Implementation and Standardisation Problems of
International Normalised Ratio in Anticoagulation
Therapy inDenmark
KARINKYNDE,
Department ofClinical Biochemistry, Roskilde County Hospital, DK-4000 Roskilde, Denmark.
INTRODUCTION
In viewofthe increasing need for monitoring anti–
coagulation therapy the International Committee
on Thrombosis and Haemostasis and the Interna–
tional Committee for Standardisation in Haema–
tology in 1985 recommended the use of the new
unit, called the International NormalisedRatio, ab–
breviated INR, as an expression for the activity of
coagulation factors Il+VII+X or the prothrombin
time.
The use of the INR calculation allowed corre–
lation with the recommended levels fortreatment
of patients in anticoagulation therapy and should
overcome the problems in comparing the results
reported by different laboratories, independent of
the source of thromboplastins in the reagents of
the various test kits.
TheCommitteeonAnalytical Qua!ity underThe
Danish Society ofClinical Chemistry introduced
regularly extern quality assessment schemes on
coagulation in 1989, and found an extreme!y high
interlaboratory variation of the prothrombin time.
Measured in arbitrary units it demonstrated a co–
efficient of variation (CV) of about 20%.
Implementation of the INR calculation in the
beginningof the nineties clearly demonstrated, that
INR-normalisation of the prothrombin time itself
was not a sufficient too! for reducing the !arge in–
terlaboratory variation. Furthermore, marked dif–
ferences among the various commercial test kits
were observed, indicating divergence among dif–
ferent sources of thromboplastins resulting in dif–
ferent INR levels.
The reasons for theobserveddiscrepancies seem
to be due to the International Sensitivity Index,
abbreviated ISI, values of the thromboplastin rea–
gents used by the laboratories. However, even la–
boratories using the same reagent had a consider-
Klinisk Kemi
i
Norden3, 1999
able variation, indicating that other factors, such
as the choice of normal plasma and type of coa–
gulometer, play an important role.
This stressed the need for each laboratory to
determination the ISI-value of the thromboplastin
used in the local analytical system. The recommen–
ded way of establishing a local ISI value using
samples from 20 normal persons and 60 persons
in stable anticoagulation therapywill mean an un–
acceptable workload for each laboratory.
METHODSANDMATERIALS
In an effort to overcome the problems of the stan–
dardisation, the Committee onAnalytical Quality
under Danish Society of Clinical Chemistry in
1992 introduced a too! in form of a national refe–
rence plasma, the so-called ISI calibrator with an
assigned INR value.This calibrator should be used
fordetermine the ISI value of the reagent in use to
the local setup of the test.
The choice of the material for calibration pur–
poseorcontrolmaterial is critical. Lyophilisedma–
terial, normally used by the manufactures of test
kits, seems unsuitable for calibration or externa)
quality assessment schemes due to matrix effects.
Thus, fresh frozen pooled material was chosen.
ISI Calibrator
The first Danish ISI calibrator was a liquid, fresh
frozen pooled plasma preparation from 20 volun–
tary patients in stable anticoagulation therapy gi–
ving their consen
t.
Two hundredmi ofbloodwere
drawn from each patient in siliconised glass tubes
containing trisodium citrate 3.8 w/v%. The tubes
were centrifugedwithin one hour at 3000G for 15
minutes. Each plasma samplewas checked for anti–
HIV,HbsAg and anti-HCV. Plasma was pooled in
61
