Here is an interesting question from Dr Larry Smith at Memorial Sloan-Kettering:

George,

What is the consensus on how to handle samples drawn from a mediport that are obviously contaminated with heparin?  We run a thrombin time to prove that they are contaminated, and request a peripheral draw.  Sometimes we neutralize the heparin with a heparinase when the patients are not on heparin.  Often the medical staff will insist on the result from the contaminated sample and object to recollecting a peripheral draw.  How do we differentiate between a patient who is actually on heparin and the sample is contaminated?

Thanks, Larry.

Here is another question from my friend and colleague, Vicki Cardone, Hematology Supervisor at Children's Hospital of Alabama, just across the street from my own University of Alabama Hospital:

Would a hemolyzed sample cause the anti-Xa value to be falsely decreased as it would give a stronger color reaction, or am I just way off?

Thanks, Vicki

I want to draw your attention to a useful new comment posted Friday, June 6 by "Stacy" on the "Checking for Clots" topic in "Specimen Collection."

Members of Pat Letender's Medlab-L have been discussing the effect of hemolysis on coagulation testing. This statement appears in CLSI standard H21-A4:

Samples that have visible hemolysis should not be used because of possible clotting factor activation and end point measurement interference. Some current instruments using an optical detector may have problems with end point determinations on samples that are icteric, lipemic, or contain substances that interfere with light transmission. Alternative methods (e.g., mechanical/electromechanical) should be considered.

There is no reference associated with this speculative statement, so just for fun, I'd like to throw a little heat into the discussion by asking if there is any authoritative evidence for rejecting hemolyzed specimens.

We thought we had heard the last of this discussion, but one question remains. While a lot of work was done supporting the use of 3.2% sodium citrate for coagulation testing specimens, little was done to generalize it to platelet testing. We are using 3.2% citrate specimens for aggregometry, PFA-100, and Accumetrics Ultegra, but I've seen very little to support its validity in those tests. Please pass on what you know:

1. Have you recorded or published studies supporting 3.2% citrate specimens for platelet testing?

2. Are you continuing to require 3.8% citrate for platelet testing?

I've been seeing several posts in Medlab-L about the need for a discard tube when collecting hemostasis specimens using butterfy needle sets. The air within butterfly set tubing occupies approximately 0.5 mL and reduces first tube collection by that volume. For a 3.0 mL total collection tube, this is a 16% reduction in total final volume. Most manufacturers assert that anything over a 10% reduction prolongs PT and PTT results as it causes a relative increase in the anticoagulant to blood ratio.

We've just posted three modules on thrombophilia and one on hemostasis blood specimen collection. Be sure to visit these and take advantage of the opportunity for continuing education credit through CACMLE. Geo.

This is a question from my colleague and neighbor, Vicki Cardone.

Do you feel slight hemolysis would compromise PT/PTT results?  If so would you recollect or add a comment to let physician know the results may be falsely decreased. We are a Children's Hospital and do not want to recollect unnecessarily. Thanks. Vicki Cardone, Children's Hospital of Alabama.

Sue Osier from Northside Hospital in Atlanta asks whether anyone is using Greiner specimen collection tubes and whether they seem to be causing any trouble on Stago equipment, specifically, the STA-R and Sapphire.

Rita Burns sent me this message on January 8: How do you handle a strong cold agglutinin for coagulation testing?  This particular sample generated dye errors on the Biomerieux (Trinity Biotech, Inc) MDA180.  After warming the sample on the unit, we were able to retrieive results with no error flags, but we are still concerned as to its effect of falsely prolonging results.