Mabel Adams

I like to grade mixed field. It gives a better record of what the tech actually saw.

Sorry, that is still on my to-do list, but moving toward the top. We have done some education, but I envision some sort of flow-chart that has a route that leads to TRALI.

It should be possible in Meditech to build a rule that looks at the cord results for weak D and at least pops up a message box to remind the tech to follow through with the KB testing rather than Fetal screen. We currently use a T test on the cord blood where we check the mom's type to make sure she is Rh neg, then we put in a comment stating that the mom needs a fetal bleed test ordered. This prints out as a message to the OB floor to know what to do next. It probably wouldn't be too hard to make it look at the cord results too and even turn out a statement that they should order a KB on the mom instead. The biggest problem with computers in this case is that the cord blood and the fetal bleed test are on two different patients so various test reflexing systems don't work.

I assumed that this was a public site which makes it searchable by Google or any other search engine. It is easy to think we are alone in our little blood bank world, but anyone can read what we post. When I Google my name, my BB talk posts are what comes up. Good to think about if you are in the job market. I've heard that some employers do online searches--probably more of an issue for young folks with their drinking pictures on Myspace than us staid and seasoned blood bankers, but hey, you never know. This maybe should make us think twice about some of our jokes (darn).

If you have a frequent need you could go to thawed plasma and keep it already thawed.

I like to repeat the story of an inadvertant Rh pos transfusion to an Rh neg young female patient. They sent the patient for a red cell exchange by pheresis, then did a Kleihauer and calculated the dose of RhIG based on residual Rh pos cells. It is such a novel approach I like to add it to people's lists of options.

I agree that the 4 hour limit stems from an open system stored at room temp, not something specific to cryo. Seems like it should be 6 hours. Our supplier is getting ready to provide pre-pooled cryo. I will try to ask them this question.

I decided to create all the ISBT product codes rather than lump them together as we did with codabar. I suspect that in the future we will be glad we separated them as ISBT becomes the standard everything is built around. Disentangling past products would be a nightmare. One additional comment about pheresis products always having separate product codes (1st & 2nd containers): our supplier says they could split a large pheresis plasma donation into 2 parts, but this would be done as ISBT divisions, not as separate product codes. So we had to be sure our system could scan in aliquots that were shipped to us. Pretty rare and some suppliers may not do this.

Sounds like it would be easier to just get a sterile connecting device rather than deal with changing from a closed product code to an open one in Meditech.

That's what I had heard too. I think it is even in the notes of our procedure.

Gil, this might be one of those cases where knowing about transfusions at other facilities makes interpretation easier. It is really easy to skip getting history--and history isn't always easy to get or accurate--but not knowing can sometimes cause confusion, as in this case, or sometimes burn us because we thought the patient was not recently transfused. I will admit to being guilty sometimes of not pursuing history if the case seems clear--but someday it will get me.

We are on 5.61 Magic. Can you tell us the DTS that allows Meditech to read the extended part of the product code barcode? I had been told that this was not possible. When we log in autos, we manually add the lower case L in Other Data so it is flagged as auto. We keep track of autos by the fact that they are the only RBCs we get that aren't LR. I understand statistics aren't working too well in MT for ISBT due to the product code changes etc. I haven't run into the problem yet--probably because of how simply we do our statistics.

The Rh decision should be based on FMEA ideas--how likely is it that the wrong Rh will be given and how bad are the consequences. Compared to ABO, the consequences are significantly less (hemolytic reaction and possible death vs. sensitization). Odds of being wrong are based on the facts that 1) we usually have it right on the first specimen, 2) 85 percent of the wrong patients will still be Rh pos and 3) we are much less likely to transfuse females with childbearing potential than older folks or young male traumas. If I am doing the statistics right, the risk, if you are issuing non-O blood but the patient type turns out to be wrong, would be ~55% if you issue A, ~80% if you issue B and ~96% if you issue AB. Those are all quite a bit higher than the Rh risk percentage of ~15%. If we wanted to do it perfectly we would genetically type all patients and donors and only give phenotypically matched units for all antigens--after we have confirmed ID via DNA -- but we have to do what is cost-effective and currently reasonable. Avoiding killing people with mistranfusion is the main goal at this point. Thankfully it is a rare error, but the consequences are so dire, it is worth a great deal of effort to prevent it. Let me know if I overlooked anything.

