in case you have a patient with an antibody identificated, and need a transfusion today. But later on (a week or a month later) he or she needs another transfusion:

Do you performe the antibody screening and full identification again, then give blood antigen negative, cross match negative

or performe the antibody screening and cross match and just if the cross match appears positive, you do the antibody identification again?

or do you do something else????

[Malcolm Needs ☆]

this is the reason for my question!

in case you have a patient with an antibody identificated, and need a transfusion today. But later on (a week or a month later) he or she needs another transfusion:

Do you performe the antibody screening and full identification again, then give blood antigen negative, cross match negative

or performe the antibody screening and cross match and just if the cross match appears positive, you do the antibody identification again?

or do you do something else????

We would perform an antibody identification every time even if the patient has not been transfused since the last time. The patient has already been shown to be a responder, by the fact that he or she has already produced an alloantibody. What you don't know is whether the patient has produced a second (or more) antibody of a different specificity, but slower than the identified antibody.

You should not rely on the cross-match under these circumstances. The red cells in the panel are in a solution designed to preserve the antigen expression, but not necessarily the oxygen carrying capacity.

The red cells in the unit are in a solution designed to preserve the oxygen carrying capacity, but not necessarily the antigen expression, and you also may not know the rest of the units phenotype (it may have heterozygous expression of the Kidd antigens, for example, whereas the anti-Jka lurking along with the known antibody may only react in vitro with Jk(a+b-) red cells; but it sure will react with Jk(a+b+) in vivo)!

I have, only today, been working on a sickle cell patient who, a week ago, had anti-E+M+Fya+Jkb; he now has an incredibly strong anti-Fy3 as well.

:eek:

Edited August 2, 2009 by Malcolm Needs

[Brenda K Hutson]

I agree; I would absolutely not rely on a compatible XM for these patients (though this was the practice in my current Hospital before I came). Their rationale was; well, you will pick it up in the XM. Then I came across an example for them.

Wherever I have worked, the policy has been to at least perform a rule-out Panel every 7 days (unless of course the antibody screen prior to that indicates something else, and/or an increase in strengh). Because of their historical practice, I decided to break them in slowly by saying they only had to do it every 10 days (which I am not comfortable with but will change it soon). So, in their initial resistence to this, one Tech. performed and Antibody Screen on a patient with a known Anti-c; so cells II and III came up. She decided that matched the Anti-c so she didn't do any further work. Several days later, we got a new specimen and I just happened to be the one working on it. I had thrown in a select cell panel and antigen negative crossmatches at the same time as the Type and Screen. We now had a different Lot# of Screening Cells and Cell I was also reacting, so I knew there was something else. In the meantime, my crossmatches with c- blood were compatible. Turned out the patient had also made an Anti-Jka. I pulled the previous specimen (which should have had a work-up) and performed a screen and panel; the Jka had been there at that time also. And the units? All Jka heterozygous.

It is scenarios like that which are helping me convince these Techs. that they need to follow my protocol.

So, in answer to your question? Perform at least a rule-out Panel every 10 days, unless they see changes inbetween.

Brenda Hutson, CLS(ASC)SBB

[lateonenite]

I go with whenever the specimen has expired, because the 72 hour expiry is the minimum time frame for potential new antibodies to form. Particularly at the moment in our lab where we do not baseline phenotype patients until an antibody is formed - for transfusion dependant patients, the phenotype is impossible.

There is an argument in our laboratory about patients with multiple antibodies - particularly one with an auto; the theory was that we could not detect any "new" antibody. But I'd have to point out that the antibody panel in those cases has a lot more to do with reaction strength than actual identity. There have been occasions where we've had to transfuse Kidd or Susan heterozygous because we couldn't get E, C, Cw, K, Kpa, Fya, Jka and S neg, so subsequent panels would show increasing reaction strengths.

If the patient is an inpatient has had no transfusions since the last specimen, then yes, a week to ten days would do (even if the patient was transfused in the previous three months), but we have had patients that have gone to a private provider, received a transfusion and formed an antibody, come back to our hospital within a week. Unlucky.

Fran SEALS

[Brenda K Hutson]

I go with whenever the specimen has expired, because the 72 hour expiry is the minimum time frame for potential new antibodies to form. Particularly at the moment in our lab where we do not baseline phenotype patients until an antibody is formed - for transfusion dependant patients, the phenotype is impossible.

