Hi George. My name is Reynaldo Caparros. I know that it has been taught and drilled in our knowledge on how we should properly fill up blue tops. This has also resulted in numerous delays in patient care. My question is, are there any studies that have been done to check the possibility that short samples can be acceptable? I saw one study that was done on Sarstedt tubes. They found out that up to 70% fill rate is acceptable. Please help. Thank you.
Hello, Mr. Caparros, and thank you for your question, which raises an important issue. The article you’ve referred to is Sena Odabasi M, Yalcinkaya Kara ZM. Are tube fill volumes below 90% a rejection criterion for all coagulation tests? Lab Med. 2024;55:442-446. doi: 10.1093/labmed/lmad108. PMID: 38104249, available from your medical library. I also suggest two additional references, first, an open-access article, Gosselin RC, Marlar RA. Preanalytical variables in coagulation testing: setting the stage for accurate results. Semin Thromb Hemost. 2019;45:433–48. doi: 10.1055/s-0039-1692700. PMID: 31291676, and also our May 22, 2025 Fritsma Factor post, Impact of Sodium Citrate Levels on Coagulation Assays.
The definitive document for hemostasis specimen management is CLSI, Collection, Transport, and Specimen Processing for Testing Plasma-based Coagulation Assays. 6th Edition, CLSI Guideline H21, 2024, available from the Clinical and Laboratory Standards Institute. H-21 maintains the 90% fill rate “unless in-house studies are performed,” and is based upon articles published prior to 2001.
Combining the information from these publications, your facility may modify the minimum fill volume requirement, provided your clot-based assays are proven to be unaffected by the citrate concentration at the new limits. The modification may help you reduce the number of “re-collects.”
Here is the open-access article that Dave McGlasson references in his comment below: McGlasson DL, Kaczor DA, Krasuski RA, et al. Effects of pre-analytical variables on the anti-activated factor X chromogenic assay when monitoring unfractionated heparin and low molecular weight heparin anticoagulation. Blood Coagul Fibrinolysis. 2005;16:173–6. doi: 10.1097/01.mbc.0000164424.90545.6e. PMID: 15795534.
- Dr. Ali Sadeghi-Khomami, author of the poster, “Impact of Sodium Citrate Levels on Coagulation Assays, reference above, also provided the following materials from the WFH:
- Citrate adjustment for HCT > 55%
- Short draw differences
Hello George, Thank you for inviting me for comment on this issue.
Although the issues discussed so far are interesting, I think that everyone in this thread have dealt with them very successfully, and I dare to say that I believe this to be true for most labs) I am sure that we all have contributed to the policies and procedures that currently exist that address these issues and they seem to be working.
So with this in mind, I would like to take this discussion in a different direction by making some thought provoking ideas that I think can make future progress:
1. Samples–Fill: We have all been around for a long time and I wonder if you remember that at one time BD had a blue bar on their label that represented proper fill (don’t know if it is still there). Seems pretty simple as a great visual aid for phlebotomy, nursing, etc. to use. Should we encourage all tube makers to use this idea? This certainly becomes very helpful in those situations where the cap is popped. And for those very difficult infant/pediatric sticks why can’t we make a better micro system to draw their blood? Our instruments can now deal with micro samples.
2. Sample Integrity: can’t we find a way to use the power of computers (AI) to help validate or disqualify results? I can think of studies to do and I am sure you all can also. And what about using the power of computer to deal with HCT issues? Surely rules could be made for a LIS to identify high or low HCTs from CBCs that would automatically send an alert to all coagulation orders identifying this. And imagine that tube companies can make tubes that could be used in these high low ranges. We would have a normal, high and low HCT tube. May sound crazy but it would sure beat adjusting citrate and as far as I know this has never been considered or looked at. And what about hemolyzed or lipemic samples? Can’t we use the power of computer capabilities to help drive decision making with these samples? Reject, accept,rerun,resolute,etc. I would love to see a brain storming session(s) on using all the new technologies and computer power available.
