Jump to content

Computer Crossmatching


NewBBSup

Recommended Posts

We will be implementing computer crossmatching this spring.  I'm starting to write the procedure for this.  Would anyone be willing to share their procedure with me?  I'm having a problem coming up with something intelligent for the purpose section.

I also have seen in previous feeds that most of you do NOT require an ABO/Rh verification on the sample if another tech wants to set up units for a computer crossmatch.  Our current policy now is if another tech crossmatches on a sample that someone else did the type and screen on they must also type the patient's sample for verification.  I'm hoping I can eliminate this for speed and ease since the computer crossmatch is supposed to be quicker.

Link to comment
Share on other sites

We do not require another tech to perform another verification of blood type just to use the computer for crossmatching.  The computer is set up to require two ABORh's.  Our policy is that we require two separate draws by two different phlebotomists when we have a patient that our computer system has never seen.  The first ABORh is performed by our automated platform on one of the samples and then a technologist performs the second ABORh on the second sample.  Our computer system, and I would hope yours too, keeps that historic information for the next time the same patient visits.  The second time that patient arrives in our institution, we only draw one sample for testing and perform the ABORh.  At this time, our computer recognizes the historic blood type and the current (i.e. Two ABORh's are present for computer crossmatching).  If the patient meets our other criteria, then we are able to perform computer crossmatch.  FDAEXM_Guidance.pdf

I am not aware of any regulatory or accreditation requirements that state testing needs to be performed the way you describe your current process.

Link to comment
Share on other sites

5 hours ago, NewBBSup said:

Our current policy now is if another tech crossmatches on a sample that someone else did the type and screen on they must also type the patient's sample for verification.

This policy is for serological crossmatches; it ensures the correct ABO for the specimen on the tech's board (rack). For electronic crossmatch the ABO is checked by the computer when the product is selected, which makes a manual ABO retype by a tech redundant.

The purpose ticks the boxes for speed, efficiency and cost.

Link to comment
Share on other sites

Electronic XM should be part of your crossmatch procedure.  EXM is just another option for crossmatching, like IS or an AHG/IGG crossmatch. 

If you're doing a whole new procedure - the purpose is to use the computer to determine ABO compatibility.  That's all you're really doing - using the computer logic instead of a immediate spin XM to determine ABO compatibility.  That's why you need multiple ABO's on the patient to qualify.

Link to comment
Share on other sites

For what it's worth, here's the grid we follow to validate our EXM process.

Please note, this needs to be done annually or whenever you change/upgrade software or hardware.

We use Sunquest, here's some abbreviations of functions/test defined so it'll make sense:

ABR = test code for ABO/Rh

BAD = BB Admin Data, patient's permanent record

AS = Antibody screen

EXM = Electronic XM

BPI = Blood product issue

Validation 7.1 EXM grid 040416.doc

Link to comment
Share on other sites

22 hours ago, NewBBSup said:

We will be implementing computer crossmatching this spring.  I'm starting to write the procedure for this.  Would anyone be willing to share their procedure with me?  I'm having a problem coming up with something intelligent for the purpose section.

I also have seen in previous feeds that most of you do NOT require an ABO/Rh verification on the sample if another tech wants to set up units for a computer crossmatch.  Our current policy now is if another tech crossmatches on a sample that someone else did the type and screen on they must also type the patient's sample for verification.  I'm hoping I can eliminate this for speed and ease since the computer crossmatch is supposed to be quicker.

Which computer system are you using?  Are these "type the patient's sample for verification" being recorded in your current computer system?

Link to comment
Share on other sites

On 1/25/2017 at 10:26 AM, NewBBSup said:

 

We will be implementing computer crossmatching this spring.  I'm starting to write the procedure for this.  Would anyone be willing to share their procedure with me?  I'm having a problem coming up with something intelligent for the purpose section.

Here is what I wrote for our purpose:

Compatibility testing is all pre-transfusion testing performed on a potential transfusion recipient and the appropriate donor blood, in an attempt to ensure that the product will survive in the recipient and induce improvement in the patient’s clinical condition; the electronic crossmatch is performed with the ABO compatibility truth tables built in a computer system. 

