gel vs tube correlation

[apugh]

We are establishing our procedure to correlate our gel results with tube results every 6 months as JACHO says we have to do. How have you handled this? How many samples did you use to correlate gel/tube testing? Any ideas anyone has would be greatly appreciated! Thank you!

[David Saikin]

It seems to me that once you validate a procedure (gel vs tube), why would you need to continue to validate? Concepts like this are what made a lot of folks drop JCAHO . . .

[bbville]

We were cited (by a different accrediting agency than Joint Comission) for not having every 6 month correlations of manual gel/tube/ProVue testing. All other departments must do this if the same test(s) is/are reported by different instruments or methods, so I guess it makes sense in that way. We now just do a days worth of work or so by all three methods.

[Lcsmrz]

I always interpreted this standard to apply to quantitative results only, between multiple analyzers or between the routine method and the backup method. For the same reason, we don't compare our automated UA dipstick results against the visual interpretation, and we watch our Rapid Streps vs cultures only to monitor the kit's sensitivity -- we had a problem previously.

Comparing gel vs tube is comparing apples to oranges. Beyond your validation, exactly what information would you gain?

[BankerGirl]

I always interpreted this standard to apply to quantitative results only, between multiple analyzers or between the routine method and the backup method. For the same reason, we don't compare our automated UA dipstick results against the visual interpretation, and we watch our Rapid Streps vs cultures only to monitor the kit's sensitivity -- we had a problem previously.

Comparing gel vs tube is comparing apples to oranges. Beyond your validation, exactly what information would you gain?

I had the same interpretation of the correlations, but I have inspectors here now that say we have to pick 20 samples twice a year and perform every gel test on each sample, then repeat it all in tube. We mainly use gel but also have tube for backup, trouble shooting, ABO/Rh retyping. I told here that we perform daily ABO/Rh testing by both methods and periodically repeat antibody screen/ID/crossmatching by tube and that is not good enough. We have to call it correlation and perform the entire test menu on the same samples.:cries:

[Eagle Eye]

method correlation is required by CAP. My understanding is that, you have to decide/define how you are going to do method correlation. If you do not pick up weak antibody by tube and pickup by gel, you can write that in your summary that due to the sensitivity of the test, reult will not be same. call CAP, I do not think you need to do 20 specimen.

[David Saikin]

Run a weak reacting antibody in gel (2+ or less) and then run it in tubes (you'll get negative rxs) . . . so much for correlation. If you do advanced ab workups you can correlate WAIHA results - I always have to fall back on tubes/PEG. There is no rule that says you have to run 20 samples - you just have to correlate.

[BankerGirl]

Well, I wrote the procedure the way she wants it, but I still think she is interpreting the regs wrong. I don't see anything that says the entire test menu needs to be performed on every sample. That is my main contention.

[BBK710]

method correlation is required by CAP. My understanding is that, you have to decide/define how you are going to do method correlation. If you do not pick up weak antibody by tube and pickup by gel, you can write that in your summary that due to the sensitivity of the test, reult will not be same. call CAP, I do not think you need to do 20 specimen.

I just got my CAP checkist with the new TRM.31450. I understand why you would need to do a correlation if you are automated and use something else for backup if it's down or for stats etc.

We are not automated, use gel for antibody screen, antibody identification and AHG crossmatches. We use tubes for ABO/Rh. If we do antibody screens or panels in tubes it is not instead of gel, it is to aid in identification/interpretation of the results in gel. So I planning to check with CAP for some clarification on this reg. Comparing gel and tube methodology is really comparing apples to oranges and seems to me to be a waste of time and money.

[BankerGirl]

I just got my CAP checkist with the new TRM.31450. I understand why you would need to do a correlation if you are automated and use something else for backup if it's down or for stats etc.

We are not automated, use gel for antibody screen, antibody identification and AHG crossmatches. We use tubes for ABO/Rh. If we do antibody screens or panels in tubes it is not instead of gel, it is to aid in identification/interpretation of the results in gel. So I planning to check with CAP for some clarification on this reg. Comparing gel and tube methodology is really comparing apples to oranges and seems to me to be a waste of time and money.

