Chris Powell at King’s Daughters’ Health asks: commercial plasma distributors offer pre-screened patient specimens for PT and PTT lot rollover. How can this product be used and still meet the CAP and CLIA requirement to collect specimens from your own patient population?
Hi, Chris, thank you for your question. I’ve modified your reference to a specific high-quality product so that I could use generic objective terminology. We’ve sent this inquiry to a member of the CAP coagulation committee: “Is there a CAP guideline that indicates it is acceptable to use a commercial set of normal donors if a laboratory cannot collect enough from their local population?” We hope to receive an answer today or tomorrow, and will post it here. Meanwhile, I’ve checked with local coagulation specialty labs and learned they routinely use commercial plasma sets for PT and PTT lot-to-lot validations and to update reference intervals with no criticism from CAP assessors. Even large tertiary care facilities find that collecting, screening, labeling, and storing normal donor sets that meet GLP standards for safety and accuracy demands high-level technical expertise and is time-consuming. It is more cost-effective to use commercially prepared products. Let’s wait to hear from CAP.
From Robert Gosselin, UC Davis Healthcare, 5-5-16: Hey GF, Not sure about commercial plasma, as I am not certain how the samples are collected. If collected in tubes, probably okay. If collected in ACD or equivalent from a blood bank, then probably not so hot as citrate ratio may be off. Better question would be how many is sufficient, and just using normal or also abnormals? Factor deficient (all or some) vs depleted samples? Too bad coag is such bad science, otherwise these questions would be easier to answer… BG
From Thomas Exner, Haematex Research, May 6: My humble opinion is that if the aim is to ensure the continuity of reagent/system specificity and sensitivity then the test plasmas to be used should be selected to cover the range of abnormalities for which the test is used. This would include over-citrated and under-citrated normal and abnormal plasmas. Probably focus on abnormals which are of most concern to the particular clinic. Commercial plasmas should be included but you should probably not rely just on one supplier. You simply do not want to see any test plasmas giving results likely to cause clinical mis-diagnosis. Of course the numbers should be kept to a workable level but enough to ensure statistical significance. Hoping this helps. Cheers from Tom Exner.
From Donna Castellone, May 5: Hi everyone, when people ask me about using this I tell them if you can’t get sufficient samples, is better than doing nothing–almost–but only after really trying to obtain samples in house. I would prefer to see them use in house patients, even if they are pooled and frozen; collected across the reportable range, filled in with some normal subjects (usually lab people), and fresh patients run over a period of time. It shouldn’t be done in one day. So it would appear most places can get this done. They used to purchase these at my institution, and I stopped it, most were normal- we had d-dimers that resulted in a 0 result. It may take some creative thinking but most places can get their own samples to do this study. Hope this helps. Donna
From Dr. Emmanuel Favaloro, May 6: Hi George, I have been following the conversation thread! Our lab looked at this ages ago; I see that you removed the identity of the specific plasma sets mentioned, but I only know of one commercial supplier, although Dave McGlasson indicated there were 2. Anyway, we found that the sets gave similar results to true individual normal plasma sets for some tests but not for other tests, so we decided to move away from these in general towards a ‘transference’ model instead. We have published the ‘transference’ model approach for PT/INR, but not for the other assays. And we never published the commercial set study data–maybe we should, given the level of interest. Happy to expand on this if there is any interest/questions arising. I guess, I should qualify to say that we (in Australia) are not burdened by the same regulatory restraints as you folk in the USA–“the land of the free” we hear, but I keep seeing it more and more as “the land of the red tape!” If your regulators have their way, you guys will be in even worse shape soon. cheers, Emmanuel Favaloro.
Additional from Bob Gosselin: I suppose, on second review, if the commercial plasma material contain both normal and abnormal samples, with the latter representing the institution’s patient population, and the samples are run on both current and new lot reagents, then any bias associated with sample (e.g. citrate) would be systematic. Then it’s merely a matter of paired-t test to see whether differences exists, and whether those differences require reference range (or HTR) change.
From George: I’d like to see a comparison of lot-to-lot validation outcomes and reference intervals established both by using commercial plasma and locally collected normal specimens. There are two manufacturers, George King BioMedical and Precision BioLogic, who produce plasma sets. The sets represent the entire analytical range.
Added by George on 5-8-16, I’ve been in beautiful Portland, Oregon since Friday with the opportunity to speak for the PALCOE group, a group of knowledgeable and dedicated Medical Laboratory Scientist.s who meet quarterly on Saturdays to discuss current In Vitro Diagnostics issues. One participant told the group that she uses commercial plasmas for lot-to-lot validation, or “rollover,” but prefers to use locally procured normal blood specimens to prepare reference intervals. Like many lab scientists these days, she provides a mean and reference interval for a series of institutionally related laboratories. Her approach is to have each participating laboratory provide specimens and to then provide a composite mean and reference interval using data from the participants.