The electronic verification of donor/recipient compatibility has many advantages.  In addition to the major benefits of increased patient safety and decreased unnecessary work, the electronic crossmatch has advantages  that include time savings, job simplification, increased throughput, decreased turnaround time, reduced patient sample requirements, and reduced handling of potentially hazardous samples, as well a avoidance of false-positive and false negative reactions associated with serological crossmatches.

Link to comment
Share on other sites

On 1/25/2017 at 0:26 PM, Malcolm Needs said:

What is a computer crossmatch?  How does the computer do a crossmatch?

I think you mean electronic issue?

Malcolm, you need a hobby, this is being a little nit picky.  :rofl:  But then I guess it's now our job to keep things set folks straight.  I do think that the common term used on this side of the pond is either electronic crossmatch or computer crossmatch but I have to agree that electronic issue is probably more accurate.

Link to comment
Share on other sites

11 hours ago, John C. Staley said:

Malcolm, you need a hobby, this is being a little nit picky.  :rofl:  But then I guess it's now our job to keep things set folks straight.  I do think that the common term used on this side of the pond is either electronic crossmatch or computer crossmatch but I have to agree that electronic issue is probably more accurate.

I suppose so John, but like Peter Issitt before me (not that I am comparing myself with the great man), I am very keen on the use of correct nomenclature.  Indeed, if my failing old memory serves me well, my first post on this site was all about that subject.

Link to comment
Share on other sites

22 minutes ago, tricore said:

This is what FDA calls it! Ah, semantics.

Guidance for Industry "Computer Crossmatch"

(Computerized Analysis of the Compatibility between the Donor’s Cell Type and the Recipient’s Serum or Plasma Type)

Also known as CAX.

I am not arguing that the FDA call it a computer crossmatch, indeed, I am sure they do.  However, what I am asking is, where in the entire procedure does a crossmatch, particularly a crossmatch performed by a computer, take place?  The answer is, it doesn't!

Link to comment
Share on other sites

2 hours ago, Malcolm Needs said:

I am not arguing that the FDA call it a computer crossmatch, indeed, I am sure they do.  However, what I am asking is, where in the entire procedure does a crossmatch, particularly a crossmatch performed by a computer, take place?  The answer is, it doesn't!

In the Meditech computer ecosystem, the crossmatch occurs when the user enters donor unit numbers into the BBK Enter/Edit results routine. Blood components are released for transfusion using the BBK Issue Units routine which is separate and distinct from the BBK Enter/Edit results routine.

Link to comment
Share on other sites

7 minutes ago, Dansket said:

In the Meditech computer ecosystem, the crossmatch occurs when the user enters donor unit numbers into the BBK Enter/Edit results routine. Blood components are released for transfusion using the BBK Issue Units routine which is separate and distinct from the BBK Enter/Edit results routine.

But in what way, EXACTLY, is that a crossmatch?  Certainly the computer's logic programmes are "selecting" a unit, but they are not performing a crossmatch.  This is why a serological crossmatch is required if a clinically significant antibody (other than ABO antibodies) are present (well, that and the fact that once an antibody has been identified, there is every chance that there are other specificities, often directed against low prevalence antigens, present in the plasma).

Link to comment
Share on other sites

2 hours ago, Malcolm Needs said:

But in what way, EXACTLY, is that a crossmatch?  Certainly the computer's logic programmes are "selecting" a unit, but they are not performing a crossmatch.  This is why a serological crossmatch is required if a clinically significant antibody (other than ABO antibodies) are present (well, that and the fact that once an antibody has been identified, there is every chance that there are other specificities, often directed against low prevalence antigens, present in the plasma).

The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching.  

Are you restricting the definition of a crossmatch to include only serological methods?  In 1987 the AABB allowed users to limit the serological crossmatch to the immediate-spin test if the antiglobulin antibody screen test was negative.  Then, in 199?, the AABB allowed users to replace the immediate-spin crossmatch with a "computer crossmatch" to detect ABO incompatibility.  Before this, the word "crossmatch" was prefaced with "major" and "minor".  So the definition of "crossmatch" has undergone many changes over the years.