There is an argument in our laboratory about patients with multiple antibodies - particularly one with an auto; the theory was that we could not detect any "new" antibody. But I'd have to point out that the antibody panel in those cases has a lot more to do with reaction strength than actual identity. There have been occasions where we've had to transfuse Kidd or Susan heterozygous because we couldn't get E, C, Cw, K, Kpa, Fya, Jka and S neg, so subsequent panels would show increasing reaction strengths.

If the patient is an inpatient has had no transfusions since the last specimen, then yes, a week to ten days would do (even if the patient was transfused in the previous three months), but we have had patients that have gone to a private provider, received a transfusion and formed an antibody, come back to our hospital within a week. Unlucky.

Fran SEALS

I would not count on a change in the strength to decide whether there are underlying alloantibodies! Warm Autos come and go and change strength. I think there are basically 2 options when you are dealing with a patient with warm autoantibodies who is being transfused: either perform adsorption studies, or give phenotypically matched. When I worked at places with an extensive refernce dept., we did the adsorptions. Now that I am at a smaller place that sends these work-ups out, I only send them out the first time. If they have not been transfused in the past 3 months, I ask them to do a complete phenotype and we just give phenotypically matched blood in lieu of a work-up. Less work, less money and a shorter turnaround time.

Brenda Hutson, CLS(ASCP)SBB

[webersl]

If we hadn't had a specimen in the last 3 days, we would start with a screen, and then do a selected cell panel to rule out everything but the previously ID'd Aby(s).

[MAGNUM]

On each subsequent Crossmatch ordered, if the Antibody screen is positive, we perform the identification because we do not ag type each unit for all antigens and although the patient received antigen negative units for the original antibody, we cannot guarantee that other antigens are present that could cause the patient to create new antibodies which happens more times than you might think. Better safe than sorry and have a delayed transfusion reaction.

[David Saikin]

We would perform a new abid when we were dealing with a new specimen . . . basically every three days.

[Brenda K Hutson]

Right; my response was only related to work-ups on patients with warm autoantibodies. If no autoantibody, I have found (by doing a phone survey once of local hospitals), that people have different intervals for which they require additional antibody detection. 3 days is the safest. In my survey (and what I am used to at places I worked), 7 seems to be the lucky number. That being said, I was shocked to find that there are places that go a month without performing additional antibody identification (unless of course they see a change in strength or a cell that doesn't fit the known antibody). As I stated earlier though, I don't think you can depend on that; at least not in my experience.

Brenda Hutson, CLS(ASCP)SBB

[lateonenite]

I would not count on a change in the strength to decide whether there are underlying alloantibodies! Warm Autos come and go and change strength. I think there are basically 2 options when you are dealing with a patient with warm autoantibodies who is being transfused: either perform adsorption studies, or give phenotypically matched. When I worked at places with an extensive refernce dept., we did the adsorptions. Now that I am at a smaller place that sends these work-ups out, I only send them out the first time. If they have not been transfused in the past 3 months, I ask them to do a complete phenotype and we just give phenotypically matched blood in lieu of a work-up. Less work, less money and a shorter turnaround time.

Brenda Hutson, CLS(ASCP)SBB

Sorry, I wasn't clear - the strengths I was referring to was not a case of attempting to identify an antibody, more to monitor a transfusion dependant patient who had received (heterozygous) antigen positive cells as nothing else was available; if the patient required blood and the panel cells were stronger on those antigens then we would have to insist on antigen negative (if possible).

The patient I'm referencing above had a non-specific auto 4+ on original adsorption with varying reaction strengths in panel cells.

We had no option to phenotype more than Rh cells as the patient was transfusion dependant (and I mean, at least every month) and had received transfusion elsewhere by the time the antibodies started to come up. Hence the debate with very specific patients like this - why bother panelling when no identity can really be made - even our external reference had to give "cannot rule out....."

[SMW]

I would hope no one is "re-identifying" an antibody that has previously been detected and identified but rather running select cells to rule-out the development of any additional, new allo-antibodies? Once an antibody has been identified (and evaluated for clinically significance), antigen negative units should be selected so it really doesn't matter if a previosuly identified anti-X is still reacting (or not). On the other hand, if you're using testing with the patient serum/plasma as the (sole) method to select antigen negative units then one would need to demonstrate that the serum/plasma is still reacting with 'X' (including reactivity with a single dose cell) the same as would be done as a control if using a reagent antisera to select the antigen neg. units.

