INR values for a fresh frozen control plasma at the
therapeuticallevel with andwithout recalculation
of ISI. It was possible to reduce the interlaborato–
ry CV with 50% by recalculating the ISI values
stated by the manufactures according to the Da-
Figure 7.
Scandinavian EQAS, JNR values ojjrozen contra/plas–
ma ajter recalculation with the Danish JSJ calibrator.
: tl---=1.1- -
----1
20f---- -
~
15
:--------~~··~~-------
::>
z
10
.
Stago
.
Nycotest
INR-Values
•
Thrombotest
O
Other
All methods: Mean=2.36 INR, N=124, CV=3.9%
Figure 8.
Danish EQAS, Nov. 1998, on coagulation jaetors
Il+VII+X at anticoagulation level.
-· -
~-
- - ::::::-=.:_·==.. --
20 - ·
15 - '
~
E
10 -
::J
z
5 - .
2,1
J.,:,~~,--
2.3
2,5
--2:J2,
9 ·-
3
,
1
ManufacturerBIS~values
INRValues
All methods: Mean=2.44 INR, N=69, CV= 6.5%
nish mode!. The mean value of the fresh frozen
plasmawas not altered but the variation decreased
from 10.5% to 3.9%.
Additionally the survey included a freeze dried
plasma sample at the therapeuticallevel, but itwas
shown, that itwas not possible to reduce the varia–
tion of thismaterial by recalculating the ISI value.
Klinisk Kemi
i
Norden 3, 1999
Figure 9.
Danish EQAS, Nov. 1998, on coagulation jaetors
Il+Vll+X at moderate reduced leveL
25 - •
20
ii
15
E
~
10 -
INRValues
All methods: Mean=l.SO INR, N=68, CV=7.0%
Comparing of laboratories using the calibra–
tionmode! with those using anotherway ofca–
libration
Figure 8 and 9 shows the distribution of INR va–
lues obtained with frozen control plasma at the
anticoagulation leve! and the moderate reduced
level respectively, found at theDanishquality ass–
essment scheme in November 1998. Those Jabo–
ratories who have ehosen to adopt the ISI values
stated by the manufactures of test kits or use an–
other kind ofcalibration seems to fit veryweil with
the users of the Danish ISI recalculation mode!.
Only a few outliers were seen.
DISCUSSION
Thecalibrationmodel wasmetwithgeneral appro–
val inDenmark.
It
is an easyway for the laborato–
ries, even the small ones, to establish a reliable
ISI value in the local set up of the test.
Until now only fresh frozen plasma have been
found suitable for the purpose. A two-points cali–
bration has been used, one calibrator at the thera–
peuticallevel and one at the normal leve!. Eighty–
seven percent of all laboratories in Denmark use
the ISI calibrator and 72%use the normal calibra–
tor.Twenty-eight of the laboratories have ehosen
to continue with their own normal plasma or a
commercial freeze dried normal plasma. This in–
dicates that some manufactures of reagents are
providing their customers with ISI values of the
thromboplastins valid for thecoagulometers inuse.
The precisionof theprothrombin time has been
stable through the years since introducing of the
65
