exlimey

Please forgive my ignorance.....but do the regulations require facilities to "perform QC" on the reagents or the assay ?

Presumably predicted to be V-VS+ ? Might want to check on the hrB status, too.

My sympathies, Darren. That's a horrible situation. It's generally accepted that the best way to minimize HUMAN error is automation (with the proviso that the systems are well designed). If a LIS allows post-dated entries, it would seem logical that "scanning" would be the best option to back-fill the missing data. And, probably a lot quicker. If one is forced to perform manual data entry, the process should probably include a second check (verification), preferably by a second individual. Either way, not fun for anyone. Good luck.

Some caution may be appropriate. Most "colds" are indeed clinically insignificant and mere laboratory annoyances. Most of these are autoantibodies that can be avoided by pre-warming methods. However, some can be clinically important - there are examples of anti-Vel that behave exact as LaurieD describes, but are IgM, cold-reactive alloantibodies that can cause serious in vivo hemolysis. I think one case reported by Jill Storry was actually fatal. Another nasty beastie in this category is anti-P+P1+Pk (anti-Tja, in the old vernacular). I would suggest that a Reference Laboratory take a look at the sample (the Blood Bank of Hawaii is close ), just to give some assurance that the troublemaker is "just a cold auto". Pre-warming without an antibody ID may be dangerous. Just my two cents.

To paraphrase Malcolm: A proven responder may have "other stuff", too. A serological crossmatch is probably the best, and sometimes, the only way to detect it.

All of the above are excellent suggestions. I will have to get up earlier to contribute.

From reading the previous comments, both old and new, it appears that the manufacturer (Ortho) does not specifically require the bubble and therefore nothing is in writing (the Directions for Use). You may be out of luck trying to find something to reference.

I find it interesting that various users have been told/advised to use a pipette in a manner that may compromise the accuracy of the volume delivered. I'm sure the pipette instructions indicate to use vertically. Thankfully, the serological assays that are used by the Transfusion Medicine field have a wide range of tolerance. The "1-drop to 1-drop" concept is horrifying to many other pathology disciplines.

Agreed. The initial results using the Screening Cells still need to be resolved. "Non-specific" probably won't be acceptable. Please clarify - "Screening cell is positive with 2 lines(2+)". Does this mean that one, or two of the Screening Cells are reactive?

Agreed. Anything less than a full phenotype is useless. We don't even do that for Screening Cells, which are arguably a lot more important.

Excellent. It's not often that the IFU comes to our rescue.

This issue - the switch to plastic - seems to bubble up every few years (pardon the minor pun). When I was a puppy in my early years, last century, labs were already tossing around the idea to avoid potentially dangerous, sharp glass tubes. When broken, the plastic used for test tubes is also sharp, possibly worse that glass, as Malcolm suggests. As others have mentioned, static is always an issue with the plastic version, rather than occasional with glass. Other than that, and in my experience, plastic test tubes tubes work almost as well as glass for serological testing. However, many "tube reagents" are not formulated for, or qualified in plastic. The Directions for Use/ Package Inserts may be restrictive. Two points - personal opinion of a cranky old man: 1. One event does not indicate a trend - changing the whole system to address a single cut-finger incident is unreasonable. 2. The various safety apparatuses (however they be mis- or confusingly named) exist to limit institutional legal liability, i.e., prevention of legal action ("please don't sue us"). The workers' actual safety is often secondary.

The classic WAA case - an autoantibody that prefers the presence of a normal e antigen on its target red cells. May or may not be compatible with e- (R2R2) cells, but often shows weaker reactivity, especially in it's early stages of development. By the time the autoantibody gets to the 4+ stage (complete, solid agglutination), there are rarely any weaker cells. The question: Do you transfuse e- (R2R2), i.e., the "least incompatible" ? From a rarity/inventory point of view, in the short term that may be manageable, but probably unsustainable long term. The decision gets more complicated when the patient is E-. One could argue that there's a significant chance that by transfusing "double-dose" E+ cells, you'll cause the patient to make anti-E. Oops. Next time around, the R2R2 option may be off the table. And ultimately, as Petz opines, there's scant evidence to support that "least incompatible" cells (e- in the above case) survive in vivo any better than random cells.