[lateonenite]

I'll confess I've never been trained to selectively panel from the positive antibody screen. We run a primary panel (usually 11 cells (CSL) or 16 cells (Immucor) with known multiples), then select cells to knock out knowns, prove likely ones (definite neg for known, or dosage with known, pos for suspected).

In our previous system also, we would report history of anti-X and then anti-Y detected in this sample, so we would have to demonstrate the disappearance of a known antibody if that was the case.

I'll add to that we never use the patient's sera to verify antigen negative status, every unit we receive must be pheotyped to confirm antigens by guidelines, regardless of the information provided by ARCBS.

Edited August 7, 2009 by lateonenite
forgot something

[John C. Staley]

Hold on to your seats because most of you aren't going to like this but.....

If we had ID'd an antibody, let's say anti-K for discussion purposes, in the past (how far in the past is not relevant), we would do an antibody screen on the current specimen, it the screen is positive only for anti-K and the AHG x-m is compatible with K negative units we do no additional testing. This is in full compliance with AABB Stds (I don't have a copy with me to quote chapter and verse).

[lateonenite]

Holy Ag's Batman!

*faints dead away*

In trauma on night shift, we will go crossmatch compatible if we can't ID, and leave it for the mythical "someone else"...but the work up still happens later.

[SMW]

So if a patient has a history of an anti-e, you run a 16-cell panel (of which 14-15 cells are likely e-positive), you now have maybe 1 or 2 results (from the e-negative cells) that have provided you any useful information? That's what makes blood bank serology so interesting...you have to think before you leap. It doesn't make sense to waste tech time and facility reagents/money, let alone wasting patient sample, by performing testing that provides no value and does not change the type of product selected for transfusion. Why not report history of anti-X, no additional antibodies detected? You're still going to select X-negative units and perform a serologic crossmatch but perhaps only needed to test a few panel cells.

As a second comment, the regulations do NOT require units to be tested with a reagent antisera to determine they are antigen negative. The regs (e.g., AABB 5.14.3) state units "...shall be prepared for transfusion that do not contain the corresponding antigen and are serologically crossmatch-compatible." The standards do not specify the method or HOW you make the determination the units are antigen-negative or even who does that testing. So if you determine the patient's serum is reacting at X-strength with a single dose expression of a particular antigen and you test that unit with the patients serum, you complied with both requirements (antigen neg and xmatch) with only one test. Think of it as an analogy to antibody detection. The regs state methods of testing shall demonstrate clinically significant antibodies but they do NOT specify what those methods are---you get to decide (and provide supporting rationale if necessary). The same for crossmatch, ABO typing, infectious disease testing, ...... Testing with reagent antisera may be your selected method to determine that a unit is antigen negative and comply with the regs, but please do state that is the requirement or imply that is the only acceptable method to meet the requirement.

Edited August 7, 2009 by SMW
-

[lateonenite]

For a single antibody, we would run a primary 11 cell panel, and the CSL's are good at separating out Rh antibodies particularly, not so terrific as things get more complicated. The 16 cell panel we use when the patient has multiple antibodies, otherwise it's generally used for selective panelling, after the primary panel.

We can't say something has disappeared if we haven't proved it - especially if our screening cells don't necessarily demonstrate the disappearance because the "new" antibody can be positive in at least one cell that crosses with the "old" antibody. I will say what we now do is only report an anything new and the other information is available in the LIS, but our testing practice has not changed.

And, I'm Australian:

The ANZSBT guidelines 2.2.2.1: "Red cells should be selected which are negative for the relevant antigen where the antibody is clinically significant [Table 1]. Antigen typing should be confirmed by the laboratory performing the pretransfusion testing." Table 1 lists by clinical significance, obviously.

Using the patient's serum is not an accepted method of confirmation of antigen-negative status.

You can quibble on the use of should up there, but GLP dictates that "should" in this instance means "must" - everyone makes mistakes, even the ARCBS.

Edited August 7, 2009 by lateonenite

[Brenda K Hutson]

Sorry, I think I am getting lost in this thread (that's what happens when you try to read with a migraine)!