The FDA allows Transfusion Services to use a qualified computer system to detect ABO incompatibility between donor and recipient in lieu of serological methods to detect ABO incompatibility, i.e., as stated by tricore, "Computerized Analysis of the Compatibility between the Donor’s Cell Type and the Recipient’s Serum or Plasma Type".

 

Edited by Dansket
changed wording
Link to comment
Share on other sites

1 hour ago, Dansket said:

The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching.  

Are you restricting the definition of a crossmatch to include only serological methods?  In 1987 the AABB allowed users to limit the serological crossmatch to the immediate-spin test if the antiglobulin antibody screen test was negative.  Then, in 199?, the AABB allowed users to replace the immediate-spin crossmatch with a "computer crossmatch" to detect ABO incompatibility.  Before this, the word "crossmatch" was prefaced with "major" and "minor".  So the definition of "crossmatch" has undergone many changes over the years.

The FDA allows Transfusion Services to use a qualified computer system to detect ABO incompatibility between donor and recipient in lieu of serological methods to detect ABO incompatibility, i.e., as stated by tricore, "Computerized Analysis of the Compatibility between the Donor’s Cell Type and the Recipient’s Serum or Plasma Type".

 

Yes, Dansket, I am restricting the definition of a crossmatch to include serological methods - otherwise it is a "selection" or "issue", and not a crossmatch.

If we are going into the history of this, Ludvig Hektoen suggested that the safety of transfusion might be improved by crossmatching blood between donors and patients to exclude incompatible mixtures in 1907, although this took a bit of time to catch on.  The AABB ruling in 1984 (not 1987) that allowed users to limit the serological crossmatch to the immediate-spin test (see the 11th edition of their manual) if the antiglobulin antibody screen test was negative followed on from papers by such luminaries as Boral and Henry (a good read) and, chiefly, the workers at Ann Arbor, including John Judd.  This ruling was confirmed by the FDA in a letter, also in 1984.

The computer issue was first allowed by AABB in 1993, but confirmed by the FDA in 1994 as long, as you say that the patient's ABO type had been confirmed twice (and found to be the same), and as long as the computer system contains logic to alert the user to discrepancies."

You may want to have a brief look at "Developments in Compatibility Testing" that I put in the library section of this site in November 2009, but I am not insisting (not that I could) or reading the AABB book by Steven Pierce and Marion Reid, "Bloody Brilliant!  A History of Blood Groups and Blood Groupers." 1st edition, 2016 (a superb book in my opinion).

I must admit to being a little confused by your first sentence, "The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching.", as in "The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching." there is human intervention by the "selection" by the human, and not the computer, means that it is not, by definition, either a "computer crossmatch" or a "computer issue"?  I am quite prepared to be persuaded otherwise.

Lastly, although I thoroughly and absolutely concede that the FDA do call it a "computer crossmatch", in no way am I prepared to concede that the FDA are correct in their nomenclature.

Link to comment
Share on other sites

3 hours ago, Malcolm Needs said:

I must admit to being a little confused by your first sentence, "The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching.", as in "The user (not the computer) in my facility selects a RBC component from the refrigerator for crossmatching." there is human intervention by the "selection" by the human, and not the computer, means that it is not, by definition, either a "computer crossmatch" or a "computer issue"?  I am quite prepared to be persuaded otherwise.

Some computer systems allow user to select the component from a list displayed on screen, very, very risky in my opinion.  My approach requires user to select component from refrigerator and scan the barcoded DIN into the BBK Enter/Edit results routine.  These same steps are done, whether a serologic or electronic crossmatch is indicated.  The only difference is that a donor tubing segment must be removed from the blood container for a serologic crossmatch.

Based on my reading of your reply, do I understand correctly that you have not personally configured or operated a computer system qualified for the computer crossmatch in a Transfusion Service?

