August 3, 20231 yr comment_86258 On 8/1/2023 at 4:10 PM, applejw said: The beauty of the requirement is that there is no magic number of samples. Exactly! We have to do it, but make the plan 'reasonable' for your facility.
August 3, 20231 yr comment_86260 On 8/1/2023 at 10:10 PM, applejw said: The beauty of the requirement is that there is no magic number of samples. True, but the point is, to make this joke of a requirement worthwhile, rather than just a box ticking exercise, there should be specified specificities, but they won't do that because they know (or, rather, should know) that they could never get sufficient of "weak" antibodies of certain specificities, and that diluted "strong" antibodies will have a completely different association constant, and so using these diluted "strong" antibodies will serve to control/compare absolutely nothing.
August 14, 20231 yr comment_86351 If we have to hunt for a sample to use for this that will give consistently comparable results, we aren't testing the method, we are testing our ability to find a suitable sample. I heard that CAP will sell you one that will consistently give the same results. If we aren't going to change anything (can't recalibrate gel!) based on the answers we get (like chemistry would), then why are we doing this? I talked TJC out of it last inspection (I probably got a little heated over the stupidity of it because I had to come in the day after a concussion to meet with them, but maybe they took pity on me and didn't cite me because of my unfiltered brain). No one has been able to explain to me any meaningful takeaway from doing this comparison. If I am ever forced to do it, we will just keep copies of sample results that we run by two methods to solve a problem and make a note that they are acceptable because we expect these differences between methods. If anyone can give me a valid use for this, I would be very appreciative.
August 19, 20231 yr comment_86393 When TJC inspected us last year, we weren't cited but she recommended that we include in our QC policy for the lab why we don't do BB correlations. I finally got around to writing this up. Did I miss anything? Is anything in error? No published evidence or immunohematology expert supports periodic method correlations. Blood bank methods aren't expected to correlate perfectly. We use their differences to avoid rouleaux, clinically insignificant, and weak warm auto reactivity so we can detect any significant alloantibodies. No method will detect 100% of significant antibodies. If they don't correlate on a given sample, there is nothing to be done to "fix" the methods. Deviation from manufacturer's instructions violates FDA and other regulations. It would be extremely complex and probably futile to try to validate all changes that would make all specimens correlate. To select a specimen for correlation testing that is expected to correlate between methods while avoiding those that won't correlate is unscientific and a waste of resources.
August 21, 20231 yr comment_86402 I like it Mabel! We all need to start using that for our CAP inspections to see if we can shove them in the right direction.
August 21, 20231 yr comment_86410 On 8/19/2023 at 2:14 PM, Mabel Adams said: Blood bank methods aren't expected to correlate perfectly. We use their differences to avoid rouleaux, clinically insignificant, and weak warm auto reactivity so we can detect any significant alloantibodies. No method will detect 100% of significant antibodies. Blood bank methods aren't expected to correlate perfectly. We use their differences to avoid rouleaux, clinically insignificant, and weak warm auto reactivity so we can better detect any significant alloantibodies. No method will detect 100% of significant antibodies. I am going to tweak the above to say we can "better detect' antibodies so it works better with the next sentence and doesn't imply that we can detect "any" significant antibodies.
August 23, 20231 yr comment_86421 We have had HFAP until they were purchased by ACHC, and both of those have cited me for the lack of correlations: one for not performing them at all, although I pointed out that we do that on every specimen without previous records, and just last year because I didn't include crossmatch tests with my Type and Screen correlations. I contested the citation again, stating that it was the exact same methodology as the antibody screen, but was unsuccessful as their standards say "the same test using different methodologies ". I gave up and added XM as well. No concept of what we do and no common sense!
January 2, 20241 yr comment_87501 I have posted my battle with CAP over this requirement. I lost and waved the white flag. I will do the bare minimum (1 sample) for antibody identification for automated gel, manual gel and tube-LISS methods. Done. They said they would revisit the standard for this year but they have not.
Create an account or sign in to comment