Trm.40660

[adiescast]

I found this question as I was performing our annual self inspection. I am curious as to how others are handling the annual re-validation of the electronic crossmatch. I would hate to have to repeat the complete validation annually. I was thinking about just demonstrating that the expected QA failures work and that it will allow the crossmatch under the appropriate situation. Any thoughts on whether that would be enough or other ideas of how to approach this?

TRM.40660

Phase II

N/A YES NO

If the facility performs electronic crossmatches, is there documentation of on-site validation prior to implementation, at the time of each alteration that may affect function, and at least annually?

NOTE: Without alteration of the system, the electronic crossmatch function would be expected to operate successfully. However, given the importance of this function, periodic (at least annual) challenges of the system are necessary, to verify detection of clinically important incompatibilities.

[bmarotto]

Well, I would not do a complete validation every year. In the guidence, the y use the term "periodic challenges of the system". I would take that to mean you would test at least one example of the following scenarios to document the system blocks electronic crossmatch.

1. Patient does not have two ABO determinations

2. Patient has a positive antibody screen with clinically significant antibody identified

3. Patient has current NEGATIVE antibody screen with a history of a clinically significant antibody

4. ABO incompatible XM combination (A to O, A to B, B to A, etc)

5. Testing on current sample is not complete

6. You manually add an antibody to a patient BB file but there is a current order and an electronic XM unit has been been issued. This simulates a case where the patient or another hospital provides antibody information after you have performed an electronic crossmatch. The system should NOT allow additional units to be electronically crossmatched.

This is off the top of my head. I am sure there are other scenarios you could test. I would be interested to hear how others decide to address this question.

[bmarotto]

Actually, a follow-up to scenario 6 would be to have a unit already electronically crossmatched before manually adding the antibody. The system should alert you if you try to issue the unit.

[bmarotto]

Here is a link to an interesting discussion of CAP TRM.40660 where several experts question the benefit of annual validation of ELXM. It sounds like the question will be reworded again in the next CAP checklist. Until then, I think we can get by with minimal annual testing.

http://www.cbbsweb.org/enf/2009/electronicxm_validationcap.php