Prior to any "crossmatching" in a computer system, user must enter results of ABO, Rh and Antibody Detection testing on the patient and ABO/Rh confirmation testing on the rbcs from a donor tubing segment.  This means results of anti-A, anti-B, anti-D, Rh control, A1 cells and B cells for the patient and anti-A, anti-B, anti-D, and anti-A,B (however it is done) on the donor rbcs are entered into the BBK Enter/Edit results routine.  The user entered results are compared to "truth tables" and ABO/Rh interpretations are generated accordingly.

It is simple process for the computer to "match" both patient and donor unit when both are ABO/Rh identical.  It is more complicated for the computer to "crossmatch" when the ABO/Rh of patient and donor unit is not identical.  "Crossmatching" is accomplished by another "truth table" that is configured by the user to allow selection of ABO/Rh compatible but not identical donor units for a patient.

In my system, if a patient qualifies for a "computer crossmatch", the system automatically enters 'NP' as the results of the immediate-spin and IgG crossmatches.  If a patient does not qualify for a "computer crossmatch", user must enter serological test results for the immediate-spin and IgG crossmatches.

BTW, I believe your dates for article by Boral and Henry are incorrect.  Their article was published in the journal Transfusion in 1984 and the AABB approved the concept of the Type and Screen in 1987.

Link to comment
Share on other sites

6 hours ago, Dansket said:

Some computer systems allow user to select the component from a list displayed on screen, very, very risky in my opinion.  My approach requires user to select component from refrigerator and scan the barcoded DIN into the BBK Enter/Edit results routine.  These same steps are done, whether a serologic or electronic crossmatch is indicated.  The only difference is that a donor tubing segment must be removed from the blood container for a serologic crossmatch.

Based on my reading of your reply, do I understand correctly that you have not personally configured or operated a computer system qualified for the computer crossmatch in a Transfusion Service?

Prior to any "crossmatching" in a computer system, user must enter results of ABO, Rh and Antibody Detection testing on the patient and ABO/Rh confirmation testing on the rbcs from a donor tubing segment.  This means results of anti-A, anti-B, anti-D, Rh control, A1 cells and B cells for the patient and anti-A, anti-B, anti-D, and anti-A,B (however it is done) on the donor rbcs are entered into the BBK Enter/Edit results routine.  The user entered results are compared to "truth tables" and ABO/Rh interpretations are generated accordingly.

It is simple process for the computer to "match" both patient and donor unit when both are ABO/Rh identical.  It is more complicated for the computer to "crossmatch" when the ABO/Rh of patient and donor unit is not identical.  "Crossmatching" is accomplished by another "truth table" that is configured by the user to allow selection of ABO/Rh compatible but not identical donor units for a patient.

In my system, if a patient qualifies for a "computer crossmatch", the system automatically enters 'NP' as the results of the immediate-spin and IgG crossmatches.  If a patient does not qualify for a "computer crossmatch", user must enter serological test results for the immediate-spin and IgG crossmatches.

BTW, I believe your dates for article by Boral and Henry are incorrect.  Their article was published in the journal Transfusion in 1984 and the AABB approved the concept of the Type and Screen in 1987.

Your post answers a lot of questions.  The definition of the computer crossmatch/computer issue seems to be much looser in the USA than in the UK, and I think that it is this that is causing the misunderstanding between the two of us.

In the UK, we are governed by Guidelines produced by the British Committee for Standards in Haematology (BCSH), now the British Society for Haematology (BSH), and the actual Guideline dealing with computer issue of blood is, Milkins C, Berryman J, Cantwell C, Elliott C, Haggas R, Jones J, Rowley M, Williams M, Win N.  Guidelines for pre-transfusion compatibility procedures in blood transfusion laboratories.  Transfusion Medicine 2013; 23 3-35. (doi:10.1111/j.1365-3148.2012.01199.x), and this guideline specifically excludes any patient or sample from electronic issue of blood if, at any stage, there has been human intervention at any stage in entering the ABO or D typing into the computer.  In other words, the grouping has to be performed by automation, this typing has to be read and interpreted by the computer on board of the automation and then transferred electronically to the patient's record by computer.  There must be two separate identical groups (one can be historical, or it can be performed on a second sample, but not the initial sample grouped twice), and there must be no clinically significant atypical alloantibody detected in the plasma, or have been detected in the plasma in the past, otherwise electronic issue/computer issue is forbidden.