You did indeed read my mind. O cells (panels) nonreactive, A cells (XMs) reactive = A2 with anti-A1.

What is the ABO group of the patient and of the crossmatched units ?

He passed peacefully in his home near Wilmington, NC.

I'm in line with the above answers. A current diagnosis would be useful, especially to give us an idea if the aforementioned blood products (platelets, plasma) may be in play. There has to be some kind of more recent stimulus.

Thanks, Sandra. As I'm sure you knew, I am aware of the answer - no way, no how are blood suppliers going to "discard" ~10% of their product. But I think it's important to consider the consequences of some of the now routine testing algorithms. No testing mean results are unknown, but once one has information, one may be required to take action. Many transfusion protocols for chronic users involve Rh (C/E) and K matching - there's another batch of donors whose (partial) phenotype is known and considered to be quite immunogenic. It goes on. I do find it interesting that your system "hides" the K+ status, but openly prints the K- attribute on the label. Another thought: If the K type of all of the patients were known, they could get the K+ units. The antigen frequencies should match up.

At that point, you could probably test the Last Wash for specificity !! You'd probably get the results you need.

I have always run the eluate and last wash in parallel, but the question did make me think. I appreciate the attempt to reduce (potentially unnecessary) work, but don't like the idea of doing two-stage testing (eluate first and then Last Wash, or the other way around). If this happens, you've lost any efficiency (and time) you believed you gained by not testing the eluate and Last Wash in parallel. However, the cautious approach to a low volume (rare) specimen may have some merit - checking the Last Wash first adds confidence that any eluate prepared from the washed cells will be more likely to be valid. The two-stage testing concept has crept into the laboratory over the last couple of decades and is completely valid for follow-up or reflex testing. But.....one of my peeves: Reagents that suggest "Immediate spin, incubate negatives". If you're running a negative control anyway and/or most of your tests will be negative (DATs with anti-Complement reagents), you'll almost always be incubating, so why bother with the Immediate Spin ? Bottom line: Do what the eluate kit manufacturer says (unless you've validated otherwise).

Hah ! Excellent point, David. I wonder how much emphasis should be put on those "microscopic" reactions, especially when the endpoint a titration is most often defined as the "last MACROscopic (or 1+) reaction"? How do the Powers-That-Be justify that little nugget ? Can you imagine resulting-out a "change in titer" based on a microscopic reaction ? Talk about piling-on to the already acknowledged confusion level.

An excellent discussion point. I think many others have similar questions and concerns. The have been several other threads on this forum with similar subject matter. As an Old Fart, I feel obliged to spout some (un-referenced) history. Most of the original work on clinical significance of antibodies in pregnancies was done in the absence of potentiators and definitely before the use of (semi)automated test systems. I think it was a "saline-IAT" using 22% albumin (BSA) as a diluent. Most of those antibodies were anti-D, for obvious reasons. There's not much out there in the literature in terms of controlled or organized studies regarding other specificities. There are a fair number of one-of-a-kind case studies, but most of the stuff is retrospective analysis of data. Basically, other than anti-D, nobody really knows what an antibody titer means, but as Ensis01 suggests, detecting a change in titer (increase) may be more important. In an era when basic tube shaking is going away, it only makes sense (we have no other option) to convert to the new techniques and equipment, but I suspect that it has the potential to further confuse an issue which already has enough confusion to (dis)satisfy everyone. I don't envy anyone handling this hairball. As a last thought...the high-powered potentiators (and techniques) used today don't reflect what's going on in vivo. Arguably, if one ignored the 22% BSA diluent, the saline-IAT is a better mimic of the in vivo scenario.

Just curious.....what does the system do with those potentially immunogenic K+ units (donors) ?

I am not aware of any LICENSED anti-Cob available in the USA. The American Red Cross system uses an unlicensed version for most of its Cob phenotyping requests. I believe Cob can be determined by the HEA BeadChip process (molecular typing). If you have a patient that needs Co(b-) red cells, I think those are your two options: PHENOtyping with an unlicensed reagent or units predicted to be Co(b-) by GENOtyping.

My personal favorite Ch/Rg confirmatory test: Strong reactions with C4d-coated cells. Old school, but really cool to see.