CAP TRM.31450 Comparability of Instrument/Method

[Ardele Hanson]

"If the laboratory uses more than one instrument/method to test for a given analyte, the instruments/methods are checked against each other at least twice a year for correlation of results."

My hosptial uses the Echo. We also have a Manual Station Solid Phase platform for antibody screens and antibody IDs. And we also have tubes,

My question is: What happens when the Echo gets a 2+ for a Jka, the manual platform gives a 1+, and the screen is totally negative in tubes?

Am I compliant? The Echo and solid phase are so much more senstive. Tubes are not going to 'catch' every single antibody - we know that.

So, what do I do? Should I not be comparing tubes to solid phase (apples to oranges)? Or just say I am okay?

Thanks for your help.

[R1R2]

Pick the samples you use for the study carefully. Pick a postiive sample that will work with all methods, like a 4+ anti-D.

[Eagle Eye]

-----My two cents-------------

ECHO has same methodology as Solid phase? If you are you using solidphase & ECHO interchangeably you must have validated/comapared you solidphase and ECHO. If you are using solidphase for ABID and ECHO for screening

& if your solid phases is not giving same reaction as ECHO, you will be questioned. Tube : we know that is less sensitive so you can explain that.

This is what we do: Run same specimen on ProVue, on Gel station and by tube. Atleast once a year I make a specimen with two antibody, one weak & one strong antibody. I had demonstrated that the antibody was picked up by ProVue, manual gel (same strength). By tube we picked up only one antibody but we explained in our summary I explained that by tube the antibody was missed due to sensitivity of the method. ...was signed off by medical director and we had no problems..Our CAP inspector liked it.

we do the same for all tests: type, screen, abid, antigen typing and crossmatch.

Tube is not our primary method for abid, but we use it time to time for the investigation of Ab. (we only use gel for crossmatching)

[dld]

Ardele-

We use manual gel, tubes with PEG as a backup, or tubes with no enhancement. I do this pointless comparison of all three methods we use. Gel is more sensitive than tubes/PEG, and PEG is more sensitive than tubes with no enhancement, so when I get 2+ in Gel, 1+ with PEG, and w+ or neg with no enhancement, I just write "correlates as expected".

Really this is a complete waste of time because we all know how the different methods should compare to each other, but I do it anyway so we don't get dinged for our CAP inspections.

Hope this helps you out.

[R1R2]

Ardele-

We use manual gel, tubes with PEG as a backup, or tubes with no enhancement. I do this pointless comparison of all three methods we use. Gel is more sensitive than tubes/PEG, and PEG is more sensitive than tubes with no enhancement, so when I get 2+ in Gel, 1+ with PEG, and w+ or neg with no enhancement, I just write "correlates as expected".

Really this is a complete waste of time because we all know how the different methods should compare to each other, but I do it anyway so we don't get dinged for our CAP inspections.

Hope this helps you out.

I agree with dlb, this is a complete waste of time, don't overthink this.

R1R2

[Ardele Hanson]

I totally agree with all of you. And, no, we do not use the tube method for ABID - maybe once in a great while one cell from one panel kit; but tubes are not used for identification of antibodies.

The reason I ask is because I work with a person who is demanding that all methods match. (Her title is System Technical Coordinator. Hospital System is 7 hosptials, only 3 use the Echo. Others use the manual solid phase station or tubes.) And it is frustrating to even try to explain anything to her. I have tried!

Thank you for your help. I will share all your posts with her.

[David Saikin]

While I agree it is a waste of time, my comparison always expects weaker reactions in tubes (using both PeG and LISS) than with gel. sometimes tubes will be negative while gel is reactive . . . that is also expected.

[MAGNUM]

So what happens when you go away from gel and use the Echo for the main platform, and tubes as your backup? Are you not testing the same platforms since the Echo and the tubes use the same reagents?

[L106]

I added the following comment to our procedure for Correlation Studies:

"It is recognized that the reactions or strength of reactions may not be exactly the same for the testing done on the Echo and testing done with PeG/tube technique. It is the responsibility of the Blood Bank Supervisor to evaluate whether any variation is acceptable."

Donna

[David Saikin]

So what happens when you go away from gel and use the Echo for the main platform, and tubes as your backup? Are you not testing the same platforms since the Echo and the tubes use the same reagents?

Yes the reagents are the same; the platform is different (automated vs manual). As intimated above and in other threads - this does not seem appropriate for blood bank but . . . we don't set the rules.