From this it can be seen that, whereas you say that your user must enter results of ABO, Rh and Antibody Detection testing on the patient and ABO/Rh confirmation testing on the rbcs from a donor tubing segment.  In the UK, such human intervention would not be allowed and, incidentally, the units are typed by automation at the blood centres (no human intervention), the computer and eye readable labels produced one at a time for each pack (human intervention here, as these must be stuck on to the units, but the Quality Control involved here is frighteningly strict) and then the units are wanded in by bar code reader at the hospitals, again, with no human intervention.  In the UK therefore, at no stage is there ANY serological crossmatch, and the computer does no crossmatch either, but does select and print the labels of suitable units, which it "chooses" (not the human) using the logic files.  It will not allow issue of different, but compatible, groups (e.g. group O to a group A).  It is all very strict in the UK.

Based on my reading of your reply, do I understand correctly that you have not personally configured or operated a computer system qualified for the computer crossmatch in a Transfusion Service?  You are absolutely correct in your assumption of never having personally configured a computer system qualified for electronic issue/computer issue, although I have operated one.  As I have often stated on this very site, I am an IT idiot, and I should be the last person let loose on configuring ANY computer system of any sort!

BTW, I believe your dates for article by Boral and Henry are incorrect.  Their article was published in the journal Transfusion in 1984 and the AABB approved the concept of the Type and Screen in 1987.  Sorry, but no, I didn't put any dates in for the Boral and Henry paper (see my posts above), but the one to which I was referring was considerably earlier than 1984 (Boral LI, Henry JB.  The type and screen:  A safe alternative and supplement in selected surgical procedures.  Transfusion 1977; 17: 163-168.), but the year 1984 for the approval by the AABB for the concept of the Type and Screen (and the 11th edition of their manual) is taken from  Steven Pierce and Marion Reid, "Bloody Brilliant!  A History of Blood Groups and Blood Groupers." 1st edition, 2016, page 482, column 1.  I do not know Steven Pierce, but I have met and chatted to Marion Reid (and Christine Lomas-Francis) on several occasions, and if Marion says that date is correct, then it is correct in my mind!  Incidentally, they also quote the same paper by Boral and Henry as do I.

Link to comment
Share on other sites

OK, at the risk of certain people becoming cross with me, I will cross over to another line of definition of the term and point out it seems as if, in the colonies anyway, that the first part of the term may be seen as crossing  over from the word "across".  

So in that sense, when performing a crossmatch, one is comparing data across two points. In the case of a serological crossmatch, it is from the antigens on the red cells across to the antibodies present in the patients serum or plasma.  In the case of a "computer" crossmatch, one is comparing data from a computer's patient record across to the information (also stored on the computer) about the red cells in the unit.

Thanks, now I have to go back to work, got all these pesky patients to deal with here...

Scott

Link to comment
Share on other sites

10 minutes ago, SMILLER said:

OK, at the risk of certain people becoming cross with me, I will cross over to another line of definition of the term and point out it seems as if, in the colonies anyway, that the first part of the term may be seen as crossing  over from the word "across".  

So in that sense, when performing a crossmatch, one is comparing data across two points. In the case of a serological crossmatch, it is from the antigens on the red cells across to the antibodies present in the patients serum or plasma.  In the case of a "computer" crossmatch, one is comparing data from a computer's patient record across to the information (also stored on the computer) about the red cells in the unit.

Thanks, now I have to go back to work, got all these pesky patients to deal with here...

Scott

Enjoyed that!

I going to stick with the other name we use for it, which is Electronic Issue!

Link to comment
Share on other sites

Of course, since almost everyone has a computerized system of some sort, all "issuing" is done "electronically",

but I'mnotgoingtobotherwiththeIScrossmatchrightnowthankyou really is too  much of a mouthful for what we are actually doing, even though it is linguistically correct.  (Linguistic, from the Latin for "pretentiously verbal".)

OK, now I am really going back to work...

Scott

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Advertisement

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.