So if I get you right this time, you are saying that the warm autoantibody is so strong that they cannot totally remove it with adsorption studies, and they have been multiply transfused so you can't get a complete phenotype? If that is what you are saying, I have one more sugesstion. Do you send you difficult work-ups to a Reference Lab? If yes, I would ask them if they can do reticulocyte separation. That way, you can get your complete phenotype and don't have to depend on a change in strength, just because you can't trype them (with your methods) and even adsorption studies are not working. If your local Reference Lab does not perform Reticulocyte Separation, you might call some well known Reference Labs in your area or your state and see if anyone does this; then send them a specimen so you can get a complete phenotype. Of course even that test has its limitations in that the patient has to be reticing.

Did I understand your issue correctly this time? If not, I will wait until my headache is gone before I put my foot in my mouth again.....

Brenda Hutson, CLS(ASCP)SBB

[L106]

Hold on to your seats because most of you aren't going to like this but.....

If we had ID'd an antibody, let's say anti-K for discussion purposes, in the past (how far in the past is not relevant), we would do an antibody screen on the current specimen, it the screen is positive only for anti-K and the AHG x-m is compatible with K negative units we do no additional testing. This is in full compliance with AABB Stds (I don't have a copy with me to quote chapter and verse).

John -

I do believe this may be the first time I haven't seen eye-to-eye with you! (They say "There's a first time for everything!") In the example you presented, let's say that Screening Cell I is Kell Positive, and your patient's plasma reacted with that one Screening Cell, so you are assuming it's Anti-Kell in the patient's plasma, right?

What if Screening Cell I happens to be Jka homozygous Positive (and the other two screening cells are either Jka Negative or Jka heterozygous Positive.) You would never know that the patient has Anti-Jka if you do not test any panel cells. Sorry, I couldn't feel comfortable transfusing a patient in that scenario. (Am I misunderstanding something?)

Donna

[lateonenite]

Sorry, I think I am getting lost in this thread (that's what happens when you try to read with a migraine)!

So if I get you right this time, you are saying that the warm autoantibody is so strong that they cannot totally remove it with adsorption studies, and they have been multiply transfused so you can't get a complete phenotype? If that is what you are saying, I have one more sugesstion. Do you send you difficult work-ups to a Reference Lab? If yes, I would ask them if they can do reticulocyte separation. That way, you can get your complete phenotype and don't have to depend on a change in strength, just because you can't trype them (with your methods) and even adsorption studies are not working. If your local Reference Lab does not perform Reticulocyte Separation, you might call some well known Reference Labs in your area or your state and see if anyone does this; then send them a specimen so you can get a complete phenotype. Of course even that test has its limitations in that the patient has to be reticing.

Did I understand your issue correctly this time? If not, I will wait until my headache is gone before I put my foot in my mouth again.....

Brenda Hutson, CLS(ASCP)SBB

Yes and yes and no to retic separation. I don't think I was clear about that one specific scenario anyway.

Running away now!

[dmpollock]

Our current policy follows (with reference). I would like someone from U Mich to verify, but I believe their practice is the same as below. Note, not applicable for warm autoantibodies.

-------------------------------------

For patients who have a history of previously identified antibodies, it is not necessary to repeat the antibody ID when they come back for subsequent transfusions, provided that ALL of the following conditions are met:

1) The antibody screen fits the history. For example, if the patient previously had Anti-K and Anti-E, the screen cells that are positive for E or K are currently positive, and the screen cells that are E neg K neg are not reacting.

2) Donor units are phenotyped negative for the patient's antibodies. In the above example the donor units must be negative for both E and K.

3) The phenotyped donor units are crossmatch-compatible at the Coombs phase as well as immediate spin compatible.

Reference:

W J Judd, S H Butch, Repeat Antibody Identification Studies: How Much Is Enough?, Transfusion, 2002—Vol. 42, Supplement.

[John C. Staley]

John -

I do believe this may be the first time I haven't seen eye-to-eye with you! (They say "There's a first time for everything!") In the example you presented, let's say that Screening Cell I is Kell Positive, and your patient's plasma reacted with that one Screening Cell, so you are assuming it's Anti-Kell in the patient's plasma, right?

What if Screening Cell I happens to be Jka homozygous Positive (and the other two screening cells are either Jka Negative or Jka heterozygous Positive.) You would never know that the patient has Anti-Jka if you do not test any panel cells. Sorry, I couldn't feel comfortable transfusing a patient in that scenario. (Am I misunderstanding something?)