[bmarotto]

What about ABO & Rh? We do both automated and tube. The primary test is done by automated solid phase. Confirmatory blood types if there is no historical type are done by tube, albeit on a different sample. We do about 15 tube confirmatory blood types a day. It seems that we have method comparison built in to our daily workflow but will CAP see it that way? I am wondering if I need to add ABO & Rh to the twice a year comparison.

Edited April 3, 2013 by bmarotto
typo!

[Deny Morlino]

Document the comparison specifically twice yearly so their is not a question as to whether it is performed or not. Remember the old addage, "If it was not documented, then you did not do it".

[Eagle Eye]

Agree. We document our tube comparision on our CORD specimen. Twice a year...actually we use split sample for this.

[tbostock]

I joke that I do this correlation to prove that the methods don't always correlate.

I don't hand pick strong ones, I just randomly pick some positives and run them on Tango (solid phase), gel, PEG, and LISS. If they don't all agree, I explain why...the apples and oranges thing. In BB, if there was a perfect method that detects everything, we would all be using it, wouldn't we?

[PAWHITTECAR]

Terri that is an excellent way to put it. I am going to use that the next time I get questioned as to why I got a positive screen (Gel) and ARC got negative (tube).

[tbostock]

Terri that is an excellent way to put it. I am going to use that the next time I get questioned as to why I got a positive screen (Gel) and ARC got negative (tube).

Here's the official lingo from my policy:

PRINCIPLE: The Blood Bank performs correlation of tests methods to evaluate the results of the same test performed with different methodologies. Methods in Blood Bank may not correlate 100% due to sensitivity and specificity of each test method. The goal of correlation of methods is not to achieve perfect correlation, but to understand the sensitivities, specificities, and limitations of the different methods used in Blood Bank. Some methods (e.g. Tango and LISS antibody screen) are not used interchangeably, but comparatively, when interfering substances (rouleaux, cold agglutinins, etc) may interfere with a particular method. No method in Blood Bank will always detect all clinically significant antibodies.

[PAWHITTECAR]

I love it!!

[nicole]

We also use the Echo with manual Capture stations as our backup, and have Liss tubes available as well. We very rarely go back to the tube methodology, only in those patients with a known adversion to solid phase. Manual Capture and the Echo do not correlate all that well even though they are the same reagents. The reactions are much stronger on the Echo. I was told that this is probably due to the heating mechanism in the Echo, it comes to temperature much sooner. All you can do is report what you get.

[geekay]

wow...thanks for the suggestion..sounds great.....

[tbostock]

P.S. I do the correlation for all tests that I do with different methods: ABO/Rh, Cord blood, DAT IgG, Antibody Screen, Weak D, AHG crossmatch.

[dalene]

Hi all,

Just finished our mock CAP inspection for this year. Question is how many samples do you run for the correlation studies?

[AMcCord]

I love it!!

Ditto!

[AMcCord]

How many samples? Well....that's a good question.

For my ABO/Rh correlations, I specify that every 2nd type on a new patient is part of the correlation. This is spelled out in my policy. Any discrepancies are investigated and documented (actually, this has only happened once and that was a weirdly interesting D antigen). So for that we have overload.

For antibody detection, we use the negative control for tube QC as pat of our correlation with the Echo. For tube QC we select a sample that had a negative antibody screen according to the Echo, then run an antibody screen on it as part of daily QC with PeG and LISS. After 4+ years of doing this, we have only had a discrepancy a couple of times - each discrepancy was the result of a Capture only non-specific antibody. This type of discrepancy is noted in the daily QC log and on the correlation report form as method dependent. To compare positive results, in each half of the year we pick a minimum of 4 antibody positive samples from the Echo and repeat them with PeG and LISS. I specified that 2 must be Rh antibodies (not RhoGAM anti-D) and 2 must be non-Rh AHG reactive antibodies like anti-K, anti-Jka, etc. If the Echo is more sensitive than the tube method, so be it - the discrepancy is method dependent and documented accordingly. I usually have 5 or 6 correlation samples, depending on what samples we get with clear cut antibodies and sufficient sample size. (We don't have a large number of antibody positive samples, so this number is small. I don't include warm autos.)

Antibody ID - anytime we do additional testing by tube to finish up an identification we start on the Echo, we have a candidate sample that could be used as method comparison, if the workup is extensive enough (not just a few selected cells). We use PeG routinely, so that covers that. If there is plenty of sample, a repeat of the PeG tested cells using LISS and saline completes the comparison, which is then documented on the correlation report form.

The whole thing is easier for us (we're not a large facility) if it's incorporated into our routine workload.

Edited April 16, 2013 by AMcCord

[marquij@caribe.net]

What would be considered acceptable criteria???

[Eagle Eye]

Let your Medical Director define and write that in policy...