Mabel Adams

John, very good questions. Very difficult to scan patient hospital band and specimen labels per Epic specimen collection protocol in the OR. Here's one worrying scenario: Label printed for wrong patient (printed outside room and brought in, printed to wrong room, left in room from prior case) > T&S collected with wrong ID > Pt in room would have no testing being done and we would do it on the one the label/order is for. Caught (eventually) when they call for blood for different patient? They could send the ID card (proxy for blood bank band under drapes) they made out for wrong patient to pick up blood and we would issue it. ID would match on ID card and unit. Unit would not scan in Epic on real patient but they could override or give anyway. Hopefully they would check it against the big display screen in OR which has the right patient on it. Also, could issue blood later for other patient (whose label was used) based on wrong blood in tube. Scenario #2: T&S drawn in pre-op or ED is wrong blood in tube. Let’s say BB banding was also improper so band is on patient going to OR. Tube & band have patient going to OR ID on them (but someone else’s blood in tube). ID Card is made up in OR. Timeout is performed. All will match but still wrong blood in tube. Nothing will catch error until ABO-incompatible transfusion reaction. A VTYPE drawn in the room would catch the error. Evidence: Baylor NW transfusion fatality last year. If we could make it so labels couldn't print easily except in right room that would probably move #1 into the acceptable risk realm. If we could tighten up the electronic documentation of collection in scenario #2, that could probably be brought into a reasonable risk realm. One of our problems is the blood bankers being able to tell for certain where the specimen was drawn and whether the patient was scanned to collect it. I think this is mostly learning curve but we would need to make sure that our people could (and would remember to) do it. I greatly appreciate your thoughts and suggestions. Please add any unfortunate scenarios I have overlooked.

Thanks for the response. Doesn't this negatively impact your O blood supply? We can do it for a while but keeping them on O through a bad MTP in OR would deplete our O red cells and our supplier is often running short (and is hours away), especially at this time of canceled blood drives. It's bad enough for the O patients. It really is a hardship for them to draw patients in OR during a bad case and then when they have to do it twice they are beside themselves. Mistakes occur for sure. But we need to have a correct blood type too.

I tried searching but this is a tricky one to find. What are your policies regarding collection of a 2nd blood type specimen in OR, especially during an emergency or really urgent case? Our OR wants to substitute the "timeout" where they double-check patient ID for a second blood type and I was tasked with finding out what others do. We currently request a repeat type only on non-O patients with no historic type or separate specimen to test. We stick with O blood until we can get the 2nd type. If you allow them to not collect the second type, do you issue type-specific blood once the T&S is done?

Another question on anti-CD-47 drugs. Does anyone know how long the interference persists after the patient goes off of the drug? For DARA it is 6 months but I haven't found it for these drugs.

Enjoy the attached from 20+ years ago. ABOincompatible.pdf

But a perspective that is always needed, I'm sure.

The adsorption/elution showed anti-A but not at immediate spin. It reacted 2-3+ with 3 sources of A1 cells (2 different vendors' reverse cells plus an A1 donor) after 30 minutes at room temperature and was negative with 2 O reagent cells. Last wash was all negative. It doesn't matter, but what subgroup do you think it is? Am?

I guess the only risks to her of giving A red cells would be if our testing is mistaken or the next time she comes in bleeding it isn't really her but someone who is O with a missing reverse antibody that we assume is her. (Definitely chasing unicorns here.) We should be able to convince ourselves it is her as long as the new sample reacts the same as now including with anti-A,B, right? No one is likely to have time to repeat the absorption/elution. I want to leave clear instructions on her record for whatever unlucky generalist has to deal with this someday. Or would we stay with O unless/until we repeated the adsorption/elution on each specimen? BTW, can we debate why they changed "absorption" to "adsorption" 20ish years ago (ad is Latin for "to"; ab is Latin for "from")? It seems like we are usually trying so absorb antibody out of a plasma sample so "ab" makes more sense to me. Here we are trying to adsorb it onto the patient cells. I was using "absorption" because it is opposite of usual, but I had the Latin prefixes backward in my mind. I vote for a patient-sample-centric universe so "absorption" for the usual warm auto workup and "adsorption" for this sort of testing. Or less highfalutin use of language altogether and just use "absorption" because it is a more common word. I'm also interested in input from blood bankers from countries where English isn't the primary language.

We have a young (first trimester pregnant) woman who forward types as O but reverse types as A in gel and tube. No recent transfusions. The usual steps to resolve a discrepancy (cold reverse, RT inc of forward) produced no new information except we got a 1-2+ reaction with anti-A,B in tube (but nothing with anti-A reagent in tube after 30 minute room temp incubation). We have only one vendor's commercial anti-A. We are going to try absorption and elution but either way we have to decide how to transfuse her to put a note in her profile. Assuming that the eluate contains anti-A would you report her as A or as O? If you report her as A would you add a special need in the computer to give only O RBCs? If we report her as O we would add a comment to give her A plasma rather than O. Whatever we do, she is going to have an unresolvable discrepant type in the computer so none of these decisions are to keep the computer happy. I want to keep her safe from getting incompatible products but still be as accurate as possible. If we call her O they will test her baby's ABO (I know, I know) which could be wrongly typed as O (if it inherited this from mom) which makes me want to avoid testing baby, but that is not enough reason to make the decision on the mom. I appreciate any input on the case or the transfusion decisions.

This is why we treat all those weakly reacting anti-D patients as though the screen is negative and give them RhIG. It is safer even if the antibody is really a newly forming anti-D. Most of these are detected at the time of delivery so this baby is not at much risk from an immune anti-D of that low a titer. If the baby has HDFN, then we will proceed accordingly. If we pick up a weak anti-D and the patient received RhIG in the prior months but there is no record of a negative screen before the RhIG was given we will sometimes follow the patient for a few months to make sure the titer is decreasing rather than rising. We even have a test called SAB Rh Type that is for an Rh type and reflexes an antibody screen if the Rh is negative. That way we have a baseline antibody screen before RhIG is given which increases the odds that the anti-D detected later is passive.

I voted, but we actually do DAT only if the baby is positive for an antigen mom could have antibody to--Rh+ babies of Rh- moms and non O babies of O moms. I have tried for years to get them to drop the testing of babies of O moms but can't budge them.

There is a new company that I saw at the AABB meeting. Their centrifuge takes the Serofuge heads and otherwise looked promising. They will let you try it for free. Drucker Diagnostics' SERO 12 Has anyone tried it?

We don't use just the cells that Ortho marks as an abbreviated panel for ruling out other specificities in the presence of passive anti-D. We add whatever other cells are required to have double-dose expression of the usual suspects. That said, we don't require double-dose K for ruling out anti-K on antibody screens so would not require it here. Also, we don't require use of double-dose E or C positive cells in the presence of anti-D, whether passive or immune, due to the low statistical risk to the patient.

US Scenario: patient has a positive gel screen so a panel is done which doesn't key out to anything. A repeat screen in PEG or 3% cells converted to 0.8% in gel comes out negative. So we conclude that it is probably an antibody to gel diluent. We can charge for the repeat screen by a different method, but does anyone charge for the antibody ID? If so, do you feel you must report an antibody ID result? Then what? Negative? Inconclusive? Antibody to reagent? We wouldn't want anything that would interfere with future electronic crossmatch in our BBIS. Similar issues can arise with other tests sometimes.