Wouldn't the patient himself be able to supply plenty of complement for any hemolytic reaction? I remember when we used to add fresh serum in testing to make sure we had complement in the system. It didn't have to come from the donor or the patient. I suppose a complement deficient person could be like an IgA deficient person and make an antibody to foreign complement and go into anaphylaxis.

We go into the Other Info screen and add the auto lower case L manually. We get all LR red cells except for auto, so that helps us tell them apart for statistics etc. We are on 5.61 magic. I haven't heard of any newer versions that can handle this. Do you know the DTS# to make this change? Watch out for Meditech statistics programs. They might not capture all your new product codes so your counts will be way down. And irradiating an aliquot?? Oh, my!

Computype has a small quantity offering too (called CLOT, I think), but it was still cheaper to buy a 26 yr. supply from Shamrock (2 roll minimum for custom--then we quit getting 7 day plts.). Price per label didn't matter to us, because we were never going to use up anyone's minimum order before they change the names of the products etc. You don't have to relabel thawed FFP if only for in-house use (but you do thawed plasma). I didn't know about Digi-trax at the time so can't compare prices.

Some define the retypes as forward only and others do a full retype. I validated (because the typing sera insert was vague) using blue top tubes for blood types. I like that I am not ruining the specimen for add-on tests like I do the CBC tube. We have done it so far due to our new requirement to have a blood type for the current admission for FFP transfusion. We are still trying to decide on a retype policy. I think we would match the Rh so as not to deplete the Rh neg pool. I liked the suggestion of sending out a marked retype tube so no one is drawing two tubes the first time and sending the second one for the redraw.

We not only don't get any paper back from nursing, we don't provide the record for the chart in the first place. Our units have only a compatibility tag. (Less chance for mix-ups.) Nursing was using a transfusion flow sheet even when we still got papers back so they didn't need a transfusion slip for thier documentation. They used to put the slips in the chart but then they were in the Labs section and the flowsheet was in Graphs. Now the ID check etc. are part of the flow sheet and everything is in one place. (Well, the computer record of BB is still part of the Labs section.) They do their own auditing.

One of our frustrations is that our reference lab still uses tube so it would do no good to send out weak gel reactions as they would seldom find anything.

These patients must have been transfused a bunch to accumulate exposures to that many low freqs! Or they have very bad luck.

There is no "acquired A antigen", just B, so I'd guess that couldn't explain the case presented. Clerical/ID errors are quite a bit more common than acquired antigens. Probably bone marrow transplants are more common, but you can usually get that history by asking.

After making sure that all the antibodies for which commercial antisera is readily available have been ruled out (with a double dose cell if at all possible), repeating the gel tests with longer incubation and doing a gel xm, we transfuse. Occasionally I drop back to LISS when I am convinced I am dealing with a weak warm auto or HTLA-like antibody. I work really hard to rule out Kidd system antibodies because weak ones often won't react with one cell that they should but the only cells reacting are pos for the Ag and the patient is neg. I don't worry so much about non-Kidds. I will admit, if even Kidds are that weak, they still probably won't do that much damage. Kidd has such a reputation, I'm careful, but I have never seen it do its delayed TRX thing.

I have just asked that the peds committee review our policy for keeping cords. We hold them a week, but no one ever seems to request that we do any testing on them so it seems like a waste of energy. Of course we keep those we tested for RhIG determination in with our other samples for at least a week.

One problem that I once had like this was when admitting used a historical patient for a new patient with the same name. All the data (except blood type) matched between history and current, it just didn't match the actual patient that was in. Also, I hear reports of uninsured patients borrowing someone's ID so they can use their medical insurance coverage. Does your admissions office do photo ID checks and compare history carefully?

Why would washing cells help with a warm auto? Just removing other possible sources of reactions???