3. Methodologies: One simple comment here. “Perhaps we should go chromogenic.“ 30 years ago I worked with a Chemistry Director who every time he saw me he would say, “In chemistry we could do coagulation better than you after all all it is is enzyme chemistry and immunology.“ I now think that he may have been correct. 🤪
Final thought: remember that automation is removing the technical skills that we all learned to deal with these issues so we should raise the bar of expectations instead of rehashing the past. I can imagine a lot of great studies…
Well I hope that this different spin encourages intellectual dialogue on how we can look to the future of making coagulation better for the lab and the people we serve.
Best regards to my colleagues, Dan Kaczor
It depends on the assay. For example, APTT tests are less affected by short collections than PT/INR. I doubt that for the Sarstedt tube study “They found out that up to 70% fill rate is acceptable” for all tests, but possibly for APTT? So, you need to assess in your lab, what fill rate is acceptable for which tests. Also, you might make a pragmatic judgement to avoid recollections (e.g., pediatrics) if a test result (e.g., antithrombin level, factor level) is normal on a tested underfilled tube (since in theory, the result will still be normal, albeit higher, in a correctly filled tube). But if you accept a blanket “70% fill rate is acceptable for all hemostasis tests” without proper validation, then that is not acceptable. I think there is good data to show that a 90% fill is acceptable for most hemostasis tests, and that is why that is accepted as a universal cut-off for recollection.
Hi everyone, This is a very interesting topic to me because I didn’t find any systematic report investigating citrate level increase in plasma and its impact on coagulation assays. I couldn’t even find clear scientific evidence why 3.2% citrate tubes are better than 3.8% citrate tubes.
To make that more interesting for you, let me ask you this simple question. What would be the outcome of citrate increase in plasma on PT, APTT, dRVVT assay results?
A. Clot time increase
B. Clot time decrease
C. It depends on the assay
From the past, we have data from clinical studies or reports for blood collection on different citrate tubes under extreme conditions but not a systematic citrate variation in plasma without other variables. We did a study reported in the poster that George linked in his response: in fact, the right answer is C. So, it is not as simple as chelating calcium of the plasma sample, citrate could also impact ionic strength or altogether eliminate calcium from reagents in the assay with a surprising outcome. In our hands, IL’s PT and APTT on MAX3 analyzers gave a shorter clotting time with increasing citrate! Weird right?
In the publication George attached, I want to call attention to the non-calcium assays for all of the variables. There was no clinical or statistical significance with any of the anti-Xa chromogenic testing for unfractionated heparin or low molecular results. This includes three different types of anticoagulant tubes: 3.8% and 3.2% sodium citrate and CTAD collection tubes using bothe 9:1 and 6:1 blood to anticoagulant ratios. I look forward to your comments.
I think there have been sufficient studies to suggest the impact of short collections of citrate tubes on hemostasis testing, and I believe Dot/Richard as well as the UK clan in Sheffield (Kitchen and other gang members) have published on this topic. The excess citrate has an effect on calcium dependent assays, whereas in other assays the issue may be a dilutional effect.
But it’s not all black and white…what if you have a short sample from a patient with an HCT of 25%. Are those results really affected? What is published is the effect of improper citrate:plasma ratio (it’s not citrate to whole blood), but to my knowledge no one has addressed short sampling of citrate tubes in anemia. Perhaps some young eager beaver can consider this project.
As a side bar, I also think short sampling tubes may lead to dilution effect, but whether that dilution effect is clinically significant remains to be elucidated and likely a local issue with stakeholder. This is especially relevant for timed collections, on vulnerable populations (neonates/kids/pregnancy/etc) so I am not a huge fan of blanket rejections, especially if the testing is not calcium dependent. If I get a short sample result for an AT of 90% when hemodilution is the concern, I will consult the ordering provider, go through the hemodilution effect (estimated bias) but a redraw is only going to be “more normal” than a short sample collection.
Bottom line: case by case basis for me…
BG
Thank you very much George. This is very informative and helpful.