For our October, 2018 Quick Question we asked, “For lupus anticoagulant testing, how do you express the DRVVT screen and confirm results?”
Here are the answers from 36 respondents:
a. Ratio of screen to confirm: 17 = 47%
b. Ratio of screen to confirm normalized to MRI: 17 = 47%
c. Difference of screen and confirm in seconds: 0
d. Difference of screen and confirm in seconds normalized to MRI: 2 = 6%
The even split between “normalizers” and “non-normalizers” illustrates a difference of opinion in our community. CLSI standard H60 requires normalization for both the CRVVT-based and the PTT-based lupus anticoagulant test profiles, quoting several sources. The “non-normalizers” quote data that show normalizing rarely affects the diagnosis but adds unnecessary mathematical complexity. The debate is likely to continue. I’ve attached article summarized in Mr. McGlasson’s comment below:
As an aside, the term “screen” is a poor choice, as a screening test is a test employed on unselected individuals, whereas lupus anticoagulant testing is typically ordered for an indication such as pregnancy loss or unprovoked thrombotic events, and is often performed in follow-up to a prolonged PTT or an initial mixing study. Looking for a better term than screen, I’ve seen “initial” or “first-tier.” What is your preference?
Kristin, how did you compute
Kristin, how did you compute the normalized ratio. Did you use the MRI or the PNP? Also, Coumadin will interfere with many aspects of the DRVVT particularly the FII and FX. Very interested in these findings you expressed.
At the University of Vermont
At the University of Vermont Medical Center we also performed our own study on the normalized ratio as it is a CLSI guideline and recommended in the manufacturer’s insert. We ran 23 samples that had an elevated INR for DRVVT screen and confirm using the standard ratio and also the normalized ratio. Nine patients that had a positive ratio with the standard ratio were no longer positive when calculated with the normalized ratio. Further investigation showed that these patients were on warfarin for afib and had no history of LA. Patients with a history of LA remained positive when calculated with the normalized ratio. Although the normalized ratio may not increase detection of LA, the normalized ratio is an important step that a laboratory can take to rule out false positive results that can occur from an elevated INR whether from Warfarin therapy, vitamin K deficiency or other factor deficiency. False positive results can end up in the medical record and lead to a diagnosis of LA and long term anticoagulation. Since we do not have control of all of the orders we receive, or access to all patient information, we feel it is important to take every step to rule out false positive results.
Email from Dave McGlasson:
Email from Dave McGlasson:
We computed the DRVVscreen/DRVVconfirm ratio three ways using reagent sets from five manufacturers. We first computed the reference interval (RI) using +/- 3SD for all five reagents from 42 lupus anticoagulant (LA) negative plasmas. The results normalized to the mean of the RI (MRI) was 1.000–1.002, normalized to the pooled normal plasma was 0.941–1.035, and the uncorrected screen/confirm ratio range was 1.014–1.064. We then tested 43 LA positive plasmas using all three calculations. MRI normalization yielded a range of 1.797–2.010, PNP normalization produced range of 1.838–1.916. The uncorrected range was 1.877–2.042. The uncorrected ratios correctly identified 42 of the 43 LA-positive patients. In performing all three computations, we found no clinically significant difference. We concluded there was no need to perform a second mathematical computation, especially considering the possibility for a manual computation error. The article is attached above.