Donna

Good point Donna and nope, you understood correctly. The theory is that if the antibody screen does not indicate a "new" antibody and all antigen negative units are AHG x-m compatible the odds are greatly in your favor that you are safe to transfuse. If I remember correctly this was first put forward by John Judd and the University of Michigan Blood Bank.

Do you perform immediate spin x-ms or electronic x-ms? If so aren't you worried about the possibility of an antibody that's not on you screening cells? Same basic idea.

[BSIPHERD]

I believe an antibody only needs to be identified once. Once you have identified a clinically significant antibody, you are obliged to respect it from then on. If an Anti-Jk(a) is identified today, you will give Jk(a) negative blood from now on. So whether it is currently detectable has only one point of significance to me – can I use the serum/plasma to screen for compatible units or do I have to go straight to reagent typing serum. So we may run one cell just to see if it’s reacting.

But Standards says you need to do antibody detection testing on every specimen collected for pretransfusion testing (and it has to be collected no more than 3 days prior to transfusion if the patient has been transfused or pregnant in the last 3 months). It also says that in patients with previously identified clinically significant antibodies, methods of testing shall be those that identify additional clinically significant antibodies.

So to me that means you have to run selected cells that are positive for additional clinically significant antibodies the patient might form, but negative for the previously identified antibody/antibodies.

Re-identification of antibodies already identified at your institution adds to health care cost and makes no difference in the way you treat the patient.

Any way, that's my 2 cents worth!

[Malcolm Needs ☆]

Good point Donna and nope, you understood correctly. The theory is that if the antibody screen does not indicate a "new" antibody and all antigen negative units are AHG x-m compatible the odds are greatly in your favor that you are safe to transfuse. If I remember correctly this was first put forward by John Judd and the University of Michigan Blood Bank.

Do you perform immediate spin x-ms or electronic x-ms? If so aren't you worried about the possibility of an antibody that's not on you screening cells? Same basic idea.

In the UK, a patient is not eligible for electronic issue if a clinically significant atypical alloantibody has every been detected in their plasma, so this would not be an issue (unless, of course, the patient switches hospitals, forgets to produce their antibody card and the antibody is no longer detectable, and then they are a candidate for an anamnestic response).

However, am I correct in thinking that you are referring to an antibody directed against a low incidence antigen here John, where the antigen is almost certainly not going to be expressed on the screening red cells? If so, that is a slightly different situation, because you are much less likely to come across, for example, a unit that is both K+ and Wr(a+) (although, of course, it can happen) than one which is K+ and Jk(a+) (to use Donna's example).

Even then though, there was a paper/editorial written some years ago now by George Garratty entitled something along the lines of "Do we need to worry about low incidence antigens" (I don't think that is by any means the correct title, but it was something like that) in which he showed amthematically that there was very little chance of a unit of blood expressing a low incidence antigen being given to a patient by electronic issue, who had an antibody directed against this same antigen, and it being strong enough to cause anything more than a minor delayed haemolytic transfusion reaction, resulting in a little jaundice and a requirement for re-transfusion as the red cells are removed from the circulation.

If I can find the actual reference amongst the detrius on my desk here at home, I'll post it.

Edited August 12, 2009 by Malcolm Needs
I know that nobody will believe this, but SPELLING!!!!!!!!!!

[Malcolm Needs ☆]

I believe an antibody only needs to be identified once. Once you have identified a clinically significant antibody, you are obliged to respect it from then on. If an Anti-Jk(a) is identified today, you will give Jk(a) negative blood from now on. So whether it is currently detectable has only one point of significance to me – can I use the serum/plasma to screen for compatible units or do I have to go straight to reagent typing serum. So we may run one cell just to see if it’s reacting.

But Standards says you need to do antibody detection testing on every specimen collected for pretransfusion testing (and it has to be collected no more than 3 days prior to transfusion if the patient has been transfused or pregnant in the last 3 months). It also says that in patients with previously identified clinically significant antibodies, methods of testing shall be those that identify additional clinically significant antibodies.

So to me that means you have to run selected cells that are positive for additional clinically significant antibodies the patient might form, but negative for the previously identified antibody/antibodies.

Re-identification of antibodies already identified at your institution adds to health care cost and makes no difference in the way you treat the patient.

Any way, that's my 2 cents worth!

I would agree that a "mini-panel" would be sufficient, rather than a full panel, although we always run one positive.

[Malcolm Needs ☆]

If I can find the actual reference amongst the detrius on my desk here at home, I'll post it.