AMEN!!!! I still feel like since we are doing it on an ongoing basis, AND documenting that we are doing it, we shouldn't have to repeat this.

[LaraT23]

They really aren't the same method when you think about it. I also agree with the statement that this applies mainly to quantitative methods. I do gel and tube antibody screens and have not had any issues with citations because it is just common knowledge that they don't correlate. I would definitely challenge that, and if you can drop JCAHO for lab accreditation. We are only CAP and the rest of the hospital is JCAHO. Works very well for us.

[Deny Morlino]

About 2 years ago the lab dropped JCAHO due to inconsistancies and large "gray" areas. We transitioned to COLA and the requirements are more stringent but also more "black and white". The hospital stayed with JCAHO up until last year and began the transition to NIAHO and ISO 9001. The fact that the regs were subject to wide and varied interpretation depending upon the inspection team was frustrating to say the least. Time is better streamlined now that we know what to expect as we constantly work to remain prepared for an inspection.

[Townsend]

I agree with all of the comments about this applying to quantitative testing and the purpose of the initial validation testing. However, that being said, with the new TRM.31450 in the BB CAP checklist, if you are CAP inspected, consider a gel vs tube correlation procedure at least semi-annually. I called CAP for clarification, and I was told that this is exactly what the new checklist item was referring to - gel vs tube (or another method, such as sold-phase). I even tried to stress the fact that the tube testing was only done as supplemental testing and not done instead of gel - but they weren't going for it!

Please reply if CAP tells any of you anything differently.... I would love to drop this testing!

Stephanie Townsend, MT(ASCP)SBB

[tbostock]

Just another unreasonable request by a regulatory agency. You are indeed comparing apples to oranges, but as Stephanie said, you're held to it if they make it a reg. We would just have lots of arguing to do if we had a 1+ in gel that was negative in tube testing, especially if they don't understand Blood Bank, which is true in some cases.

[Eagle Eye]

I agree with all of the comments about this applying to quantitative testing and the purpose of the initial validation testing. However, that being said, with the new TRM.31450 in the BB CAP checklist, if you are CAP inspected, consider a gel vs tube correlation procedure at least semi-annually. I called CAP for clarification, and I was told that this is exactly what the new checklist item was referring to - gel vs tube (or another method, such as sold-phase). I even tried to stress the fact that the tube testing was only done as supplemental testing and not done instead of gel - but they weren't going for it!

Please reply if CAP tells any of you anything differently.... I would love to drop this testing!

Stephanie Townsend, MT(ASCP)SBB

We got exact same response. I also argued in same direction and was told that if you get negative result with tube you can summarize and state /or ptove and state that tube is less sensitive than Gel. Basically your medical director signs off on the review.

[clmergen]

I am finalizing this procedure right now. I stated that we would run 10 samples in tube and on the Echo. Differences needed to be resolved and documented (rouleaux, weak antibody, etc). The one hospital with 2 Echos will have a QC comparison done. I am not doing any statistical analysis or anything at this point because of the low number of data points. I will do some kind of system comparison and have it signed off by the various medical directors.

This is being required in Micro also, it is a CMS requirement and CAP wants to keep their CLIA standing.

[Eagle Eye]

The number of specimen to do correlation are based on your policy. When we called CAP we were told that we even run only two specimen and we will be OK.

[RWOOD57]

With regulations like these, I can see why reform to the system is needed.

[csjuarez]

In response to the new CAP Checklist item (TRM.31450) we have begun using the sample from our CAP Transfusion Medicine-Automated (JAT) survey after we have run them on the ProVue and the results are submitted to CAP. We test those same samples in the tube for ABO/Rh and Antibody Screen and also manual gel. Since we get the survey 3x a year, we have our bases covered.

[huntilla]

The number of specimen to do correlation are based on your policy. When we called CAP we were told that we even run only two specimen and we will be OK.

We were just inspected by the CAP last month and I have run two to three samples every six months for method comparison and the inspector was satisfied by our process. We only compare the type and screen resultsfrom our Echo to the tube method. Anytime we have a murky antibody ID we are confirming results in the tube method for resolution, I did not receive any negative feedback regarding lack of documentation for Antibody ID comparison.

There will only be new and more onerous regulations developed as time goes on.

North Texas Blood Banker